@Vunce
Everyone hears differently, everyone has his own sense for aesthetics.
I wanted to present objective measurable and repeatable results, instead of imposing my own impressions or worse - opinions.
I consider this approach as fair.
Also, good measurements usually corelate to my perception of sound. So, for me the amplifier with the best numbers among these I've compared sounds the best. For me. Some like tube amplifiers with even more distortion, I don't, but I respect the multitude of different perceptions.
...
I wasn't about to discuss tastes, I just want to solve the problem that increases the THD.
Anyone?
Ideas, please.
Everyone hears differently, everyone has his own sense for aesthetics.
I wanted to present objective measurable and repeatable results, instead of imposing my own impressions or worse - opinions.
I consider this approach as fair.
Also, good measurements usually corelate to my perception of sound. So, for me the amplifier with the best numbers among these I've compared sounds the best. For me. Some like tube amplifiers with even more distortion, I don't, but I respect the multitude of different perceptions.
...
I wasn't about to discuss tastes, I just want to solve the problem that increases the THD.
Anyone?
Ideas, please.
@Signal lost 
Thank you very much for your reply. I agree with you.
However, I know that this board is capable attaining very low THD - just observe results published by @xrk971 and @lykkedk.
I have decided to build this amplifier for two good reasons: THD and more importantly great community that has gathered here in this thread. I Highly appreciate very friendly and cooperative athmosphere in this thread.
Now, I have a problem which I obviously can not solve without some advice from these who have built Xmass Amp.
I would like to have at least a hint where to look.
Thanks
Thank you very much for your reply. I agree with you.
However, I know that this board is capable attaining very low THD - just observe results published by @xrk971 and @lykkedk.
I have decided to build this amplifier for two good reasons: THD and more importantly great community that has gathered here in this thread. I Highly appreciate very friendly and cooperative athmosphere in this thread.
Now, I have a problem which I obviously can not solve without some advice from these who have built Xmass Amp.
I would like to have at least a hint where to look.
Thanks
Well...
If you look at this thread here https://www.diyaudio.com/community/threads/behringer-umc-202hd-for-measurements.341309/post-7034259
I did some measurement's with and without snubber's on the transformer. - The THD numbers was better when transformer was snubbed.
I also used the schematic / layout from post #186 https://www.diyaudio.com/community/threads/xmas-amp-dibyas-tda7293-by-jhofland.365002/post-6488134
But I don't think the differences are very important measurement-wise!
Are you using the same transformer for all the test's ?
(Oh... I did not forgot to measure the ripple with my scope btw..
I will post here when I try it out)Jesper.
Thanks Jasper,
I will study the links that you've posted. There must be some change since the 1st version. Your fine amplifier and your results are a great motivation to persist.
Yes, I use the same power supply to make sure that meassurements are done on the consistent basis.
I have several amplifiers and I plan to use several external power supplies connected to amplifiers via XLR connections so I will be able to run all my amplifiers as dual mono blocks. I have several ideas, including the use of snubbers. Power supply is a heart of any amplifier, very important part that plays a prominent role. Amplifier just modulates power from the power supply.
Thanks again dear Jasper for encouraging me to persist with this project.
I wish you a pleasant evening.
I will study the links that you've posted. There must be some change since the 1st version. Your fine amplifier and your results are a great motivation to persist.
Yes, I use the same power supply to make sure that meassurements are done on the consistent basis.
I have several amplifiers and I plan to use several external power supplies connected to amplifiers via XLR connections so I will be able to run all my amplifiers as dual mono blocks. I have several ideas, including the use of snubbers. Power supply is a heart of any amplifier, very important part that plays a prominent role. Amplifier just modulates power from the power supply.
Thanks again dear Jasper for encouraging me to persist with this project.
I wish you a pleasant evening.
There should not be a difference between the 1st and later versions with regards to THD. Changes were made to size of input capacitor I think. Otherwise same.
Here is what I got for 10Vrms into 10ohms. You are running 5.5ohms here but I don’t think it should change that much. Sometimes, THD in 0.025% range may have to do with your dummy load resistor. Are you using a wirewound or cement filled load? They tend to be high distortion. The best are using a high power metal thin film series parallel array or an EBG thin planar non inductive resistor (like UHP-300).
Good luck finding the source of the distortion. OTOH, it is dominant 2nd order and monotonically descending higher orders. Whereas the other one is dominant odd orders and 5th is higher than 3rd. This might sound metallic or fatiguing over long listening sessions.
Let me continue your line of reasoning:
If the crossover and speaker distortion dominates, then, everything we do here is pointless. Then, what is the point of Xmass Amp too if its parameters do not matter under the weight of speaker distortion?
Fortunately, all little details make difference however small, the difference will exist. Sum all these small differences and that will make a notable difference.
Now, let us make a simple experiment: find some old recording made on some crappy tape recorder from seventies. You will be able to distinguish voices of people you know. It is the capability of human mind to distinguish small differences in sound. Quite amazing.
Therefore it makes sense to be "obsessed".
Please note: No offence meant, just different point-of-view.
I do appreciate your efforts to create something new and I am sure I will be able to extract the maximum performance from your design. "Obsessed" people at least do their the best to attain the objective. They do not surrender so easily.
I spent some time with the amp last night and this morning while working in the wood shop. It was enjoyable to listen to. I seem to like listening to the system where it resides. One of the speakers (DIY) wobbles a little on the concrete floor, so I had to fold over a piece of cardboard to stuff under one corner. Should I let that bother me? It doesn't. But please chase your own idea of excellence as that is the foundation of all we do.
I would like to express my deepest gratitude to Dibya, Jhofland and X for designing this amp and moreover - providing it as an open source project to the community!
Thanks to all the rest, who provided ideas for improving the design in sonic and stability aspects.
