Borbely tells us that the minimum bias for mosFET amplifiers is 500mA irrespective of the number of output devices.
That gives one 1Apk of ClassA current to the output.
That gives one 1Apk of ClassA current to the output.
And the very good sounding Goldmund amplifiers runs with very low bias current just as the very good sounding Perraux.
Koldby
I dont now what happen with the others new models Goldmund about low bias ,with Mimesis 9.2 everything go hot VAS-OUTPUT .I cant think T03 with low bias....
Look at the heatsinks in goldmund amps.....
Even mimesis must have pretty low bias not to burn everything around it...
My own design Audio Mirror TFA 3 was a 2x150 watt hybrid tube / MOSFet amp with BIG heatsinks and it ran really hot with 100mA pr FET.
Even mimesis must have pretty low bias not to burn everything around it...
My own design Audio Mirror TFA 3 was a 2x150 watt hybrid tube / MOSFet amp with BIG heatsinks and it ran really hot with 100mA pr FET.
Those too.
But also , and much more so, the Telos series..
I have a clone in house.
2x350 Watt and a heatsink smaller than my HOT running TFA3..
And it stays cool.
My Audio Mirror was with 2SK135/2SJ50 TO-3.
Anyway look at the heatsinks and the poweroutput.
No class A 1 amp peak here that is for sure.
Goldmund Mimesis 8 power amplifier | Stereophile.com
Mimesis-8 runs 80 mA bias TOTAL for all the MOSFets and have a +/- 62,5 V supply.
But also , and much more so, the Telos series..
I have a clone in house.
2x350 Watt and a heatsink smaller than my HOT running TFA3..
And it stays cool.
My Audio Mirror was with 2SK135/2SJ50 TO-3.
Anyway look at the heatsinks and the poweroutput.
No class A 1 amp peak here that is for sure.
Goldmund Mimesis 8 power amplifier | Stereophile.com
Mimesis-8 runs 80 mA bias TOTAL for all the MOSFets and have a +/- 62,5 V supply.
Cool stays in Denmark in Greece everything goes hot.
"""Mimesis-8 runs 80 mA bias TOTAL for all the MOSFets and have a +/- 62,5 V supply.""
""The 8 offers just two pairs per channel of complementary output MOSFETs (Hitachi K134/J49). These carry a modest bias current of around 80mA total.""
it means 40 ma per mosfet i cant think 20ma per mosfet!! mush more to hear this sound in case T03
Anyway my amp runs 130-145 ma per mosfet at +/- 70V
Nagys he says this amp go really hot .
"""Mimesis-8 runs 80 mA bias TOTAL for all the MOSFets and have a +/- 62,5 V supply.""
""The 8 offers just two pairs per channel of complementary output MOSFETs (Hitachi K134/J49). These carry a modest bias current of around 80mA total.""
it means 40 ma per mosfet i cant think 20ma per mosfet!! mush more to hear this sound in case T03
Anyway my amp runs 130-145 ma per mosfet at +/- 70V
Nagys he says this amp go really hot .
mosFETs do not get the benefit of optimal ClassAB bias.
You need to run them with a lot of bias to get minimum crossover distortion. The harder you run them the lower the crossover distortion, right up until they never leave ClassA.
Then use feedback to attenuate that crossover distortion.
There are many experts here that tell us to run mosFETs hard just for that reason.
There is an alternative, as used by R.Cordell and others.
You need to run them with a lot of bias to get minimum crossover distortion. The harder you run them the lower the crossover distortion, right up until they never leave ClassA.
Then use feedback to attenuate that crossover distortion.
There are many experts here that tell us to run mosFETs hard just for that reason.
There is an alternative, as used by R.Cordell and others.
Did you ever listen to an amp with these different biases.
I know about the crossover dist. min. this very old knowledge, but that dosent equal best sound. And it also depends on how the frontend is designed.
