I assume you are using FQA9P25 and not IRFP9240.
If the driver is only providing 130mA, then the other device is providing the rest, i.e. 370mA.
To have maximum efficiency and power, you want them close to 1:1.
Morde had 215mA to start with and got close to 250mA after some resistor changes.
The amp will still work as it is.
But it is quite a bit off design values.
So I would at least want to check if the other FET is actually delivering 370mA.
And whether it will respond to the ratio change (increase R11 and decrease R13 while keeping the sum constant).
And as said before, put in a protection circuit.
Patrick
If the driver is only providing 130mA, then the other device is providing the rest, i.e. 370mA.
To have maximum efficiency and power, you want them close to 1:1.
Morde had 215mA to start with and got close to 250mA after some resistor changes.
The amp will still work as it is.
But it is quite a bit off design values.
So I would at least want to check if the other FET is actually delivering 370mA.
And whether it will respond to the ratio change (increase R11 and decrease R13 while keeping the sum constant).
And as said before, put in a protection circuit.
Patrick
I've done a few tests but believe I have inconclusive findings.
Tests suggested I perform to see if Q4 has excessive current
R15 tests - triple checked on both channels, and at 8Vac P-P input, measures 70mVac across R15 on both channels.
Place a 0.1r resistor from -Vs and Q4 drain test
- linked a resistor from -Vs to the Q4 drain (didn't have 0.1r) so I tried a 0.22r and 0.33r on both channels. I only get 5-7mVac (with the 0.33r which is 15-17mA) I know that doesn't make sense, because the current has to go somewhere? But I ran the test multiple times (4 times on one channel and twice on the other) and different ways connecting the resistor to the -Vs and drain of Q4 - everything had same result
I also tried some voltages & temperatures to see if Q4 is current hogging enough to heat the mosfet unequally
- I took out the left channel (only because that offset is the most stable and quick to settle, so noob thinking that it is slightly better matched or built somehow) I checked all voltages and compared to the right channel, they both match
- ran the same 8V p-p input for 30 minutes and took both channels temperatures directly at the bolt and right below the bolt holding the FQA's to the heatsink. Both channels and both output mosfets were within +/-5C across all 4 mosfets.
- checking the bias, the offset and played some real music again. All checked out and the music plays fine.
At this point - not sure what other testing I can do to figure out this current variance in the drivers (Q4/Q5)? Not sure I have the tools (or skill) to do too much more testing (O-Scope, DMM, Capacitor Meter and Transistor Tester)
Tests suggested I perform to see if Q4 has excessive current
R15 tests - triple checked on both channels, and at 8Vac P-P input, measures 70mVac across R15 on both channels.
Place a 0.1r resistor from -Vs and Q4 drain test
- linked a resistor from -Vs to the Q4 drain (didn't have 0.1r) so I tried a 0.22r and 0.33r on both channels. I only get 5-7mVac (with the 0.33r which is 15-17mA) I know that doesn't make sense, because the current has to go somewhere? But I ran the test multiple times (4 times on one channel and twice on the other) and different ways connecting the resistor to the -Vs and drain of Q4 - everything had same result
I also tried some voltages & temperatures to see if Q4 is current hogging enough to heat the mosfet unequally
- I took out the left channel (only because that offset is the most stable and quick to settle, so noob thinking that it is slightly better matched or built somehow) I checked all voltages and compared to the right channel, they both match
- ran the same 8V p-p input for 30 minutes and took both channels temperatures directly at the bolt and right below the bolt holding the FQA's to the heatsink. Both channels and both output mosfets were within +/-5C across all 4 mosfets.
- checking the bias, the offset and played some real music again. All checked out and the music plays fine.
At this point - not sure what other testing I can do to figure out this current variance in the drivers (Q4/Q5)? Not sure I have the tools (or skill) to do too much more testing (O-Scope, DMM, Capacitor Meter and Transistor Tester)
Well, you know yourself all these does not make sense.
But the heat and the output DC (almost impoosible to make mistake) says the circuit works normally.
And it plays music.
So more like your measurement rather than circuit problem.
> R15 tests - triple checked on both channels, and at 8Vac P-P input, measures 70mVac across R15 on both channels.
8Vp-p at input or output ? What is the dummy load resistance ?
> linked a resistor from -Vs to the Q4 drain
The MOSFET drain is connected directly on the PCB to -Vs.
