TDA2050 vs LM1875 performance tests to see which is the BEST
Вставка
- Опубліковано 30 бер 2021
- These two ICs are my favorite audio chips produced. I wanted to take a deep dive into their performance to see which one stood out as the best.
Patreon www.patreon.com/user?u=12254451
Mail: PO box 362, Vandalia, OH 45377
Email: johnaudiotech@gmail.com
Thank You! - Наука та технологія
Sir your channel deserves ten times subscribers.
Great video John!! Always like it when I get the notice that you posted a new video!
Hmm, you and John should collaborate on an affordable, complete stereo amp project based around one of these chips ;).
I always Love your videos and I think there will hardly be any teacher like you who will explain smaller details with ease to their students even in most premier engineering colleges and universities and that too with practicals!
Great test, thanks for sharing your findings and motivating us to play with these chips
Tank you for experience. I loved your demonstrative mode to amplifiers. Yes very good perfec video. Big like for you
Thanks John,we can see your hard work
Thanks for the superb comparison!
Nice work John - really interesting.
man, i just saw the first 10 seconds of video, paused it, liked, subscribed and now writing this. here is what im looking for and i know that by that intro bro
Hi John,
Being an engineer that has worked in a variety of disciplines. It was back in 2004 while pondering with a hunch that someday amplifiers will be class D based and driven by Class A FETs or Vacuum tube preamps. Audio knowlege to me then was strictly repairing Guitar amps and other audio equipment 'from a Schematic' for people in my spare time. I used to listen to and maintained 70's 80's era component stereo equip.
With several of my designs, (AB and a Couple D) with tube pre and a DAC, I hear detail in music that my old school equipment does not exhibit.
I still love the 80's "Ted Nugent" audio quality but my personal taste ranges from Classical, blues, jazz fusion to the DefTones.
About 4 years ago I dove into this SS niche area as well a Vacuum tube audio design.
It was way out of my engineering expertise but now I'm an addicted experimenting and learning everyday !
BTW: You are indeed humble. As far as a ' Tech ' I think a bit more than that John. What other technical skills resides in your mind ? I can't even begin to imagine.
John you are my #1 guy to watch on UA-cam. I have been working with these Class D and AB modules as well as my own designs from class A tube preamps to TDA2030 / 2050 Chips. I have no idea how many variations of the TDA3116D2 based Chinese boards I have. As you know the operative word there is VARIENCE.
Not much one can do with a poorly laid out PCB, while others with some changes are winners for me.
Coincidently They turn out to be the ones in your top 5.
I buy and modify MANY of the Chinese Kits and boards and replace inferior parts to USA Caps, Semiconductors etc. I built a dual TDA2030 with bipolar supply and it actually sounds wonderful even with the controversial Triode preamp running at low voltage. Yes, I play with those and some do look good on the scope Sine tests and FFT (mine is a high end Hantek) and add comfort and warmth to the incoming music.
To save ground-up design time, I buy a particular 6J1 Fever preamp kit, modify the heck out of them and run them at around 100 VDC or more to the plate. The others I design and build are 12AU7 family based with about 240-260 VDC to the plate, cathode biased. My stockroom has over 4000 Vacuum tubes and many USA NOS premium 12a*7 including several rare 7316's, Yes they work in place of a 12AU7 and seem to be a very high quality and rugged 12AU7's though not on any cross refence sheet. They do sound great ! as well as some 5963's in the computer "Relay" family.
(often my designs and PV/Grid design choice is based on my transformer inventory. For single ended OTs I bite the bullet)
It is time order LM1875's (Not Fake) and build a board to try in comparison to the TDA's. Where did you get the Lm1875
We have very similar interests in this niche, and our test equipment and bench are very similar. There are also 2 cats here Mr. B and Mr. Pnut they jump on the bench to assist me....with reorganizing my test leads, jumpers, grab-on clips etc.. I don't tell them they are doing it all wrong. Lol
With that said John,
You have taught many things to expand my knowledge and enhance my testing capabilities using said time and frequency domain test equipment, and so much more !
