I think I had the SK117s back to front. I now have a much better set of results.
There's quite a range on the values of Idss and I can't match them up as nicely as I'd like. Q2x, Q3x and Q5x should be 3-4mA and these are the lowest values I have:
3.00mA
3.27mA
4.40mA
4.46mA
4.65mA
5.15mA
Which leaves 5.43mA and 5.66mA for Q7 (4-5mA)
Do those values look okay? As there's quite a wide range for Q2x, Q3x and Q5x which are the best values of R3x and R6x to use? I have 9k and 6.2k, and 33V and 27.7R respectively, but I have or can get other values.
There's quite a range on the values of Idss and I can't match them up as nicely as I'd like. Q2x, Q3x and Q5x should be 3-4mA and these are the lowest values I have:
3.00mA
3.27mA
4.40mA
4.46mA
4.65mA
5.15mA
Which leaves 5.43mA and 5.66mA for Q7 (4-5mA)
Do those values look okay? As there's quite a wide range for Q2x, Q3x and Q5x which are the best values of R3x and R6x to use? I have 9k and 6.2k, and 33V and 27.7R respectively, but I have or can get other values.
The 3.00 & 3.27 IDSS ones can be put to Q3x positions along with 9k1 R3x*. Their lower Vgs (off) is better serving there under just one Vbe=VDS.
The 4.40 & 4.46mA IDSS ones can be put to Q2x positions along with 33R R6x. That value is going to degenerate them enough not to get much hot due to the large VDS they see.
The 4.65mA & 5.15mA ones can be put to Q5X positions where the about 140mW Q4x dissipation they are going to impress on it is still manageable by that 500mW PCmax (at 25C Ta) 150C Tj device (derate PCmax for higher ambient)
5.43mA & 5.66mA are fine for Q7 use still leaving enough voltage for the input stage's cascode base bias system (Q7+ four LEDs) to work well after the R15 C7 local filter drop (IDSS*R15).
*Because such a combination may prove marginal, you could use 5k VR2x or 12K R3x right from the start or to replace when not reaching high enough RAIL DC for proper setup.
The 4.40 & 4.46mA IDSS ones can be put to Q2x positions along with 33R R6x. That value is going to degenerate them enough not to get much hot due to the large VDS they see.
The 4.65mA & 5.15mA ones can be put to Q5X positions where the about 140mW Q4x dissipation they are going to impress on it is still manageable by that 500mW PCmax (at 25C Ta) 150C Tj device (derate PCmax for higher ambient)
5.43mA & 5.66mA are fine for Q7 use still leaving enough voltage for the input stage's cascode base bias system (Q7+ four LEDs) to work well after the R15 C7 local filter drop (IDSS*R15).
*Because such a combination may prove marginal, you could use 5k VR2x or 12K R3x right from the start or to replace when not reaching high enough RAIL DC for proper setup.
Dear Salas,
Have been without a my phono system for over six months while waiting for my Pete Riggle tonearm. Thought it would be a quick delivery so I went ahead and removed the ET - did not feel like putting it back together and just waited. The chance you take when you buy from a craftsman and not a factory.
During this time I rebuilt the power supply and since they were laying about I made a raw supply with an LCLC filter using these giant DUCATI 1500 uF PP caps. Before it was an LC using the same choke and cap. I know it is overkill.
Now for my problems - on one board I seem to have no adjustment with VR2x - I am getting just under 35 volts and I can set TP1/TP2 to 3.6 volts so I figure this is fine (?). VR1 works as one would expect.
Just for fun, since I have made a few complementary buffers using 2SK170/2SJ74 where you must adjust for zero output bias, I thought I would see what I get after the capacitor in the phono circuit and the offset is rather large - around 200 mV. This seems impossible to me, but my knowledge of electronics is meager. This is a 3.3 uF cap.
I have not connected the phono amp to an amplifier yet - a little concerned that this is a problem.
Does this indicate anything to you? And if so, what do you think is happening?
