[NuClear235]

Yes, did talk about it thru email. Thak You for all tips.I made this ask to forum, if anybody did not the same problem.
-jumper YES
-YES -27V at stabilied DC part (AC 24V/6A source)
-NO -rezistor 22R I have tampered (as in MiniMax) - I have 300 ohm headphones
-NO -I did it as it is printed on PCB I will repair it
-YES 1k trimmers
-YES 100R VishayDale
-YES 2.2R
-YES PF1
-Tubes 12EA, bias 13,5V
-Buffer BIAS 230mV between TB1L and TB2L. Current I dont know.

So I will connect output rezistor 22R and rotate FETs...

 

[tomb]

Yes, did talk about it thru email. Thak You for all tips.I made this ask to forum, if anybody did not the same problem.
-jumper YES
-YES -27V at stabilied DC part (AC 24V/6A source)
-NO -rezistor 22R I have tampered (as in MiniMax) - I have 300 ohm headphones
-NO -I did it as it is printed on PCB I will repair it
-YES 1k trimmers
-YES 100R VishayDale
-YES 2.2R
-YES PF1
-Tubes 12EA, bias 13,5V
-Buffer BIAS 230mV between TB1L and TB2L. Current I dont know.

So I will connect output rezistor 22R and rotate FETs...

 
That's your answer.  Note that it's possible they've been burned out, too.  Transistors can take a lot of abuse, but soldering them in backwards is often not recoverable.  The silkscreen is correct if 2N5087/88's are used, but the reversal is required if the Toshiba JFETs are installed.  There's no good method to show it two ways on the PCB silkscreen, so documenting it thoroughly on the MOSFET-MAX website was the choice.
 
All of this is detailed here:
 

 
Anyway, I hope the JFETs are OK and you notice an immediate difference. 

 

[NuClear235]

I changet them. Amp plays, so I hope they are OK. Or bad will play too?
Sound stage is better and sound is not so boring. It is big step of improvement.
Thank You for detailed solution. It was mea culpa.

I changed tubes to 12FK6 - fo it is amp build. 12AE I have to MiniMax.
Buffer BIAS set to 250mV. Heatsinks are colder than I expected.
What still stay is that sound is like "speakers playing thru fat textil". So we said it our language-I dont know English equivalent. Highs are not crystal clear... BJT MiniMax still rulez with airy sound.

 

[tomb]

I changet them. Amp plays, so I hope they are OK. Or bad will play too?
Sound stage is better and sound is not so boring. It is big step of improvement.
Thank You for detailed solution. It was mea culpa.

I changed tubes to 12FK6 - fo it is amp build. 12AE I have to MiniMax.
Buffer BIAS set to 250mV. Heatsinks are colder than I expected.
What still stay is that sound is like "speakers playing thru fat textil". So we said it our language-I dont know English equivalent. Highs are not crystal clear... BJT MiniMax still rulez with airy sound.

 
Well, here's another language issue

 :

-NO -rezistor 22R I have tampered (as in MiniMax) - I have 300 ohm headphones 

I don't really understand what you're saying there.  If you have higher resistance at the output than 22R, that could be a reason you're missing the highs.  Stick with 22R - you might even try less, but I would hesitate to go below 10R.
 
There is an issue using 12FK6 tubes with 300 ohm headphones, too.  The 12FK6 is the lowest gain tube available in the Millett Hybrid circuit.  There may not be enough voltage swing at that load impedance to properly run the headphones.  This is especially true with the MOSFET-MAX, because the MOSFETs eat up a couple of volts just to attain a turn-on state.  You may have noticed this when biasing.  Unlike BJT's which react to bias changes immediately, you can turn the trimmers quite a few turns (assuming they're set low in their range) before the MOSFETs even react with a higher current.
 
You should put either 12FM6 or 12AE6 in there if you're using 300 ohm headphones.  That said, it may very well be that the MiniMAX is better with higher-impedance phones.  The MOSFET-MAX is really for low-impedance phones - especially like planars - or flea speakers.  It can really put out the current, but may suffer from lack of voltage swing with high impedance loads.
 
EDIT: One example of this is several years ago, the two best competing DIY headphone amps were the PPA and the M3.  These two amps were very similar in all respects, except one was based on a BJT buffer, the other used MOSFETs.  Many people think to this day that the PPA was better with Senns with its discrete BJT output than the M3 with MOSFETs.  In any other area, the amps were equivalent or perhaps the nod went to the M3 with lower impedance loads.  So, it's a debate with a rich history, BJT's vs. MOSFETs.
 
Finally, if the heat sinks are cool, crank it up.  You should be able to hit 150ma (330mV) if you really want to push it.  It may not maintain that when you case it up, but at least you would've tried all the variables.

