I'm sorry it didn't work out, but maybe it will work out someday. I have absolutely no idea about electronics or how to do this kind of work, although I have a better idea now than I did before I watched your video. Well done, Tony!
Sorry about my delayed response. Far too many electronic and honey do projects. While my setup may be capable of measuring T/S Parameters, I'd never done it before and learned a lot from your video. What I found quite annoying were the resonances in our family rooms and cars when setting up or replacing speakers. A sweep test of my new 1995 Mercedes C-280 found a resonance around 45 Hertz, caused by a loose speaker mount. A 35 Hertz resonance was caused by 2 plastic wire clamps an assembler had left on the rear sub-deck. At resonance, they bounced like Mexican jumping beans. Same with some loose hardware in my 2011 Camry Hybrid, after I'd replaced a speaker with voice coil drag. Velcro, duct tape, pieces of foam, RTV and cable ties are great for damping resonances. The most recent were the two metal straps I'd used to earthquake-proof our 7 foot high grandfather clock. Without them, you could easily push the clock over with a finger. There was a strange noise when the clock chimed or the subwoofer hit the right note. Neither the clock person nor I could locate the noise. That's because I'd disconnected the straps so he could look at the rear of the clock. When I reconnected the straps. I realized one of them was resonating and kissing the top of the clock enclosure. A strip of fuzzy Velcro under both straps stopped the racket. Velcro also works for bookshelves of rattling teacups and under the bottom of rattling picture frames. Also keeps them from tilting.
Those are great and innovative solutions. Great job. My biggest resonant sources are two glass cabinet doors. That is, each cabinet door has a pane of glass in a wooden frame. I'm sure a little hot glue will fix it, it's just low on my list of priorities. I usually forget about them until I'm playing bass and they start rattling. But at that point, I'd usually prefer to just play and ignore the rattle rather than fix it. I'd make a mental note about it, then promptly forget. :D
glad i found this video, i have 2 mid-range woofer with no parameters, the only parameters that i have is that the RMS power is 80W, and the impedance is 8 ohm, i've tried this once, but i didnt do a great job to calculate it, and i already made the enclosure for it sadly, probably gonna re-measure it and try to fix the enclosure
Good luck to you. I hope it all works out. My numbers ended up kinda weird, but the cabinet still sounds pretty good, just not very loud. But the amp has a DI, so I guess it doesn't matter all that much. However, I haven't tried it with an actual drummer, so I have no idea if I can even hear it. Thanks for watching.
have you heard about software called WinISD, its basically a software to design a box based on the parameters, but more complex than the website version
I have not heard of that software. Even so, my goal was to keep the project as inexpensive as possible. And even if it were free, I am very wary of downloading software from random websites. Besides, the website versions were good enough to show me the speaker was probably junk.
its free, i've download it, and i already scanned it with my bitdefender anti-virus and its fine, yeah you're right, there's no point to use complicated software for such a cheap speaker tho, if you curious about the software, you can check it here www.linearteam.org/ . Anyway i already measure my speaker, it's turn out my speaker Qes is worse than yours, mine about 6.22, its cheap tho for a 5" woofer with no brand, around $3-4, its so hard to get a good quality speaker without a lot of money here, so yeah maybe i just stick with this woofer for awhile.. , there's an alternative for the rice, you can use kinetic sand, its not sticky and easy to clean.
Not that it means a lot, but on the Low Resistance Range (typically 200 Ohms) both Digital and Analog meters have test lead errors as much as 1/2 Ohm. The easiest way to compensate is to short the probe tips together and note the reading as "Test Lead Resistance". Subtract it from all real measurements. If you change the meter Range to 2k or higher, lead Resistance is negligible. Another way to measure the Voice coil resistance is to connect a fresh 1.5V D battery (NO HIGHER VOLTAGE!!!) to the cone through that same 10 Ohm resistor and a series-connected multimeter on the 2 Amp DC Range. For a 2 Ohm Voice Coil, the meter should read around (1.5V / 12 Ohms) or 125 milliAmperes on the 200 milliAmpere DC Range. The internal Voltage drop of the meter and its internal fuse may give a slightly lower Current value. No Big Deal! Just record the measured Current. Then take a second meter on the 2 Volt (or 2000 mV) DC Range and connect it DIRECTLY across the speaker's Voice Coil Terminals. Use Ohm's Law, R = E/I to calculate the Voice Coil Resistance. What you've just done is taken a 4 Terminal Kelvin measurement, one of the most accurate ways to measure Resistance, especially low Resistance. The Current reading already includes the 10 Ohm resistor's Resistance, plus the Current meter's internal and lead resistance. The lead resistance of the Voltage-reading meter is inconsequential. What's left to affect your result is only the accuracy of the two meters. One other point - the most accurate range for a Digital meter is the 200 millivolt range. Most meters use the Dual Slope technique, where the internal Battery Voltage, the internal Rs and Cs, the internal dividers, shunts, adjustments, ramps and clock frequency cancel, leaving only the internal Voltage Reference. If properly factory-tweaked, the 200 milliVolt range can easily be 10 times better than the stated accuracy.
Excellent! This is great information. I will definitely use this in other projects I plan on doing. Especially the note about the digital meters and errors at low resistance. Thank you for watching and commenting.
While generally not as accurate, old Analog Ohmmeters, like my Micronta 43-491 (Radio Shack in Amsterdam) were subject to battery aging and had a Zero adjustment, also perfect for zeroing lead resistance. Whenever I worked on world-wide CTs or MRIs in the 70s, 80s and 90s, I always scouted the local Radio Shacks as soon as I arrived in town. You can't believe how many multi-million $$$ scanners were diagnosed with, and actually contained Radio Shack parts.
The described technique can also be used to measure the 10ohm resistor, or any low value resistor for that matter. If you have a lab PSU, it is very easy to set some reasonable current limit, then measure the voltage drop directly on the resistor or resistive-anything. If you don't trust the current reading of the PSU, then either include an amper meter in series too, but is cruical to always measure the voltage drop on the device under test, as close as possible. You can measure miliohms this way if you can pump out amps from your PSU (and the device can tolerate it, eg. you measuring a cable). I'd like to point out that measuring AC voltage might be inaccurate with multimeters if the signal is way off of 50/60 Hz. Might be accurate, but you can only be sure, if the datasheet guarantees it in some frequency range, or if you check it with either a known accurate flat response function generator or an another known accurate voltmeter. It's not a problem in this subwoofer range, but might be higher up in the audio spectrum.
I agree with Gabor. Whenever you're doing low resistance measurements, the 4 terminal or Kelvin connection overcomes lead and contact resistance by ensuring no substantial current flows through the Digital Multi-meter's test leads. When you measure a low value resistor using the DMM, by itself, the small, but significant current the DMM applies to the resistor-under-test also flows through the DMM's test leads, causing an (I)(R) product Voltage drop. By itself, the DMM can't separate the lead's (I)(R) drop from the resistor-under-test (I)(R) drop. You have to subtract the test lead (I)(R). On an old analog meter, you can usually re-zero or null the test lead (I)(R) drop. Or you can use the Kelvin connection. en.wikipedia.org/wiki/Four-terminal_sensing Lord Kelvin is the same engineer/physicist who gave us the first two Laws of Thermodynamics. The Kelvin temperature scale is also named after him. 0 degrees K ~ - 273.16 degrees Celsius or Absolute Zero, where (at least in 2022) all molecular or entropic motion ceases and all pure resistances become noiseless. en.wikipedia.org/wiki/William_Thomson,_1st_Baron_Kelvin
As far as an update, it's still going fine. Haven't had any reason to disassemble the cabinet I put it in and start over. Though it might be too quiet, but I'm not sure if the amp is too weak or if the box is limiting the output. If I ever need to get it louder, I might re-do this video and get more accurate parameters.
Obviously, the measurement error is due to the use of rice as an additional weight. The rice rattled audibly. This means that the additional weight of 50 grams was not firmly coupled to the diaphragm of the loudspeaker. The measurement was effective with far less than 50 grams and therefore the resonance frequency was much too high. With this wrong frequency one receives a much too high value for Qts in the calculations. For a correct measurement it is absolutely necessary to use a compact mass as additional weight and to ensure that it holds firmly to the diaphragm. Modelling clay is ideal for this purpose. In my experience, these cheap loudspeakers have a Qts of about 0.8 to 1.3. This makes it impossible to achieve a linear frequency response without manipulating the frequency response of the amplifier. Nevertheless, it will be possible to achieve a passable sound in a suitable cabinet. With instrumental loudspeakers, linearity is also less important. Often it is the objective errors that make these speakers sound attractive.
