I used to design and repair test and control equipment at a major Quartz Crystal manufacturer. This was "not recently". We did everything from growing the quartz through building TCXO, OCXO, DIL oscillators, SAW devices and Filters. Our DIL oscillators used a real thick-film modules, with the chip bonded to it, and connections made with bond wire. Partly because (in big volume) it was cheaper, but also because it allowed you much more space to "do things properly" - with one of those things being proper decoupling. Dividers were quite common, partly because it is hard to make good crystals at certain frequencies, and there was also a "size" thing - good 1MHz crystals are quite physically big. Did you get a sample of the gas when you opened it? Mostly they're filled with Nitrogen with a tiny bit of helium. This is part of the controlling the ageing, the helium makes it easy to check them for leaks very thoroughly. The mounts were crucial, get them wrong, and you get a lousy, rotten xtal. How you hold the crystal changes its characteristic. Get it wrong and you'll get a horrible product with issues like "band breaking". For crystals in conventional packages - those two clips means you hold it in just the right way, and with just enough "spring". It's much hard with DILs and that's why you often see three posts. If you were seeing substantial shifts and changes in ageing following a shock, then bluntly, your crystals were a bit crap. The phenomenon has several names in the industry, we called it "sleepy crystals". It really was (and still it) one of the industry's dirty little secrets. It is a sign of a production problem. I'll explain that later. In extreme cases you would have a product sat in a drawer for a few months where the Oscillator wouldn't start at power up. If you (very carefully) took it apart, you would find the ESR was through the roof, the Q was pathetic and if you did get it to start oscillating, it was likely to have been on a "band break" - so the frequency would be slightly (but significantly) offset. Bang it on the table - and all would be back to normal again, until it was left for ages. One of my first post graduation jobs was to design something which could detect this problem without the whole "waiting a month or two to see what happened" thing. I designed a system, which used the then standard ISO methods of measure frequency and ESR, except at very low drive levels. You would then gradually increase the drive level, watching to see how F and ESR change, until you are at maximum power, then you wound the drive level back to the zero again. You plotted the results on a graph, and you would get a fairly classical hysteresis curve, and the bigger the area inside the graph, the dodgier the crystals. You only had to leave the crystals alone for a day or two, per-test to get a clear result. I'm being called out on an errand now, but I will return later and explain where the phenomenon comes from.
"Did you sample the gas?" Yeah, he was hoovering above it with a GC syringe, and injected it into his pocket GC. LOL. I'm certain he didn't sample the gas. Ha.
Wow, that really took me back. About 20 years ago when I was working for NEC at a PC factory in MA. We were getting our PC motherboards (MBs) from SCI Technology in Tampa FL. We started seeing a lot of PCs fail on the assembly line. Some of the MBs were DOA, but most (90%) were just working badly. After looking at some clock signals on one of the poorly working MBs, I zeroed in on one of those little metal oscillator modules. The clock was off-frequency, so I replaced it and the MB worked! We powered up about 20 of the failing PCs and I used an RF spectrum analyzer (with a probe antenna) to observe the signals from the boards. Since these boards had six oscillator modules, I mapped out the RF profile from a good MB and compared frequencies emitted by the failing MBs. Some of the failed MBs had more than one failing oscillator modules. The failure pattern of which module and the shift in frequency was random. I decided it was a QC problem with the oscillator modules. However, since some of the off-freq modules came from different vendors, I deduced these modules were somehow damaged at the SCI factory in Tampa. SCI said Nope, no way.. I had a collection of failed modules, a microscope with a Polaroid camera and a dermel tool.. I did a presentation for the NEC VP of manufacturing and was quickly on a plane to Tampa. The photos of the little cracks in the quartz disks told the story. Because I'm a Ham Radio guy, I knew that dropping a radio Crystal 4" onto a desktop would break it. So I knew what to look for at the SCI factory. I watched the guy that loaded the tubes for the PCB parts loading machines. Nothing unusual. Then late in the afternoon, I saw him unloading the plastic tubes from the module vendor, he was dumping them into a hard plastic tray, from a height of about 6 inches!! Afterwards, he hand-loaded them into the special tubes for the parts-loaders. He didn't believe he was breaking them. I shown him the pictures. Another Case closed.. :)
Awesome story, thanks for sharing. Reminds me of (IIRC) an Intel story about bad yield on their chips. They couldn't figure out what was happening so sent someone to follow the wafers from the source. Turns out a very diligent goods-inwards person was opening the sealed wafer boxes and counting the wafers in the (not particularly clean air) loading dock to make sure they got every wafer they paid for.
So what happened to the Goods-inwards Person? Fired? or just given a friendly warning to "Please dont unseal the packages" with the understanding that the persons intention was good, but the result wasnt?
In the late 1970s, Intel made the 8251 USART and NEC used their own version of the NEC8251. After seeing double letters being printed (during line testing) on the Spinwriter serial RS232 printers, I telexed NEC Tokyo describing the failures. Weeks later came the reply from Japan. "Intel 8251 has handshaking error in RTS CTS logic. Our NEC8251 is exact copy, so it also has error. We will create new NEC8251A without Intel error"! And within a month, we got tubes of replacement USARTs..
I don't know what SCI did, but the cracked quartz oscillator elements problem disappear the afternoon I got to Tampa. I got to spend the weekend with my father-in-law before heading back to BeanTown..
I don't know jack shit about what this component is about or what to use it for, but was hooked with your narration style and the way you tear down baby electronics components, sorta reminds me of how I used to tear down my toys when I was a kid to see how it works 🙏🏻
Nice video. BTW, your viewers might find this interesting: this device is a thick-film hybrid. Meaning, the traces are not etched, like a PCB. Rather, they are made by silkscreening on a conductive paste and then firing it in a kiln. And what looks like solder is probably conductive epoxy. I don't recall if you ever did a video explaining hybrid circuits. I used to design stuff like this back in the '80s.