It is time I become a Johnaudiotech Patreon since you so kindly , humbly, and without profit share this well presented, and easy to understand information that truly help folks like me.
Wow, long comment but just had to do it with Thanks !
…and again,
Thank you so much John !
👍👍
Great audio channel!!!
Very nicely done, simple to understand, and really informative! You have a talent for making things simple, that others make complicated!
You get a sub-i-script-ion! :)
JOHN! Excellent!
John, have you tried the TDA2050L chips by Unisonic Technologies Co.? The claimed specs are exactly the same as for original TDA2050 by ST.
Great explanation with comparison ....really appreciate....Pls do some experiment with STK family ICs , such as 4191, 465.... etc..
You are AWESOME MAN
I appreciated the daily dose style intro!
Very interesting .
Good work!
Love your channel...
Thanks for the informative video John. The main take away I get from this is of the two, a TDA2050 is preferable if you don't plan to over drive it. Imo, a given level of IMD sounds way worse than the same level of THD. Also, this video demonstrates that Thiel networks can be beneficial.
Great video Sir,
Could you do a video of an ideal amplifier circuit assembly from your experience 🙏
Hey thanks for the video's
What timing, I just fully rebuilt my 5.1 stereo, it used to be all SMD parts, now it's all through-hole (long story).
I replaced the output section with TDA2030A. The center speaker bridge I made, was having a lot of 1-2MHz oscillations, that for now I mostly seemed to have tamed by adding a small bypass cap to 1 of the inverting pins. Although, I'm still picking up 99MHz radio too, but it's not shielded yet either, I have to make a new box and wire it nicely.
So far the overall balance might need tweaking, but sounds ok. It was cheap to start with, IDK what a good system should sound like.
Now I'm tempted to call that a trail run, and remake the whole thing again, and improve my layout and wiring, and use all new legit parts, not ebay's cheapest LOL
That was great test. I used to think lm1875 is better than tda 2050 now my mouth is shut 😅 will start investing in tda2050. But orginal IC s are hard to find.
There are new chips called TDA2050L by different manufacturer Unisonic Technologies Co.. But I have not yet seen tests of them.
Just got ten of those mailed to me today . They are there. Laying on shelves of local suppliers.
7:50 Even with no speaker you can hear the harmonics of the 10kHz tone when it clips. I remember mine doing the same thing during testing. Sounds like a little ESC for a remote control car. Informative video, BTW!
Hi John,
I've been wondering, with that Siglent scope and a Siglent signal generator, would you be able to do Bode plots?
I can't help but wonder: would it not be more convenient to have a Siglent function generator and cheaper tracking power supply than the Siglent power supply and a Feeltech?
Great video! A very interesting comparison. Thank you,.
Cheers
Mark
I'm wondering how the reactive performance would be when connected to something that more closely resembles a loudspeaker. I'd think capacitive would be rare. I'd expect it to be inductive. Also curious about output impedance/dampening factor. That would be an indication of how the amp deals with the back EMF created by the speaker and how well it can control the movement.
THANK YOU, thank you for this recording, can you record a similar test with LM4780 vs LM3886? Regards. DZIĘKUJĘ, dziekuję za to nagranie, czy może Pan nagrać podobny test z LM4780 vs LM3886 ? Pozdrawiam.
Appreciated for your measurement. Have you ever tried Inverting with buffer one? It might be lower distortion.
At 7:13 you show 10.29 v RMS @ 29.00 V pk-pk. The last time you showed the power supply was +/- 12 VDC. I am interested in rail-to-rail performance, or how close the linear output range comes to actual supply voltage. Did you do a off-camera change to supply voltage, and if so do you remember what value? Thanks.
Thanks for showing this interesting comparison, could you post a link to the boards you are using?
I'd like to see your impression of high output ICs.
Could these be configured into the Footnote Guitar Amp Board from Parts Express to give it more output for an 8 inch 4 Ohm Jensen Green 25W Speaker?