Now that you mention it, that makes complete sense. I didn't have any clay on-hand, however. Though I agree that in music, especially rock based music, such errors are attractive to some listeners. The Kinks were famous (infamous, maybe?) for poking holes in the speaker cones to get the sound they were after. I guess at that point T/S parameters really aren't worth much. Thanks for watching.
@depunkt Yeah, I probably should have used some clay. But I didn't have that available. If you are trying to get the T/S parameters then you can use this method (substitute the clay for rice). Thanks for watching.
In my opinion, this may be an older video, but it's relavancy is just as applicable, as when it was produced; & it has been really well explained. I cannot understand why it only has a 'like' rating of ~1.45%, per views. I have acquired a couple of older Sony 8" subwoofers, pulled from cheap audio equipment. In attemtping to research the design requirements, for a custom subwoofer, I quickly discovered the issues surrounding using drivers, with unknown parameters, which led me here, eventually. And, I have found out, that Sony do NOT share their tech: what a surprise. Even though I know the speaker models, any technicals are not to be had, on the internet. As someone who is not an audio guy, but has some electronics' ability, I preferred a video that was presented in plain speak. Your idea of using a phone, with a frequency app is just what the doc ordered 👌. Given my Denon AVR is bluetooth capable, I am able to simplify the connections, somewhat 😵💫: headphone jack>jack to 3.5mm jack>3.5mm to RCA adaptor cable. Then, a cut off RCA cable, with a 10ohm resistor soldered to the positive? I think I'll just do the lazy way, & use the supplied spreadsheet. Thanks, mate.
Thank you. I think a lot of people disliked that I used rice instead of clay or playdoh to weight the cone. But other than that I think most other things went pretty well. There is another video where I attempt to design an enclosure for this speaker that didn't go very well due a misprint of a formula on the website. I should go back and fix it, but it's pretty low on my priority list.
@@HandlebarWorkshops Thanks for letting me know, you are still monitoring this video. I have a question, that may/may not relate to your measurement parameters: the amplifier you showed, does it fit the stipulated requirements -ie. 1-10W RMS, & an output impedance of
@@citizengkar7824 While doing research for this video, I did not read anything about output impedance, nor about the RMS. This method isn't meant to be (nor can it be) extremely accurate, but doable for someone to do once or twice without investing in expensive tools. It's meant to get close enough to get a good (but not necessarily a HiFi) sound out of a speaker. After all, since it is a fairly unknown speaker, you probably didn't pay much for it anyway. You just want to make it sound as good as possible without spending too much more money. It would be cheaper to buy a new speaker with known parameters than to buy all of the scientific tools to do this at a high level of accuracy.
@@HandlebarWorkshops I do agree all you've said. Though, it does state those requirements, on the Elliott site, you've linked. "1.1 Measuring Re, Fs, Qes, Qms and Qts To measure these parameters using the method outlined below, you'll need to have the following items: A power amplifier, rated at 1-10 Watts (RMS) or thereabouts (must have low output impedance
That Elliott site has changed since I used it for this video. For a while it was a 404, then when it came back one of the formulas was different. I haven't gone back to see if that new formula would result in a better end result with this speaker. I don't remember any of that list being there, but it's been years, so my memory of all that may be a bit off.
You're welcome. When I made this, there were no other videos on this subject that I could find. I thought it needed to more available. This was also my very first youtube video. Now I occasionally see other videos show up in my feed by more established channels. So even if mine wasn't the best, it was the first. :D Thanks for taking the time to watch and comment.
The real test of a subwoofer is how much you can feel its infrasonic (sub 20 Hz) output without hearing it. 30 years ago, I tested what was one of the best subwoofers on the market - an 18 inch servo'd Velodyne, in the listening room of a high end audio dealer in San Mateo, CA. I didn't have my SPL meter at the time. At minimum volume, I connected my CD player to the Velodyne Input and selected the 16 Hz CD test track. I turned up the volume until I heard something, then backed off a few dB. There was no audible sound coming from the sub, but the windows of the listening room were vibrating. Fortunately, nothing else in the room was audibly resonating. I WAG'd the sub's SPL at maybe 120 dB. At infrasonic frequencies, you should be able to feel and even see the output of a good subwoofer without hearing it. Anything you hear is either harmonic distortion or some nonlinearly-resonating object in the room. You may see things like windows or paintings vibrating - OK as long as you can't hear them. I periodically sweep our listening room for irritating resonances, which I usually damp with pieces of hidden Velcro. Some things need caulking. During one sweep of our new place, I heard a rasping noise from an adjacent room. It was the metal straps I'd used to earthquake-proof our grandfather clock. Some Velcro under the straps killed the noise. After sweeping a few rooms, you learn to exercise "a bit of healthy engineering paranoia". Ask any pilot - zero or insufficient paranoia is dangerous. Too much becomes "engineering or pilot paralysis" - the fear of making the wrong decision or making no decision at all.
I use this speaker in the cabinet of my bass guitar amp. This is by no means a high quality amp. My friend who gave it to me said it was from a cheap 80's keyboard amp that got ruined when his basement flooded. I wasn't looking for high fidelity nor infrasonic frequencies (I'm really not that good of a bass player... yet). But the guitars hanging on my wall will audibly resonate with the low E on the bass. As will the glass panels in my cabinet. It's not annoying enough to me to spend the time fixing all the rattling. Especially knowing that I will be remodeling the room in the not too distant future. And I think some part of me likes it. Though I may not be very good, I can say my bass playing is "room rattling". One of the main things I love about live music is being able to feel the kick drum and bass. I wonder if that Velodyne could recreate that.
My calculated Qes was 34.04. On the elliott sound there is a note "Qes = Qms / (r0 - 1) Note: While this is supposed to be the correct formula, many have found it gives Qes that's much too high". An alternate formula is also given Qes = (Qms/(r0-1))*(Re/(Rs+Re)). By using this formula your new Qes should be 1.4 and mine is 3.4, In my case it is clear I have to use the alternate formula but in yours?
I realized that the formula I was using was incorrect. And the website that gave me those formulas updated them after my video came out. Either way, I just went ahead and made a cabinet for it by copying a commercially available cab. It may not be the best for that speaker, but it's fine for what I need.
I start talking about it at 20:08 (I know, it's a long and dry video and difficult to focus on. Sorry about that.) Vr is Ir * Rs. Ir is calculated just before Vr and Rs is the actual measured value of the physical resistor in the circuit.
@@HandlebarWorkshops Don't be sorry please. Your explanation is very easier to understand. Thanks for the reply.
3 роки тому+1
Hello!, thank you, very interesting video!, I think there is a problem using rice for the added mass, because momentarily it would be in the air, you could probably do the test with clay or playdho!
thank you for the tutorial video, Please measure a speaker that has published specifications, then compare the two results together, what do you think?
That would be ideal, but speakers can be expensive. This one was free and my method got close enough. If I can find a cheap enough speaker, I will do just as you suggest. However, I still have to go back and do some re-calculations. The ESP website put out some new equations that should help me get an even better result. I just have to find the time to do it.
In the future if you would ever like to build anymore speaker cabinets or maybe a pair of studio monitor speakers.There is a device Parts Express sells call Dayton Dats V3. This device measures all T/S parameters in about one second. It's also supplies frd which is frequency response and zma which is impedance curve files so you can build passive crossovers accurately. It also can function as an oscilloscope a signal generator truly awesome little device for $130. There is also software like bass box Pro which you can import frd and zma files to and build 100% accurate enclosures.
That is true, but I was trying to see how close I could get on a limited budget. That Dayton Dats is more expensive than a new speaker. I guess it would be worth the expense if I were dealing with unknown speakers on a daily basis, but so far, this is the only one I've ever come across. Thanks for watching.
I prepared a similar setup but my speaker is rather small, a 52mm dia speaker with 4ohms (5w). I am measuring the 10ohm resistor's voltage across its legs and the voltage is climbing instead of declining when the driving freq. comes close to resonance freq. I plotted the output voltage and saw that it makes a double hill. What is going on?
