M-125 volume problem - fixed !

Hello,

I am looking for some debugging advice to help me narrow down something that has been bothering me.  My right monoblock is not as loud as my left one.  I've been using the gain on my sp14 to compensate (about 4 clicks), but I want to figure out why it is softer.  My conclusions so far:
- it's not just room acoustics - at low volume listening a foot from the speakers the right output is obviously softer
- it is not the speaker or speaker cable - I swapped them and it follows the monoblock
- it's not the input - I switched the rca cables
- it is not the output tubes - swapped all 4 KT-88s between monoblocks
- it is not the driver tubes - swapped them
- voltages at all 8 pins on the 4 output tubes are all within published range.  

I don't hear any audio artifacts.  They both sound great to my ears, it is only a volume difference.  But I want to isolate what is different and am asking for advice on how to do that.  Are there measurements I should be taking at the speaker binding posts?  Could a bad soldering job there cause this?  Could it be a wrong/bad component on the board?  

Thank you

Has there always been a volume difference? If so, maybe it's a mis-wired output tap. Try the other one.

Also check the Triode/UL switch for backwards.

You didn't mention bias.

If the problem is new, then safely trace the circuitry for a bad joint. But with voltages being checked as ok on the output tubes, seems like things are good up to there. Gotta be something with the output trans?

Can you post photos of the internals of both?

Thank you for the response.
- The problem is not new, the right has always played software than the left
- I switched taps and heard no difference
- I switched the right monoblock to triode mode and the volume dropped further as I would expect. So the switch is not wired backwards.
- both monoblocks have autobias boards, but initially they did not. The autobias board did not introduce this problem. Bias is very close to .46 for all 8 KT-88 tubes. Autobias instructions recommended .45 - i set it for just a bit above that. previously they were set to .5. No impact to this problem.
I can post pictures later - can't do that right now.

Thanks again

What output caps do you have installed? Those would be my next target.

All parts are stock. Do you mean the c6 and c7 russian pio caps?

If you have a scope you can measure the signal output at different points inside each monoblock and compare to find where the signal strength is dropping. My guess is that your problem is at the binding posts. Why? I ran into the same problem when I first built my monoblocks. Everything inside the two amps checked out OK, but the signal at the speaker posts was different. I soldered the transformer outputs to the binding post again and got some improvement. I was still not satisfied and decided to make the huge upgrade to WBT binding posts. The Dayton Audio post might be just as good, but since my speakers have WBT terminals, I spend the $ for my amps to match the quality on my speakers... my problem was solved. Yours might be too. I always thought that the binding posts connections could be better and that they are too close to the chassis for my liking. Please note that the post provided with the M-125 kit are not bad/cheap quality, but if you are looking for superior sound and willing to spend the extra cash, I recommend to everyone to upgrade your speaker binding posts on your M-125s. Good luck

Thanks AmpedUp.  Coincidently I decided yesterday I did not like the looks of black binding post connections or the looseness of the banana plug fit.  I replaced the post but it had no effect.  And since the problem exists with both the 4 and 8 ohm posts, I am assuming that both red posts don't have the same problem.

Here are pictures of the right (quieter) monoblock.  Will post pictures of the left as well.  You will see from the pictures I am new to this and my wiring management is not very good.  I am definitely open to all criticism about the quality of my work if you think it may be contributing to the issue.  

Right (quieter monoblock): https://www.amazon.com/photos/shared/Yj_ZzY6gSQisj3VARqHn2g.CiwRCnhuyGlwM5axgrDizv

Note - you will see the resister on V2 looks different.  It is a 1000 ohm 1 watt resister like it is supposed to be, and the problem predates my needing to replace this resister.  I have Vishay resisters coming and will replace it will a physically larger resister for better heat dissipation.

Left (louder monoblock): https://www.amazon.com/photos/shared/YMwofMnHRiqbbkeP-N30Ng.3_tyshaeQ4gW9THdsT610Y

IntarsiaGuy wrote:All parts are stock. Do you mean the c6 and c7 russian pio caps?

Yeah, I had a bad Russian cap that took out a couple of 6550 tubes. I had a hard time figuring it out because it was never sudden. After replacing them I had no further problems.

I recommend these as a replacement.
https://www.partsconnexion.com/product49901.html

Thanks for the recommendation. I bought some for backups and might replace if other debugging does not lead me anywhere. BTW, do you know what I should be seeing for resistance between P5 of the output tube sockets and ground? This is from the other thread I am on for this issue.

Pin-5 to Ground on both of mine measure 115kΩ before creeping up. On my M-125 Pin-5 and Pin-6 measure -62.4VDC when powered.

Have you considered that the louder amp may be the one that has the problem?

