I know indra , solution per se (as we already know) elegant and effective , but puzzling is what is culprit of hysteric behavior of dB's amp ........ and both channels the same
for now , I found just two minor changes ocassionaly needed - increasing either R18 or P2 (will send 20K in future) for proper offset , and decrease of R3 , to lower turn on current surge in buffer when using Borbely WCF
all other problems happened, were due to operator frenzy 🙂
name it ....... whatever you can and can't imagine : rotated opto , rotated 7805, mixed places of BD critters , too much diodes mounted , Cinemag pushed in pads for Jensen , mixed trimpots....
though , Greedy Boyz are still much more clever , than I was making my 3rd , 4th or 5th amp .......
for now , I found just two minor changes ocassionaly needed - increasing either R18 or P2 (will send 20K in future) for proper offset , and decrease of R3 , to lower turn on current surge in buffer when using Borbely WCF
all other problems happened, were due to operator frenzy 🙂
name it ....... whatever you can and can't imagine : rotated opto , rotated 7805, mixed places of BD critters , too much diodes mounted , Cinemag pushed in pads for Jensen , mixed trimpots....
though , Greedy Boyz are still much more clever , than I was making my 3rd , 4th or 5th amp .......

Last edited:
forgot to say - agree about gate leakage , thought about that but somehow (silly , dumb , old?) didn't mention ........ probably looking for something "more" drastic
dB - care to measure - DVM on mVdc range - across gate resistor of SITs ?
dB - care to measure - DVM on mVdc range - across gate resistor of SITs ?
"dB - care to measure - DVM on mVdc range - across gate resistor of SITs ?"
will do - R21 only or include the flying resistor too ?
"puzzling is what is culprit of hysteric behavior of dB's amp ........ and both channels the same"
"all other problems happened, were due to operator frenzy"
greedy idiot I may be in the end 🙂 - the symmetry is what puzzles me too , I must be a very good idiot and reliably did the same thing on both sides
I will grab a bunch of measurements to help try and clarify the electrical behaviour of the amp.
Last night I was able to use P1 to crank up the bias easily to 2.4 A at 19.7V on the rails. I wanted to make sure that P1 was still functioning in one direction at least - I could then again easily lower it back down to 1.8A which is the lowest it would go with that Rail voltage.
"High line voltage? Mislabeled donuts?"
?? I have used Avel Lindberg transformers many times without issue - they are clearly well matched as both channels are the same . I will measure the line voltage but it is normally around 120VAC without much fluctuation - the grid where I live is pretty good.
Avel rates their primaries at 115V ( Primaries: 115V + 115V. Frequency: 50/60 Hz ) That might explain the slight bump
thanks for all the input to try and help me resolve this - truly appreciated.
..dB
will do - R21 only or include the flying resistor too ?
"puzzling is what is culprit of hysteric behavior of dB's amp ........ and both channels the same"
"all other problems happened, were due to operator frenzy"
greedy idiot I may be in the end 🙂 - the symmetry is what puzzles me too , I must be a very good idiot and reliably did the same thing on both sides
Not when the output of opto is saturated. Measure opto output voltage between R21 & R20 (pin 3 & 4) to verify, if voltage is < 0.5V then opto output has saturated and no longer able to regulate Iq.
I will grab a bunch of measurements to help try and clarify the electrical behaviour of the amp.
Last night I was able to use P1 to crank up the bias easily to 2.4 A at 19.7V on the rails. I wanted to make sure that P1 was still functioning in one direction at least - I could then again easily lower it back down to 1.8A which is the lowest it would go with that Rail voltage.
"High line voltage? Mislabeled donuts?"
?? I have used Avel Lindberg transformers many times without issue - they are clearly well matched as both channels are the same . I will measure the line voltage but it is normally around 120VAC without much fluctuation - the grid where I live is pretty good.
Avel rates their primaries at 115V ( Primaries: 115V + 115V. Frequency: 50/60 Hz ) That might explain the slight bump
thanks for all the input to try and help me resolve this - truly appreciated.
..dB
We were wondering for the reason you have 25.4v rail, too high for a 18V winding loaded at 2A+. On regular Mosfets and bipolars, a bit higher rail only means a bit higher dissipation, but Iq of VFET is a function of supply voltage, at a certain rail voltage the biasing opto hits minimum Vce and no longer able to regulate....I have used Avel Lindberg transformers many times without issue - they are clearly well matched as both channels are the same ...
Donald
Just checked my M2 with Toshibas - I get 21V with a 1.8A draw.
Does your M2x use the same supply? or do you have a separate supply for it?
What does it measure?
Hope that helps
Bob
Just checked my M2 with Toshibas - I get 21V with a 1.8A draw.