I'm not an electronics guy but I love your videos. I find this fascinating how when we build a device each individual component has so much detail and knowledge behind it. Standing on the shoulders of giants indeed! Even Wozniak was just taking complex components already designed by someone else and plugging them together in a novel way. All around the world there are a small number of experts on just one topic I guess - capacitor construction, chip fabrication and of course crystal clock circuits ☺️ it means that ordinary guys can take a raspberry pi for example and all that work has already been done for you so that we can concentrate on the target application, software and maybe very particular hardware for the application in hand without having to learn about oscillator circuits or PCB fab!
That internal bypass cap IS necessary, since these are high frequencies here (8MHz) vs. the 'user' frequency of 1MHz which needs to be bypassed as well.
Absolutely fantastic video dave - one of your best imo. What I like about this is not just the sweet tear down of those crystals, but the way you've demonstrated how bypass capacitors have an effect (and how distance from the connection makes all the difference), but also your mention about vibration too. I've come across odd ball issues where a system was suffering jitter because the can was not grounded (on a 2 pin crystal), and I've also seen issues where another system had jitter or something because the can wasn't supported (floating) and was susceptible to vibrations from road traffic 20 yards from the house lol. People think these things aren't relevant, but they are factors.
I totally agree that this video was one of the best yet too! Such an amazing fundamental, no-fuss look into something that so many of us have wondered about.
The voltage seen due to trace length is V=-L di/dt. Ie if your trace is long it has a larger inductance and hence the current drawn or returned through it will generate a voltage. Keep all power and gnd traces short or use planes !!
+CDC Remember you could skip the extra stuff. Just see what you need/want. ua-cam.com/video/_4-Sx-T6VBc/v-deo.html sci-toys.com/scitoys/scitoys/radio/am_transmitter.html
I would find it interesting to see temperature compensated (TCXO) , oven heated (OCXO) and voltage controlled (VCXO) oscillators aswell, in comparison to the regular oscillator.
Thanks for going the extra mile and figuring out what that noise actually was. I had no idea it was a 8Mhz crystal with a divide-by circuit. Good stuff!
The reason for using an 8MHz crystal and divider is that if you used a 1MHz crystal, the crystal would be WAY to big. About the lowest frequency you can get in an HC-18 package is a little bit less than 4MHz, and they were pushing it when you got one at 3.579..MHz (NTSC colorburst frequency). Now you know.
I'm so happy that you're back into the interesting stuff again! No debunking videos, just fun and interesting tests and experiments. And equipment shootouts of course! You can almost feel that you actually enjoy making the videos again. Thanks man!
You do realise that a lot of people like the debunking videos? And the recent pocket multimeter shootout wasn't that popular. Many won't find this video interesting either. People subscribe for many different reasons.
I didn't watch the multimeter shootout video. I've already got 4 multimeters that I'm happy with - a bench meter on the shelf above my bench,, a good handheld meter, current clamp meter and a basic 2000 count meter.
Everyone here should have a small stash of oscillator modules. Much easier to get your test circuit running when you can rely on the oscillator working without much fuss. Often you can find them on sale from reliable sellers as well.
Internal cap is just presumably intended to only absorb the current spikes from the CMOS state transitions within that chip, which is why you need to start playing around to get things right when you have a load connected to the output. I've personally never had much luck putting bypass caps on things. I do it anyway though. The signals usually still look like they're strong enough to take satellites out of orbit...
Hi Dave This channel is one of the best. I have used these oscillators in many of my designs and always treated it as a black box. Thank you for the informative tear down. Keep up the good work. Thank you.
I’ve taken a few of these apart that used two discrete smd rf transistors to form an oscillator and buffer. The resistors were part of the ceramic substrate and the caps np0 (presumably) standard chip caps.
Dave more videos like that. We need people with work experience to share their life experience and achievements here. Very nice video. Comments were even better. You the man.
Really neat. At 15:45, and previous close ups, it looked like the lower left pin was poorly soldered to the ceramic PCB. Also, I wonder how the metal lid affects the performance? Perhaps if the crystal circuit is very weak, it's susceptible to external noise? Back in my manufacturing consulting days, I had a client who was a contract manufacturer, and their client had them doing some 500 piece build. They were getting about 50% yield. Long story short, their client fouled up the PCB design a little and had the CM driving some heat sinks on the PCB with a mallet. Half of the time they were completely assembling the PCB hammering the heat sinks on, then soldering the crystals, the other half of the time they were soldering the crystals then hammering the heat sinks on.
The resonance frequency of the springs should be in the low kHz range at maximum; I suspect it to be even below 1 kHz. This is so far below the quartz oscillation frequency that the springs won't be excited; the ends won't even oscillate at any significant amplitude. They behave just like a solid mounting post. I think they act as a vibration and shock damper to prevent the quartz disk from breaking due to mechanical shock.
I cut open crystal packages in the 1960s and saw springs in some of them as well. This is the first time I've seen crystals with little divider chips. Cool! Thanks!
Yeah, I took one of those apart once. All that was in it was the crystal, and a 74LS90 in an SOIC14 package, all on a tiny PCB. Big deal; I way hoping for earth-shaking alien technology! Bummer.
their is earth shaking alien tech in there.. the rock that vibrates precisely when u put a little current thru it lol thats crazy ass shit if u ask me :p
(@2:41) - check out the piss-poor solder connection at the lower left! That ought to fail the next time someone drops it on the floor! 😃 The others don’t look too bueno, either!
3:41 - Maybe it's the lighting, but nothing appears to be connected to Pin 1! (3:38) Update: Ah, it 'could' be used as a tri-state enable (never seen one) or simply not connected.