So did you cut (desolder) the drain free from the PCB -Vs before placing 0R33 in between ?
How is the scope probe connected in both cases ?
Did you measure both at the same time (I assume 2 ch scope) ?
Scope is AC coupled ?
If you want to double check with multimeter, you need to lower AC frequency to 100Hz.
And multimeter will give you V_ac values in RMS, not pk-pk.
Patrick
But the heat and the output DC (almost impoosible to make mistake) says the circuit works normally.
And it plays music.
So more like your measurement rather than circuit problem.
> R15 tests - triple checked on both channels, and at 8Vac P-P input, measures 70mVac across R15 on both channels.
8Vp-p at input or output ? What is the dummy load resistance ?
> linked a resistor from -Vs to the Q4 drain
The MOSFET drain is connected directly on the PCB to -Vs.
So did you cut (desolder) the drain free from the PCB -Vs before placing 0R33 in between ?
How is the scope probe connected in both cases ?
Did you measure both at the same time (I assume 2 ch scope) ?
Scope is AC coupled ?
If you want to double check with multimeter, you need to lower AC frequency to 100Hz.
And multimeter will give you V_ac values in RMS, not pk-pk.
Patrick
You are really good at pointing out where I went wrong - which I appreciate, if you didn't I would keep going around in circles.
> R15 tests - yes it is output of 8V p-p using 8-ohm 200w resistors on both channels
>MOSFET drain was not disconnected - I just put the resistor in between the connector for -Vs and the drain pin, that's probably the issue as is typical - my setup/measuring technique. Will re-look at the board/schematic and see what the best course of action is to possibly test again
>Scope / AC coupled - you went right over my head. It's a 2-channel, 100MHz scope, that I generally push the "Auto Set" button on and go from there.
> I didn't trust the scope, so I did try a DMM but didn't change the sine wave from 1k.
As you mentioned - all signs point to it working, maybe not 100% perfect, or possibly a different H2/H3 than designed, but I like it as is. I should have the PSU's in another week and I will try with more capacitance and see if that sounds better. I am also quite happy running it SE (RCAs), but need to add the speaker protection so I feel confident using it with my more expensive speakers. Then I can give it a real audition.
Thanks again Patrick - this really is a nice amp (project) and I've learned/learning a lot along the way.
> R15 tests - yes it is output of 8V p-p using 8-ohm 200w resistors on both channels
>MOSFET drain was not disconnected - I just put the resistor in between the connector for -Vs and the drain pin, that's probably the issue as is typical - my setup/measuring technique. Will re-look at the board/schematic and see what the best course of action is to possibly test again
>Scope / AC coupled - you went right over my head. It's a 2-channel, 100MHz scope, that I generally push the "Auto Set" button on and go from there.
> I didn't trust the scope, so I did try a DMM but didn't change the sine wave from 1k.
As you mentioned - all signs point to it working, maybe not 100% perfect, or possibly a different H2/H3 than designed, but I like it as is. I should have the PSU's in another week and I will try with more capacitance and see if that sounds better. I am also quite happy running it SE (RCAs), but need to add the speaker protection so I feel confident using it with my more expensive speakers. Then I can give it a real audition.
Thanks again Patrick - this really is a nice amp (project) and I've learned/learning a lot along the way.
Got tired of waiting for mouser to stock up on 2SK209GR so picked up 50 from ARROW and set about matching Vgs @ 2mA last night
I got at least 2 identical quads of tightly matched jfets and another 2 matched quads
Just need to screen a few of the J113's I have for ids and will be good to start putting it together.
..dB
I got at least 2 identical quads of tightly matched jfets and another 2 matched quads
Just need to screen a few of the J113's I have for ids and will be good to start putting it together.
..dB
You have the hardest part done, all 20 of my J113s had high enough Ids to use.
Good luck! Very nice sounding amp IMO. I finally got to listen on a good set of speakers for a few consecutive hours this weekend.
Good luck! Very nice sounding amp IMO. I finally got to listen on a good set of speakers for a few consecutive hours this weekend.
Squeezing out chunks of time is my challenge 🙂 . I was surprised that these little jfets mostly had an ids of 11mA , so even if all of my J113 fail to make the cut, I will parallel 2 of the 2SK209s for a substitute if it comes to that. Good to hear that you have managed to get a real sense of what the amp is capable of - I am looking forward to putting it all together too. My dilemma / decision I need to make is whether I run this as a balanced mono or just stick to the original plan for stereo. I suspect balanced will negatively impact the desired H2 / H3 ratio and am leaning towards a good old classic first watt style amp.
