Hello,I'm a novice of class D amplifier.
I have been DIYing a class D amp board based on TDA7498.
From the datasheet,I learned that it outputs 100W power on the condition that
6 ohm,THD+N 10%,36V DC power supply.I'm wondering more these conditions.
First, is there any less complicated methods to discern the THD+N 10% situation, just from the oscilloscope screen?
Second, is there any standard about test condition of amp's nominal output power? I mean the amp board.As we know, audio amp can't always output the nominal max power,or it will be damaged by overheat.
I'm looking forward to the skilled practitioner's help.
Thanks🙂
I have been DIYing a class D amp board based on TDA7498.
From the datasheet,I learned that it outputs 100W power on the condition that
6 ohm,THD+N 10%,36V DC power supply.I'm wondering more these conditions.
First, is there any less complicated methods to discern the THD+N 10% situation, just from the oscilloscope screen?
Second, is there any standard about test condition of amp's nominal output power? I mean the amp board.As we know, audio amp can't always output the nominal max power,or it will be damaged by overheat.
I'm looking forward to the skilled practitioner's help.
Thanks🙂
What do you want to learn about your amp?
You can view the distortion waveform by using the scope in "ADD" and invert one channel. One channel is the input sinewave and the other channel is the output waveform.
You adjust the gain of the vertical amps until you get the "best null" and that corresponds more or less to the waveform of the distortion. If it is 10% of the full output,that is quite a bit of signal...
There are a variety of tests that can be used to determine how well an amp is working, including looking at the current & voltage vs. load, and one could include a reactive load if you are fearless... but you might not like what you see...
I'd not worry about it, unless there is a reason to verify the performance. Just crank it up to full power into a load (6 or 8 ohm) and see that it makes the requisite waveform up to clipping.
_-_-bear
You can view the distortion waveform by using the scope in "ADD" and invert one channel. One channel is the input sinewave and the other channel is the output waveform.
You adjust the gain of the vertical amps until you get the "best null" and that corresponds more or less to the waveform of the distortion. If it is 10% of the full output,that is quite a bit of signal...
There are a variety of tests that can be used to determine how well an amp is working, including looking at the current & voltage vs. load, and one could include a reactive load if you are fearless... but you might not like what you see...
I'd not worry about it, unless there is a reason to verify the performance. Just crank it up to full power into a load (6 or 8 ohm) and see that it makes the requisite waveform up to clipping.
_-_-bear
Thanks for your help.
In fact, I'm facing up with a problem about output power now.
The current test phenomenon is that individual channel is capable of outputing 83W RMS or so.I use a 40V DC SWPS and 2 7.5 ohm resistant loads so that 83W is derived by 25*25/7.5.
While the max out power of each channel reduced down to 38W RMS when both of channels worked.
When approaching that boundary state, the coil sounds continuously and the output waveform turns back to a straight line intermittently.
The deficiency of power supply loop layout has been ruled out by adding additional power wire experiment.
And I've choose the coil with 6A saturation current.
Have you meet with the similar problem before?
Hoping for some suggestions sincerely .
In fact, I'm facing up with a problem about output power now.
The current test phenomenon is that individual channel is capable of outputing 83W RMS or so.I use a 40V DC SWPS and 2 7.5 ohm resistant loads so that 83W is derived by 25*25/7.5.
While the max out power of each channel reduced down to 38W RMS when both of channels worked.
When approaching that boundary state, the coil sounds continuously and the output waveform turns back to a straight line intermittently.
The deficiency of power supply loop layout has been ruled out by adding additional power wire experiment.
And I've choose the coil with 6A saturation current.
Have you meet with the similar problem before?
Hoping for some suggestions sincerely .
Distortion rises very quickly when an amplifier clips, so the difference in power between 0.1% THD and 10% THD is not substantial in any good amplifier, maybe just 1dB or 2dB. Also, THD involves testing with continuous tones, which usually results in more power supply sag than when playing music.
However, a bad class D amplifier may as well exhibit 1% THD or more well below clipping. Giving the power output at 10% THD hides this phenomena.