As long as the spring pressure exceeds the oscillation frequency vibration it's a superior setup ..think about it the contacts are spring loaded so dropping it won't throw the values off as much if at all
We had a supplier rev up a sub-assembly... our application didn't require ethernet, so they took the (big, expensive) chip off the board. This meant they didn't have a clock source for the rest of the board, so they dead-bugged on a DIP oscillator... and proceeded to have real squirrely problems... sometimes the boards would be good, sometimes not. I got a free trip to Ottawa to help with the problem... turned out that putting bypass caps on the deadbugged DIP oscillator cured the issue. If you read the literature they recommend bypassing them!
I've teared down dozens of these and those are all the same. Crystal, op-amp and couple of passives. That N.C pin is used variously, sometimes with tuning cap, sometimes something else.
Also noticed that. 13:00 looks like a sum of the 8Mhz and some other sine wave... for instance, sin(x)+sin(4x+(pi/2)) gives a similar plot. There's maybe some parasitic RLC swinging together with the crystal..
The DaveCAD background is actually a real photo of a Post-It note with my hand drawn DaceCAD, I then use Smoothdraw and the Wacom tablet to draw on top.
Datasheet on the oscillator ship from the second oscillator (the one with the busted resonator) - it's s seiko part. www.npc.co.jp/en/products/xtal/clock-oscillator/5054-5056-series/
Hey guys. The tech behind Crystal Oscillators is pretty new to me, so I was hoping to get a couple of basic questions answered, for which I was I was struggling to find answers online. 1) How exactly are the fixed output frequencies of an SPXO set? We have many 'standard frequencies' such as 10,12,16,20 MHZ, etc. Is it only due to the way the crystal is cut, or does it also depend on the other supporting circuitry within the Oscillator? 2) How does the circuitry for a programmable crystal oscillator (Such as Epson's) differ from a fixed frequency oscillator? I know these involve 'blank' oscillators that can be programmed to a desired frequency using a handler. Does it contain any circuitry such as programmable non-volatile memory? Thanks in advance!
there is only one question left) what is a quartz resonator after all? Is it a filter? a frequency allocator or is it a frequency converter? what frequency signal should I send so that the resonator is excited and begins to resonate. From physics, we know about resonances at matching frequencies, well, if quartz is 4 kilohertz, then I have to send a signal to it exactly 4 khz, no more and no less, and eventually get a stable frequency of 4 khz and its harmonics. But now the question is not about harmonics. I know there are filters that pass through themselves only a narrow band of frequency to which they are created, but quartz seems to emit itself. Hence the question: WILL I BE ABLE TO MAKE QUARTZ OSCILLATE 4 KHZ BY APPLYING 1 KHZ OR 11 KHZ???? OR, WILL I BE ABLE TO SMOOTHLY CHANGE THE FREQUENCY WITHIN WIDE LIMITS AND QUARTZ WILL CLEARLY BEND ITS FREQUENCY, WITHIN WHAT LIMITS OF ITS POWERS?) If we tap on an acoustic guitar, we will hear the resonance frequency of the guitar body! it turns out that a short pulse was applied and the body responded with a pulse frequency of 800 hertz. How many pulses should be applied to quartz so that it resonates continuously?
Dave, I wonder if the 0 ohm resistor is there to configure the divider. Maybe try remove it? It didn't look like it was needed for jumping over another trace on the ceramic substrate.
The substrate is expensive to design relative to the simple oscillator so there are likely more applications for the ceramic substrate than just the device in the video - configurable substrate allows reuse between product lines.
Technically the quartz IS THE OSCILLATOR. It is called a QUARTZ CRYSTAL OSCILLATOR. The electronics drive the quartz as you should know. The electronics are not a resonant circuit. THE CRYSTAL RESONATES. The analogy is that a clapper strikes the bell, but it is the bell that resonates. Perhaps contemporary usage has smeared the older form of language into indistinction.
I don't know what this is, but I guess I'm learning something. Never know if I ever need to ever know this, never know. Hey, I know ABC and 123 because I paid attention and now I know how to use it. Thank you professor.
Have you tried the circuit again before cracking the capacitor? Maybe the exposed quartz disc is sensible to EMR, and the metallic capsule works as a faraday cage.
Sorry, I thought you only tested the oscillator with both the top of the package AND the bypass capacitor removed, not just without the metallic package.
Hi Dave, I got a long unanswered question i would like to know regarding termination resistor, hopefully you can make a video explain the following. 1) what exactly the purpose of us putting a termination resistor in a transmission line, what if we don't put termination resistor on that? will it damage our circuit or causing any problems? 2) I saw lots of articles mention about the termination resistor of 50 Ohm; 75 Ohm or even 150 Ohm ect... how do we know which is the correct termination value resistor we should put in a particular transmission line or a particular circuit? Thanks
I wonder if the higher frequency ripple is due to L/C resonance in the transmission line on the output. It is terminated with a dummy load, but there are no guarantees that the XO is driving at 50R or that the transmission line itself will be matched. Just a thought.
I've always wanted to tear these apart but I don't have too many of them and have always wondered if they had a bypass cap in and run out as well, kickass video man! As usual, awesome awesome stuff! It's always fascinating to look inside of stuff whether it's something from the 50s or whether it was in last year's cell phone model LoL :)
Can you connect the little buffer to its own power supply? That way you can see the effect while the buffer is properly decoupled. Nice video. Maybe people now also understand how important it is to decouple properly with boost converters etc.
You should try to show the voltage drops in the GND traces too, the effect of these is shifting the GND at the device(s) which obviously is not a good thing either.
Any chance that 3rd connection to the piezo is to double (or change) the frequency? I'm thinking 1st harmonic (=2 nodes;1 antinode) vs 2nd harmonic (=3 nodes; 2 antinodes) - like a guitar string??