..dB
..dB
There are much better balanced amps to buils then bridging 2 of these, e.g. the F5X.
The whole point of the J2 or UDNeSS is that it is singled ended with highish even harmonics.
Using them in balanced destroyed the purpose.
Patrick
The whole point of the J2 or UDNeSS is that it is singled ended with highish even harmonics.
Using them in balanced destroyed the purpose.
Patrick
Thanks for confirming my assumptions Patrick. I have enjoyed following the progress of your F5X project, it is beyond beautiful in its execution with attention to all levels of technical detail. My humble homage to the F5 is an F5V3 turbo which is still in early assembly phase..dB
Hello all. Progress is slow but every step forward is closer to a working amp 🙂
I have a quick question which I cannot see an answer for.
Resistor 11a on the schematic does not seem to be used by anyone. It would in theory occupy the space under R11 (labelled R12a)
I just have the trimmer to install and then am ready wire it up for a test.
Thanks for any insight to R11a
..dB
I have a quick question which I cannot see an answer for.
Resistor 11a on the schematic does not seem to be used by anyone. It would in theory occupy the space under R11 (labelled R12a)
I just have the trimmer to install and then am ready wire it up for a test.
Thanks for any insight to R11a
..dB
I guess you meant R12a / R14a ?
https://www.diyaudio.com/forums/att...602423076-udness-dont-semisouths-img_2305-jpg
They are there to allow you to fine trim dynamic current ratio.
UDNeSS, or You don't need Semisouth's
The values provided in the schematics are based on the FQA FETs used by Morde.
You can change the current ratio by changing the ratio of R11//12 and R13//14.
Best to play with it in Spice first to get an idea.
Patrick
https://www.diyaudio.com/forums/att...602423076-udness-dont-semisouths-img_2305-jpg
They are there to allow you to fine trim dynamic current ratio.
UDNeSS, or You don't need Semisouth's
The values provided in the schematics are based on the FQA FETs used by Morde.
You can change the current ratio by changing the ratio of R11//12 and R13//14.
Best to play with it in Spice first to get an idea.
Patrick
Thanks Patrick.
I thought that was the case based on the table you added to the PDF. I was just surprised to see that R11a is listed in the BOM and based on the schematic it is in parallel to R11
Seems I cannot load images from my phone, will just leave it out and see if i need to trim. I used the same FETS as Morde so let's hope it all works without additional tweaking.
..dB
I thought that was the case based on the table you added to the PDF. I was just surprised to see that R11a is listed in the BOM and based on the schematic it is in parallel to R11
Seems I cannot load images from my phone, will just leave it out and see if i need to trim. I used the same FETS as Morde so let's hope it all works without additional tweaking.
..dB
Last edited:
R11a in the schematics = R12a on the PCB.
I did not have space to place it under R11.
And was too lazy to change document. 🙂
Patrick
I did not have space to place it under R11.
And was too lazy to change document. 🙂
Patrick
boards populated and power supply assembled, now just to tap the heatsinks and we will have another UDNess in the wild
I added a trimmer for setting the bias, hopefully I have given myself enough range.
I added a trimmer for setting the bias, hopefully I have given myself enough range.
If that trimmer is parallel to R11,12,13, or 14 then it will not survive for long.
The easiest way to trim bias is to adjust R6.
R12a, 14a should be used to adjust current ratio.
And they want to be 0.4W minimum.
See Appendix 2.1 of post #89.
Patrick
The easiest way to trim bias is to adjust R6.
R12a, 14a should be used to adjust current ratio.
And they want to be 0.4W minimum.
See Appendix 2.1 of post #89.
Patrick
So you drilled through the PCB to install the trimmer for R6.
I strongly suggest you to replace it later with fixed resistors after trimming.
The power resistors next to it will get hot.
And trimmers will drift with temperature, even if they survive the heat.
Patrick
I strongly suggest you to replace it later with fixed resistors after trimming.
The power resistors next to it will get hot.
And trimmers will drift with temperature, even if they survive the heat.
Patrick
I have confirmed thru measurements with ltspice that an H2 of +90 degrees is perfect negative phase second harmonic.So -120 degrees is ?