Well, a 52mm diameter is about 2 inches. The whole point of the T/S parameters is to design and build a cabinet that tuned to the speaker so it can properly reproduce the low frequencies. As such, it mainly applies to bass guitar amp speakers and sub-woofers. I'm not sure such a small speaker has enough mass to require a tuned cabinet. Nor would it be able to reproduce the low frequencies accurately enough to require a tuned cabinet.
@@HandlebarWorkshopsThanks for the answer. I am a newbie. So how is it possible to design a speaker enclosure for small speakers then? Generally, small speakers come with a sealed enclosure but there are bass reflex speakers too for instance. How to design those?
Honestly, I don't know how to design a cabinet for that small of a speaker. I'm rather new to all of this as well. I'm just trying to figure it all out as I go. Until I recorded that video, I had never really done anything like that before, nor have I've done it since. However, I would assume that the size of the cabinet wouldn't matter with such a small speaker. A port (a hole cut into the speaker baffle - sometimes with a tube - to allow air to escape) may or may not help depending on the size of the cabinet. A speaker that size will not move much air (that's how speakers work, the paper cone pushes air which causes pressure waves that our ears and brains interpret as sound). And since air is rather springy, if the enclosure is rather large, it won't matter at all. I would suggest you experiment. I don't think you can get the T/S parameters for such a small speaker. Make some enclosures out of cheap wood at different sizes and see what sounds the best. You can even record it and post a link here.
Thiele/Small worked for me, but only assuming the mfg specs are accurate! Sadly, there are statistical significant variations with mass produced speakers
I have no idea what the sampling rate of my meter is. I bought it used without a manual. I guess I could go look it up online, but I don't think it's necessary. Using an oscilloscope would be much better, but they're rather expensive and most people (like me at the time) don't have them. I was trying to do this in a way that was relatively easy and cheap with tools that a lot of people have or have access to. I've since purchased a used scope but I already put the speaker in a cabinet and had been playing it often.
First Point - the integration time of a Dual Slope meter is 100 milliseconds, intentionally 6 cycles at 60 Hz and 5 cycles at 50 Hertz. If you can remember from Calculus, the integral of a sine wave over a full cycle is zero. 100 milliseconds is the lowest integration time that's an integer of both 50 and 60 Hertz. Why? According to Sampling theory, 50 and 60 Hertz will both fall in the bottoms a deep notches. That way, you can do a very low-level DC measurement in the presence of a significant line voltage pickup and your DC measurement will hardly be affected. The display will barely flicker. A wonderful invention. I believe Weston Instruments may have held the original Patent. Second Point - I came to this website because I managed to install a 12" shallow Subwoofer in a 36" by 18" by 5" , 1 .1 cubic inch enclosure that fits PERFECTLY under the tonneau bar of my 2016 Toyota RAV4. I love classical organ and tried to get down to 16 Hz, without spending $$$$. I was able to get down to 24 Hz before the cone started breaking up - unlike my TV room's (modified) Sunfire True Subwoofer that will eek out a 115 dB SPL at 16 Hz. I need a subwoofer with a breakup-resistant, stiff Aluminum cone and about a 2 cubic foot enclosure. What I also discovered - don't expect decent bass from a shallow subwoofer. Physically, the cone is too shallow. You need a real, deep, stiff conical cone to minimize breakup, multi-mode, doubling (2nd Harmonic) or tripling (3rd Harmonic) or higher-order Harmonic distortion. So, I've just found a reasonably-priced Harmony HA-A122, 12" Anodized aluminum cone subwoofer, with no published Thiele-Small parameters. Their prime retailers have no TS data. I haven't yet reached Harmony's engineers. I have a small electronics lab in my garage. I designed medical electronics for 40 years before recently retiring. I thought it would be fun to do my own TS stuff if there were no published data, and I had to buy a sub to test it myself. Instead of clay, we have some of our granddaughter's leftover Play Doh. Since she's 17, I need it more than she does. The Doh is still flexible after 13 years. A full Play Doh container fits the vent of a legacy Radio Shack 12" subwoofer, allowing closed v ported frequency response comparisons. Third Point - a Linkwitz Filter is excellent for open-loop flattening of a subwoofer's response in a sealed enclosure. It removes the resonant hump and significantly raises the bass response if your sub cone doesn't distort and your amplifier power is sufficient. A vented enclosure can also be done, but it's very tricky and requires much more amplifier power.
Thanks for all of the information. I hope that Harmony gets as low as you want it. I'm surprised you can even hear 16Hz. Or are you just feeling it? That said, I wonder if my video might be a bit too remedial for you. Sounds like you have a pretty decent setup and can probably put a video with much more technical information than this aimed at an engineering or very experienced audience. This video was more aimed at musicians. ;)
i`ve got it! They have changed the domain, so after some digging around i found the new page, it`s this one sound-au.com/tsp.htm hope you can take advantage if this.
I have a set of 6 inch speakers and was wondering if it was necessary to make a tuned cabinet since i dont have an amp available to use to measure those parameters?
For optimal bass response, you could try to design a tuned cabinet, but I don't think it would be worth the time and effort. If you look at commercial high end speakers that use the same size, you'll find that all of the cabinets are around the same dimensions. I would just use measurements similar to those and call it good. That's what I ultimately did with my bass cabinet since this video didn't give me useful results. Thanks for watching.
It's just a frequency generator app. I'm not even sure that that particular one exists anymore. Just grab the highest rated one off the app store or google play.
One question. When doing this, the amp should have a linear output frequency between 20 to 20k corect? If you happen to choose an amp that is not good outputing low frequencies the result will not be accurate, correct? So what are the recomendations for the amp? Class AB, D?
Well, in this case, I was trying to get a good approximation using items most people would have easy access to, or possibly already own. I just used my stereo receiver with the EQ set to flat. I have no idea what it's frequency response is like along the spectrum since I don't own a spectrum analyzer. Maybe the lows should have been boosted, I don't know. I believe that both classes of amps can be designed in such a way to output a flat response to all frequencies in the audible spectrum, only that the class AB would probably cost more. They would also most likely be lab equipment (like a frequency generator) since such a flat response is usually not very pleasing to listen to. Thanks for watching.
If you're testing a speaker, the Class of the amplifier is probably unimportant, because the Total Harmonic Distortion (THD) or Linearity of the speaker is, at minimum, 10 times worse, and likely more. If the Amplifier's Frequency Response isn't flat, you can make up for it with a Radio Shack or similar Sound Level Meter that's been tweaked (on paper) for flat frequency response. Or you can use Reciprocity to measure and correct for it. B & K makes excellent, professional Sound Level Meters. Frequency Response is usually not considered a nonlinearity. It's just a frequency-dependent loss. True amplitude nonlinearity creates harmonic frequencies that don't exist in the original source material. What a lot of people consider "good" bass is simply nonlinearity-based harmonics, overtones or multi-mode cone breakup. You generally can't hear 16 Hz fed to a good subwoofer. If you do, it's likely 32, 48, 64, 80 Hz or higher-order cross-products.
That measuring tape is wonky. It takes into account the length of the plastic box. The number shown on the tape is the distance from the plastic tab to the end of the plastic box. Really confusing, but that was all I had at the time that could span that distance. Thanks for watching.
That was one suggestion that I saw, but I didn't have any modeling clay. And I figured most people were more likely to have rice than modeling clay, as well.
I dunno if this might be the reason for the surprising numbers, but loose rice is not good. The added known mass should directly and consistently alter the mass of the driver (piston). The mass should remain affixed to the cone during the measurement; the combined mass of the rice (or other loose additional mass) and cone will only be stable at rest - a state that never occurs during the measurement. Sticking stuff to the cone is obviously not desirable, but necessary for accurate results. It might be worth trying to find if anyone has figured out a solution to this problem - either mathematical (an algorithm that accounts for a loosely added mass) or practical. Thanks for walking through the process. I have to do this at some point down the road. Familiarizing myself with the procedure in advance :)
I hadn't thought about that. Interesting. I was looking to get a uniform distribution of weight, and I thought something like the rice would be both economical and easier to spread out. Good luck with your T/S Parameters journey. If you figure something out, let me know. Thanks for watching.
