Is there any formula to calculate the Vref based on the Idss of the njfet and the reference resistor in v1.1? For the same Vref, higher Idss requires higher resistor value, right?
Anyone has the figure of the input impedance of the Q3 BC550C?
Has anyone thought of using a JFET instead of bipolar in place of Q3? lower noise, higher impedance? What is the problem with a JFET?
Anyone has the figure of the input impedance of the Q3 BC550C?
Has anyone thought of using a JFET instead of bipolar in place of Q3? lower noise, higher impedance? What is the problem with a JFET?
Last edited:
V_out should be about 0.6 + Idss_Q6 * (R7 + Rtrim)
Yes, for same Vref, higher Idss requires higher Rtrim
Q3 vs jfet: I certainly thought of it, but found other areas of improvement so I addressed those first.
Ikoflexer,
Yes I derived the same formula but the result did not match, hence the question.
In my working example, Idss=7.3mA, so 0.6 + 0.0073 * (2200 + 200) = 18.12 (V). I only got 13V.
By the way, if that is the formula, for a given voltage, the higher the Idss would require lower Rtrim.
Edit: The above figures are approximate. I have not measured the actual current in this leg. The Idss was measured independantly, i.e. outside the working circuit. Perhaps that is where things can go wrong.
Regards,
Bill
Yes I derived the same formula but the result did not match, hence the question.
In my working example, Idss=7.3mA, so 0.6 + 0.0073 * (2200 + 200) = 18.12 (V). I only got 13V.
By the way, if that is the formula, for a given voltage, the higher the Idss would require lower Rtrim.
Edit: The above figures are approximate. I have not measured the actual current in this leg. The Idss was measured independantly, i.e. outside the working circuit. Perhaps that is where things can go wrong.
Regards,
Bill
Last edited:
Under just Vbe derate your 9-10V measured 2sk170 idss by 30-35% as I have wrote here some times before.
Most important is to make sure that you have some significant current going through the shunt mosfet, Q4. If there isn't enough current there, your output voltage will sag. A quick test is to reduce the value of R1 a bit, say, to 10R. If your output voltage goes up, then it's clear, the shunt current was set too low.
OK. I won't try the BF244B then, as it will give 15mA current. I will select a lower Idss from a batch of 2sk170.
From the Futurlec batch, I have 5 x 2sk170 with Idss between 3.x mA to 6.x mA. Could I use those?
The reason is, from the tests I have done so far, film capacitors give a far better subjective sound than electrolytic capacitors. I will report when the tests are all done. So I would like to have the highest possible R7+Rtrim to push the corner frequency down.
In any case, I am very very very pleased with the result. With the film cap alone without the electrolytic cap, it has the absolutely the best sound I have ever heard. And I have a solution even if I can't get the corner frequency down to a reasonable value. This is because my speakers are active so I can use this very best sounding ccs-shuntreg with film caps only to drive only the tweeter and the midrange (The PSSR of the OPA627 is only 88dB at 5kHz), and have a second ccs-shuntreg with electrolytic caps (or even a LM317/337) to drive the bass and subs, and let the 118dB PSSR (up to 100Hz) of the OPA627 to deal with the bass.
From the Futurlec batch, I have 5 x 2sk170 with Idss between 3.x mA to 6.x mA. Could I use those?
The reason is, from the tests I have done so far, film capacitors give a far better subjective sound than electrolytic capacitors. I will report when the tests are all done. So I would like to have the highest possible R7+Rtrim to push the corner frequency down.
In any case, I am very very very pleased with the result. With the film cap alone without the electrolytic cap, it has the absolutely the best sound I have ever heard. And I have a solution even if I can't get the corner frequency down to a reasonable value. This is because my speakers are active so I can use this very best sounding ccs-shuntreg with film caps only to drive only the tweeter and the midrange (The PSSR of the OPA627 is only 88dB at 5kHz), and have a second ccs-shuntreg with electrolytic caps (or even a LM317/337) to drive the bass and subs, and let the 118dB PSSR (up to 100Hz) of the OPA627 to deal with the bass.
Thanks. Allright I will go through the trouble of installing a 1R resistor and measure the current of Q4.
Individual tuning pays always. You can use idss that is not adversely enlarges pinch off. Up to 8.5mA-10mA tested at 9V I have seen is doing ok. Don't go thinking with OPA pssr because in a 627 pre first thing V1 builder Marinos got, was bass dynamics in his posts early in another thread and in this one.
And we talk VS large batteries.
That's why I keep repeating. Every new regulator I test with a passive load. In v1.1 the current through R1 should be roughly equal to about 20mA + current through shunt mosfet + current through passive load. Therefore, it better be the case that the current through the shunt mosfet is adequate.
I run 'em hottest practical. Its not only the shunt currents relationships. It livens up Mosfets gm. Double or tripple the audio part's bias current is good. In my phono I run 250mA to feed 45ma. He he. Of course when you go for amplifiers, the heat penalty will not allow such royal luxuries.
Double or tripple the audio part's bias current is good. In my phono I run 250mA to feed 45ma. He he. Of course when you go for amplifiers, the heat penalty will not allow such royal luxuries.Individual tuning pays always. You can use idss that is not adversely enlarges pinch off. Up to 8.5mA-10mA tested at 9V I have seen is doing ok. Don't go thinking with OPA pssr because in a 627 pre first thing V1 builder Marinos got, was bass dynamics in his posts early in another thread and in this one.
So could I try the ones measured at 3.5mA, for example?
I run 'em hottest practical. Its not only the shunt currents relationships. It livens up Mosfets gm.
This correlates well with my measurements. Running them at higher current and higher current decreased the output impedance more and more. In practical terms, the ripple at the output that was induced by a sine current load decreased more and more as the shunt current was increased. This makes the supply stiffer and stiffer. Good stuff, to run them hot
This correlates well with my measurements. Running them at higher current and higher current decreased the output impedance more and more. In practical terms, the ripple at the output that was induced by a sine current load decreased more and more as the shunt current was increased. This makes the supply stiffer and stiffer. Good stuff, to run them hot
It looks like I should get a power amp chassis to build some shunts with 0.8A currents to drive my 200-350mA load.
And just to drive nutnut nutier, it slightly alters subjective tone with each fold increase given the interaction with the audio application served. Better tune a Bosendorfer piano, its easier. The more you tweak, the more it subtly alters Mr. Nut.
Better tune a Bosendorfer piano, its easier. The more you tweak, the more it subtly alters Mr. Nut.Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.