If you look closely, there's no electrode on that 3rd position. The only crystals where you can do tricks like that are the three electrode "hybrid" crystals. these have two modes of operation, one where the two ends are in phase, and one where they are antiphase, and the resonant frequencies are slightly different. They're actually quite fun, but you need to dump silver on four different places to make them work the way you want. (Re) Building the assembly for measuring that was the very first job I had to do on my first day. All done with double balanced mixers - and current controlled attenuators, and as phase inverters.
Look closely - as Baker above said, it's not connected. The purpose is obviously mechanical: It's to keep the disc from flexing and cracking the solder joints.
The frequency of the crystals that I used to use for Ham radio, was controlled by the thickness of the crystal. We used to take them out of their cases increase their frequency, by carefully grinding them thinner. Found a video! ua-cam.com/video/j7CS4A8wB1E/v-deo.html My buddy didn't have a flat glass to use as a grinding surface, so he used his VOM (Volt Ohm Meter) glass face.. Soon, he was on the correct freq, but his meter was a frosty white! He had to spit on the glass to read the VOM! He gave sold that meter $50 meter for $5, and I sprayed the glass with Krylon clear coat. Used it for about 15 years! Anyways, the 'Rock' as we called them has a clean 'fundamental' frequency. If you want to double it, you can use resonate circuits. Or divide it with flip-flops..
@0:30 .. Surprised. I have an ancient colorburst crystal in a miniature tube (cant recall if it's 7 or 9 pin) envelope - looks identical. I guess I always assumed the guts of the more modern cans were much smaller.
You might have tried to break off the white plastic at the bottom just to see if they are in fact acting like rivets to hold the metal pieces together...
Be interesting to see if the resonator is actually going at a multiple of 1Mhz and they are using a divider to produce an accurate 50-50 duty cycle. The ripple would suggest that it may be going at 8mhz
Wow, I thought about doing this before just never did, however very interesting. Wonder what is the extra flang holding the crystal for, dampening parasitic oscillation? Hmm
There are precisely 8 pulses per cycle visible on the scope, isn't this an 8MHz crystal with a frequency divider? edit: never mind, he rants about it a minute later
I can't believe that he hacked the top off of the can instead of sawing the two halves apart at the weld between the two container halves. The internal bypass capacitor is understandable as the power input pin is not supposed to be a signal output pin. Ron W4BIN
I used to design and repair test and control equipment at a major Quartz Crystal manufacturer. This was "not recently". We did everything from growing the quartz through building TCXO, OCXO, DIL oscillators, SAW devices and Filters.
Our DIL oscillators used a real thick-film modules, with the chip bonded to it, and connections made with bond wire. Partly because (in big volume) it was cheaper, but also because it allowed you much more space to "do things properly" - with one of those things being proper decoupling.
Dividers were quite common, partly because it is hard to make good crystals at certain frequencies, and there was also a "size" thing - good 1MHz crystals are quite physically big.
Did you get a sample of the gas when you opened it? Mostly they're filled with Nitrogen with a tiny bit of helium. This is part of the controlling the ageing, the helium makes it easy to check them for leaks very thoroughly.
The mounts were crucial, get them wrong, and you get a lousy, rotten xtal. How you hold the crystal changes its characteristic. Get it wrong and you'll get a horrible product with issues like "band breaking".
For crystals in conventional packages - those two clips means you hold it in just the right way, and with just enough "spring". It's much hard with DILs and that's why you often see three posts.
If you were seeing substantial shifts and changes in ageing following a shock, then bluntly, your crystals were a bit crap. The phenomenon has several names in the industry, we called it "sleepy crystals". It really was (and still it) one of the industry's dirty little secrets. It is a sign of a production problem. I'll explain that later.
In extreme cases you would have a product sat in a drawer for a few months where the Oscillator wouldn't start at power up. If you (very carefully) took it apart, you would find the ESR was through the roof, the Q was pathetic and if you did get it to start oscillating, it was likely to have been on a "band break" - so the frequency would be slightly (but significantly) offset.
Bang it on the table - and all would be back to normal again, until it was left for ages.
One of my first post graduation jobs was to design something which could detect this problem without the whole "waiting a month or two to see what happened" thing.
I designed a system, which used the then standard ISO methods of measure frequency and ESR, except at very low drive levels. You would then gradually increase the drive level, watching to see how F and ESR change, until you are at maximum power, then you wound the drive level back to the zero again. You plotted the results on a graph, and you would get a fairly classical hysteresis curve, and the bigger the area inside the graph, the dodgier the crystals. You only had to leave the crystals alone for a day or two, per-test to get a clear result.
I'm being called out on an errand now, but I will return later and explain where the phenomenon comes from.
Very interesting stuff! Posting a reply so I can hopefully get notified of your followup comment.
Awesome detail, thanks!
We used to fill and hermetically seal underwater modules with nitrogen for thermal conductivity. Was easy to use tester to sniff for leaks.
Andrew Baker - That was a fascinating read - thanks! ☺️
"Did you sample the gas?" Yeah, he was hoovering above it with a GC syringe, and injected it into his pocket GC. LOL. I'm certain he didn't sample the gas. Ha.
Wow, that really took me back. About 20 years ago when I was working for NEC at a PC factory in MA.
We were getting our PC motherboards (MBs) from SCI Technology in Tampa FL.
We started seeing a lot of PCs fail on the assembly line. Some of the MBs were DOA, but most (90%) were just working badly. After looking at some clock signals on one of the poorly working MBs, I zeroed in on one of those little metal oscillator modules. The clock was off-frequency, so I replaced it and the MB worked!
We powered up about 20 of the failing PCs and I used an RF spectrum analyzer (with a probe antenna) to observe the signals from the boards. Since these boards had six oscillator modules, I mapped out the RF profile from a good MB and compared frequencies emitted by the failing MBs. Some of the failed MBs had more than one failing oscillator modules. The failure pattern of which module and the shift in frequency was random.