IntarsiaGuy wrote:he autobias board did not introduce this problem.  Bias is very close to .46 for all 8 KT-88 tubes.  Autobias instructions recommended .45 - i set it for just a bit above that.  previously they were set to .5.
Thanks again

I just realized something I'd like to understand.
When I measure Bias while there are four tubes in place, the Bias is 1V. If I only had one pair of tube, two tubes in place, the Bias would be adjusted to 0.5V. Both of these populations will result in 50mA of Current per tube.

I don't know how the AutoBias board operates, so please forgive my ignorance on this. When you say all 8 KT-88 tubes are at .46, is this V or mA? In other words, is this number a per tube spec, and if so, is it for Amps or Volts?

ttocs, others have mentioned, and i am certainly open to the louder amp being the issue. Wen powered, my pin 5 and 6 are right around -60.0 volts on both amps. But my pin 5 to ground are not reading anything near 115K and that might be pointing to a problem.

When I reread the instructions about biasing with 2 versus 4 tubes, I contacted Erhard audio to understand if I should be biasing at .5 or 1 volt. Since the 4 tubes are biased independently with the autobias board, they should be at .5 and not doubled up to 1. Pin 1/8 are all right around .462 VDC.

I am not familiar with the driver boards for the M125 but have you checked the resistor locations/values between the two amps? Maybe some components were installed in the wrong place during the build causing a gain difference between driver boards. Does this board have the CCS? If so maybe one is weak (or strong?).

I have measured 2 differences with the amps. The resister at R28 on the driver board (same resister installed in both as far as stripes indicating they are 10 ohm resisters)
The right (quieter) amp reads 1.4 ohms, the left reads 10.0 ohms.

And measuring from pin 5 of the output tube sockets to the associated GRD on the autobias board, V1 and V2 on the right (quieter) amp read 1K. The other 6 readings are all 100K.

So this gives me some things to work on...

100 ohm instead of 100K ohm resisters on the right autobias board. Fixing that now.

Replaced, all 8 measure at about 100K now. Looking into R28...

Lifted one side of R28 and it measured fine. Suspect the LM334 is bad.

Sounds like you're getting close. Good job!

Thanks - part on order, will report back once I've replaced and tested

The LM334 arrived.  I installed it in the right monoblock, and that was the issue.  My sound stage is no longer skewed off to the left.  The two monoblocks play at the same volume.  This volume delta was really bothering me and I am delighted to have it fixed.  Thank you for the help and advice.

Thanks for sharing IntarsiaGuy. It was fun rooting for you to figure this out and maybe learn something along the ride. Congrats...

I am not one to question a tried and true design that has been working for years, but I recall a few instances where people have replaced the LM334 IC to fix a problem. I have a technical question about this - so let us engineers and techs have some fun here... LM334s are very temperature sensitive but there is a design trick using a diode and resistor in parallel to R28 to nullify any temp change (See page 8 on the Data Sheet)

https://www.ti.com/lit/ds/symlink/lm334.pdf?HQS=dis-mous-null-mousermode-dsf-pf-null-wwe&ts=1636648151083&ref_url=https%253A%252F%252Fwww.mouser.com%252F

Would adding this to the build improve the design similar to the yellow sheet mod? Does the max operating temp of the LN334 at 185 degrees ever come in jeopardy? I do not know if heat is causing any problems with the LM334 but wanted to ask all the smart people that participate in this forum.

Mike

AmpedUp wrote:

Would adding this to the build improve the design similar to the yellow sheet mod?  Does the max operating temp of the LN334 at 185 degrees ever come in jeopardy?  I do not know if heat is causing any problems with the LM334 but wanted to ask all the smart people that participate in this forum.

Adding the resistor / diode combination will stabilize the current drift due to temperature.

If going to the effort, replacing the plastic TO-92 package with a metal TO and adding a heat sink would reduce the drift from self heating.

I've never used anything other than 12AU7 as my driver tube, but I did have one go on me and that may have been when the LM334 went too.  I did see a recommendation for installing the LM334 on the bottom of the board to help isolate it from heat - not sure if that is good advice or not.

As implemented, the LM334 is going to do a lot of self heating which impacts the current which impacts the self-heating heat which impacts the current...

People typically place monoblocs side by side. In the M-125, the left amp LM334 is heated by the radiation from the right amp power tubes.
If you can feel, or better yet measure a temperature of more than a few degrees, difference between the right side of the left and right amp chassis, the amps are too close.

Mounting the LM334 under the board will be a more constant thermal environment, not subject to the vagaries of open windows and aircon.
The amps should still be physically separated to reduce heating by thermal radiation.
Raising the amps permits better air flow.

Note that the above are all small but definitely measurable.

Note to Tubes4HiFi: The M-125 could easily be mirrored, just bend half the chassis the other way.
Novice builders should build the Left [power tubes left] amp to familiarize and then do the Right.