THANKS and take care,
Have been without a my phono system for over six months while waiting for my Pete Riggle tonearm. Thought it would be a quick delivery so I went ahead and removed the ET - did not feel like putting it back together and just waited. The chance you take when you buy from a craftsman and not a factory.
During this time I rebuilt the power supply and since they were laying about I made a raw supply with an LCLC filter using these giant DUCATI 1500 uF PP caps. Before it was an LC using the same choke and cap. I know it is overkill.
Now for my problems - on one board I seem to have no adjustment with VR2x - I am getting just under 35 volts and I can set TP1/TP2 to 3.6 volts so I figure this is fine (?). VR1 works as one would expect.
Just for fun, since I have made a few complementary buffers using 2SK170/2SJ74 where you must adjust for zero output bias, I thought I would see what I get after the capacitor in the phono circuit and the offset is rather large - around 200 mV. This seems impossible to me, but my knowledge of electronics is meager. This is a 3.3 uF cap.
I have not connected the phono amp to an amplifier yet - a little concerned that this is a problem.
Does this indicate anything to you? And if so, what do you think is happening?
THANKS and take care,
Hi Rick
When a VR2x isn't responding (but used to respond) its possible that Q3x has died to a short. By measuring RDS (drain to source resistance) with the Ohm meter you can check for that even without desoldering it first. If its a build that checks OK for non shorted RDS but still behaves like non adjusting then its possible that the particular Q3x has too high an IDSS for the chosen R3x value and you may try a lower R3x value.
Complementary JFETs need symmetric supplies. You probably got those already. Their zero levels should be referenced to the same potential as the previous stage's zero level. There could also be added a large value (say 1Meg) resistor to ground at the capacitor's output referencing to signal's zero level and slowly discharging any residual. So for the capacitor to be empty of charge when connecting and disconnecting the output to other stages or equipment.
When a VR2x isn't responding (but used to respond) its possible that Q3x has died to a short. By measuring RDS (drain to source resistance) with the Ohm meter you can check for that even without desoldering it first. If its a build that checks OK for non shorted RDS but still behaves like non adjusting then its possible that the particular Q3x has too high an IDSS for the chosen R3x value and you may try a lower R3x value.
Complementary JFETs need symmetric supplies. You probably got those already. Their zero levels should be referenced to the same potential as the previous stage's zero level. There could also be added a large value (say 1Meg) resistor to ground at the capacitor's output referencing to signal's zero level and slowly discharging any residual. So for the capacitor to be empty of charge when connecting and disconnecting the output to other stages or equipment.
Thanks, yet again, Salas for all of the help AND for the phono amp.
So I can better understand the cap conundrum - the cap will pass some DC voltage but NO current?
Can't wait to get home and check that FET. Hope I have some laying about even though my instinct is I need to change the resistor.
With gratitude,
So I can better understand the cap conundrum - the cap will pass some DC voltage but NO current?
Can't wait to get home and check that FET. Hope I have some laying about even though my instinct is I need to change the resistor.
With gratitude,
Well, the good news is the FET is not shorted.
I measured the one in the other (adjustable) channel and it does have a higher resistance, if that means anything at all ...
I am using 6.19k resistors in both channels., What would you recommend I use, instead? PLEASE.
THANKS and take care,
I measured the one in the other (adjustable) channel and it does have a higher resistance, if that means anything at all ...
I am using 6.19k resistors in both channels., What would you recommend I use, instead? PLEASE.
THANKS and take care,
Dear Salas,
I have settled my one channel's PS voltage level.
But my question this time is: is it normal for just connecting one's multimeter to cause the TP1/TP2 readings to vary? Should one have to wait, seemingly forever, for a steady reading?
When you take this measurement on your phono stage do you immediately get a solid reading?
Since I am not getting a solid reading (I did think I was earlier and have found that was not the case) where is the first place to look for a problem?
AS always, the 2SK369s are amazingly sensitive to temperature changes. This being the beginning of the warm season in Atlanta means ceiling fans. I have tried to insulate them from breezes short of placing in a box which would make taking the measurement difficult.
I get the feeling this measurement is not a life or death critical thing but would like to get the the two channels the same.
Which is worse - a little high or a little low?