I decided it was a QC problem with the oscillator modules. However, since some of the off-freq modules came from different vendors, I deduced these modules were somehow damaged at the SCI factory in Tampa. SCI said Nope, no way..
I had a collection of failed modules, a microscope with a Polaroid camera and a dermel tool..
I did a presentation for the NEC VP of manufacturing and was quickly on a plane to Tampa.
The photos of the little cracks in the quartz disks told the story.
Because I'm a Ham Radio guy, I knew that dropping a radio Crystal 4" onto a desktop would break it. So I knew what to look for at the SCI factory.
I watched the guy that loaded the tubes for the PCB parts loading machines. Nothing unusual.
Then late in the afternoon, I saw him unloading the plastic tubes from the module vendor, he was dumping them into a hard plastic tray, from a height of about 6 inches!! Afterwards, he hand-loaded them into the special tubes for the parts-loaders. He didn't believe he was breaking them. I shown him the pictures. Another Case closed.. :)
Awesome story, thanks for sharing.
Reminds me of (IIRC) an Intel story about bad yield on their chips. They couldn't figure out what was happening so sent someone to follow the wafers from the source.
Turns out a very diligent goods-inwards person was opening the sealed wafer boxes and counting the wafers in the (not particularly clean air) loading dock to make sure they got every wafer they paid for.
So what happened to the Goods-inwards Person? Fired? or just given a friendly warning to "Please dont unseal the packages" with the understanding that the persons intention was good, but the result wasnt?
Great events often come from little causes.
In the late 1970s, Intel made the 8251 USART and NEC used their own version of the NEC8251.
After seeing double letters being printed (during line testing) on the Spinwriter serial RS232 printers, I telexed NEC Tokyo describing the failures. Weeks later came the reply from Japan. "Intel 8251 has handshaking error in RTS CTS logic. Our NEC8251 is exact copy, so it also has error. We will create new NEC8251A without Intel error"!
And within a month, we got tubes of replacement USARTs..
I don't know what SCI did, but the cracked quartz oscillator elements problem disappear the afternoon I got to Tampa. I got to spend the weekend with my father-in-law before heading back to BeanTown..
Fascinating video Dave. Crystals and resonators belong to the dark arts, just like fortune telling and antenna design.
The rabbit hole goes very deep!
I don't know jack shit about what this component is about or what to use it for, but was hooked with your narration style and the way you tear down baby electronics components, sorta reminds me of how I used to tear down my toys when I was a kid to see how it works 🙏🏻
Nice video. BTW, your viewers might find this interesting: this device is a thick-film hybrid. Meaning, the traces are not etched, like a PCB. Rather, they are made by silkscreening on a conductive paste and then firing it in a kiln. And what looks like solder is probably conductive epoxy. I don't recall if you ever did a video explaining hybrid circuits. I used to design stuff like this back in the '80s.
good info thanks :)
Interesting info!
Fascinating
I'm not an electronics guy but I love your videos. I find this fascinating how when we build a device each individual component has so much detail and knowledge behind it. Standing on the shoulders of giants indeed! Even Wozniak was just taking complex components already designed by someone else and plugging them together in a novel way. All around the world there are a small number of experts on just one topic I guess - capacitor construction, chip fabrication and of course crystal clock circuits ☺️ it means that ordinary guys can take a raspberry pi for example and all that work has already been done for you so that we can concentrate on the target application, software and maybe very particular hardware for the application in hand without having to learn about oscillator circuits or PCB fab!
That internal bypass cap IS necessary, since these are high frequencies here (8MHz) vs. the 'user' frequency of 1MHz which needs to be bypassed as well.
Absolutely fantastic video dave - one of your best imo. What I like about this is not just the sweet tear down of those crystals, but the way you've demonstrated how bypass capacitors have an effect (and how distance from the connection makes all the difference), but also your mention about vibration too. I've come across odd ball issues where a system was suffering jitter because the can was not grounded (on a 2 pin crystal), and I've also seen issues where another system had jitter or something because the can wasn't supported (floating) and was susceptible to vibrations from road traffic 20 yards from the house lol. People think these things aren't relevant, but they are factors.
I totally agree that this video was one of the best yet too! Such an amazing fundamental, no-fuss look into something that so many of us have wondered about.
The voltage seen due to trace length is V=-L di/dt. Ie if your trace is long it has a larger inductance and hence the current drawn or returned through it will generate a voltage. Keep all power and gnd traces short or use planes !!
"Hi, it's time for a quick 2 minute teardown" * video goes on for 17 minutes *
Welcome to the EEVBlog
+CDC Remember you could skip the extra stuff. Just see what you need/want.
ua-cam.com/video/_4-Sx-T6VBc/v-deo.html
sci-toys.com/scitoys/scitoys/radio/am_transmitter.html
“Welcome to the EEVBlog” That made me laugh. Yeah; he’s the chatty know-it-all neighbor you don’t want to punch in the mouth!
Classic Dave trickery
Sorry it cut into time for all the rest of UA-cam offerings?
I would find it interesting to see temperature compensated (TCXO) , oven heated (OCXO) and voltage controlled (VCXO) oscillators aswell, in comparison to the regular oscillator.
Thanks for going the extra mile and figuring out what that noise actually was. I had no idea it was a 8Mhz crystal with a divide-by circuit. Good stuff!
The reason for using an 8MHz crystal and divider is that if you used a 1MHz crystal, the crystal would be WAY to big. About the lowest frequency you can get in an HC-18 package is a little bit less than 4MHz, and they were pushing it when you got one at 3.579..MHz (NTSC colorburst frequency).
Now you know.
I'm so happy that you're back into the interesting stuff again! No debunking videos, just fun and interesting tests and experiments. And equipment shootouts of course! You can almost feel that you actually enjoy making the videos again. Thanks man!
