So I spent a few weeks trying to see if anybody has encountered opamp over shoot and ringing on the TDA72XX series.
The design is as per the data sheet. Overshoot and ringing was observed on various TDA PCBs.
So after weeks of searching I could not find a single bit of info that somebody has been able to achieve this.
I did find a few posts on here where others said they are facing the same problem. These guys have purchased all their parts from reputed companies so that eliminates my doubt about my TDAs not being original.
I have tested all my TDAs. To see if the power output is at least ok and meets the specs of the orignal IC and they do.
If anybody has been able to build a TDAXX and get a clean signal without over shoot and ringing.
It would be nice to see a pic of the scope screen as well as how you achieved this.
Before I throw in the towel on these efforts and move to building and testing PCBs for the LM3886 I though Id make this one last ditch attempt to figure this out. Any inputs appreciated.
My test setup. TDA7294. Square wave pulse sent into the input pin from my phone or function generator or pc I tested all 3. Scope sitting on the output pin. (Tested both with and without speaker load of 8 Ohm speaker.
The design is as per the data sheet. Overshoot and ringing was observed on various TDA PCBs.
So after weeks of searching I could not find a single bit of info that somebody has been able to achieve this.
I did find a few posts on here where others said they are facing the same problem. These guys have purchased all their parts from reputed companies so that eliminates my doubt about my TDAs not being original.
I have tested all my TDAs. To see if the power output is at least ok and meets the specs of the orignal IC and they do.
If anybody has been able to build a TDAXX and get a clean signal without over shoot and ringing.
It would be nice to see a pic of the scope screen as well as how you achieved this.
Before I throw in the towel on these efforts and move to building and testing PCBs for the LM3886 I though Id make this one last ditch attempt to figure this out. Any inputs appreciated.
My test setup. TDA7294. Square wave pulse sent into the input pin from my phone or function generator or pc I tested all 3. Scope sitting on the output pin. (Tested both with and without speaker load of 8 Ohm speaker.
Square waves contain many harmonics.
What is the frequency of the overshoot? If above 25kHZ, use an inductive filter to remove it as it is not audible, just a waste of energy. but ... having said that, I would expect overshoot with a square wave.
What is it like with a sine wave at say 15kHZ? If perfect, ignore the square wave data you have as it is meaningless.
What is the frequency of the overshoot? If above 25kHZ, use an inductive filter to remove it as it is not audible, just a waste of energy. but ... having said that, I would expect overshoot with a square wave.
What is it like with a sine wave at say 15kHZ? If perfect, ignore the square wave data you have as it is meaningless.
Im not sure how to calculate the frequency of the over shoot. I am trying various filters and cap values between the pins of the supply line. To see what gives me the clearest signal. The square wave comes thru but at the peak it rings a bit and then goes flat. It just bugs my OCD more than it bugs the quality of the music.
Sine wave is absolutely clean. At any freq I test from 20Hz to 20Khz. And its flat. There is no hum and music sounds pretty decent if a bit tinish. I am thinking of doing the recommended tweaks like upping the input cap to 2 Uf and adding 2 UF between positive and negative rails and 1 uf between +/- rails and ground to see if it helps. Also upping the boostrap cap from 22uf to 47uf.
Clean sine waves something I have not been able to achieve with any of my tone control boards or Graphic Equalizer boards.
Sine wave is absolutely clean. At any freq I test from 20Hz to 20Khz. And its flat. There is no hum and music sounds pretty decent if a bit tinish. I am thinking of doing the recommended tweaks like upping the input cap to 2 Uf and adding 2 UF between positive and negative rails and 1 uf between +/- rails and ground to see if it helps. Also upping the boostrap cap from 22uf to 47uf.
Clean sine waves something I have not been able to achieve with any of my tone control boards or Graphic Equalizer boards.
There is no need to calculate - just use your oscilloscope and zoom in to see the frequency of the ringing region. In my case overshot is sharp but the ringing region is very short, with just few fading cycles. I have observed this on three different boards.
This is from two TDA7293 in Modular configuration, ie one as Master, the other as Slave.
Display is 20kHz Sq wave into 8 ohms at 10p-p.
Display is 20kHz Sq wave into 8 ohms at 10p-p.
Thanks @Michael Chua
Could you please publish schematics of that amplifier with values of components. That would help us to identify the cause for ringing.
Could you please publish schematics of that amplifier with values of components. That would help us to identify the cause for ringing.
@Michael Chua
Thank you very much for your useful reply.
I have gone directly to my bench and your solution worked. I didn't have 20p capacitor on stock but I had 30p - overshoot and ringing are gone.
Now I will completely revamp my boards to have all values according to your schematics.
I do indeed apreciate your preparedness to help. I have almost thrown three pairs of board intended to my new three-way design with active crossover.
@chinoy
I do also appreciate your initiative to solve this problem.
In primary school and later as an undergraduate student I loved mathematics very much and solving mathematical problems. As a graduate student I've discussed mathematics with my professor and he told me: "The problem is to find the problem. Once you identify it properly, the solution is there!"
Indeed: The problem is to find the problem.
Thanks guys! I've almost thrown away in the bin boards that are otherwise just fine.
Thank you very much for your useful reply.
I have gone directly to my bench and your solution worked. I didn't have 20p capacitor on stock but I had 30p - overshoot and ringing are gone.
Now I will completely revamp my boards to have all values according to your schematics.
I do indeed apreciate your preparedness to help. I have almost thrown three pairs of board intended to my new three-way design with active crossover.
@chinoy
I do also appreciate your initiative to solve this problem.
In primary school and later as an undergraduate student I loved mathematics very much and solving mathematical problems. As a graduate student I've discussed mathematics with my professor and he told me: "The problem is to find the problem. Once you identify it properly, the solution is there!"
Indeed: The problem is to find the problem.
Thanks guys! I've almost thrown away in the bin boards that are otherwise just fine.
Dito to what I am seeing. Sharp over shoot followed by just a few fading cycles. I have observed this on 3 different boards and two different chips one is marked Made in Singapore and the other is just regular marking both ring the same.
I too feel adding a 10-30 Pf cap may solve the problem. I have one tone control board based on the rockola tone control that uses a 10 pf cap and it is the only board with zero ringing. (Its not a TDA board its a tone control board). But I look at the TDA IC as just one large opamp.
On the input side at the audio jack earth. On the output side the Neg terminal of the speaker. If no speaker then the earth of the speaker out. Never to the earth of the PSU.
Nothing wrong with it. Just never struck me to use it.
Maybe Ill give it a shot. In my mind I want to see the exact same signal my speaker is seeing.
Square wave testing is a convenient way to check stability. If the amplifier rings a lot, its poles are poorly damped, so it must be at the edge of instability. Whether the ringing is audible is irrelevant in this case.
Indeed,
but there is even more: from square wave response we may obtain valuable information on frequency response of the device under test which is even more important information.
In my view, optimization of power amplifier is a trade-off between several objectives including the most significant: THD and frequency response. Neglecting one or another leads just to the extremes of a single parameter. Quite notable are extremely low THD amplifiers where frequency response is traded-off for lower THD values. But at one point harmonics become inaudible whilst the whole design goal was sacrificed for having insignificantly low harmonic distortion. Interestingly, for my taste, the best sound comes from an amplifier with perfect square wave response at even 5Hz and lower. But, of course, distortion must be held as low as possible, but not at any cost.