Thanks for backing up your design choices with measurements and leaving the snake oil behind!
I can, without a second thought, say that this is the most informative thread regarding the TDA7293. I learned a lot about the considerations behind selecting and properly sizing the input and FB caps. Learned much more! I'm happy that I went through 48 pages discussion.
I will use the provided files to simplify the design by eliminating the servo. With all my respects to Dibya and Jhofland, I would like to take advantage of their improvements over the datasheet circuit, while still maintaining the circuit easy to build, cost-effective, relatively small footprint and of course maintain the sonic quality. I need an "universal", easy to replicate amp to be implemented in various small projects (mostly for my non-golden-eared friends), which is capable of working on wide range of voltages, depending on the output power requirements.
What do you guys think about modular operation? In theory each chip should see twice less load impedance, further reducing the THD. Less thermal dissipation per chip as well. The supply voltage can be a bit higher when using 4 ohms load while still being in the SOA. What are the drawbacks of modular implementation besides the increased price and larger PCB?
What do you think is better - the local electrolytic caps on the board to be individual pair for each chip or just double size pair of capacitance, shared between the two chips? The bootstrap cap should be enlarged as well, according to the datasheet. Anything else? Overall, do you think it worths adding a second TDA?
Would be happy to hear your opinion.
Kind regards!
Thanks to all the rest, who provided ideas for improving the design in sonic and stability aspects.
Thanks for backing up your design choices with measurements and leaving the snake oil behind!
I can, without a second thought, say that this is the most informative thread regarding the TDA7293. I learned a lot about the considerations behind selecting and properly sizing the input and FB caps. Learned much more! I'm happy that I went through 48 pages discussion.
I will use the provided files to simplify the design by eliminating the servo. With all my respects to Dibya and Jhofland, I would like to take advantage of their improvements over the datasheet circuit, while still maintaining the circuit easy to build, cost-effective, relatively small footprint and of course maintain the sonic quality. I need an "universal", easy to replicate amp to be implemented in various small projects (mostly for my non-golden-eared friends), which is capable of working on wide range of voltages, depending on the output power requirements.
What do you guys think about modular operation? In theory each chip should see twice less load impedance, further reducing the THD. Less thermal dissipation per chip as well. The supply voltage can be a bit higher when using 4 ohms load while still being in the SOA. What are the drawbacks of modular implementation besides the increased price and larger PCB?
What do you think is better - the local electrolytic caps on the board to be individual pair for each chip or just double size pair of capacitance, shared between the two chips? The bootstrap cap should be enlarged as well, according to the datasheet. Anything else? Overall, do you think it worths adding a second TDA?
Would be happy to hear your opinion.
Kind regards!
Last edited:
Glad you found this thread useful StambaMamba.
I think Dibya has already made a parallel version of the TDA7293 on his own before this thread. As shown in a picture on Post 1:
https://www.diyaudio.com/community/attachments/dibya-8ch-tda7293-jpg.904733/
Maybe he can chime in on how it sounds. I think it just needs local PSU rail caps and output load balancing power resistors - maybe 0.22ohm 3W would be sufficient.
I think Dibya has already made a parallel version of the TDA7293 on his own before this thread. As shown in a picture on Post 1:
https://www.diyaudio.com/community/attachments/dibya-8ch-tda7293-jpg.904733/
Maybe he can chime in on how it sounds. I think it just needs local PSU rail caps and output load balancing power resistors - maybe 0.22ohm 3W would be sufficient.
Hi X, U shared LM3886 board i made to compare with TDA7293 , i will share TDA7293 one later when i am return home in Summer [ I am moved to another City for my Uni]
I will suggest discreate amp for higher power requirement , Slew Rate of TDA7293 is not enough under parallel mode for almost double the power.
I will suggest discreate amp for higher power requirement , Slew Rate of TDA7293 is not enough under parallel mode for almost double the power.
Thanks for chimin' in, X and Dibya. Indeed halving the already low slew rate of the 7293 in parallel mode is a downside. Probably I should't expect a single 15 pin chip to be the panacea I wish...it has its restrictions after all...
Anyway I'm curious to see the Dibya's tda7293 parallel implementation in the summer...whenever it is... hehe (currently winter in Europe).
Meanwhile I'll try to contribute with a decent pcb design of a single-chip board with all your changes, but without the dc servo.
Anyway I'm curious to see the Dibya's tda7293 parallel implementation in the summer...whenever it is... hehe (currently winter in Europe).
Meanwhile I'll try to contribute with a decent pcb design of a single-chip board with all your changes, but without the dc servo.
While trying to lower distortion on one of my TDA builds I read this post for ideas.
Have a doubt. He has some caps between the positive and ground, and negative and ground. On opamps we call these power decoupling caps. On opamps they recommend using 100nf. Why have you used 1Uf here.
My original build I had 100nf between these pins under the TDA.
My question is why 1 Uf. And why not 10 Uf ?. I have some tiny 50v caps which are all the same size from 100nf to 1 uf to 10 uf. Before I start swapping these out to see how it impacts the circuit. If somebody could share the logic.
Have a doubt. He has some caps between the positive and ground, and negative and ground. On opamps we call these power decoupling caps. On opamps they recommend using 100nf. Why have you used 1Uf here.
My original build I had 100nf between these pins under the TDA.
My question is why 1 Uf. And why not 10 Uf ?. I have some tiny 50v caps which are all the same size from 100nf to 1 uf to 10 uf. Before I start swapping these out to see how it impacts the circuit. If somebody could share the logic.
Newer TDA7293 with higher slew rate will not be happy with 100nf , want it to oscillate? Go for 100nf
That's exactly the problem I also want to solve:
How to prevent the TDA7293 overshoot with ringing?
This is the problem I experience with three different circuits.