I have an idea. How about some magnets on the front and back of the cone. I have seen people using coins and blutac before. I think stong masking tape and supergluing coins on is a better idea.
While measuring the Fs voltage is decreasing when you increase the frequncy but in next step voltage is increasing when you increasing the frequency ... how does it working like this
That is the resonance frequency, the frequency where the the voice coil in the speaker has to do the least amount of work (lowest voltage) to move the cone. It all depends on what the cone is made of, how big the magnet is, what kind of glue is used to attach the paper cone to the metal structure of the speaker and how it all works together (resonates). I am at a loss to come up with a good analogy, at the moment, but think of holding a pool noodle at its center point. If you move your hand back and forth, the ends start to wiggle. If you experiment with the speed of your hand, you will find a frequency where the ends flop around in a very consistent pattern and it will be very easy for you to keep that pattern going - this is the resonance frequency of the pool noodle. If you try to speed up or slow down, either you have to use more strength in your arm or the ends of the pool noodle get chaotic.
I'm doing a very quick and dirty job here. Just looking for good enough. In this case, any capacitance introduced by the alligator clips is negligible. Thanks for watching.
You'd have to measure it to see, but probably not. Most cellphones require some kind of powered speakers (much like computer speakers) that have an amplifier built into the speaker enclosure that you have to plug into the mains outlet of your house. The current provided by the cellphone is probably not high enough to drive a speaker by itself.
I am just using a stereo receiver as an amp. But you could use a computer sound card or even a guitar/bass amp (as long as it is not a tube based amp - using a tube based amp without a proper speaker load will burn the tubes out). Thanks for watching!
You can convert a sealed enclosure to a vented enclosure with a Passive Radiator (drone). The bass will be extended, but 1) the time response (dynamics) won't be as clean and 2) the very bottom end response will drop off 24 dB per Octave v. 12 dB per Octave, which is essentially the same as point 1).
The reason for the resistor is to split the voltage so as not to blow the speaker accidentally. It also makes a very convenient way of calculating the current, measured voltage divided by the measured resistance. Speakers are an impedance load, that is their 'resistance' changes with the frequency. The resistor is always a static 10.4 Ohms and can be used as such when calculating current any any frequency.
Thanks for the reply. I haven't got a resistor atm but I tried to find the resonant frequency but I don't see the voltage drop. Would doing the test without one cause this to happen?
Be very careful with that. You risk blowing the speaker without the resistor. If you're measuring across the speaker, you would would be looking for a voltage maximum - the opposite of what you'd look for across the resistor. The voltage would be set by the amp and then shared between the resistor and the speaker. So when the voltage is at a minimum across the resistor, it's at its max across the speaker and vice-versa.
That's great, makes perfect sense. Would you mind if I message you direct. I'm new to this and have some other questions you might be able to help with? Thanks again
if you measured the speaker resistance with the amp and resister connected, then you got the wrong value. the rice bouncing around is not cool for the loaded mass test.
I should have just run out and grabbed some Play-Doh. It is readily accessible to most people and pretty cheap. And that was the ultimate point of the video series - to allow people to figure out the T/S parameters with easily accessible (off the shelf) and inexpensive items. Thank you for watching and commenting.
@@HandlebarWorkshops how is my doh idea 😁.. after testing , you can eat it too.. I think I should make a video with that ... The second part will be , me enjoying music and eating a pancake made from the doh .. .. Crazy scientist ... 🤣
From what I was able to find online, they did manufacture speakers, mainly for off and store brand amps and speakers. The actual cabinet was long gone by the time the speaker was given to me, but I suspect it was an 80's era Guitar Center house brand. Thanks for watching.
The speaker looks like a 15" / 38.1cm. Maybe you measured the diameter wrong? The "surround" is the crimp at the edge of the cone, i think (being not a native English speaker).
Great video, I’m going to do these measurements myself on some pioneer speakers. You definitely measured the diameter wrong in the video. The diameter is closer to 375mm. You are not within 0,1g of 50g the scale shows 50g at 49,5g rounding up so you are within 0,1g of 49,5g. :) Rref calculated should be 506,22mV ( Vs*Rs/(Rs+Re) with Measurements at 2 Octaves above Fs maybe wrong 292mV? Finally Qes = Qms/(r0-1)*(Re/(Rs+Re)) witch with a Vref of 506,22 would give you a Qes of 0,94 and thereby a Qts of 0,84 (below 1) :) Vas = 360 Litres I hope this helps you.
Thanks for all of the info. You are absolutely correct on the diameter and weight of rice. It's a 15" speaker which would translate to 381mm. Even if it's not exactly 15", 322mm was way off. But I am confused as to where you came up with "Rref" (which I think you meant to type "Vref" since you call it that later on and the units are in volts not ohms). And your Qes formula is different from what I was using. Why? I'm glad you came up with a much saner value for Qts than I did (which leads me to believe you are correct), but I can't blindly accept that without and explanation. Thanks for watching.
I just put your measurements into the sheet you refer to, there the formula for Qes is as I wrote above :). Your measurement for Vref was the only thing off, and when I put in 292mV I got the same values that you got.
Ah,I see. The Elliot Sound Products page was updated in June of 2018, three months after I posted this video. It looks like the Qes formula was updated with the note, "While this is supposed to be the correct formula, many have found it gives Qes that's much too high," then gives the formula that you used. It also looks like he updated the formula for r0 as well. I'll add a sticky comment to highlight the changes. Thank you so much!
I can't see a reason why you'd want/need to do that for a tweeter. These parameters dictate the size (volume - cubic feet or meters) of the cabinet needed so that it works with the big speaker to resonate properly. Usually tweeters are just stuck in a cabinet with little consideration about the tweeter other than how much space/volume the tweeter takes up and thus needs to be added to the size of the cabinet. That said, I'm not sure it's possible, A lot of tweeters I've seen are fully enclosed with a horn. Without an easily accessible speaker cone, you won't be able to reproduce my rice/clay step. Thanks for watching!
Thank you for the update, but I can't find Rod Elliott's article on that site. The link in the description still works. Could you provide a direct link to Rod Elliott's article on diyaudioprojects.com, please? I will update my links.
I'm sorry it didn't work out, but maybe it will work out someday. I have absolutely no idea about electronics or how to do this kind of work, although I have a better idea now than I did before I watched your video. Well done, Tony!
Thank you. I'm actually building a cab for it right now. I hope to have that video up in a couple weeks.
Sorry about my delayed response. Far too many electronic and honey do projects.
While my setup may be capable of measuring T/S Parameters, I'd never done it before and learned a lot from your video.
What I found quite annoying were the resonances in our family rooms and cars when setting up or replacing speakers.
A sweep test of my new 1995 Mercedes C-280 found a resonance around 45 Hertz, caused by a loose speaker mount. A 35 Hertz resonance was caused by 2 plastic wire clamps an assembler had left on the rear sub-deck. At resonance, they bounced like Mexican jumping beans. Same with some loose hardware in my 2011 Camry Hybrid, after I'd replaced a speaker
with voice coil drag. Velcro, duct tape, pieces of foam, RTV and cable ties are great for damping resonances.
The most recent were the two metal straps I'd used to earthquake-proof our 7 foot high grandfather clock. Without them, you could easily push the clock over with a finger. There was a strange noise when the clock chimed or the subwoofer hit the right note. Neither the clock person nor I could locate the noise. That's because I'd disconnected the straps so he could look at the rear of the clock.
When I reconnected the straps. I realized one of them was resonating and kissing the top of the clock enclosure. A strip of fuzzy Velcro under both straps stopped the racket.
Velcro also works for bookshelves of rattling teacups and under the bottom of rattling picture frames. Also keeps them from tilting.
Those are great and innovative solutions. Great job. My biggest resonant sources are two glass cabinet doors. That is, each cabinet door has a pane of glass in a wooden frame. I'm sure a little hot glue will fix it, it's just low on my list of priorities. I usually forget about them until I'm playing bass and they start rattling. But at that point, I'd usually prefer to just play and ignore the rattle rather than fix it. I'd make a mental note about it, then promptly forget. :D
Great video ! I really liked how show every step of the measurements, it make things really clear and understandable.
