I have to confess doing something bad. I used the chassis from an amp using the TP Mono boards. I had used single output winding EI transformers.
Found some dual secondary toroidal, 160 va, 24 volt surplus on eBay. Big mistake. They ended up being 28 volts. Have been running them. Rectified dc was 39 - 40 volts.
Ran it for the last few months. Worked perfect. Did seem to generate more heat even when not playing.
Decided to replace the power transformers to take the stress off the LM3886TF power chips. Looked at transformers that would fit in the case for the power transformer. Wanted to use 22 volt, what I found I liked was another 24 volt.
Installed these two Transformers and loaded it measured 26 volt AC and 37 volts dc. Amp does not get as warm.
BUT I swear it sounded better with voltage rails maxed out. This is not output power related, the speakers are very efficient and it never outputs 1 watt.
Am I fooling myself?
Found some dual secondary toroidal, 160 va, 24 volt surplus on eBay. Big mistake. They ended up being 28 volts. Have been running them. Rectified dc was 39 - 40 volts.
Ran it for the last few months. Worked perfect. Did seem to generate more heat even when not playing.
Decided to replace the power transformers to take the stress off the LM3886TF power chips. Looked at transformers that would fit in the case for the power transformer. Wanted to use 22 volt, what I found I liked was another 24 volt.
Installed these two Transformers and loaded it measured 26 volt AC and 37 volts dc. Amp does not get as warm.
BUT I swear it sounded better with voltage rails maxed out. This is not output power related, the speakers are very efficient and it never outputs 1 watt.
Am I fooling myself?
Last edited:
Interesting result. Early in the development days of FE, Bob or Dario tried transformers of the same voltage, but different VA. If I remember correctly, Bob found that higher VA (about 300 VA) sounds better than lower VA (180 VA, maybe).
Is it possibly power rather than voltage? I'm guessing no, because you likely have move combined VA in the two transformers than the single high voltage one.
Interesting mystery.
Is it possibly power rather than voltage? I'm guessing no, because you likely have move combined VA in the two transformers than the single high voltage one.
Interesting mystery.
I suspect the LM3886TF may run a little higher bias through the output transistors at higher voltage. It sure ran hotter even when idling.
I feel better since backing off the voltage. Think that long term 40 volt rails was going to kill the chips. Even running the rails at +/-37 is a little high.
Hi lokason,
Please post the specific components that weren't available. Often times Dario or some of us will know of alternatives. Also, double check that you are using the v1.5 or later BOM, assuming your boards are V1.5 or later. Components do become obsolete over time, so the older BOM has more unavailable parts.
Jac
Hi Lehmanhill !
It was the R2 ( CMA02040X1002GB300 on the bom : Mouser Electronics, Inc. France ) but i've been able to bought it from RS with the 2 toroidal transformer and the lm3886TF ( wich was not available before the end of april on mouser
) so everything is good !
Thank you !
It was the R2 ( CMA02040X1002GB300 on the bom : Mouser Electronics, Inc. France ) but i've been able to bought it from RS with the 2 toroidal transformer and the lm3886TF ( wich was not available before the end of april on mouser
) so everything is good !Thank you !
Regulated Power Supply Zeners
At one point, Dario noted that, although its not necessary to match components, matching the 14V zeners in the voltage regulator (D105, D205) will reduce DC offset. I never paid much attention to this until last year when I built some old boards with a jfet opamp and found high DC offset. Although I never definitively found the reason for the DC offset, the zeners could have been part of the problem.
At present, I am building a pair of v1.72 boards and decided to buy extra zeners and match them to maximize my chances of a low DC offset. I measured the zener voltage at 5 mA for a batch of 16 zeners. What I found was a variation in zener voltage of nearly 1V. I was surprised how high that variance was. I think I will be matching zeners in the future.
Jac
At one point, Dario noted that, although its not necessary to match components, matching the 14V zeners in the voltage regulator (D105, D205) will reduce DC offset. I never paid much attention to this until last year when I built some old boards with a jfet opamp and found high DC offset. Although I never definitively found the reason for the DC offset, the zeners could have been part of the problem.
At present, I am building a pair of v1.72 boards and decided to buy extra zeners and match them to maximize my chances of a low DC offset. I measured the zener voltage at 5 mA for a batch of 16 zeners. What I found was a variation in zener voltage of nearly 1V. I was surprised how high that variance was. I think I will be matching zeners in the future.
Jac
Hello,
I've a stupid question. My two lm3886 are on the same heatsink and not in a metal box.