However, a bad class D amplifier may as well exhibit 1% THD or more well below clipping. Giving the power output at 10% THD hides this phenomena.
Are you sure your SMPS can deliver enough power to feed two of your amps continuously ?
Regards
Charles
Regards
Charles
Yes.The rating output power of the SWPS is 600W.
I'm doubting one of the self-protection functions has always been activated.
I'll collect more experimental data including pins' voltage and output waveform.
I guess there may be solutions inside many clues above.
I'm doubting one of the self-protection functions has always been activated.
I'll collect more experimental data including pins' voltage and output waveform.
I guess there may be solutions inside many clues above.
Hello,it's been several weeks before I posted the threads lastly.
And my TDA7498 amp board still can't reach its max power as the datasheet described.
It always was activated by its self-protection since the diagnostic pin output the positive pluses.
I've fed the board with 40V and mounted a heatsink with a fan on the slug of the chip.
So it is not the UVP (under voltage protection) and OTP(Over temperature protection) that lead to the intermittent output power.
Besides that,OVP(Over Voltage Protection) has also been excluded because the waveform from the PVCC pins haven't reach the absolute maximum ratings 44V.
At last,it seems that some unaccountable reason aroused the OCP(Over Current Protection)
The typical Iovp from the datasheet is 6A.
While the real current never reached more than 5A(20V/6.7ohm=2.9A RMS).
In fact,I am doubting there's something unknown that limited the maximum output current.
Another potential reason is the saturation current of the coil.I've already replace it with
an alternative one whose Isat is nearly 9A.
Looking forward to help.
And my TDA7498 amp board still can't reach its max power as the datasheet described.
It always was activated by its self-protection since the diagnostic pin output the positive pluses.
I've fed the board with 40V and mounted a heatsink with a fan on the slug of the chip.
So it is not the UVP (under voltage protection) and OTP(Over temperature protection) that lead to the intermittent output power.
Besides that,OVP(Over Voltage Protection) has also been excluded because the waveform from the PVCC pins haven't reach the absolute maximum ratings 44V.
At last,it seems that some unaccountable reason aroused the OCP(Over Current Protection)
The typical Iovp from the datasheet is 6A.
While the real current never reached more than 5A(20V/6.7ohm=2.9A RMS).
In fact,I am doubting there's something unknown that limited the maximum output current.
Another potential reason is the saturation current of the coil.I've already replace it with
an alternative one whose Isat is nearly 9A.
Looking forward to help.
How are you measuring the 44v? Or as close as you come?
You need a scope - since that pin likely looks at *peak* not RMS or average level!
😀
_-_-bear
You need a scope - since that pin likely looks at *peak* not RMS or average level!
😀
_-_-bear
Very Simple Way to measure amp power
We usually meausure rated power for one amp with sinewave signal(1Khz for full range amp),sustained signal added until over tempreture protection or over 5 minutes.
We usualy make amp burn-in one by one channel , full power 20minutes or half power 24 hours.
IEEE suggest EIAJ standard.we call it burst mode.
We usually meausure rated power for one amp with sinewave signal(1Khz for full range amp),sustained signal added until over tempreture protection or over 5 minutes.
We usualy make amp burn-in one by one channel , full power 20minutes or half power 24 hours.
IEEE suggest EIAJ standard.we call it burst mode.
Hi all, I'm a relative beginner in this kind of amplifiers...
Regarding of the first post in this thread:
I also bought an amplifier based on tda7498 from sure electronics (AA-AB32189). It is meant to be installed in portable music box powered with li-po batteries, 10 cells in series which is around 40V, 4400mAh capacity and this battery is capable to deliver 50 amps of current at least.
Now, there is one thing that bothers me badly:
When measuring output power to 8ohm speakers with sinewaves, I get numbers which are about the same as specifications claims, and thats good! But, when playing music, output signal get some kind of protection cuts way before amplifier output power capabilities. To be exact, those cuts in output are about 0.1s long and they are occuring when overall louder note is reproducing, like drum snare. The more I amplify, the more often it cuts the signal. There are no THD when this happens.