Does your M2x use the same supply? or do you have a separate supply for it?
What does it measure?
Hope that helps
Bob
Hi Bob - I built my M2X with SMPS supplies - so no comparison unfortunately.
I cannot recall the unloaded voltage but I think it was close to what I am getting which means that under load, the voltage is not sagging
18VAC x 1.414 = 25.4V DC = this is the exact line voltage I am getting with a full load which perhaps speaks to the quality of the regulation of the Avel transformers
- I should have gone with smaller / cheaper transformers 15+15 which would have yielded rail voltages of 21.21VDC - this might be the easier solution ultimately if we do not find the problem with my build.
..dB
I cannot recall the unloaded voltage but I think it was close to what I am getting which means that under load, the voltage is not sagging
18VAC x 1.414 = 25.4V DC = this is the exact line voltage I am getting with a full load which perhaps speaks to the quality of the regulation of the Avel transformers
- I should have gone with smaller / cheaper transformers 15+15 which would have yielded rail voltages of 21.21VDC - this might be the easier solution ultimately if we do not find the problem with my build.
..dB
Hi Bob - I built my M2X with SMPS supplies - so no comparison unfortunately.
I cannot recall the unloaded voltage but I think it was close to what I am getting which means that under load, the voltage is not sagging
18VAC x 1.414 = 25.4V DC = this is the exact line voltage I am getting with a full load which perhaps speaks to the quality of the regulation of the Avel transformers
- I should have gone with smaller / cheaper transformers 15+15 which would have yielded rail voltages of 21.21VDC - this might be the easier solution ultimately if we do not find the problem with my build.
..dB
that's having nothing with Donut regulation
effective DC voltage is matter of rectifier and filter (cap bank ) not Donut voltage up and down
what we are interested in is - exact AC voltage at defined load and resulting DC at same defined load
so , please check that and write here
btw. silly question - your DVM is trustworthy > battery state OK and non-suspicious?
and another one - where ( on which position) you're monitoring Iq .... can't remember that you mentioned in-rail 0R1 , so it must be R in CRC ?
"btw. silly question - your DVM is trustworthy > battery state OK and non-suspicious?"
not so silly - a year ago I had a old Fluke do some strange things 🙁- replaced the battery - problems continued 😕- it wasn't worth the cost they wanted to repair it - buried the meter - using different one - all is good 😀
point is that I went down the wrong path because of not checking the basics
not so silly - a year ago I had a old Fluke do some strange things 🙁- replaced the battery - problems continued 😕- it wasn't worth the cost they wanted to repair it - buried the meter - using different one - all is good 😀
point is that I went down the wrong path because of not checking the basics
No time to get all the data last night but checked all the DMMs ( I use 4 cheapo $10 ones , one average one and one very good one ( sperry) - they all measure the same, so I am confident it is not a measurment artifact - but good to check !!
The line voltage at home is 120.5V AC
I have to pull the front panel off the amp to get the unloaded and loaded AC/DC so will check that in time.
My psu is fairly standard FW approach - I use single 33000uF caps instead of the 2 15000uF for CRC
0R1 - I initially used a separate 0R1 3W resistor but after you mentioned using the R in the CRC , I started using that instead and was getting the same readings.
..dB
The line voltage at home is 120.5V AC
I have to pull the front panel off the amp to get the unloaded and loaded AC/DC so will check that in time.
My psu is fairly standard FW approach - I use single 33000uF caps instead of the 2 15000uF for CRC
0R1 - I initially used a separate 0R1 3W resistor but after you mentioned using the R in the CRC , I started using that instead and was getting the same readings.
..dB
Once I have all the data I need , I am leaning toward your patch to drop the voltage to the SIT. I did find a source for the 15V transformers (but didn't spend much time looking ) - that would be option 2 ..dB
Yes, gather more data for sure before making any change, as they say, measure twice but cut once. Good luck.
Edit. When you reduce the voltage to the SIT, make sure to have Iq at least 10% above minimum.
Edit. When you reduce the voltage to the SIT, make sure to have Iq at least 10% above minimum.
Last edited:
Donald
When you stuffed your boards did you do one at a time or the pair together?
I populated both at the same time, symmetrically part by part
Are the heat sinks hot when operating? the 9240s ? the THF51s?
you betcha
I think I have the same heatsinks but am planning to use two per channel. Probably overkill on my part?
Best
Bob
I think you will be happy not to have to run the air conditioner while listening 🙂
..dB
😀😀😀
My workspace is a mess - normally would not stack
mine is lately lesser than usual .....it's simply easier to work

see - almost done some smd soldering , so packing and sending right after (few days of ) State Holidays
Attachments
- Home
- Amplifiers
- Pass Labs
- Babelfish M25, SissySIT - general building tips and tricks