Thank you. Glad I could help.
glad i found this video, i have 2 mid-range woofer with no parameters, the only parameters that i have is that the RMS power is 80W, and the impedance is 8 ohm, i've tried this once, but i didnt do a great job to calculate it, and i already made the enclosure for it sadly, probably gonna re-measure it and try to fix the enclosure
Good luck to you. I hope it all works out. My numbers ended up kinda weird, but the cabinet still sounds pretty good, just not very loud. But the amp has a DI, so I guess it doesn't matter all that much. However, I haven't tried it with an actual drummer, so I have no idea if I can even hear it.
Thanks for watching.
have you heard about software called WinISD, its basically a software to design a box based on the parameters, but more complex than the website version
I have not heard of that software. Even so, my goal was to keep the project as inexpensive as possible. And even if it were free, I am very wary of downloading software from random websites. Besides, the website versions were good enough to show me the speaker was probably junk.
its free, i've download it, and i already scanned it with my bitdefender anti-virus and its fine, yeah you're right, there's no point to use complicated software for such a cheap speaker tho, if you curious about the software, you can check it here www.linearteam.org/ . Anyway i already measure my speaker, it's turn out my speaker Qes is worse than yours, mine about 6.22, its cheap tho for a 5" woofer with no brand, around $3-4, its so hard to get a good quality speaker without a lot of money here, so yeah maybe i just stick with this woofer for awhile.. , there's an alternative for the rice, you can use kinetic sand, its not sticky and easy to clean.
If it's a sealed enclosure, don't forget to try the Linkwitz Filter before anything too expensive or time consuming.
Not that it means a lot, but on the Low Resistance Range (typically 200 Ohms) both Digital and Analog meters have test lead errors as much as 1/2 Ohm. The easiest way to compensate is to short the probe tips together and note the reading as "Test Lead Resistance". Subtract it from all real measurements. If you change the meter Range to 2k or higher, lead Resistance is negligible.
Another way to measure the Voice coil resistance is to connect a fresh 1.5V D battery (NO HIGHER VOLTAGE!!!) to the cone through that same 10 Ohm resistor and a series-connected multimeter on the 2 Amp DC Range. For a 2 Ohm Voice Coil, the meter should read around (1.5V / 12 Ohms) or 125 milliAmperes on the 200 milliAmpere DC Range. The internal Voltage drop of the meter and its internal fuse may give a slightly lower Current value. No Big Deal! Just record the measured Current.
Then take a second meter on the 2 Volt (or 2000 mV) DC Range and connect it DIRECTLY across the speaker's Voice Coil Terminals. Use Ohm's Law, R = E/I to calculate the Voice Coil Resistance. What you've just done is taken a 4 Terminal Kelvin measurement, one of the most accurate ways to measure Resistance, especially low Resistance. The Current reading already includes the 10 Ohm resistor's Resistance, plus the Current meter's internal and lead resistance. The lead resistance of the Voltage-reading meter is inconsequential.
What's left to affect your result is only the accuracy of the two meters.
One other point - the most accurate range for a Digital meter is the 200 millivolt range. Most meters use the Dual Slope technique, where the internal Battery Voltage, the internal Rs and Cs, the internal dividers, shunts, adjustments, ramps and clock frequency cancel, leaving only the internal Voltage Reference. If properly factory-tweaked, the 200 milliVolt range can easily be 10 times better than the stated accuracy.
Excellent! This is great information. I will definitely use this in other projects I plan on doing. Especially the note about the digital meters and errors at low resistance.
Thank you for watching and commenting.
While generally not as accurate, old Analog Ohmmeters, like my Micronta 43-491 (Radio Shack in Amsterdam) were subject to battery aging and had a Zero adjustment, also perfect for zeroing lead resistance. Whenever I worked on world-wide CTs or MRIs in the 70s, 80s and 90s, I always scouted the local Radio Shacks as soon as I arrived in town. You can't believe how many multi-million $$$ scanners were diagnosed with, and actually contained Radio Shack parts.
HA! That's awesome. I miss Radio Shack. :(
The described technique can also be used to measure the 10ohm resistor, or any low value resistor for that matter. If you have a lab PSU, it is very easy to set some reasonable current limit, then measure the voltage drop directly on the resistor or resistive-anything. If you don't trust the current reading of the PSU, then either include an amper meter in series too, but is cruical to always measure the voltage drop on the device under test, as close as possible.
You can measure miliohms this way if you can pump out amps from your PSU (and the device can tolerate it, eg. you measuring a cable).
I'd like to point out that measuring AC voltage might be inaccurate with multimeters if the signal is way off of 50/60 Hz. Might be accurate, but you can only be sure, if the datasheet guarantees it in some frequency range, or if you check it with either a known accurate flat response function generator or an another known accurate voltmeter. It's not a problem in this subwoofer range, but might be higher up in the audio spectrum.
I agree with Gabor.
Whenever you're doing low resistance measurements, the 4 terminal or Kelvin connection overcomes lead and contact resistance by ensuring no substantial current flows through the Digital Multi-meter's test leads.
When you measure a low value resistor using the DMM, by itself, the small, but significant current the DMM applies to the resistor-under-test also flows through the DMM's test leads, causing an (I)(R) product Voltage drop. By itself, the DMM can't separate the lead's (I)(R) drop from the resistor-under-test (I)(R) drop. You have to subtract the test lead (I)(R). On an old analog meter, you can usually re-zero or null the test lead (I)(R) drop.
Or you can use the Kelvin connection.
en.wikipedia.org/wiki/Four-terminal_sensing
Lord Kelvin is the same engineer/physicist who gave us the first two Laws of Thermodynamics. The Kelvin temperature scale is also named after him. 0 degrees K ~ - 273.16 degrees Celsius or Absolute Zero, where (at least in 2022) all molecular or entropic motion ceases and all pure resistances become noiseless.
en.wikipedia.org/wiki/William_Thomson,_1st_Baron_Kelvin
Thanks for detailed video, I'll be very glad to see any updated one too.
As far as an update, it's still going fine. Haven't had any reason to disassemble the cabinet I put it in and start over. Though it might be too quiet, but I'm not sure if the amp is too weak or if the box is limiting the output. If I ever need to get it louder, I might re-do this video and get more accurate parameters.
Obviously, the measurement error is due to the use of rice as an additional weight. The rice rattled audibly. This means that the additional weight of 50 grams was not firmly coupled to the diaphragm of the loudspeaker. The measurement was effective with far less than 50 grams and therefore the resonance frequency was much too high. With this wrong frequency one receives a much too high value for Qts in the calculations. For a correct measurement it is absolutely necessary to use a compact mass as additional weight and to ensure that it holds firmly to the diaphragm. Modelling clay is ideal for this purpose.
In my experience, these cheap loudspeakers have a Qts of about 0.8 to 1.3. This makes it impossible to achieve a linear frequency response without manipulating the frequency response of the amplifier. Nevertheless, it will be possible to achieve a passable sound in a suitable cabinet. With instrumental loudspeakers, linearity is also less important. Often it is the objective errors that make these speakers sound attractive.
Now that you mention it, that makes complete sense. I didn't have any clay on-hand, however.
Though I agree that in music, especially rock based music, such errors are attractive to some listeners. The Kinks were famous (infamous, maybe?) for poking holes in the speaker cones to get the sound they were after. I guess at that point T/S parameters really aren't worth much.
Thanks for watching.
@depunkt Yeah, I probably should have used some clay. But I didn't have that available.
If you are trying to get the T/S parameters then you can use this method (substitute the clay for rice).
Thanks for watching.
In my opinion, this may be an older video, but it's relavancy is just as applicable, as when it was produced; & it has been really well explained. I cannot understand why it only has a 'like' rating of ~1.45%, per views.
I have acquired a couple of older Sony 8" subwoofers, pulled from cheap audio equipment. In attemtping to research the design requirements, for a custom subwoofer, I quickly discovered the issues surrounding using drivers, with unknown parameters, which led me here, eventually.
And, I have found out, that Sony do NOT share their tech: what a surprise. Even though I know the speaker models, any technicals are not to be had, on the internet.
As someone who is not an audio guy, but has some electronics' ability, I preferred a video that was presented in plain speak. Your idea of using a phone, with a frequency app is just what the doc ordered 👌. Given my Denon AVR is bluetooth capable, I am able to simplify the connections, somewhat 😵💫: headphone jack>jack to 3.5mm jack>3.5mm to RCA adaptor cable. Then, a cut off RCA cable, with a 10ohm resistor soldered to the positive?