I tried different preamps (like mp3 players and laptops), lower voltage power supplies, different settings on input gain switch, but I always get this problem.
To be odd enough, I have two 2x25w amplifiers also from Sure and they are doing the same thing. What am I doing wrong?
I can't understand how is possible that this tda7498 amplifier can draw 5A from 40v battery on sinewaves (with no noticeable THD) but can not draw more than 0.5A before signal cuts on music are occuring.
Please help, I am desperate, my "2x100w amplifier" can't give me more than 2x10w on music!
Regarding of the first post in this thread:
I also bought an amplifier based on tda7498 from sure electronics (AA-AB32189). It is meant to be installed in portable music box powered with li-po batteries, 10 cells in series which is around 40V, 4400mAh capacity and this battery is capable to deliver 50 amps of current at least.
Now, there is one thing that bothers me badly:
When measuring output power to 8ohm speakers with sinewaves, I get numbers which are about the same as specifications claims, and thats good! But, when playing music, output signal get some kind of protection cuts way before amplifier output power capabilities. To be exact, those cuts in output are about 0.1s long and they are occuring when overall louder note is reproducing, like drum snare. The more I amplify, the more often it cuts the signal. There are no THD when this happens.
I tried different preamps (like mp3 players and laptops), lower voltage power supplies, different settings on input gain switch, but I always get this problem.
To be odd enough, I have two 2x25w amplifiers also from Sure and they are doing the same thing. What am I doing wrong?
I can't understand how is possible that this tda7498 amplifier can draw 5A from 40v battery on sinewaves (with no noticeable THD) but can not draw more than 0.5A before signal cuts on music are occuring.
Please help, I am desperate, my "2x100w amplifier" can't give me more than 2x10w on music!
Peak Signals
Hi Brighter.
Someone with specific knowledge of your amplifier may be able to confirm if what I'm saying is applicable, but what I suspect is:
You have turned the gain up so music signals are amplified to the point where *average* output level is 0.5A or whatever you are reading i.e. well short of the amplifiers published rating.
However music signal is not constant and that signal will be made up of mostly low level but short high intensity peaks (such as when those snares hit). Typically peak levels can be 10x average levels so at those short term peaks you are probably driving the amplifier above its capacity and a protection circuit is kicking in to mute the amp.
If you want to be able to reproduce those peak signals louder you will essentially have to either get a bigger amplifier or use more efficient speakers. If you want to get louder overall sound levels (not just those instantaneous peaks) you could do either of the above or apply some sort of signal compression to reduce the dynamic range of your music.
As a test I suggest finding some non dynamic music i.e. stuff recorded with lots of compression to maximise overall sound pressure. If I am right you should both be able to play this much louder to the ear and measure a greater actual output before the amp cuts off.
There are other more esoteric possible causes, and if you are running close to rated max power supply voltage and my suggested playing of compressed music source makes it worse rather than better we can suspect bus pumping or no doubt other things those more knowlegable than me will chip in with. But try the simple test first.
Hi Brighter.
Someone with specific knowledge of your amplifier may be able to confirm if what I'm saying is applicable, but what I suspect is:
You have turned the gain up so music signals are amplified to the point where *average* output level is 0.5A or whatever you are reading i.e. well short of the amplifiers published rating.
However music signal is not constant and that signal will be made up of mostly low level but short high intensity peaks (such as when those snares hit). Typically peak levels can be 10x average levels so at those short term peaks you are probably driving the amplifier above its capacity and a protection circuit is kicking in to mute the amp.
If you want to be able to reproduce those peak signals louder you will essentially have to either get a bigger amplifier or use more efficient speakers. If you want to get louder overall sound levels (not just those instantaneous peaks) you could do either of the above or apply some sort of signal compression to reduce the dynamic range of your music.
As a test I suggest finding some non dynamic music i.e. stuff recorded with lots of compression to maximise overall sound pressure. If I am right you should both be able to play this much louder to the ear and measure a greater actual output before the amp cuts off.
There are other more esoteric possible causes, and if you are running close to rated max power supply voltage and my suggested playing of compressed music source makes it worse rather than better we can suspect bus pumping or no doubt other things those more knowlegable than me will chip in with. But try the simple test first.
