I ordered the BOM thermistors to replace mine. I tried the ones I had installed detached from the heatsink, but got the same resulting low bias after start-up. If understand the tuning procedure, I should let the amp come to complete temperature stabilization before adjusting bias with the new thermistors installed?
Steve
Yes it is a trial/error process depending how close you are to the intended stabilized target bias. You may have to perform a few boot up/adjustment to get consistent bias. You have very big heatsinks thus longer to have temperature stabilized.
In the mean time to obtain your new ordered thermistors you can try to increase R7/R8 to 10kohm instead of actual 4K7 (ensure this is what you have installed at this time and not 470 ohm...).
Also ensure you have good thermal paste or pad between output mosfet and heatsink and pressure with mounting screw is well applied.
Fab
In the mean time to obtain your new ordered thermistors you can try to increase R7/R8 to 10kohm instead of actual 4K7 (ensure this is what you have installed at this time and not 470 ohm...).
Also ensure you have good thermal paste or pad between output mosfet and heatsink and pressure with mounting screw is well applied.
Fab
Last edited:
A Few Biasing Questions
Fab:
R7/R8 are 4.7K ohm. I already removed my thermistors so I can not try your 10K suggestion. Something I could try should switching to your original thermistors not work.
A few questions about bias and offset. When you say "simultaneously adjust P1/P2, I assume you mean alternating small adjustments of each, not physically trying to adjust both at the same time (too much of a coordination test for me). Is it normal behavior that when you adjust either P1 or P2, BOTH voltages across R15 and R16 should change? When I was was adjusting, they seemed to alway stay within 1 mV of each other. Basically you need to adjust both to get the offset and bias voltage correct? Is there only a single adjustment condition where both bias voltages match and the offset is near zero? I ask because on several occasions when I was adjusting I did turn one of the pots a couple of turns instead of alternating between the two. Is there any way at this time to adjust the pots to a known condition (voltage measurement somewhere else) before trying to adjust for the final bias and offset? Or is it even necessary?
And my question about offset is, Where should I be measuring it? I have been measuring as close to the output and ground on the board as possible, and the offset is slightly higher than when measured at the speaker binding posts. Maybe attributed to resistance thru my speaker protection board and wiring?
Thanks for the help.
Steve
Fab:
R7/R8 are 4.7K ohm. I already removed my thermistors so I can not try your 10K suggestion. Something I could try should switching to your original thermistors not work.
A few questions about bias and offset. When you say "simultaneously adjust P1/P2, I assume you mean alternating small adjustments of each, not physically trying to adjust both at the same time (too much of a coordination test for me). Is it normal behavior that when you adjust either P1 or P2, BOTH voltages across R15 and R16 should change? When I was was adjusting, they seemed to alway stay within 1 mV of each other. Basically you need to adjust both to get the offset and bias voltage correct? Is there only a single adjustment condition where both bias voltages match and the offset is near zero? I ask because on several occasions when I was adjusting I did turn one of the pots a couple of turns instead of alternating between the two. Is there any way at this time to adjust the pots to a known condition (voltage measurement somewhere else) before trying to adjust for the final bias and offset? Or is it even necessary?
And my question about offset is, Where should I be measuring it? I have been measuring as close to the output and ground on the board as possible, and the offset is slightly higher than when measured at the speaker binding posts. Maybe attributed to resistance thru my speaker protection board and wiring?
Thanks for the help.
Steve
Effectively the adjustment of P1 and P2 is done one after the other in small increments and for the setting of BIAS the method is very simple.
1: Set a multimeter on the output of the amplifier to measure the output offset and disconnect the speaker protection.
2: Set a second multimeter to one of the points indicated in the manual (R15 or R16) for setting the BIAS.
During the entire adjustment procedure you will not move your multimeters anymore.
Now you turn in small increments P1 and P2 one after the other so as to have the desired value on the BIAS multimeter and so as to have a null offset on the output multimeter, the settings of P1 and P2 act between them.
1: Set a multimeter on the output of the amplifier to measure the output offset and disconnect the speaker protection.
2: Set a second multimeter to one of the points indicated in the manual (R15 or R16) for setting the BIAS.
During the entire adjustment procedure you will not move your multimeters anymore.
Now you turn in small increments P1 and P2 one after the other so as to have the desired value on the BIAS multimeter and so as to have a null offset on the output multimeter, the settings of P1 and P2 act between them.
Project16,
Thanks for detailing and clearing up the method in which to set bias and minimize dc offset
.
On another note, can you, ggetzoff, and pinnocchio (Do), check your amplifiers this weekend and verify if the bias is within spec? 