I think I'll just do the lazy way, & use the supplied spreadsheet.
Thanks, mate.
Thank you. I think a lot of people disliked that I used rice instead of clay or playdoh to weight the cone. But other than that I think most other things went pretty well.
There is another video where I attempt to design an enclosure for this speaker that didn't go very well due a misprint of a formula on the website. I should go back and fix it, but it's pretty low on my priority list.
@@HandlebarWorkshops Thanks for letting me know, you are still monitoring this video.
I have a question, that may/may not relate to your measurement parameters: the amplifier you showed, does it fit the stipulated requirements -ie. 1-10W RMS, & an output impedance of
@@citizengkar7824 While doing research for this video, I did not read anything about output impedance, nor about the RMS.
This method isn't meant to be (nor can it be) extremely accurate, but doable for someone to do once or twice without investing in expensive tools. It's meant to get close enough to get a good (but not necessarily a HiFi) sound out of a speaker.
After all, since it is a fairly unknown speaker, you probably didn't pay much for it anyway. You just want to make it sound as good as possible without spending too much more money. It would be cheaper to buy a new speaker with known parameters than to buy all of the scientific tools to do this at a high level of accuracy.
@@HandlebarWorkshops I do agree all you've said. Though, it does state those requirements, on the Elliott site, you've linked.
"1.1 Measuring Re, Fs, Qes, Qms and Qts
To measure these parameters using the method outlined below, you'll need to have the following items:
A power amplifier, rated at 1-10 Watts (RMS) or thereabouts (must have low output impedance
That Elliott site has changed since I used it for this video. For a while it was a 404, then when it came back one of the formulas was different. I haven't gone back to see if that new formula would result in a better end result with this speaker.
I don't remember any of that list being there, but it's been years, so my memory of all that may be a bit off.
Hey thanks for taking the time to do this!
You're welcome. When I made this, there were no other videos on this subject that I could find. I thought it needed to more available. This was also my very first youtube video.
Now I occasionally see other videos show up in my feed by more established channels. So even if mine wasn't the best, it was the first. :D
Thanks for taking the time to watch and comment.
The real test of a subwoofer is how much you can feel its infrasonic (sub 20 Hz) output without hearing it.
30 years ago, I tested what was one of the best subwoofers on the market - an 18 inch servo'd Velodyne, in the listening
room of a high end audio dealer in San Mateo, CA. I didn't have my SPL meter at the time. At minimum volume, I
connected my CD player to the Velodyne Input and selected the 16 Hz CD test track. I turned up the volume until I heard
something, then backed off a few dB. There was no audible sound coming from the sub, but the windows of the listening
room were vibrating. Fortunately, nothing else in the room was audibly resonating. I WAG'd the sub's SPL at maybe 120 dB.
At infrasonic frequencies, you should be able to feel and even see the output of a good subwoofer without hearing it.
Anything you hear is either harmonic distortion or some nonlinearly-resonating object in the room. You may see
things like windows or paintings vibrating - OK as long as you can't hear them.
I periodically sweep our listening room for irritating resonances, which I usually damp with pieces of hidden Velcro.
Some things need caulking. During one sweep of our new place, I heard a rasping noise from an adjacent room. It was
the metal straps I'd used to earthquake-proof our grandfather clock. Some Velcro under the straps killed the noise.
After sweeping a few rooms, you learn to exercise "a bit of healthy engineering paranoia". Ask any pilot - zero or
insufficient paranoia is dangerous. Too much becomes "engineering or pilot paralysis" - the fear of making the wrong
decision or making no decision at all.
I use this speaker in the cabinet of my bass guitar amp. This is by no means a high quality amp. My friend who gave it to me said it was from a cheap 80's keyboard amp that got ruined when his basement flooded. I wasn't looking for high fidelity nor infrasonic frequencies (I'm really not that good of a bass player... yet). But the guitars hanging on my wall will audibly resonate with the low E on the bass. As will the glass panels in my cabinet.
It's not annoying enough to me to spend the time fixing all the rattling. Especially knowing that I will be remodeling the room in the not too distant future. And I think some part of me likes it. Though I may not be very good, I can say my bass playing is "room rattling".
One of the main things I love about live music is being able to feel the kick drum and bass. I wonder if that Velodyne could recreate that.
The only problem is that the rice bounces, try taping a weight to the speaker center!
Playdough works great
A guy could use playdough as his weight so it doesn’t bounce all over.
Yes. That is totally a valid option. I just didn't have any on hand and didn't want to go buy some for this single use.
Thanks for watching!
Interesting! But I wonder if even 'Documented' speakers T/S should be individually tested, too?
i have new speakers with the datasheet, i will try your step, if the parameters is same you are great mr!
Since you have all of the T/S Parameters, you can skip to about 23:00 and use what you already have to start designing a cabinet.
Thanks for watching.
My calculated Qes was 34.04. On the elliott sound there is a note "Qes = Qms / (r0 - 1) Note: While this is supposed to be the correct formula, many have found it gives Qes that's much too high". An alternate formula is also given Qes = (Qms/(r0-1))*(Re/(Rs+Re)). By using this formula your new Qes should be 1.4 and mine is 3.4, In my case it is clear I have to use the alternate formula but in yours?
I realized that the formula I was using was incorrect. And the website that gave me those formulas updated them after my video came out.
Either way, I just went ahead and made a cabinet for it by copying a commercially available cab. It may not be the best for that speaker, but it's fine for what I need.
When finding Hz (below/above Fs), i can not understand voltage (Vr=0.166). Where it came from? please help.
Thanks in advance.
I start talking about it at 20:08 (I know, it's a long and dry video and difficult to focus on. Sorry about that.) Vr is Ir * Rs. Ir is calculated just before Vr and Rs is the actual measured value of the physical resistor in the circuit.
@@HandlebarWorkshops Don't be sorry please. Your explanation is very easier to understand. Thanks for the reply.
Hello!, thank you, very interesting video!, I think there is a problem using rice for the added mass, because momentarily it would be in the air, you could probably do the test with clay or playdho!
Yeah, it didn't even occur to me at the time.
Thanks for watching and the comment.
thank you for the tutorial video,
Please measure a speaker that has published specifications, then compare the two results together, what do you think?
That would be ideal, but speakers can be expensive. This one was free and my method got close enough. If I can find a cheap enough speaker, I will do just as you suggest.
However, I still have to go back and do some re-calculations. The ESP website put out some new equations that should help me get an even better result. I just have to find the time to do it.
In the future if you would ever like to build anymore speaker cabinets or maybe a pair of studio monitor speakers.There is a device Parts Express sells call Dayton Dats V3. This device measures all T/S parameters in about one second. It's also supplies frd which is frequency response and zma which is impedance curve files so you can build passive crossovers accurately. It also can function as an oscilloscope a signal generator truly awesome little device for $130. There is also software like bass box Pro which you can import frd and zma files to and build 100% accurate enclosures.
That is true, but I was trying to see how close I could get on a limited budget. That Dayton Dats is more expensive than a new speaker. I guess it would be worth the expense if I were dealing with unknown speakers on a daily basis, but so far, this is the only one I've ever come across.
Thanks for watching.
@@HandlebarWorkshops I totally agree. A lot of money, for a once, or twice-off.
I prepared a similar setup but my speaker is rather small, a 52mm dia speaker with 4ohms (5w). I am measuring the 10ohm resistor's voltage across its legs and the voltage is climbing instead of declining when the driving freq. comes close to resonance freq. I plotted the output voltage and saw that it makes a double hill. What is going on?
Well, a 52mm diameter is about 2 inches. The whole point of the T/S parameters is to design and build a cabinet that tuned to the speaker so it can properly reproduce the low frequencies. As such, it mainly applies to bass guitar amp speakers and sub-woofers. I'm not sure such a small speaker has enough mass to require a tuned cabinet. Nor would it be able to reproduce the low frequencies accurately enough to require a tuned cabinet.
@@HandlebarWorkshopsThanks for the answer. I am a newbie. So how is it possible to design a speaker enclosure for small speakers then? Generally, small speakers come with a sealed enclosure but there are bass reflex speakers too for instance. How to design those?
