Cool I will go visit your post and continue the discussion there. Rather than spoil this thread.
I will try adding all possible variations of caps to see if anything helps.
The unit sounds fine. But my OCD will not let me enjoy it till I see a clear signal on the scope lol.
No hum. No noise. But scope says No lol. I had one friend who told me to stop sending square waves to the amp. As that is like sending it a d/c signal and it can blow the chip. But I only see the ringing when sending square pulses.
Your friend is wrong. Investigating square wave response helps understand many important aspects of amplifier performance: frequency response, stability issues, phase shifts, treble boost, low frequency attenuation etc. I start listening if and only if the amplifier meassures good. Listening isn't adequate substitute for instrumental analysis.
I have left four surplus pairs of Xmas Amp to give away but I will donate them just after I make sure some issues can be rectified.
Otherwise it is just waste of time.
I agree with you. I guess the friend just didnt want me blowing up more chips.
But I have happy to hear that even chips purchased form 100% legit places are having these problems.
I too have a plethora of PCBs for the TDA. So I am hopeful at least one of them will work as intended.
Ps: Are we testing the right way ?.
i.e. how do you scope the output ?. Initially I was just scoping the output from the Chip Amp. But realised this may not be the right way to do it. So now what I do is hook up a speaker to the output and then scope the results.
Its a cheap 8 ohm speaker so Im ok even if it pops. But I feel it makes more sense to test with the speaker in circuit.
Could somebody verify what is the right way to test. ? As of right now I test with speaker without speaker and I chk the FFT analysis. i.e. join both the earths of the probe and put one probe to speaker + and one to speaker ground. Turn on math function FFT.
But my first objective is to see what frequencies give me a clean square pulse.
My basic circuit is as per the data sheet nothing extra nothing modified. That is my base. Then I start trying different things to see if it helps or makes the problem worse. My transformer is kept 6 ft away from the amp.
I was able to fix it to a point where even at full volume I have zero hum. With no input and with my phone providing input. So some progress at least. Now only the over shoot and ringing needs to be sorted.
When I test with Sine waves. When I zoom into the amplified sine wave output I see a lot of tiny little steps making up the sine wave output. Is this normal or does it indicate a problem. ?.
i.e. how do you scope the output ?. Initially I was just scoping the output from the Chip Amp. But realised this may not be the right way to do it. So now what I do is hook up a speaker to the output and then scope the results.
Its a cheap 8 ohm speaker so Im ok even if it pops. But I feel it makes more sense to test with the speaker in circuit.
Could somebody verify what is the right way to test. ? As of right now I test with speaker without speaker and I chk the FFT analysis. i.e. join both the earths of the probe and put one probe to speaker + and one to speaker ground. Turn on math function FFT.
But my first objective is to see what frequencies give me a clean square pulse.
My basic circuit is as per the data sheet nothing extra nothing modified. That is my base. Then I start trying different things to see if it helps or makes the problem worse. My transformer is kept 6 ft away from the amp.
I was able to fix it to a point where even at full volume I have zero hum. With no input and with my phone providing input. So some progress at least. Now only the over shoot and ringing needs to be sorted.
When I test with Sine waves. When I zoom into the amplified sine wave output I see a lot of tiny little steps making up the sine wave output. Is this normal or does it indicate a problem. ?.
Just a little build update...
Slooooovly
I'am trying to follow some "correct" grounding schem. (Seen in this thread by @xrk971 at one of the first pages)
I connected the 2 GND points on the PCB corner's, and the heatsinks --- >Star-GND. Star-GND is connected through CL60||Wima 22nF to PSU-GND
I catched some nice ceramic 3mm. isolator's for the TDA, right now testing it.
The 3. picture shows the connection from PSU-GND through CL60 NTC || Wima 22nF
Jesper. <----> Listning to MoonSafari right now
Slooooovly
I'am trying to follow some "correct" grounding schem. (Seen in this thread by @xrk971 at one of the first pages)
I connected the 2 GND points on the PCB corner's, and the heatsinks --- >Star-GND. Star-GND is connected through CL60||Wima 22nF to PSU-GND
I catched some nice ceramic 3mm. isolator's for the TDA, right now testing it.
The 3. picture shows the connection from PSU-GND through CL60 NTC || Wima 22nF
Jesper. <----> Listning to MoonSafari right now
While assambly it all together, I do test the setup with my THD rig.
I'am struggling some gnd-loop's problems.
I have tried to follow the attached schematic, and as i showed in my previous post I have attached an CL60+mkt at the PSU-gnd connection to the chassis.
Quistion is... Do I best attach the speaker cables from the plugs on the PCB directly to bindingpost's
Also do I do the same with the input plug's directly to the RCA's.
E.G. not involving any "star"-gnd to those connection's ?
Any tips? -Hint's...
Will continue the hunt later.
Jesper.
I'am struggling some gnd-loop's problems.
I have tried to follow the attached schematic, and as i showed in my previous post I have attached an CL60+mkt at the PSU-gnd connection to the chassis.
Quistion is... Do I best attach the speaker cables from the plugs on the PCB directly to bindingpost's
Also do I do the same with the input plug's directly to the RCA's.
E.G. not involving any "star"-gnd to those connection's ?
Any tips? -Hint's...
Will continue the hunt later.
Jesper.
So i did a lot of stuff, ending up with some "GoodOld" plain "Y"-GND
Nothing was better than that, so I decided to stop here, having a very nice amplifier, absolutely quietly amp...
Sry... a' little showoff here :
The pcb, rightside of the "Y"-GND is the speaker protection.
The little switch is on/off for the frontpanel led's.
Left & Right channel's are more or less identical, measurement wise, so only showing most Left channel here:
Plain 1V sine into Xmas Amp. ---> NonInductive 8ohm ---> Behringer soundcard.