Should I link the two HSGND to the heatsink or just one ?
Another stupid question, I've mounted the BD139-16 in the same orientation on the two board ( and noticed it when measuring the voltage... >.< ) is it possible that I fried something behind ?
Grim,
I've a stupid question. My two lm3886 are on the same heatsink and not in a metal box.
Should I link the two HSGND to the heatsink or just one ?
Another stupid question, I've mounted the BD139-16 in the same orientation on the two board ( and noticed it when measuring the voltage... >.< ) is it possible that I fried something behind ?
Grim,
Try Panelhead's suggestion first, but if that doesn't solve it, please tell us what voltages you have on both boards. Also, does the relay click?
The BD-139-16 is in the voltage regulator for the opamp. It is easy to fry. I have burn out theses transistors by accidentally shorting two pins with my voltage meter probe.
Jac
The BD-139-16 is in the voltage regulator for the opamp. It is easy to fry. I have burn out theses transistors by accidentally shorting two pins with my voltage meter probe.
Jac
Hi, the relay click.
I've put the BD-139-16 in the right direction,
Now, when i measure the voltage in output, with the input shorten : it's ok.
With nothing on the input, it's unstable :
20190324_124116.mp4 - Google Drive
Some picture of the board :
P1000699.JPG - Google Drive
P1000703.JPG - Google Drive
P1000700.JPG - Google Drive
P1000702.JPG - Google Drive
NB : i've a hum when i plugged a speacker with nothing in input.
it's still here with a potentiometer in input when the level is not anymore @0
Measuring the + input give me unstable value like in output.
About the part 11 FIRST POWER UP " the value are a little lower ( like 14V instead of 14.6V)
grim,
I've put the BD-139-16 in the right direction,
Now, when i measure the voltage in output, with the input shorten : it's ok.
With nothing on the input, it's unstable :
20190324_124116.mp4 - Google Drive
Some picture of the board :
P1000699.JPG - Google Drive
P1000703.JPG - Google Drive
P1000700.JPG - Google Drive
P1000702.JPG - Google Drive
NB : i've a hum when i plugged a speacker with nothing in input.
it's still here with a potentiometer in input when the level is not anymore @0
Measuring the + input give me unstable value like in output.
About the part 11 FIRST POWER UP " the value are a little lower ( like 14V instead of 14.6V)
grim,
You should not have to short the input. But shorting the input will lower the noise.
If shorting the input gives single digit dc offset everything is working. Plus your relay is closing, another good sign.
If the relay is open you will measure very low dc voltage on the output.
I have measured many versions of My_Ref with the input open. All had low level dc offset.
If shorting the input gives single digit dc offset everything is working. Plus your relay is closing, another good sign.
If the relay is open you will measure very low dc voltage on the output.
I have measured many versions of My_Ref with the input open. All had low level dc offset.
My bad,... i was using a 100Kohm alp... It's clearly better with a 50kohm... ^^' no more hum in the speaker \o/ ( i'll see to use a 10k if i found one like the 100k and the 50k )
Thank you lehmanhill Panelhead and ClaveFremen
Thank you lehmanhill Panelhead and ClaveFremen
Hello all,
I finished to assembling two PCB rev 1.7.
Everything was fine until I connected the input: I heard a loud "poc" on one of my speakers. After inspection, I forgot to connect the PGND on one PCB . This is what happens when you work late at night on your DIY projects!
Since then, the relay no longer activates on the PCB where the connection has been forgotten and the LM3886 is now hot.
I checked the voltages indicated on the PCB and found nothing wrong.
I have already replaced LM318, D1, D2, D2, D3 and LM3886 without success. I measure a DC voltage of -34V on the LM3886 output. 
Do you have any suggestions for troubleshooting?
The positive side is that the second PCB is functional...
Thank you for your help!
I finished to assembling two PCB rev 1.7.
Everything was fine until I connected the input: I heard a loud "poc" on one of my speakers. After inspection, I forgot to connect the PGND on one PCB . This is what happens when you work late at night on your DIY projects!
Since then, the relay no longer activates on the PCB where the connection has been forgotten and the LM3886 is now hot.
I checked the voltages indicated on the PCB and found nothing wrong.
I have already replaced LM318, D1, D2, D2, D3 and LM3886 without success. I measure a DC voltage of -34V on the LM3886 output.
Do you have any suggestions for troubleshooting?
The positive side is that the second PCB is functional...
Thank you for your help!