Honestly, I don't know how to design a cabinet for that small of a speaker. I'm rather new to all of this as well. I'm just trying to figure it all out as I go. Until I recorded that video, I had never really done anything like that before, nor have I've done it since.
However, I would assume that the size of the cabinet wouldn't matter with such a small speaker. A port (a hole cut into the speaker baffle - sometimes with a tube - to allow air to escape) may or may not help depending on the size of the cabinet. A speaker that size will not move much air (that's how speakers work, the paper cone pushes air which causes pressure waves that our ears and brains interpret as sound). And since air is rather springy, if the enclosure is rather large, it won't matter at all.
I would suggest you experiment. I don't think you can get the T/S parameters for such a small speaker. Make some enclosures out of cheap wood at different sizes and see what sounds the best. You can even record it and post a link here.
Thiele/Small worked for me, but only assuming the mfg specs are accurate! Sadly, there are statistical significant variations with mass produced speakers
Out of interest, what is the sampling rate of your meter? wouldn't an oscilloscope be a better instrument to make measurements for Voltage?
I have no idea what the sampling rate of my meter is. I bought it used without a manual. I guess I could go look it up online, but I don't think it's necessary.
Using an oscilloscope would be much better, but they're rather expensive and most people (like me at the time) don't have them. I was trying to do this in a way that was relatively easy and cheap with tools that a lot of people have or have access to. I've since purchased a used scope but I already put the speaker in a cabinet and had been playing it often.
First Point - the integration time of a Dual Slope meter is 100 milliseconds, intentionally 6 cycles at 60 Hz and 5 cycles at 50 Hertz. If you can remember from Calculus, the integral of a sine wave over a full cycle is zero. 100 milliseconds is the lowest integration time that's an integer of both 50 and 60 Hertz. Why? According to Sampling theory, 50 and 60 Hertz will both fall in the bottoms a deep notches. That way, you can do a very low-level DC measurement in the presence of a significant line voltage pickup and your DC measurement will hardly be affected. The display will barely flicker.
A wonderful invention. I believe Weston Instruments may have held the original Patent.
Second Point - I came to this website because I managed to install a 12" shallow Subwoofer in a 36" by 18" by 5" , 1 .1 cubic inch enclosure that fits PERFECTLY under the tonneau bar of my 2016 Toyota RAV4. I love classical organ and tried to get down to 16 Hz, without spending $$$$. I was able to get down to 24 Hz before the cone started breaking up - unlike my TV room's (modified) Sunfire True Subwoofer that will eek out a 115 dB SPL at 16 Hz. I need a subwoofer with a breakup-resistant, stiff Aluminum cone and about a 2 cubic foot enclosure.
What I also discovered - don't expect decent bass from a shallow subwoofer. Physically, the cone is too shallow. You need a real, deep, stiff conical cone to minimize breakup, multi-mode, doubling (2nd Harmonic) or tripling (3rd Harmonic) or higher-order Harmonic distortion.
So, I've just found a reasonably-priced Harmony HA-A122, 12" Anodized aluminum cone subwoofer, with no published Thiele-Small parameters. Their prime retailers have no TS data. I haven't yet reached Harmony's engineers. I have a small electronics lab in my garage. I designed medical electronics for 40 years before recently retiring.
I thought it would be fun to do my own TS stuff if there were no published data, and I had to buy a sub to test it myself. Instead of clay, we have some of our granddaughter's leftover Play Doh. Since she's 17, I need it more than she does. The Doh is still flexible after 13 years. A full Play Doh container fits the vent of a legacy Radio Shack 12" subwoofer, allowing closed v ported frequency response comparisons.
Third Point - a Linkwitz Filter is excellent for open-loop flattening of a subwoofer's response in a sealed enclosure. It removes the resonant hump and significantly raises the bass response if your sub cone doesn't distort and your amplifier power is sufficient. A vented enclosure can also be done, but it's very tricky and requires much more amplifier power.
Thanks for all of the information. I hope that Harmony gets as low as you want it. I'm surprised you can even hear 16Hz. Or are you just feeling it?
That said, I wonder if my video might be a bit too remedial for you. Sounds like you have a pretty decent setup and can probably put a video with much more technical information than this aimed at an engineering or very experienced audience. This video was more aimed at musicians. ;)
Hi sir , seems the website can't be found have any links
i`m having the same problem, have you got any help for this?
i`ve got it! They have changed the domain, so after some digging around i found the new page, it`s this one sound-au.com/tsp.htm hope you can take advantage if this.
Good on you man doingvthings properly. Like to see the final build. Thanks for sharing
The final build was a couple of videos later. ua-cam.com/video/2lGkpTvJKPQ/v-deo.html
Thanks for watching.
Cone mass could be easily determined by sacrificing the speaker! Cut the surround and the spider half-way around and weigh the cone!
LOL! not an option here!
weight is not the same as mass.
I have a set of 6 inch speakers and was wondering if it was necessary to make a tuned cabinet since i dont have an amp available to use to measure those parameters?
For optimal bass response, you could try to design a tuned cabinet, but I don't think it would be worth the time and effort. If you look at commercial high end speakers that use the same size, you'll find that all of the cabinets are around the same dimensions. I would just use measurements similar to those and call it good. That's what I ultimately did with my bass cabinet since this video didn't give me useful results.
Thanks for watching.
Hi nice video. What phone app did you use
It's just a frequency generator app. I'm not even sure that that particular one exists anymore. Just grab the highest rated one off the app store or google play.
One question. When doing this, the amp should have a linear output frequency between 20 to 20k corect? If you happen to choose an amp that is not good outputing low frequencies the result will not be accurate, correct? So what are the recomendations for the amp? Class AB, D?
Well, in this case, I was trying to get a good approximation using items most people would have easy access to, or possibly already own. I just used my stereo receiver with the EQ set to flat. I have no idea what it's frequency response is like along the spectrum since I don't own a spectrum analyzer. Maybe the lows should have been boosted, I don't know.
I believe that both classes of amps can be designed in such a way to output a flat response to all frequencies in the audible spectrum, only that the class AB would probably cost more. They would also most likely be lab equipment (like a frequency generator) since such a flat response is usually not very pleasing to listen to.
Thanks for watching.
If you're testing a speaker, the Class of the amplifier is probably unimportant, because the Total Harmonic Distortion (THD) or Linearity of the speaker is, at minimum, 10 times worse, and likely more. If the Amplifier's Frequency Response isn't flat, you can make up for it with a Radio Shack or similar Sound Level Meter that's been tweaked (on paper) for flat frequency response. Or you can use Reciprocity to measure and correct for it. B & K makes excellent, professional Sound Level Meters.
Frequency Response is usually not considered a nonlinearity. It's just a frequency-dependent loss. True amplitude nonlinearity creates harmonic frequencies that don't exist in the original source material. What a lot of people consider "good" bass is simply nonlinearity-based harmonics, overtones or multi-mode cone breakup. You generally can't hear 16 Hz fed to a good subwoofer. If you do, it's likely 32, 48, 64, 80 Hz or higher-order cross-products.
Thanks for the info Ron. I only understood about half of what you wrote, but hopefully it will help other people out.
12:56 - What exactly did you measure there? The diameter of the cone seems to be around 370 millimeters, not 322 as you said in the video.
That measuring tape is wonky. It takes into account the length of the plastic box. The number shown on the tape is the distance from the plastic tab to the end of the plastic box. Really confusing, but that was all I had at the time that could span that distance.
Thanks for watching.
@@HandlebarWorkshops Okay, thanks for clarifying. :)
Great video have you tried using modelling clay instead of rice
That was one suggestion that I saw, but I didn't have any modeling clay. And I figured most people were more likely to have rice than modeling clay, as well.
I dunno if this might be the reason for the surprising numbers, but loose rice is not good.
The added known mass should directly and consistently alter the mass of the driver (piston). The mass should remain affixed to the cone during the measurement; the combined mass of the rice (or other loose additional mass) and cone will only be stable at rest - a state that never occurs during the measurement. Sticking stuff to the cone is obviously not desirable, but necessary for accurate results. It might be worth trying to find if anyone has figured out a solution to this problem - either mathematical (an algorithm that accounts for a loosely added mass) or practical.