 

[NuClear235]

I had RB14 replaced by 0R. Now I am trying 22R. Maybe lower volume and other sound signature changed.
I check my older headphones Koss Porta Pro (60 ohm) and Sennheiser HD570 (120 ohm).
They are more compatible, that is true.
Give more Bias to buffer is not improving sound quality at my amp.
I check it again by oscilloscope - all is perfect.
Maybe I am not compatible with mosfet sound. it still sounds like all instruments on one pile and covered by wool - as cheap headphone amp in mobile phone. Only bass is deep and dynamics is not bad too. I thought is problem in high-mids - they make space in recording.

Thank You for info about M3 and PPA.

 

[tomb]

I had RB14 replaced by 0R. Now I am trying 22R. Maybe lower volume and other sound signature changed.
I check my older headphones Koss Porta Pro (60 ohm) and Sennheiser HD570 (120 ohm).
They are more compatible, that is true.
Give more Bias to buffer is not improving sound quality at my amp.
I check it again by oscilloscope - all is perfect.
Maybe I am not compatible with mosfet sound. it still sounds like all instruments on one pile and covered by wool - as cheap headphone amp in mobile phone. Only bass is deep and dynamics is not bad too. I thought is problem in high-mids - they make space in recording.

Thank You for info about M3 and PPA.

 
The MOSFETs could be oscillating with 0R at RB14.
 
That's probably true about more bias.  Current bias does hit a point of diminishing returns.  I just thought it was one more thing to try.
 
That said, it could very well be your tubes, too.  They can make all the difference between transparency and not.

 

[NuClear235]

Sound is not perfect, but is powerful.
I connected it to my fullrange speakers Fostex and it plays.
Volume is enough for office listening, so probably I will get this amp to my office.
Or I am thinking about change buffer to BJT, I dont know yet...

 

[tomb]

Yep.  I knew it could drive speakers.

 You may need some bigger capacitors on the output to get full bass response, but it's usually a reasonable compromise in using the stock capacitor sizes because desktop speakers can't reproduce deep bass, anyway.  

 

[tomb]

Sound is not perfect, but is powerful.
I connected it to my fullrange speakers Fostex and it plays.
Volume is enough for office listening, so probably I will get this amp to my office.
Or I am thinking about change buffer to BJT, I dont know yet...

 
Just waiting to hear your results after your e-mail expressing worry that you haven't been biasing things correctly.  Obviously if that's the case, it will directly affect the sound quality of the amp and any satisfaction with what you're hearing.  I've stated several times before, I believe the MOSFET-MAX - properly configured - has superior sound to the BJT version (MiniMAX).  Anyway, just to help others the complete instructions for biasing are here:
MOSFET-MAX Setup and Biasing

 

[scottie4442]

ok, I have tried looking everywhere, I am trying to build a mosfet max and need either a quad matched 2sj74/2sk170 or replacements, as I am going to build this in mosfet DB.  I have looked everywhere on google and ebay, the only place I can find them are from China and have read the warnings about the fakes from china.  Anyone have any ideas where I can get a quad match or what replacements I can use for these. thanks for any help

 

[tomb]

  ok, I have tried looking everywhere, I am trying to build a mosfet max and need either a quad matched 2sj74/2sk170 or replacements, as I am going to build this in mosfet DB.  I have looked everywhere on google and ebay, the only place I can find them are from China and have read the warnings about the fakes from china.  Anyone have any ideas where I can get a quad match or what replacements I can use for these. thanks for any help

Order them from bdent.com (B&D Enterprises, they're in New Jersey):
http://www.bdent.com/2sk170bl-2sj74bl-toshiba-audio-transistor.html

 

[scottie4442]

Thanks, I will try them. I want to setup my MOSFET max for my internet radio that I am building, just some flea speakers, for my desktop at work. I have seen that some people on here mention running the MOSFET max with 12aj6's but you would need high PS voltage, any idea where I could get a list of changes to run them?

 

[tomb]

Thanks, I will try them. I want to setup my MOSFET max for my internet radio that I am building, just some flea speakers, for my desktop at work. I have seen that some people on here mention running the MOSFET max with 12aj6's but you would need high PS voltage, any idea where I could get a list of changes to run them?

 
Nope - the 12AJ6 is rated for 30V - same as the other Millett Hybrid tubes.  You should be able to pop them right in there and bias to 13.5V.
 
Trouble is, the 12AJ6 is a unicorn.  I've never seen one anywhere, nor have I ever heard of anybody having one. 

 

[schubert]

The Tube Store has them, or claims to have them, anyway.
http://www.thetubestore.com/Tubes/12A-Types/12AJ6

 

[scottie4442]

I have 14 of them and saw that they were supposedly compatible, I will try them out and post results, I also have 12ae6 and 12fm6 so I will try to compare them all.