You do realise that a lot of people like the debunking videos? And the recent pocket multimeter shootout wasn't that popular.
Many won't find this video interesting either.
People subscribe for many different reasons.
I didn't write it to be offensive, quite the opposite. Just nice to see you being jovial and enjoying the exploration.
I didn't watch the multimeter shootout video. I've already got 4 multimeters that I'm happy with - a bench meter on the shelf above my bench,, a good handheld meter, current clamp meter and a basic 2000 count meter.
@@simontay4851 hah !! U were just afraid you'd end up buying a 5th !!
😅😅😅😅😅
Everyone here should have a small stash of oscillator modules. Much easier to get your test circuit running when you can rely on the oscillator working without much fuss. Often you can find them on sale from reliable sellers as well.
Internal cap is just presumably intended to only absorb the current spikes from the CMOS state transitions within that chip, which is why you need to start playing around to get things right when you have a load connected to the output.
I've personally never had much luck putting bypass caps on things. I do it anyway though. The signals usually still look like they're strong enough to take satellites out of orbit...
Very cool Dave. I used a couple of these about 35 years ago. Never considered what is inside. Thanks for the insights - very informative.
Actually 1-2MHz crystal is extremely difficult to make, so these 1MHz osc resonates at higher frequency, like 4MHz, and divides down to 1~2MHz
Hi Dave
This channel is one of the best. I have used these oscillators in many of my designs and always treated it as a black box. Thank you for the informative tear down. Keep up the good work. Thank you.
After all these years you've been teaching the world. Thank you so much.
I’ve taken a few of these apart that used two discrete smd rf transistors to form an oscillator and buffer. The resistors were part of the ceramic substrate and the caps np0 (presumably) standard chip caps.
You did a great job adding that image of a teared down xtal on Wiki, it helped the internet grow !
Dave more videos like that. We need people with work experience to share their life experience and achievements here. Very nice video. Comments were even better. You the man.
Really neat. At 15:45, and previous close ups, it looked like the lower left pin was poorly soldered to the ceramic PCB. Also, I wonder how the metal lid affects the performance? Perhaps if the crystal circuit is very weak, it's susceptible to external noise?
Back in my manufacturing consulting days, I had a client who was a contract manufacturer, and their client had them doing some 500 piece build. They were getting about 50% yield. Long story short, their client fouled up the PCB design a little and had the CM driving some heat sinks on the PCB with a mallet. Half of the time they were completely assembling the PCB hammering the heat sinks on, then soldering the crystals, the other half of the time they were soldering the crystals then hammering the heat sinks on.
As usual Dave, your videos are explained in great detail and clarity. It's great to watch and learn.
The resonance frequency of the springs should be in the low kHz range at maximum; I suspect it to be even below 1 kHz.
This is so far below the quartz oscillation frequency that the springs won't be excited; the ends won't even oscillate at any significant amplitude. They behave just like a solid mounting post.
I think they act as a vibration and shock damper to prevent the quartz disk from breaking due to mechanical shock.
The capacitor between power and ground is meant to add extra power during flip flop operation. Decoupling capacitor.
Fantastic. The hint and prove with 8MHz is genious. Cheers!
Gravity must be higher there, here this was more than 2 MINUTES. LOL. jk, very interesting video. Didn't knew how are those inside.
But there are multiple 2-minute teardowns in this video. Plus exposition for context.
Please use spacetime curvature, gravity is soooooo 1686.
I the "country" 1 mile = 10 and a couple minutes = 15-20
😂😂
Is all Australia there. The gravity points up.
One of those option resistors is open collector output. Pulls the output up. Another one is divide by 16 option.
I cut open crystal packages in the 1960s and saw springs in some of them as well. This is the first time I've seen crystals with little divider chips. Cool! Thanks!
Excellent video, a must see for anyone designing digital circuits
Yeah, I took one of those apart once. All that was in it was the crystal, and a 74LS90 in an SOIC14 package, all on a tiny PCB. Big deal; I way hoping for earth-shaking alien technology! Bummer.
How do you know 74LS90 and other digital tech isn't ALIEN... Must ask Fox Smolder...
their is earth shaking alien tech in there.. the rock that vibrates precisely when u put a little current thru it lol thats crazy ass shit if u ask me :p
The alien contribution is the bloody Magic Smoke u let out !
😡😠😡😠
(@2:41) - check out the piss-poor solder connection at the lower left! That ought to fail the next time someone drops it on the floor! 😃 The others don’t look too bueno, either!
Looks like it might have gotten hit with the dremel taking the case off. If not, yeah I need a job for that QA team lol... and a book to read.
I just realized Dave is the Steve Irwin of electronics
And sounds like that Ozzy Man fella, without the F-bombs lol
@@Shnick "We'll be opening this oscillator package today with this dremel and... ohhh! Welcome to destination f*$(ed!"
Well, let's hope Dave does not meet the electronics equivalent of a Stingray then ;-)
3:41 - Maybe it's the lighting, but nothing appears to be connected to Pin 1! (3:38)
Update: Ah, it 'could' be used as a tri-state enable (never seen one) or simply not connected.
As long as the spring pressure exceeds the oscillation frequency vibration it's a superior setup ..think about it the contacts are spring loaded so dropping it won't throw the values off as much if at all
Such an interesting video. Lots of effort made by Dave and lots of things learnt by us.
Hello! Dave
You are such a genius, many thanks for this explanation, I appreciate always your efforts.
Thank you again.
I have attended this interesting demonstration in suspense mood . Your approaches differs. Many thanks !
We had a supplier rev up a sub-assembly... our application didn't require ethernet, so they took the (big, expensive) chip off the board. This meant they didn't have a clock source for the rest of the board, so they dead-bugged on a DIP oscillator... and proceeded to have real squirrely problems... sometimes the boards would be good, sometimes not. I got a free trip to Ottawa to help with the problem... turned out that putting bypass caps on the deadbugged DIP oscillator cured the issue. If you read the literature they recommend bypassing them!