Thanks for walking through the process. I have to do this at some point down the road. Familiarizing myself with the procedure in advance :)
I hadn't thought about that. Interesting. I was looking to get a uniform distribution of weight, and I thought something like the rice would be both economical and easier to spread out.
Good luck with your T/S Parameters journey. If you figure something out, let me know.
Thanks for watching.
I have an idea. How about some magnets on the front and back of the cone. I have seen people using coins and blutac before. I think stong masking tape and supergluing coins on is a better idea.
While measuring the Fs voltage is decreasing when you increase the frequncy but in next step voltage is increasing when you increasing the frequency ... how does it working like this
That is the resonance frequency, the frequency where the the voice coil in the speaker has to do the least amount of work (lowest voltage) to move the cone. It all depends on what the cone is made of, how big the magnet is, what kind of glue is used to attach the paper cone to the metal structure of the speaker and how it all works together (resonates).
I am at a loss to come up with a good analogy, at the moment, but think of holding a pool noodle at its center point. If you move your hand back and forth, the ends start to wiggle. If you experiment with the speed of your hand, you will find a frequency where the ends flop around in a very consistent pattern and it will be very easy for you to keep that pattern going - this is the resonance frequency of the pool noodle. If you try to speed up or slow down, either you have to use more strength in your arm or the ends of the pool noodle get chaotic.
Handlebar Workshops thanx for attention sir..
Handlebar Workshops is there any other way for talk to you like instagram or facebook
i have a question!
doesn't the inductance of the alligator clips effect the calculation?
I'm doing a very quick and dirty job here. Just looking for good enough. In this case, any capacitance introduced by the alligator clips is negligible.
Thanks for watching.
Can i use my mobile direct as an amplifier which giving Vs=0.572V ??
You'd have to measure it to see, but probably not.
Most cellphones require some kind of powered speakers (much like computer speakers) that have an amplifier built into the speaker enclosure that you have to plug into the mains outlet of your house. The current provided by the cellphone is probably not high enough to drive a speaker by itself.
Handlebar Workshops can i use PAM8403 class D amplifire which have 5w output
Which ic based amplifire is used and
I am just using a stereo receiver as an amp. But you could use a computer sound card or even a guitar/bass amp (as long as it is not a tube based amp - using a tube based amp without a proper speaker load will burn the tubes out).
Thanks for watching!
Yes, it was a sealed enclosure originally.
You can convert a sealed enclosure to a vented enclosure with a Passive Radiator (drone). The bass will be extended, but 1) the time response (dynamics) won't be as clean and 2) the very bottom end response will drop off 24 dB per Octave v. 12 dB per Octave, which is essentially the same as point 1).
Great video. What's the reason for using the resistor and could it be tested without it?
The reason for the resistor is to split the voltage so as not to blow the speaker accidentally. It also makes a very convenient way of calculating the current, measured voltage divided by the measured resistance.
Speakers are an impedance load, that is their 'resistance' changes with the frequency. The resistor is always a static 10.4 Ohms and can be used as such when calculating current any any frequency.
Thanks for the reply. I haven't got a resistor atm but I tried to find the resonant frequency but I don't see the voltage drop. Would doing the test without one cause this to happen?
Be very careful with that. You risk blowing the speaker without the resistor.
If you're measuring across the speaker, you would would be looking for a voltage maximum - the opposite of what you'd look for across the resistor. The voltage would be set by the amp and then shared between the resistor and the speaker. So when the voltage is at a minimum across the resistor, it's at its max across the speaker and vice-versa.
That's great, makes perfect sense. Would you mind if I message you direct. I'm new to this and have some other questions you might be able to help with? Thanks again
Sure, you can email me at additola(at)gmail.com
Tnx for this very informative video
Thank you for watching!
if you measured the speaker resistance with the amp and resister connected, then you got the wrong value.
the rice bouncing around is not cool for the loaded mass test.
Just your mixer grinder is necessary or buying some maida flour will work though ..
I should have just run out and grabbed some Play-Doh. It is readily accessible to most people and pretty cheap. And that was the ultimate point of the video series - to allow people to figure out the T/S parameters with easily accessible (off the shelf) and inexpensive items.
Thank you for watching and commenting.
@@HandlebarWorkshops how is my doh idea 😁.. after testing , you can eat it too.. I think I should make a video with that ... The second part will be , me enjoying music and eating a pancake made from the doh ..
..
Crazy scientist ... 🤣
I'd watch that! >D
@@HandlebarWorkshops I will try , I am inspired.. ♥️😇
Awesome! Inspiring people to do crazy stuff is why I started this channel.
That company probably didn't manufacture that speaker, they just rebadged it. Probably a entry level Eminence or RCF would be my guess.
From what I was able to find online, they did manufacture speakers, mainly for off and store brand amps and speakers. The actual cabinet was long gone by the time the speaker was given to me, but I suspect it was an 80's era Guitar Center house brand.
Thanks for watching.
Sir may i have your any other contact i need your help
GOOD STUFF! thanks! :)
Um? That 10 ohm resistor is in series, not "parallel" with the speaker!
I assume you are talking about what I say @ 4:59? You're absolutely correct. I misspoke and didn't catch it.
Thanks for watching.
I'm an engineering major, too!
The speaker looks like a 15" / 38.1cm. Maybe you measured the diameter wrong? The "surround" is the crimp at the edge of the cone, i think (being not a native English speaker).
You are absolutely correct. I've also heard it called a "seal" or "gasket". Thank you for watching.
Great video, I’m going to do these measurements myself on some pioneer speakers. You definitely measured the diameter wrong in the video. The diameter is closer to 375mm.
You are not within 0,1g of 50g the scale shows 50g at 49,5g rounding up so you are within 0,1g of 49,5g. :)
Rref calculated should be 506,22mV ( Vs*Rs/(Rs+Re) with Measurements at 2 Octaves above Fs maybe wrong 292mV?
Finally Qes = Qms/(r0-1)*(Re/(Rs+Re)) witch with a Vref of 506,22 would give you a Qes of 0,94 and thereby a Qts of 0,84 (below 1) :)
Vas = 360 Litres
I hope this helps you.
Thanks for all of the info. You are absolutely correct on the diameter and weight of rice. It's a 15" speaker which would translate to 381mm. Even if it's not exactly 15", 322mm was way off.
But I am confused as to where you came up with "Rref" (which I think you meant to type "Vref" since you call it that later on and the units are in volts not ohms). And your Qes formula is different from what I was using. Why?
I'm glad you came up with a much saner value for Qts than I did (which leads me to believe you are correct), but I can't blindly accept that without and explanation.
Thanks for watching.
I just put your measurements into the sheet you refer to, there the formula for Qes is as I wrote above :). Your measurement for Vref was the only thing off, and when I put in 292mV I got the same values that you got.
Ah,I see. The Elliot Sound Products page was updated in June of 2018, three months after I posted this video. It looks like the Qes formula was updated with the note, "While this is supposed to be the correct formula, many have found it gives Qes that's much too high," then gives the formula that you used. It also looks like he updated the formula for r0 as well.
I'll add a sticky comment to highlight the changes.
Thank you so much!
Your weight was 49.5 gram
At 49.4 the scale will display 49 gramms, once it hits 49,5 gramms it will jump to 50 gramms.
Thanks!
You could have powdered or crushed that rice to make the diy food grade clay ..
How if meusurement ts for tweeter sir .. thanks
I can't see a reason why you'd want/need to do that for a tweeter. These parameters dictate the size (volume - cubic feet or meters) of the cabinet needed so that it works with the big speaker to resonate properly.
Usually tweeters are just stuck in a cabinet with little consideration about the tweeter other than how much space/volume the tweeter takes up and thus needs to be added to the size of the cabinet.
That said, I'm not sure it's possible, A lot of tweeters I've seen are fully enclosed with a horn. Without an easily accessible speaker cone, you won't be able to reproduce my rice/clay step.
Thanks for watching!
Well, I'll be dipped in shit!
the ESP website has moved to - diyaudioprojects.com/
Thank you for the update, but I can't find Rod Elliott's article on that site. The link in the description still works.
Could you provide a direct link to Rod Elliott's article on diyaudioprojects.com, please? I will update my links.
there must have been a problem with the server (sound.whsites.net/tsp.htm) when I was looking last week, it's working now.