Last edited:
Thanks for Your answer goosewing!
Yes, there are some songs which I can amplify to max without any cutoffs. And yes, those songs are recorded with low dynamics. And I hate low dynamics...
So, whats the use of an amplifier which can do 100w on sinewave only🙁
The amp can't even get warm when dynamic music is played, why is that fan cooler on top I'm wondering...
However, is there any way to bypass this protection "feature" even if it costs magic smoke to go out?
Yes, there are some songs which I can amplify to max without any cutoffs. And yes, those songs are recorded with low dynamics. And I hate low dynamics...
So, whats the use of an amplifier which can do 100w on sinewave only🙁
The amp can't even get warm when dynamic music is played, why is that fan cooler on top I'm wondering...
However, is there any way to bypass this protection "feature" even if it costs magic smoke to go out?
Last edited:
I performed simple test, in cooledit I made simple 1khz sinewave with -20dB nominal volume, 1 minute long. Every ten seconds I gained the signal in cooledit by +20dB for short period of time. In this test I used only one chanell with 8 ohm speaker on it.
Initial current draw (when -20dB was playing) was 0.4A and voltage across this speaker was 10V, so it is around 12w.
According to my calculations, when first +20dB peak strikes, amp will go crazy, cut, THD, go bye-bye, I don't know what!
So I waited to see...
And, to my surprise, nothing happened! No signal cut. I nearly lost my ears because I'm doing the test with 100db/1w PA midrange!
Than I put some music and turned down volume where average current draw is only 60mA! That is exactly what this amp draws when no input!
And again there are those st*pid cuts again!!!
I don't get it, I do not hear that kind of dynamics in this song! It is driving me insane!!!
Please help me🙁
Initial current draw (when -20dB was playing) was 0.4A and voltage across this speaker was 10V, so it is around 12w.
According to my calculations, when first +20dB peak strikes, amp will go crazy, cut, THD, go bye-bye, I don't know what!
So I waited to see...
And, to my surprise, nothing happened! No signal cut. I nearly lost my ears because I'm doing the test with 100db/1w PA midrange!
Than I put some music and turned down volume where average current draw is only 60mA! That is exactly what this amp draws when no input!
And again there are those st*pid cuts again!!!
I don't get it, I do not hear that kind of dynamics in this song! It is driving me insane!!!
Please help me🙁
I did another test, from one song I isolated a guitar hitch which gets cut when reproduced with moderate amplify rate. One moment later I pasted 1khz sinewave which is at least 15dB louder. Than I multiplicated this sequence several times, saved and copied to my sansa mp3 player. Hooked mp3 to tda7498 and played.
Again the same, quieter guitar hitch (which represents music) gets cut and 15dB louder sinewave don't! Here is the screenshot of the waveform:
What's going on in that amp, for heaven's sake?
Again the same, quieter guitar hitch (which represents music) gets cut and 15dB louder sinewave don't! Here is the screenshot of the waveform:
What's going on in that amp, for heaven's sake?
Last edited:
It was seeming all so consistently explainable until you posted those graphs.
So what is the difference between music and sinewaves, the music is more complex and has harmonics for a start. When there are multiple frequencies playing the voltages of the different frequencies sum when peaks coincide so you get short term much higher peaks compared to a sine wave. You are operating close to max voltage for the amp so it is possible high spikes are an issue.
It is all getting a bit speculative here, but you could try removing a couple of cells from your supply to lower absolute output. This would reduce absolute output powere, but would cause high voltage transients to clip rather than exceed protection levels, and would increase the ratio of achievable average sound level to maximum level i.e. would effectively compress the music and overall make it sound louder on dynamic material *if* protection on high voltage transients is the problem.
Warning - this is about the limit of my knowledge. Time for the experts to step in if this is a dead end.
So what is the difference between music and sinewaves, the music is more complex and has harmonics for a start. When there are multiple frequencies playing the voltages of the different frequencies sum when peaks coincide so you get short term much higher peaks compared to a sine wave. You are operating close to max voltage for the amp so it is possible high spikes are an issue.