Very good video, next time please compare lm3886 with tda7294.
Have you checked out the layout on the TDA2050 Boards on E-Bay - DC 12-24V 5W-120W TDA2050 Mono Audio Power Amplifier Board Module 1 Channel?
Yeah.. I still using TDA2050 for Sub woofer amplifier in my 2.1 amplifier and TDA 2004A for Satelit.
Feeeltech cracks me up. I had the same one but it said kooooolertron. I have another koooooolertron now that's very nice but I need to learn to use it. It's a rebranded juntek. Haha
If one of these components gets damaged will there be output from let's say a guitar voltage? Or does it affect the output?
Thank you
John, did you ever tried testing but with reactive load that would emulate, say 2way bass reflex loudspeaker. Some 30° phase shift makes the 4 ohm look like 2.5 ohm to the amplifier. PITA, when one tries to design really competent passive crossover network.
I tell you, when I was younger and my ears were good, I loved what the Erno Borbely amps could do - unfortunately parts are NLA. Even his not very complex preamps sounded much better than, ... something based on LM4562.
Better to sacrifice bandwidth than having the amp turn into cristal - and BTW I have no idea what some of the Hi-End audio cable firms think when they sell cables based on Litz, or similar - designs like this usually exhibit very capacitive load. Even Nelson Pass mentioned that often cables like that can make amp go crazy.
Sir john i need help
The tda7293 out put ground can replace to negative
What about inserting a Notch Filter in the feedback loop to remove the oscillation.
Where did you get the amp pcb from ?? Is there a schematic
I am also making a amplifier circuit using TDA2050 with +-20V source. use 4xTDA2050 for 2 channel BTL. With Pre chip opamp RA3678
I've been thrashing a counterfeit TDA2050 on 24V from a rechargeable vaccum cleaner battery.
It's fed by a siren generator and is going into an 8 ohm horn speaker.
I haven't blown it up yet but darn it gets loud. I might have to short the output to see if it dies. I have 10 more counterfeit chips
Did you ever find a source for those boards with that layout?
Do you know of an alternative to the obsolete TDA2003V?
Hey men, could you do a dirt cheap ebay full chinese DAC test?
I was surprised that the 1875 had the performance issues over the 2050. The 2050 is kind of the current hog but sounds good on normal room volume in a center woofer config. I gave up on the 2030's a couple years ago except for small speaker use
Thanks for the comparison, they both seem like prime candidates for driving computer speakers. Could you demonstrate/sketch how the resistor coil was introduced into the circuit to smooth that square wave? And if I understood your explanation, does that only take effect during significant clipping?
The coil/resistor is just in series with the output, so nothing fancy. In regards to clipping, this test was not about clipping, but about adding capacitance (charge storage) to the output at various output levels. It's unlikely an amp will see these levels of capacitance with typical speakers.
@@antibrevity thanks!
What about 8ohm load? I'm not surprised. We all know lm1875 is not good into 4ohm load.
Bro where can I get lm 1875 original ic??
Has anyone tried copying the Cathode Drive Amplifier that drives the heated cathode of a CRT which should have a band of 1MHZ
A little second harmonic distortion can be a great thing for audio! For example Neve's Silk circuit is based on harnessing it. How much is too much in your opinion?
I struggle to hear it even at a few percent in music, however, I prefer the amplifier to minimally distort the audio.
@@JohnAudioTech Good point. Color is prob left to pre-amps. Thank you for your videos!
Famous Dave Jones "Wriggle Wriggle Wriggle Yeah" ~
I have a Kustom dart 10fx guitar amp. And it does not have any output when I hook up the guitar (power is on)
Could it be this IC the problem?
Anyone know?
Thank you
Unfortunately, there's no way to tell just by your description. Lots of things could make it not have output.
You might try opening it up and examining the circuitboard and components closely, looking for burn marks, blown chips, and burnt components. But you need some equipment and experience to actually naill down the problem and fix it.
