brighter!
This is the question!
A badly designed crossover can cause this kind of problem (very low impedance at a frequency). Test it with dummy load, and tell what your speaker system is like! Or measure impedance curve!
First of all I want to thank You guys for helping me!
Now, mp3 player does not cut the signal because I tryed it on my home integrated amplifier. To be sure I even tried different input sources like iphone4, pc sound card, etc. None of them solved cut signal problems.
I even changed power source to 32v dc from HP printer, tried two 12v lead acid batteries in series, and always the same. This guitar hitch gets cut at around 1w output power.
Now, I will make mp3 file with those guitar hitches and upload it on rapidshare so that it can be downloaded and examined, here it is:
I also tryed my other amp (2x25w with tpa3123d2 chip, from sure electronics) and signal cut happens at about the same rate of output power - 1w, no matter what power source I use or what input source I use.
I'm going mad...
Now, mp3 player does not cut the signal because I tryed it on my home integrated amplifier. To be sure I even tried different input sources like iphone4, pc sound card, etc. None of them solved cut signal problems.
I even changed power source to 32v dc from HP printer, tried two 12v lead acid batteries in series, and always the same. This guitar hitch gets cut at around 1w output power.
Now, I will make mp3 file with those guitar hitches and upload it on rapidshare so that it can be downloaded and examined, here it is:
PHP:
https://rapidshare.com/files/3893113934/guitar_hitch_test.mp3I also tryed my other amp (2x25w with tpa3123d2 chip, from sure electronics) and signal cut happens at about the same rate of output power - 1w, no matter what power source I use or what input source I use.
I'm going mad...
There are many standard test procedures.
The first important component of any test procedure is correct mains supply voltage to match the rated voltage requirement in the specification.
The first important component of any test procedure is correct mains supply voltage to match the rated voltage requirement in the specification.
@pafi
all of my experiments was done with 8ohm midrange without any passive crossover. To be sure that possible low impedance curve of this particular speaker is not affecting the cuts I tried different 8ohm speaker. Result is apsolute the same! When I amplify to around 1w output power, signal gets cut.
I even tried one 4ohm speaker and again I get the same experience, only this time those cuts happen at around 2w. So I guess it is somewhat current related.
all of my experiments was done with 8ohm midrange without any passive crossover. To be sure that possible low impedance curve of this particular speaker is not affecting the cuts I tried different 8ohm speaker. Result is apsolute the same! When I amplify to around 1w output power, signal gets cut.
I even tried one 4ohm speaker and again I get the same experience, only this time those cuts happen at around 2w. So I guess it is somewhat current related.
Too high impedance can be a problem also, because output filter can produce resonance at very high freq.
(You may try to find the problematic frequency.)
(You may try to find the problematic frequency.)
Yeah! I found it!🙂
I made a sinewave sweep from 8khz to 12khz and played it. Around 9-9.5khz amp is starting to cut signal like crazy!
No matter what kind of speaker I connect (4ohm, 6ohm or 8ohm) it cuts the signal around 9-9.5khz. Although I don't know what the impedance characteristics of those speakers are, I can for sure assume that they are'nt the same.
To be 100% sure, I connected dummy 4ohm load (resistor 10w) and measured voltage across resistor. And yes, there are cuts also at 9-9.5khz!
So, member kevinkr was right about oscilating!
What now? How to avoid that?
I made a sinewave sweep from 8khz to 12khz and played it. Around 9-9.5khz amp is starting to cut signal like crazy!
No matter what kind of speaker I connect (4ohm, 6ohm or 8ohm) it cuts the signal around 9-9.5khz. Although I don't know what the impedance characteristics of those speakers are, I can for sure assume that they are'nt the same.
To be 100% sure, I connected dummy 4ohm load (resistor 10w) and measured voltage across resistor. And yes, there are cuts also at 9-9.5khz!
So, member kevinkr was right about oscilating!
What now? How to avoid that?
.... until the amplifier has been replaced by manufacturer, if that option is open to you.
As a sanity check you did say you had similar problem with a different design Sure amplifier as well? Does that one also respond exactly the same to the sine wave sweep?
Measure the frequency response of the amplifier using something like ARTA with your 4 ohm dummy load.