I've teared down dozens of these and those are all the same. Crystal, op-amp and couple of passives. That N.C pin is used variously, sometimes with tuning cap, sometimes something else.
I didn't really understand bypass caps but this definitely got me some good understanding of it and about EMF interference, good video ty.
(15:14) Might the resonator frequency be 32.000MHz? The ripple looks like it's 4X the frequency of the signal you're probing.
I showed the 8MHz waveform on the resonator pin
Also noticed that. 13:00 looks like a sum of the 8Mhz and some other sine wave... for instance, sin(x)+sin(4x+(pi/2)) gives a similar plot.
There's maybe some parasitic RLC swinging together with the crystal..
Thanks for cutting up one and showing us all.
Thank you Dave! I really appreciate your explanations of how electronic works.
As an engineer at Abracon, I loved this video
The bumps also allow cleaning solvent to flow under the package for defluxing.
12:06 - WHOAWWW! DaveCAD 3D! New Version is out, it seems :)
The DaveCAD background is actually a real photo of a Post-It note with my hand drawn DaceCAD, I then use Smoothdraw and the Wacom tablet to draw on top.
Pretty neat solution :) I was mentioning the perspective - you never got that before ;)
Datasheet on the oscillator ship from the second oscillator (the one with the busted resonator) - it's s seiko part. www.npc.co.jp/en/products/xtal/clock-oscillator/5054-5056-series/
Thanks, I was too lazy to search :->
Hey guys. The tech behind Crystal Oscillators is pretty new to me, so I was hoping to get a couple of basic questions answered, for which I was I was struggling to find answers online.
1) How exactly are the fixed output frequencies of an SPXO set? We have many 'standard frequencies' such as 10,12,16,20 MHZ, etc. Is it only due to the way the crystal is cut, or does it also depend on the other supporting circuitry within the Oscillator?
2) How does the circuitry for a programmable crystal oscillator (Such as Epson's) differ from a fixed frequency oscillator? I know these involve 'blank' oscillators that can be programmed to a desired frequency using a handler. Does it contain any circuitry such as programmable non-volatile memory?
Thanks in advance!
there is only one question left) what is a quartz resonator after all? Is it a filter? a frequency allocator or is it a frequency converter? what frequency signal should I send so that the resonator is excited and begins to resonate. From physics, we know about resonances at matching frequencies, well, if quartz is 4 kilohertz, then I have to send a signal to it exactly 4 khz, no more and no less, and eventually get a stable frequency of 4 khz and its harmonics. But now the question is not about harmonics. I know there are filters that pass through themselves only a narrow band of frequency to which they are created, but quartz seems to emit itself. Hence the question: WILL I BE ABLE TO MAKE QUARTZ OSCILLATE 4 KHZ BY APPLYING 1 KHZ OR 11 KHZ???? OR, WILL I BE ABLE TO SMOOTHLY CHANGE THE FREQUENCY WITHIN WIDE LIMITS AND QUARTZ WILL CLEARLY BEND ITS FREQUENCY, WITHIN WHAT LIMITS OF ITS POWERS?) If we tap on an acoustic guitar, we will hear the resonance frequency of the guitar body! it turns out that a short pulse was applied and the body responded with a pulse frequency of 800 hertz. How many pulses should be applied to quartz so that it resonates continuously?
AFAIK in these oscillator modules they always use dividers and/or PLL multipliers to ensure an exact 50% duty cycle.
Dave, I wonder if the 0 ohm resistor is there to configure the divider. Maybe try remove it? It didn't look like it was needed for jumping over another trace on the ceramic substrate.
Likely. And those other two pads too.
The substrate is expensive to design relative to the simple oscillator so there are likely more applications for the ceramic substrate than just the device in the video - configurable substrate allows reuse between product lines.
Technically the quartz IS THE OSCILLATOR. It is called a QUARTZ CRYSTAL OSCILLATOR. The electronics drive the quartz as you should know. The electronics are not a resonant circuit. THE CRYSTAL RESONATES. The analogy is that a clapper strikes the bell, but it is the bell that resonates.
Perhaps contemporary usage has smeared the older form of language into indistinction.
Is it possible the springs are clips? Perhaps the crystal disk would be added after the PCB is built and then soldered on to make assembly easier.
Amazing teardown of oscillator.
What does the AT stand for in AT cut?
I think it's because it's one a pair of cuts which have similarities, and were discovered together. One is called "AT" and the other is called "BT".
ua-cam.com/video/b--FKHCFjOM/v-deo.html
Maybe, you can communicate with the IC with AT command for trimming ! - Lame joke
What does the T stand for in AT?
Dave, Can you make a video about other types of oscillator? TCXO, OCXO, VCXO and SAW resonators. What are their uses?? Thanks!
Definitely worthy of a thumbs up. This was fascinating!
I don't know what this is, but I guess I'm learning something. Never know if I ever need to ever know this, never know. Hey, I know ABC and 123 because I paid attention and now I know how to use it. Thank you professor.
Have you tried the circuit again before cracking the capacitor? Maybe the exposed quartz disc is sensible to EMR, and the metallic capsule works as a faraday cage.
Yes, I showed it working in the video!
Sorry, I thought you only tested the oscillator with both the top of the package AND the bypass capacitor removed, not just without the metallic package.
Hi Dave,
I got a long unanswered question i would like to know regarding termination resistor, hopefully you can make a video explain the following.
1) what exactly the purpose of us putting a termination resistor in a transmission line, what if we don't put termination resistor on that? will it damage our circuit or causing any problems?
2) I saw lots of articles mention about the termination resistor of 50 Ohm; 75 Ohm or even 150 Ohm ect... how do we know which is the correct termination value resistor we should put in a particular transmission line or a particular circuit?