18:50 how does that coil on the output affect frequency response?
Inductance is far too small to affect the audio frequency range.
Super!
Would love a walk through of your daily driver setup, the one you listen to for pleasure....great tutorial...cheers.
I plan to do so when I review my speakers.
I'd like to know what became of some of the speakers you built from the old days, like the folded horns or the plexiglass sub.
Hi John,
Another question: why do we need to use non inductive resistors when speakers are inductive?
What would the same test look like if you hooked up your speakers?
And finally, wouldn't the goal be to have a flat frequency response with the speaker rather than an ideal load?
I guess that was actually three questions...
Cheers
Mark
amplifiers are normally tested with a non inductive load to check frequency response and power. Each speaker will have a different reactance that will make it hard to draw a conclusion (what to standardize on?). The low output impedance of these solid state amps will give a flat output voltage across the spectrum.
@@JohnAudioTech Thank you John. I guess that it makes sense.
Really great tutorial.
Cheers
Ma
LM1875 has higher distortion than TDA2050, interesting. LM1875 is very famous in the audiophile DIY community, maybe because they love distortion.
They do, same reason to use Tubes.
Given that Lm1875 is more forgiving for design errors, and clips bit more amicably then its no wonder it suits for DIY. 1% distortion sounds way better than any occilation.
Distorsion showned on the grap have nothing to do with human hearing and also not showing the sound character of the sound. For that reason 1875 had better and more natural sound, distorsions showed and measured here are not important on low volume where chip can handle the load.. If you check tda2050 it had a peek on 3tr harmonic where human ears are most sensitive while 1875 had the 2nd harmonics mostly domminant. There are lot of differences how they will sound like aout of regular measurements
Is there already an RC-Series-Element with 2.2 Ohms and 0,47µF between the Output of the TDA and Signal-Ground? This is important for frequency stability. Higher Resistor Values and lower capacitor Values leads to oscillating. Also the bypass capacitors at the positive and negative Power Supply Rail must not be too small, as this can also lead to oscillation. Recommended Values are 100nF and 220µF in parallel at both rails towards Ground for split-Supply. For Single-Supply the Values are 100nF and at least 1,000µF in parallel between positive and negative of the Power Supply. However, these values only apply to the TDA2050. The LM1875 needs different values.
I think it's always the best way to read the original datasheet for these IC's to mount the proper component Values.
I don't know, if or how the other components can mess up the frequency stability.
The next Question is, wich Type of Power Supply do you use - with a simple Transformer and linear regulated, or switch mode power supply? I think the Amplifier IC's don't like "dirty" Voltage from a SMPS.
I put in the 2.2 ohm and 470nf on the TDA2050 board as designated on the datasheet. I use 220nf 100v X7R ceramic across the rails. I put those in after finding the oscillation (but made no difference. I used the 100uf supplied by the kit and 6,800uf caps as shown in the video. The regulated supply is the linear type.
John sounds like JZ! ;0)))
Lm1875 works better with 8ohm load and tda2050 designed for 4ohm load
I'm in love with the lm1875 high frequency response. So silky and delicate.
I have Bliesma T34-B for testing - very interesting tweeter, and images like crazy, great dynamics but still sounds like metal dome. No smoothness of the silkies or ribbon.
the 1st comment everyone's waiting for....
Thanks John, a little bit disappointed since I'm LM1875 fan :)
Me too. When I tried the second LM1875 and saw the same distortion, my heart sank. .076% with a 4 Ohm load is still not bad. Should be less at 8 Ohms.
@@JohnAudioTech LM1875 was designed to have higher distortion to please the "audiophile". High quiescent current, good clipping behavior, barely audible/just inaudible distortion just the way audiophiles like.
LM3886 is much better than either of them. Gotta make a video of LM3886 VS TDA7293/7294.
Also TDA2030 its durable as well very popular in most audio system until now
เราเลือกแล้วจะใช้ LM1875 เป็นตัวขยายเสียง
Lm1875 with 0.076%? 😯
The 2050 is better where it counts. I know John was rooting for the 1875. Can't win 'em all the time.