Every time I have run a Mosfet with really high bias I get inferior sound, regardless of frontend design. 100 mA is to my ears the absolute max. to get decent sound from mosfets. I do not care for class A mosfet sound. Bipolar yes - Mosfet no--IMHO.
It actually puzzles me that the Goldmund Telos is sounding as good as it does, despite its pretty low bias setting..
LOL Yes very cold in Denmark..
The original Mimesis, that Nagy has used as inspiration (not to say cloned) would not survive 140 mA and 70V supply with the heatsink it was equipped with IMHO.
I ran double 2sk135/2sj50 with 100mA an 62V with BIG heeatsinks and that amp got hot!!
Let me share my own thoughts as well.
Firstly I'm very thankful to Nikos for assisting me(through PM 😀) to set my amp to different bias
After listening to different bias which needs me to adjust the DC offset as well, the higher the bias the higher the DC offset, this is my observation to my amp:
My conclusion is to stick to the original what Nagys recommendation. If I made it High bias I'm loosing the control of the bass at the same time soundstage and dynamic is not that good to my ears.
With the original bias I can feel the separation of the instruments the bass is totally in control and the vocals are really detailed.
Again this is my observation and I admit I'm not an audiophile in any aspect but I know if the amp I'm listening is good or bad 😉.
Firstly I'm very thankful to Nikos for assisting me(through PM 😀) to set my amp to different bias
After listening to different bias which needs me to adjust the DC offset as well, the higher the bias the higher the DC offset, this is my observation to my amp:
My conclusion is to stick to the original what Nagys recommendation. If I made it High bias I'm loosing the control of the bass at the same time soundstage and dynamic is not that good to my ears.
With the original bias I can feel the separation of the instruments the bass is totally in control and the vocals are really detailed.
Again this is my observation and I admit I'm not an audiophile in any aspect but I know if the amp I'm listening is good or bad 😉.
Koldby you talking about GOLDMUND series with T03 Mosfet?
So here is the original Mimesis 9.2 that is the basis of Nagys design.
+/- 80V supply....😱
And agian look at the heatsinks. They are small!
140 Ma ... They will not survive long, and If I know Goldmund original amolifiers, NONE of them were hot when running.
Attachments
This subjective evaluation is in line with my circuit analysis.
The circuit is not "flexible" (note the high power supply voltage and the small 10k feedback resistor). It is already on the "sweet spot". You cannot change one variable without seriously affecting the other variables. Worse, this is already close to being instable.
Ηι Κoldby
i have the same heatsink 300mm 200mm 40mm yes its small for 145 ma
For the history i have made some small mods
i change the zener 6,2V
i run the Vas at 5.0 ma per side very stable ,i have turn the legs and all are face to face and with caught clip, and my Mpsa 93-43 now runs at 52-56 C, before mods they run at 6,8 ma 65-68 C really really hot .Thats why the Junie have offset the Vas MPSA are not in caught clip. Now the offset is very very very stable
i change R20 with 470Ω at 70-115ma per mosfet .
i have the same heatsink 300mm 200mm 40mm yes its small for 145 ma
For the history i have made some small mods
i change the zener 6,2V
i run the Vas at 5.0 ma per side very stable ,i have turn the legs and all are face to face and with caught clip, and my Mpsa 93-43 now runs at 52-56 C, before mods they run at 6,8 ma 65-68 C really really hot .Thats why the Junie have offset the Vas MPSA are not in caught clip. Now the offset is very very very stable
i change R20 with 470Ω at 70-115ma per mosfet .
Attachments
Last edited:
Actually I have asked one of my freinds to amke the PCB s for this amp, so I can compare it with my Telos. Could be very interesting!
Do you use +/- 80V and 140 mA with those heatsinks 😱
In Greece I would guess you would have to have a very good air condition system in your listening room...
That confirms that Goldmund are using feedback to reduce errors in a high distortion amplifier.
Everybody here knows that one should design for low distortion before applying high feedback.