Thanks
Why are ceramic PCBs used for these purposes? What's the advantage compared to, say, ordinary FR4?
My guess would be the superior low thermal expansion of ceramic helps the crystal stability given the rigid mounts.
I wonder if the higher frequency ripple is due to L/C resonance in the transmission line on the output. It is terminated with a dummy load, but there are no guarantees that the XO is driving at 50R or that the transmission line itself will be matched. Just a thought.
I've always wanted to tear these apart but I don't have too many of them and have always wondered if they had a bypass cap in and run out as well, kickass video man! As usual, awesome awesome stuff! It's always fascinating to look inside of stuff whether it's something from the 50s or whether it was in last year's cell phone model LoL
:)
Very interesting Dave-so you are studying for you Radio Ham licence ?
If you use the Doofenshmirtz Evil Incorporated tune as intro.. these will just perfect!
Can you connect the little buffer to its own power supply? That way you can see the effect while the buffer is properly decoupled.
Nice video. Maybe people now also understand how important it is to decouple properly with boost converters etc.
Fun fact...the OCXO in spectrum analyzers are also mounted on springs!
Do these really have a crystal inside that shakes like crazy in its own tiny little world?
Yes, you could start off by Steve Mould's "How a quartz watch works", why 32 kHz etc :)
Theres a datasheet for that oscillator IC 613N1 they could be purchased separately
You should try to show the voltage drops in the GND traces too, the effect of these is shifting the GND at the device(s) which obviously is not a good thing either.
Any chance that 3rd connection to the piezo is to double (or change) the frequency? I'm thinking 1st harmonic (=2 nodes;1 antinode) vs 2nd harmonic (=3 nodes; 2 antinodes) - like a guitar string??
I don't believe so, but it would be nice to get an expert opinion from a xtal manufacturer on that
If you look closely, there's no electrode on that 3rd position.
The only crystals where you can do tricks like that are the three electrode "hybrid" crystals. these have two modes of operation, one where the two ends are in phase, and one where they are antiphase, and the resonant frequencies are slightly different. They're actually quite fun, but you need to dump silver on four different places to make them work the way you want. (Re) Building the assembly for measuring that was the very first job I had to do on my first day. All done with double balanced mixers - and current controlled attenuators, and as phase inverters.
Look closely - as Baker above said, it's not connected. The purpose is obviously mechanical: It's to keep the disc from flexing and cracking the solder joints.
The frequency of the crystals that I used to use for Ham radio, was controlled by the thickness of the crystal. We used to take them out of their cases increase their frequency, by carefully grinding them thinner. Found a video!
ua-cam.com/video/j7CS4A8wB1E/v-deo.html
My buddy didn't have a flat glass to use as a grinding surface, so he used his VOM (Volt Ohm Meter) glass face.. Soon, he was on the correct freq, but his meter was a frosty white! He had to spit on the glass to read the VOM!
He gave sold that meter $50 meter for $5, and I sprayed the glass with Krylon clear coat. Used it for about 15 years!
Anyways, the 'Rock' as we called them has a clean 'fundamental' frequency.
If you want to double it, you can use resonate circuits. Or divide it with flip-flops..
@0:30 .. Surprised. I have an ancient colorburst crystal in a miniature tube (cant recall if it's 7 or 9 pin) envelope - looks identical. I guess I always assumed the guts of the more modern cans were much smaller.
Willing to bet that the bypass resistor is used for switching the frequency division to match the crystal.
12'19: Did not know you wrote your own CAD software. DaveCAD. Love it.
You might have tried to break off the white plastic at the bottom just to see if they are in fact acting like rivets to hold the metal pieces together...
Always wanted to look inside one of those!
What if you put fine iron filings on that copper "ground" surface? Will it form the actual path of the current?
Right...... the 8MHz inside crystal is devided by 8 to make 1MHz Crystal oscilator..
Are these sealed with a vacuum? Or maybe inert gas like dry nitrogen?
I can't believe how good your videos are! Legend
Great surgery Mr Jones, like the orthographic DaveCad.
Very educational. Thank you Dave!
I've always wanted to pull one of these modules apart, Dave, So thanks for doing it for me. ;)
Be interesting to see if the resonator is actually going at a multiple of 1Mhz and they are using a divider to produce an accurate 50-50 duty cycle.
The ripple would suggest that it may be going at 8mhz
Didn't watch the whole video did you?
So in a critical design you might avoid using those modules and instead put in a crystal with a known frequency..
By any chance do you know how to use a 3-pin oscillator/ crystal? I've got a couple of them, but I'm not sure how to use them, thanks for your help
Wow, I thought about doing this before just never did, however very interesting. Wonder what is the extra flang holding the crystal for, dampening parasitic oscillation? Hmm
Sweet techbits, brain feels better! Thanks! So, any cap is better than no cap!
What would happen if you would put your finger on the crystal?
The "cushioning" coils supporting the crystal , the material is silver ? , interesting video !
You can simply measure the capacitance of the oscillator to know if there is one in them.
There are precisely 8 pulses per cycle visible on the scope, isn't this an 8MHz crystal with a frequency divider?
edit: never mind, he rants about it a minute later
i wonder if that jumper actually enables the divider
I can't believe that he hacked the top off of the can instead of sawing the two halves apart at the weld between the two container halves. The internal bypass capacitor is understandable as the power input pin is not supposed to be a signal output pin. Ron W4BIN
Excelente video Saludos desde Argentina
wonder what capacitor size would be best to filter that 8Mhz out
Hi Dave. Can you measure frequency directly on the crystal pins? Maybe crystal frequency is 32 megs?
I did, I thought that was pretty clear?
I know it’s an old video, newbie question, can I use the open cans and replace the crystal with sockets to use it as a a crystal tester?