It is all getting a bit speculative here, but you could try removing a couple of cells from your supply to lower absolute output. This would reduce absolute output powere, but would cause high voltage transients to clip rather than exceed protection levels, and would increase the ratio of achievable average sound level to maximum level i.e. would effectively compress the music and overall make it sound louder on dynamic material *if* protection on high voltage transients is the problem.
Warning - this is about the limit of my knowledge. Time for the experts to step in if this is a dead end.
Make sure that the amplifier is not oscillating on complex musical signals - I've run into something similar where relatively high frequency program content (10kHz) caused the (marginally stable) amp to oscillate and this triggered the amp protection circuits. No hint on 1kHz sine wave signals at all, but 10kHz sines did cause a problem.
Sometimes these sorts of behaviors are dependent on the load connected to the output of the amplifier as well. (Will or won't do it with resistive or speaker loads respectively.)
Sometimes these sorts of behaviors are dependent on the load connected to the output of the amplifier as well. (Will or won't do it with resistive or speaker loads respectively.)
Last edited:
It appears that the entire music signal has been cut by the same amount (rather then a musical peak being limited) whereas the sine wave is completely unaffected. Why this is I don't know, but I don't think it's the amplifier that would be causing this. The amplifier itself has a fixed gain system and it does not appear to have a dynamic compression system.
If you exceed the voltage rail limits, but the current limits are not exceeded then you will get clipping of the waveform and gross distortion will result, this isn't happening here. If you exceed the current limits the datasheet implies that the amplifier chip will mute instead of clipping the wave forms, again this isn't happening here.
The only possibility I can think of is that the mp3 player has a compression system in place that is designed to protect your ears from music that is too loud. Maybe the sine wave is tricking it in some way, I don't know, but the problem does indeed seem to be quite strange.
As another test I'd recommend you chop up the music signal so that it goes 1 second guitar, 1 second sine wave, 1 second guitar, 1 second sine wave etc, alternating quickly between the two.
I take it the upper music signal is as measured from the output of the amplifier?
If not, what you need to do is record the output of the amplifier and the output of the mp3 player and directly compare them.
If you exceed the voltage rail limits, but the current limits are not exceeded then you will get clipping of the waveform and gross distortion will result, this isn't happening here. If you exceed the current limits the datasheet implies that the amplifier chip will mute instead of clipping the wave forms, again this isn't happening here.
The only possibility I can think of is that the mp3 player has a compression system in place that is designed to protect your ears from music that is too loud. Maybe the sine wave is tricking it in some way, I don't know, but the problem does indeed seem to be quite strange.
As another test I'd recommend you chop up the music signal so that it goes 1 second guitar, 1 second sine wave, 1 second guitar, 1 second sine wave etc, alternating quickly between the two.
I take it the upper music signal is as measured from the output of the amplifier?
If not, what you need to do is record the output of the amplifier and the output of the mp3 player and directly compare them.
I would also guess that it is a combination of the signal spectrum and some unfortunate reaction of the outputfilter/load combination.
More interesting than the signal that is going in is the signal that is coming out of the amp - especially shortly before it cuts out.
Regards
Charles
More interesting than the signal that is going in is the signal that is coming out of the amp - especially shortly before it cuts out.
Regards
Charles
questions
Ah, I just went back and read your original post more carefully, picking up on the 'meant to' be powered by batteries and a second amp having the same problem.
1) If you change source i.e. play music direct from CD player or uncompressed wave file on PC do you get the same problem?
2) What are you actually using for a power source now? Are you able to measure voltage on the input rails while this is going on?
3) Do you get the same behavior with a different pair of speakers or resistive load?
Ah, I just went back and read your original post more carefully, picking up on the 'meant to' be powered by batteries and a second amp having the same problem.
1) If you change source i.e. play music direct from CD player or uncompressed wave file on PC do you get the same problem?
2) What are you actually using for a power source now? Are you able to measure voltage on the input rails while this is going on?
3) Do you get the same behavior with a different pair of speakers or resistive load?
- Status
- Not open for further replies.