AS much trouble as I am having (again) I am surprised how little there is about biasing on the thread. SO it must be me!
All help and counsel is appreciated.
Thanks and take care,
I have settled my one channel's PS voltage level.
But my question this time is: is it normal for just connecting one's multimeter to cause the TP1/TP2 readings to vary? Should one have to wait, seemingly forever, for a steady reading?
When you take this measurement on your phono stage do you immediately get a solid reading?
Since I am not getting a solid reading (I did think I was earlier and have found that was not the case) where is the first place to look for a problem?
AS always, the 2SK369s are amazingly sensitive to temperature changes. This being the beginning of the warm season in Atlanta means ceiling fans. I have tried to insulate them from breezes short of placing in a box which would make taking the measurement difficult.
I get the feeling this measurement is not a life or death critical thing but would like to get the the two channels the same.
Which is worse - a little high or a little low?
AS much trouble as I am having (again) I am surprised how little there is about biasing on the thread. SO it must be me!
All help and counsel is appreciated.
Thanks and take care,
It's not you, and there is actually quite a bit oof discussion about this topic. I assume your configuration is for low output MC, is that right? When set up for a MM or HOMC the extra source degeneration swamps the SK369's temperature sensitivity. I had a devil of a time getting the tp1-tp2 consistent and correct.
Best advice is clip some alligator leads to the test points and put the lid on your box, at least loosely fastened in place. Attach your meter to the other end. Turn on the amp and watch the meter. It might start slightly negative then cliiiimb. Monitor the reading for at least an hour, though in warm conditions it should stabilize sooner than that. However, it will drift up and down. The hard part is you have to remove the lid to adjust, which throws off the reading, and it is easy to overshoot.
I don't think error up or down matters, you are setting the operating DC V to a mid point. It doesn't matter whether you clip the positive or negative half of the signal, it's still clipping. Luckily the overload margin is pretty good so youvare unlikely to clip either half unless it is way off.
Best advice is clip some alligator leads to the test points and put the lid on your box, at least loosely fastened in place. Attach your meter to the other end. Turn on the amp and watch the meter. It might start slightly negative then cliiiimb. Monitor the reading for at least an hour, though in warm conditions it should stabilize sooner than that. However, it will drift up and down. The hard part is you have to remove the lid to adjust, which throws off the reading, and it is easy to overshoot.
I don't think error up or down matters, you are setting the operating DC V to a mid point. It doesn't matter whether you clip the positive or negative half of the signal, it's still clipping. Luckily the overload margin is pretty good so youvare unlikely to clip either half unless it is way off.
Dear nezbleu and Salas,
Thanks for the recommendation and verification of the advice given.
Further thanks to nezbleu for explaining what we are doing with this adjustment. Never occurred to me that 3.6V was midway between the voltage of the LED string. Another embarrassing admission. Another reason DIYAudio is such a great place to be! Especially when the thread is one started by the inimitable and generous Mr. Salas.
I have been using those leads with the "clips" on the end.
I have been letting it run for days! What I have not done is put it in a box!
Before complaining any more I will put it in a box!
THANKS again, to you both.
Thanks for the recommendation and verification of the advice given.
Further thanks to nezbleu for explaining what we are doing with this adjustment. Never occurred to me that 3.6V was midway between the voltage of the LED string. Another embarrassing admission. Another reason DIYAudio is such a great place to be! Especially when the thread is one started by the inimitable and generous Mr. Salas.
I have been using those leads with the "clips" on the end.
I have been letting it run for days! What I have not done is put it in a box!
Before complaining any more I will put it in a box!
THANKS again, to you both.
One other thought about bias setting. It would be easier with an analog meter. With modern digital meters it is very easy to get fixated on the wrong digits. The goal is 3.6V, but anything in the range of 3.5-3.7 is fine, and probably more like 3.3-3.9. if it is drifting but staying in the right range you have no worries. Obsessing about changes from 3.548 to 3.612 is not helpful. If you had an old analog meter and saw the needle gently moving around 3.6 +/- 0.1 you wouldn't care about the second digit after the decimal! 