Perfect sound, bad measurements

[ivo]

Hi!

This is my first post on this forum but my experience with tripath amps started actually in July 2008 with building an amp4. No problems so far (very fine amp for low power!). Now Ive built 2 trupath amps for biamping my Jordan jx92s/sub combination. The amps actually worked perfect from first startup on!! The first measurements were as follows:
-VPP/VNN +/-50V on both
-IPP/INN ~ 90mA on amp1 and ~70mA on amp2
-IN10 ~ 200mA on both
After about 8 hours of exiting listening I realized that the 3020 got kind of hot (~ 55C) and the heatsinks where a bit warm (~45C) but toroids were cold. So I but them back to my bench and measured the following:
-VPP/VNN +/-50V on both
-IPP ~ 100mA on amp1 ~80ma on amp2
-INN ~ 300mA on amp1 and ~280mA on amp2!!!!!!
-IN10 ~ 180mA on both
Double checked all soldering and component placement, everything seems ok.

Any idea where to look first? Did you check your IPP/INN after an extended usage of your amp? Any changes?
Your help is much appreciated!!!!

Cheers
Ivo

[V-bro]

I presume this is on Truepath?

------X-max. can be several meters on any driver, too bad it can be done only once...------

[ivo]

of course! only the best [^]

[audiophool]

>Any idea where to look first? Did you check your IPP/INN after an extended usage of your amp? Any changes?

I've been monitoring mine for 18 months. No significant change from first startup.

>INN ~ 300mA on amp1 and ~280mA on amp2!!!!!!

Check your DC offset.

Rgds.

[ivo]

Tim, thank you for your answer
I can adjust DC-Offset to below 10mV. I can not see, measure or hear any other effect only INN is to height.
In theory what could cause the INN to go height without affecting IPP? Which part of the circuitry defines actually the amount of current thru the MOSFET at idle? Maybe if I understand this better I will be able to make some more measurements (have a cheap Velleman PC-Scope)
Cheers
Ivo

[audiophool]

Is the idle temperature of both heatsinks the same on each amp? If yes then the excessive INN must be flowing through both low side MOSFETs on each amp, ie 4 MOSFETs total. This would suggest a systemic error on your part. Could it be something odd about the way that you are measuring INN and there is in fact nothing wrong at all: Did you inadvertently measure INN + IN10 on your second round of measurements?


Did you find that the offset had changed after 8 hours when you noted your idle INN had?

Provided the offset is zeroed, the idle current is entirely the result of shoot-through current. That's the current that flows from VPP to VNN through both MOSFETs during the period when both MOSFETs are on during switching between them. Also that's why IPP and INN should be the same.

If the offset is not zeroed, and you have a load connected, some DC current will flow through the load and this will result in an imbalance between IPP and INN.

Are you conducting your bench test without any load tested?

If there is an at-idle imbalance between IPP and INN and you have no load connected then the imbalance current MUST be going somewhere and you need to find out where but doing this can be tricky. This schematic should help you: http://docs.google.com/Doc?id=dfrxcj...09024957414133


Rgds.

[ivo]

Yes, this looks very odd to me as well.

>>Is the idle temperature of both heatsinks the same on each amp?

Yes, all the low side MOSFETS get hot.

>> Could it be something odd about the way that you are measuring INN and there is in fact nothing wrong at all:

I hope not. I actually measured voltage over a 1 Ohm Resistor either by Multimeter or Scope to determine Current as I have no functioning Amp-Meter. (Shame on me)
Could that be a problem?

>>Did you find that the offset had changed after 8 hours when you noted your idle INN had?

No, DC-Offset is rock stable.


>>Are you conducting your bench test without any load tested?

Yes

>>If there is an at-idle imbalance between IPP and INN and you have no load connected then the imbalance current MUST be going somewhere and you need to find out where but doing this can be tricky. This schematic should help you: http://docs.google.com/Doc?id=dfrxcj...09024957414133

So I will try to find out where the current is going to.


Could it be an influence from VN10 as it is reverenced to VNN?

Cheers
Ivo

[audiophool]

quote:>Yes, all the low side MOSFETS get hot.

How do you know that the low side MOSFETs get hotter than the high side MOSFETs? Have you removed the heatsinks? Could you paste some photos of what you are doing?

I said:

quote:>> Could it be something odd about the way that you are measuring INN and there is in fact nothing wrong at all...

This is what I mean:

quote:>>If there is an at-idle imbalance between IPP and INN and you have no load connected then the imbalance current MUST be going somewhere and you need to find out where but doing this can be tricky. This schematic should help you: http://docs.google.com/Doc?id=dfrxcj...09024957414133

So I will try to find out where the current is going to.

quote:Could it be an influence from VN10 as it is reverenced to VNN?

Yes, as per the pic above.

Rgds.

[ivo]

quote:How do you know that the low side MOSFETs get hotter than the high side MOSFETs? Have you removed the heatsinks? Could you paste some photos of what you are doing?

Yes I removed the heatsinks for a short time.

quote:Could it be something odd about the way that you are measuring INN and there is in fact nothing wrong at all...

I connected VN10 ground and VNN directly at the power supply side and measured each line individually.


There is one other thing that came to my mind. If I read this from TA3020 reference board document:

No damage will occur to the TA3020 if either the V5or VN10 collapse before VPP or VNN, assuming the mute is asserted before the supplies start to discharge.

Then I have to assume that a damage could occur to the TA3020 (or the MOSFETs?) if I remove power WITHOUT mute asserted. Is that correct? If so, then I have to check for a faulty TA3020 or MOSFET because I did exactly this on both amps.

cheers
ivo

[audiophool]

quote:Originally posted by ivo

...
There is one other thing that came to my mind. If I read this from TA3020 reference board document:

No damage will occur to the TA3020 if either the V5or VN10 collapse before VPP or VNN, assuming the mute is asserted before the supplies start to discharge.

Then I have to assume that a damage could occur to the TA3020 (or the MOSFETs?) if I remove power WITHOUT mute asserted. Is that correct? If so, then I have to check for a faulty TA3020 or MOSFET because I did exactly this on both amps.

Good find, Ivo, however, no. This statement in the datasheet is misleading. No damage will occour to the anything on the Truepath regardless of how you switch the power supplies.

Keep looking!