First measure it with a swept sine stimulus and see what happens. Do this at a number of different output powers, say 1/10th, 1/4, 1/2, 1, 2, 5 and 10 watts.
Next try doing the same thing but using an MLS stimulus.
The swept sine has the advantage that it only sends 1 frequency through the amplifier at a time. The MLS signal does the opposite.
Also measure the frequency response, over a range of output voltages and without any load attached to the amplifier. I am aware that the output filter of the class D amplifier wont work precisely as intended under this situation, rather I want to see if the limit is coming in with relation to the current the amplifier is having to provide, or if its related simply to the voltage level. Alternatively you could do this with an 8 ohm load.
If the tpa3123d2 does exactly the same thing as the TDS7498 then it implies that it's not the amplifier that is at fault.
First measure it with a swept sine stimulus and see what happens. Do this at a number of different output powers, say 1/10th, 1/4, 1/2, 1, 2, 5 and 10 watts.
Next try doing the same thing but using an MLS stimulus.
The swept sine has the advantage that it only sends 1 frequency through the amplifier at a time. The MLS signal does the opposite.
Also measure the frequency response, over a range of output voltages and without any load attached to the amplifier. I am aware that the output filter of the class D amplifier wont work precisely as intended under this situation, rather I want to see if the limit is coming in with relation to the current the amplifier is having to provide, or if its related simply to the voltage level. Alternatively you could do this with an 8 ohm load.
If the tpa3123d2 does exactly the same thing as the TDS7498 then it implies that it's not the amplifier that is at fault.
And could you also watch what happens with the supply voltage (measured directly at the board - not at the supply) when the amp behaves like that ?
Regards
Charles
Regards
Charles
I must apollogize because I have slightly misslead You guys. When I discovered signal cuts related to 9khz (and no matter what speaker or dummy load attached), that was when I tested the tpa3123d2 amp. I have to further test its ARTA behaviour and everything else, voltage and current wise.
I'm now back to the tda7498. Here is no signal cuts that is oscilation/frequency related but rather because my of PA speakers. This amp is recomended to drive 8ohm only and I hooked that PA midrange that I have to it. Prior to making conclusions here, I didn't perform impedance testing on that speaker because I thought that I did it in the past and that everything was within normal. But now I see that it wasn't. This speaker has incredible impedance drop around 17khz and by that frequency impedance is less than 1ohm! Never saw something like that before😱 No wonder that amp is going crazy!
So, before I decided to make a passive xo for that particular speaker I recorded a song with FFT filter that kills all around above mentioned 17khz. Cuts went away!
It is strange that those "cut signal behaviours" are manifesting equally on both amps but with different cause! And all of the time I thought it was the same cause
Sorry guys for making confusion here, I was scattered all around with a bunch of problems that I have to deal with and above all my lack of concentration and knowledge...
Now the open question remains tpa3123d2 and why is that oscilating there.
I now have a whole lot of informations that will help me in my further testing and all that thanks to You guys with Your merciful willingness to help a newcomer! Thanks a bunch!
I'm now back to the tda7498. Here is no signal cuts that is oscilation/frequency related but rather because my of PA speakers. This amp is recomended to drive 8ohm only and I hooked that PA midrange that I have to it. Prior to making conclusions here, I didn't perform impedance testing on that speaker because I thought that I did it in the past and that everything was within normal. But now I see that it wasn't. This speaker has incredible impedance drop around 17khz and by that frequency impedance is less than 1ohm! Never saw something like that before😱 No wonder that amp is going crazy!
So, before I decided to make a passive xo for that particular speaker I recorded a song with FFT filter that kills all around above mentioned 17khz. Cuts went away!
It is strange that those "cut signal behaviours" are manifesting equally on both amps but with different cause! And all of the time I thought it was the same cause
Sorry guys for making confusion here, I was scattered all around with a bunch of problems that I have to deal with and above all my lack of concentration and knowledge...
Now the open question remains tpa3123d2 and why is that oscilating there.
I now have a whole lot of informations that will help me in my further testing and all that thanks to You guys with Your merciful willingness to help a newcomer! Thanks a bunch!
- Status
- Not open for further replies.