TDA7293 composite amplifier done right
Hello guys,
Has anyone seen that website - Low Distortion Composite Amplifiers LM1875 TDA7293 TDA7294
There is a lot of info, circuits, PCB, etc. concerning composite amplifiers and TDA7293 in particular.
Hello guys,
Has anyone seen that website - Low Distortion Composite Amplifiers LM1875 TDA7293 TDA7294
There is a lot of info, circuits, PCB, etc. concerning composite amplifiers and TDA7293 in particular.
Thanks for posting that!
Very interesting page!
Not too far from what we´ve implemented in the prototype.
Vast differences between controlling opamps and resulting THD but that is to be expected depending on choice and individual? compensations.
Dropped a mail to the site owner to see which opamps he used and if they are individually compensated.
Anyways results are quite good!
Our prototype works with both LM4562+OPA1656 but without THD-measurements there´s hardly any proof of how well the composite works.
Have to see how far I get with the cheap soundcards that I have.
Very interesting page!
Not too far from what we´ve implemented in the prototype.
Vast differences between controlling opamps and resulting THD but that is to be expected depending on choice and individual? compensations.
Dropped a mail to the site owner to see which opamps he used and if they are individually compensated.
Anyways results are quite good!
Our prototype works with both LM4562+OPA1656 but without THD-measurements there´s hardly any proof of how well the composite works.
Have to see how far I get with the cheap soundcards that I have.
post 31 by joensd : I do not see why it would be unfair to parallel the TDA as its done with the LM3886 as well , even by Tom in his Modulus 686. Only difference is that it is much more critical with the LM whereas with the TDA it is easy and simple to do . The LM has a higher possible dissipation and current than the TDA , the TDA higher rail-voltage but easy to parallel so why not use the advantage this provides , especially for loads of 4 or even 3 ohms . Lots of subwoofers dip to that. Plus the fact that two TDA cost roughly the same or less than one LM. Paralelling three of them one is on the safe side no matter what even with the rails at 50 volts , and beeing able to reduce the heatsinks to a more reasonable size at the same time .
I just got ten of them from China and yes , I measured them one by one to make sure they are not fakes , for a total of 15 dollars, shipping included
1pcs/lot TDA7293 TDA7293V ZIP-15 IC In Stock ....Total: US $14.68
so with those 10 I can experiment and try a few things now.
post 42 from WLA : thanks for that info something to jump on and devour , cheers
I just got ten of them from China and yes , I measured them one by one to make sure they are not fakes , for a total of 15 dollars, shipping included
1pcs/lot TDA7293 TDA7293V ZIP-15 IC In Stock ....Total: US $14.68
so with those 10 I can experiment and try a few things now.
post 42 from WLA : thanks for that info something to jump on and devour , cheers
Last edited:
Hi JD,
Did you hear from the guy?
I'm very curious how to find the best possible Op Amp.
I'm thinking that at the end of the day I'm going to order a finished PCB and find the answer in the easiest way.
In some ways its early in others its not . These things tend to be slow at first and then to speed up . We all have our lives dictating how much time and when we can invest in this .You are more than welcome to add your grain to it if you want .
I do not think there will be any "best" op-amp but rather a number of suitable ones , depending on taste as they do sound slightly different.......................................
I do not think there will be any "best" op-amp but rather a number of suitable ones , depending on taste as they do sound slightly different.......................................
Last edited:
Yes, I heard from KalTecs; not sure if this is a company or not.
They said it is best to have the Open Loop Gain and Phase VS Frequency function of the TDA7293 and that LM4562 and OPA1656 are great opamps.
(I mentioned that I tried both in our prototype)
So not much information given which is OK.
You obviously need an opamp that is good enough to perform error correction.
As each opamp probably needs an individual compensation that´s quite a task but doable.
Take your favourite opamp and compensate measuring THD each time you progress.
Most opamps we talk about here are quite capable BTW but an old 741 probably won´t do.
I do have a working prototype (working as in stable til full output voltage with 8Ohm load and capacitances in parallel to the load up to 1µF).
That doesn´t mean much though as I haven´t measured THD as I stated before. In its current state the controlling opamp might be so slow that the error correction is only barely working.
See graphs from the project linked in post #42.
One of the chosen controlling opamps hardly improves the THD in comparison with a single/non-composite TDA7293 which means that either the opamp is not capable enough to do error correction or it isn´t properly compensated.
So yes, early but maybe not so early.
And again, this is not a circuit to wildly swap opamps in.
You will have to perform another stability evaluation with each swap (and following changes in compensation) and measure THD.
I think the above should answer that too and I already stated that I can´t perform THD measurements at the moment.
And yes, as Mike said, this is a hobby and time for it limited.
If you can measure THD and want to join the testing club I could share the gerbers for the prototype. I need to ask FauxFrench first though as this is his circuit.
Last edited:
Hey Bernd, you sound like my old boss back when I was designing PCBs.
Wait, maybe you are my old boss, who knows? ;-)
Sorry but the last paragraph from post 52 is still valid.
Circuit works just fine and is very stable but no THD measurements.
I worked on speakers a bit but got some symmetric SMPS that I might use for that amp.
I should really try and use a cheap USB-card that I have somewhere for THD-meas. but that might take another while.
Which free SW do you recommend for THD-measurements?
Wait, maybe you are my old boss, who knows? ;-)
Sorry but the last paragraph from post 52 is still valid.
Circuit works just fine and is very stable but no THD measurements.
I worked on speakers a bit but got some symmetric SMPS that I might use for that amp.
I should really try and use a cheap USB-card that I have somewhere for THD-meas. but that might take another while.
Which free SW do you recommend for THD-measurements?
I am very sure that i am not your boss, Jens, be it the old one or the new one...
If you take cheap USB soundcard you will probably only see, that the soundcards quality is worse than the amps quality. As you have seen with my measurements, it is the soundcard that is the limit - and my soundcard is not a cheap one. At first you have to test the soundcard itself in a loop - mute all parts of the card that are not in use and then find a level where the THD reading is smallest. My soundcard limit is the DAC or the lowpass behind it, because with an external signal, the level that shows no distortion that can be put in to the cards ADC is about 12dB higher.
The software i use is ARTA, as you might have seen on my measurements. If you only want to measure you can use it in its demo mode for free. The only limit in this mode is that you are not allowed to save your measurements. But it is possible to export your measurement as a picture ore a .csv-file. The measurements THD vs frequency i made with STEPS. This is like LIMP also a part of the ARTA software.
If you take cheap USB soundcard you will probably only see, that the soundcards quality is worse than the amps quality. As you have seen with my measurements, it is the soundcard that is the limit - and my soundcard is not a cheap one. At first you have to test the soundcard itself in a loop - mute all parts of the card that are not in use and then find a level where the THD reading is smallest. My soundcard limit is the DAC or the lowpass behind it, because with an external signal, the level that shows no distortion that can be put in to the cards ADC is about 12dB higher.
The software i use is ARTA, as you might have seen on my measurements. If you only want to measure you can use it in its demo mode for free. The only limit in this mode is that you are not allowed to save your measurements. But it is possible to export your measurement as a picture ore a .csv-file. The measurements THD vs frequency i made with STEPS. This is like LIMP also a part of the ARTA software.