2.8V sine.
Now 5V sine.
And 10V sine.
Average crosstalk (Sine's created in Audacity into both channel's).
Again non audible!
THD vs. Voltage show's what the other measurement's also show's... rising Harmonics with voltage.
Gain is very nice I must say the same on both channels... Cool.
Jesper.
Nothing was better than that, so I decided to stop here, having a very nice amplifier, absolutely quietly amp...
Sry... a' little showoff here :
The pcb, rightside of the "Y"-GND is the speaker protection.
The little switch is on/off for the frontpanel led's.
Left & Right channel's are more or less identical, measurement wise, so only showing most Left channel here:
Plain 1V sine into Xmas Amp. ---> NonInductive 8ohm ---> Behringer soundcard.
2.8V sine.
Now 5V sine.
And 10V sine.
Average crosstalk (Sine's created in Audacity into both channel's).
Again non audible!
THD vs. Voltage show's what the other measurement's also show's... rising Harmonics with voltage.
Gain is very nice I must say the same on both channels... Cool.
Jesper.
Hi Jesper,
Nice build and super looking bench! Please share more bench photos in my What’s on the Bench Tonight thread.
We have some identical items in common on our benches. But mine is absolutely messy compared to yours.
Nice build and super looking bench! Please share more bench photos in my What’s on the Bench Tonight thread.
We have some identical items in common on our benches. But mine is absolutely messy compared to yours.
Hi Jesper,So i did a lot of stuff, ending up with some "GoodOld" plain "Y"-GND
Nothing was better than that, so I decided to stop here, having a very nice amplifier, absolutely quietly amp...
Sry... a' little showoff here :
View attachment 1161950
View attachment 1161952
The pcb, rightside of the "Y"-GND is the speaker protection.
The little switch is on/off for the frontpanel led's.
View attachment 1161931
Left & Right channel's are more or less identical, measurement wise, so only showing most Left channel here:
Plain 1V sine into Xmas Amp. ---> NonInductive 8ohm ---> Behringer soundcard.
View attachment 1161935
2.8V sine.
View attachment 1161937
Now 5V sine.
View attachment 1161936
And 10V sine.
View attachment 1161938
Average crosstalk (Sine's created in Audacity into both channel's).
Again non audible!
View attachment 1161939
THD vs. Voltage show's what the other measurement's also show's... rising Harmonics with voltage.
View attachment 1161941
View attachment 1161942
Gain is very nice I must say the same on both channels... Cool.
View attachment 1161945
Jesper.
Nice build, congratulations (although minus infinity squared for the tidiness of your bench).
I'm also de-bugging ground loops at the moment, trying to use X's diagram.
Please can you describe, or point me to, the Y-GND schema you landed on?
This is my first build, and that's not an approach I've come across yet. I decided on a dual mono build, which is double the fun to learn on.
Regards, Haden
HI @Hadox
I'am using a testrig as you can see in my post to measure the amplifier, where groundloops also can be seen.
I was never really able to actually hear it with my ear's.
I use as I wrote a plain old GND "Y" connection, where all GND's are connected in a single point. This setup was giving me the best result's measurement wise that is.
(All input ground (RCA's) + output ground (loudspeaker) + PSU ground's + Heatsink ground + chassis ground etc...)
Good luck hunting this further
I'am using a testrig as you can see in my post to measure the amplifier, where groundloops also can be seen.
I was never really able to actually hear it with my ear's.
I use as I wrote a plain old GND "Y" connection, where all GND's are connected in a single point. This setup was giving me the best result's measurement wise that is.
(All input ground (RCA's) + output ground (loudspeaker) + PSU ground's + Heatsink ground + chassis ground etc...)
Good luck hunting this further
Folks:
I'm designing a compact integrated amp for one of my daughters around Dibya and Jhofland's circuit and plan on using a Galaxy 330 x 280 chassis. My one concern is the heatsink. The layout I've come up with would involve mounting the two TDA7293 amplifiers about 5" (125mm) apart on a flat 6061 aluminum bar measuring 3/4" (19mm) deep, 3" (75mm) tall and 10" (250mm) long. The aluminum bar would itself be mounted to one of the Galaxy's side panels (which diyAudio jokingly calls a "quasi-heatsink").
My question: will an aluminum bar that size (plus any additional heat sinking supplied by the side panel) suffice? The toroid will be a 300VA unit with 25V secondaries.
Thank you for your advice!
Regards,
Scott
I'm designing a compact integrated amp for one of my daughters around Dibya and Jhofland's circuit and plan on using a Galaxy 330 x 280 chassis. My one concern is the heatsink. The layout I've come up with would involve mounting the two TDA7293 amplifiers about 5" (125mm) apart on a flat 6061 aluminum bar measuring 3/4" (19mm) deep, 3" (75mm) tall and 10" (250mm) long. The aluminum bar would itself be mounted to one of the Galaxy's side panels (which diyAudio jokingly calls a "quasi-heatsink").
My question: will an aluminum bar that size (plus any additional heat sinking supplied by the side panel) suffice? The toroid will be a 300VA unit with 25V secondaries.
Thank you for your advice!
Regards,
Scott
Yes, I asked Gianluca about that. He said that while Modushop has a 4U single heatsink chassis, there is nothing in a 2U Galaxy size. He indicated that if there was a demand for such a creature Modushop would consider it, but I suspect demand would be low. So I'm heading in a different direction.
Back to the original question: can anyone confidently tell me if a 3/4" x 3" x 10" flat aluminum bar would provide adequate heat sinking for the Dibya/Jhofland pcbs?
Regards.
Back to the original question: can anyone confidently tell me if a 3/4" x 3" x 10" flat aluminum bar would provide adequate heat sinking for the Dibya/Jhofland pcbs?
Regards.