In post 1995, is that your calculated bias or the bias actually measured on both sides of the amp across the source resistors?
First of all thank you for trying to help me. I can check the output of the diff pair only tomorrow. It would be nice if you can return 🙂
Just for thinking about it meanwhile, do you have an idea what could be the problem?
Since your bias is correct, offset both absolute and relative is within range, I would suspect that the fault lies within the drive portion of the circuit. Look for unmatched feedback or ac gain.
Distortion is apparrent in all three of your traces above but becomes obvious at the higher power levels.
Distortion is apparrent in all three of your traces above but becomes obvious at the higher power levels.
gain is shelving down with frequency ;
it's possibly coincided with output mosfet's capacitance ......
what's current through input LTP ?
it's possibly coincided with output mosfet's capacitance ......
what's current through input LTP ?
Zen, can you explain this "LTP" I'm not good in abbreviations.
Do you mean it's a given that the power output decrease with rising frequency?
We are talking about 20kHz, not 100 or 200.
LTP = Long Tailed Pair. What ZEN is pointing at is that the current should be measured on the FETS in the LTP. If the current is too low, the capacitance presented by the output devices would attenuate the high frequecy. The attenuation would get worse as the frequecy increases.
Could you point me to the correct schematic for the "Green" boards. It has been several years since I built my x100's. (Like in 2004).
IIRC Mine is built with IRP244's
Could you point me to the correct schematic for the "Green" boards. It has been several years since I built my x100's. (Like in 2004).
IIRC Mine is built with IRP244's
I'm no novice but how do I measure this?
I use the boards used here
Aleph-X 100w Amplifier Construction Notes
Schematic
http://www.facstaff.bucknell.edu/esantane/movies/axschem.jpg
Board
http://www.facstaff.bucknell.edu/esantane/movies/pcb-ext.jpg
Thanks for the link to the schematic. My start of this build was way back on post #16.
You need to know (measure) the equivalent resistance of R24 with R26 and V2. Then measure the voltage drop across them and for grins measure across D1. The current is calculated using OHM's law. D1 sets up the voltage reference for the CCS FET Q6 and should be different by a couple of volts. The difference in volts should be within range of the Vgs spec sheet for the IRF9610. While you are in the area, measure the voltage across R17 which should allow you to calculate the current through the zener D1.
You can also measure the current of Q5 and Q7 by measuring the voltage across R23 and R25. The latter gets a little complicated with the current sunk by Q9 and Q4.
You need to know (measure) the equivalent resistance of R24 with R26 and V2. Then measure the voltage drop across them and for grins measure across D1. The current is calculated using OHM's law. D1 sets up the voltage reference for the CCS FET Q6 and should be different by a couple of volts. The difference in volts should be within range of the Vgs spec sheet for the IRF9610. While you are in the area, measure the voltage across R17 which should allow you to calculate the current through the zener D1.
You can also measure the current of Q5 and Q7 by measuring the voltage across R23 and R25. The latter gets a little complicated with the current sunk by Q9 and Q4.
I'll do tomorrow, please check back.
In my amp:
R19/29 is 10k
McMillan is 2.2k
R12/24 is 680 ohms setting AC current gain to app 69%
I have 12 Fet's per Mono Block
just measure voltage across R23 and R25
that will show us current through Q6,Q5,Q7
Q9 and Q4 aren't conducting , hopefully , except as current limiters in clipping condition
that will show us current through Q6,Q5,Q7
Q9 and Q4 aren't conducting , hopefully , except as current limiters in clipping condition
Last edited:
OK, couldn't resist
R23/R25
4.76V and 4.78V means app. 0.012A
R17=10k
R24=330
R26=270
R20=221
V2=200 adjusted to ? (need to check)
One thing to point out is to check the components value vs. actual. Although most of the errors have been me in the past grabbing the wrong part. I do have on my quality shame board two resistors marked as 330R that were actually 10k.
I will keep checking back, but will be gone for the next few days so I have not abandoned your effort. I am sure others like Zen Mod who are much better at this than me will chime in.
I will keep checking back, but will be gone for the next few days so I have not abandoned your effort. I am sure others like Zen Mod who are much better at this than me will chime in.
Last edited:
if you're sure that you have there 390R and up - that's OK
that gives 600uA through zener ; waaay too low
make that 1K5 ..... or even better make current through zenner in range of 5mA
counting Iz=(Upos - Uz)/R17
if PSU is +/-15V , then R17= (15-9)/5mA = 1K2
it's possible that PSU is sagging on high level of power and zenner is shutting down , with so big value of R17
Pass DIY Addict
Joined 2000
Paid Member
Looks like I'm a few days late to the party here, though you've received some great advice already!
A few things that stand out to me as potential things to check:
1) Obvious question, but I have to ask: are the fets well matched?
2) Another obvious thing (already asked by Zen Mod) is the measured value of resistors in the circuit vs. values indicated in the schematic. I learned long ago to measure, measure, measure each and every part. This stemmed from ordering 220ohm resistors and receive 220kohm resistors for a past project.
3) How do things behave if you re-set AC Current Gain back to "normal" (50%) ?
4) Bias current sounds very low for using 12 mosfets per mono. Your bias and rail voltage appear a bit low for that many fets. What is the actual power dissipation for EACH fet? Measure voltage drop across each source resistor, divide this by the value of the source resistor, and multiply by rail voltage. Ideally, each fet should be running in the 20-30w range. If your dissipation is much lower than this, perhaps you want to parallel fewer fets for the output stage?
A few things that stand out to me as potential things to check:
1) Obvious question, but I have to ask: are the fets well matched?
2) Another obvious thing (already asked by Zen Mod) is the measured value of resistors in the circuit vs. values indicated in the schematic. I learned long ago to measure, measure, measure each and every part. This stemmed from ordering 220ohm resistors and receive 220kohm resistors for a past project.
3) How do things behave if you re-set AC Current Gain back to "normal" (50%) ?
4) Bias current sounds very low for using 12 mosfets per mono. Your bias and rail voltage appear a bit low for that many fets. What is the actual power dissipation for EACH fet? Measure voltage drop across each source resistor, divide this by the value of the source resistor, and multiply by rail voltage. Ideally, each fet should be running in the 20-30w range. If your dissipation is much lower than this, perhaps you want to parallel fewer fets for the output stage?