No you can't win 'em all. I was dissapointed to see that. .076% distortion with a 4 Ohm load is still not too bad. Should be less at half the current with an 8 Ohm load.
@@JohnAudioTech I could tell you were rooting for that 1875. I could see how crestfallen you were examining the distortion too. But yeah no one is ever going to hear .076% distortion. As I understand it the threshold of audible distortion is close to 3% You should do a video about that.
unfortunately those ICs have been discontinued
You know most of the scopes these days have math functions where you can have it compute the wattage in real time. Just save the profile and your good to go. My Rigol can do this, I'm sure the Siglent can as well.
I need to sit down and fiddle around with the scope to see what it can do. I really only use it during video production.
@@JohnAudioTech It's all good, just thought you might find it interesting and perhaps useful.
Always LM 1875 still best coz of their protection future!
Unfortunately, both these amazing Tda2050 and Lm1875 chips are being phased out by the respective companies and the markets are now flooded with fakes which are actually copies of TDA2030A but printed as Lm1875 and Tda2050 or Tda2040
ST Micro canned the TDA2050 in September 2013. The LM1875 is still in production as far as I know.
so how do we know which one is fake?
I have a TDA2050 purchased long ago & I'm sure it is genuine but the ones I see online is something else.
Is buying them from reputable sellers like Mouser a good idea?
@@dmcbohol Tda2050, 2040 should not be purchased now as some manufacturers are making copies of Tda2030A and printing them as Tda2050 and 2040 selling these at higher prices. Also the fake Lm1875 chips look same like Tda2030A thereby raising doubts that it's also a copy of Tda2030A. I think only Tda2030A should be purchased as it's still available as good quality even if it's not manufactured by the original company as it's very popular still.
@@amitghosh6966 absolutely right bro.
@@dmcbohol Mouser will be a safe bet, they are good at assuring their supply flows aren't bringing fake parts.
TDA2050 is the best, if you can find one these days, I mean the original one.
So these tiny bits of metal and plastic create sounds that users can manipulate?!?! Wow! I will never understand how a $1 calculator made of metal and plastic can do complex math in an instant. Above my pay grade!
No, it's SEMICONDUCTORS, metal and plastic :)
And a fair bit of past physics research and painstaking engineering!
Ani plastik, ani metal nic nie tworzy. Tylko kryształ krzemu.
I prefer LM1875 for better analog warm bass.
TDA205 you using from 10 years 😰😧😧😧
I think the higher distortion numbers of 1875 is only for before-clipping output. For most relaxed listening use, it performs pretty much identical to 2050. I have tried both for a bedroom TV amplifier and did not notice any difference between the two - same source, preamp, supply transformer, PCB except for boucherot
I varied the signal level a bit to see if the distortion characteristics changed with both chips. It did not change.
I've seen that LM1875 has a distortion step when crossing zero level from positive to negative output and backwards. That's typical for class B amplifiers.
No, obviously not! the LM1875 is a much worse amplifier. Worse not only in terms of distortion in the sound band of different frequencies, but also because of much worse made output transistors, which have lower output current efficiency (even lower than the minimum value specified by the manufacturer in the datasheet. I took measurements at various loads) Transistors in LM1875 can not well driving a strong inductive loads - speaker.
The LM1875 is a newer amp, but sadly it is a disgusting manipulation and hypocritical failed IC from Texas Instruments. Much worse than the popular TDA2040, TDA2050. Despite slightly more power, I can only compare it to the older, basic TDA2030.
I thought it was a skunk going across the screen! Yikes😮😅
0:20 YOU ARE SHOWING LM1875 authentic chip and 4:00 you are starting testing an not so authentic lm1875 chip which came with a 2$ kit you got from China, the tests show that the fake looking chip is not so amazing .
Cute cat, 2050 is more likely to clip and sound awful, i know that.
quit waving that pencil