Although I expect a small variance, the differences that sledwards (0.8A vs 1.25A) is pretty large and has me somewhat concerned. He used a 4U heatsink and the rest of you used 3U, right? He used a different thermistor with the exact same specs except for directly attaching it to the heatsink (which makes perfect sense!) and the rest of you have the thermistor which isn’t even touching the heatsink! I am perplexed by how this system works as I have a feeling it is not a closed looped system. It seems to require some fiddling and readjustments.
Thanks,
Anand.
Thanks for detailing and clearing up the method in which to set bias and minimize dc offset
.On another note, can you, ggetzoff, and pinnocchio (Do), check your amplifiers this weekend and verify if the bias is within spec?
Although I expect a small variance, the differences that sledwards (0.8A vs 1.25A) is pretty large and has me somewhat concerned
. He used a 4U heatsink and the rest of you used 3U, right? He used a different thermistor with the exact same specs except for directly attaching it to the heatsink (which makes perfect sense!) and the rest of you have the thermistor which isn’t even touching the heatsink! I am perplexed by how this system works as I have a feeling it is not a closed looped system. It seems to require some fiddling and readjustments.Thanks,
Anand.
Last edited:
I will also be able to check on my USSA-5 but I do not have any problem of big variation on the BIAS and the behavior on the measurements is in conformity with the data of the manual of Fab at the last news.
For information I use the thermistors of the BOM, and they do not touch the heatsinks of 3U.
For information I use the thermistors of the BOM, and they do not touch the heatsinks of 3U.
My USSA5 amp has 3,37U (15cm) heatsink and has a stable bias with the small expected variance. There are many ussa5 amplifier builts that exist. This is the same principle of bias temperature compensation as the Pass F5 amp and gives same results ( I have made an F5 amp). The thermistor transfer function effect is reduced by the proper resistor value that is added in series where the equivalent resistor is in parallel with pots P1/P2 and another resistor (R5/R6). The transfer function is also affected by the distance from the heat contributor. In this case this is the vertical Mosfet drivers that require "compensation" and not the lateral Mosfet output (they have the proper temperature coefficient). In fact the thermistor effect is used also to reduce the DC stabilization time after power up.
What is not correct for bias adjustment, for example, is to make the bias adjustment with heatsink placed horizontally because they will be a lot warmer this way than once installed vertically as they are in the amp in real position. Also the 2,8ma value for current source must be used for calculated bias scheme.
Dc bias Variation is not critical in class A push-pull amps as it is in single ended class A.
Fab
What is not correct for bias adjustment, for example, is to make the bias adjustment with heatsink placed horizontally because they will be a lot warmer this way than once installed vertically as they are in the amp in real position. Also the 2,8ma value for current source must be used for calculated bias scheme.
Dc bias Variation is not critical in class A push-pull amps as it is in single ended class A.
Fab
Last edited:
My Bias is very stable on both USSA 3 and 5. NTC are the ones listed in the BoM and they're touching the heatsinks.
I would let the amp sit for 30 minutes, make some final adjustments then wait another 30 minutes and check if at the proper value. If not, keep on adjusting and waiting 30 minutes between each adjustments. If you have a very large heatsink, wait 45 minutes.
Once everything is set properly, let it sit another hour and check to make sure it is stable (which it should be anyways). On your next power on, I would wait a 30m to 1h before measuring the bias. It is a Class A amp and needs to warm up properly. The bias should be set before that time but it does not hurt to wait a little. Mine has been running for several months without any issues whatsoever, just pure good music!
All the best!
Do
I would let the amp sit for 30 minutes, make some final adjustments then wait another 30 minutes and check if at the proper value. If not, keep on adjusting and waiting 30 minutes between each adjustments. If you have a very large heatsink, wait 45 minutes.
Once everything is set properly, let it sit another hour and check to make sure it is stable (which it should be anyways). On your next power on, I would wait a 30m to 1h before measuring the bias. It is a Class A amp and needs to warm up properly. The bias should be set before that time but it does not hurt to wait a little. Mine has been running for several months without any issues whatsoever, just pure good music!
All the best!
Do
Success!
Fab/Project16/ggetzoff:
Well, it turns out the my spade mount lug thermistor IS NOT equivalent to the the one on the BOM. Changed out thermistors, let it fully warm up and biased both channels to 62mV and less than 1mV offset. Let is sit for a while and turned it off. Let it cool down to room temperature and turned it back on.
Results 61 to 62mV on each channel, 2.8mV offset on right, 0.5mV on left. Time for some listening.
Thanks,
Steve
Fab/Project16/ggetzoff:
Well, it turns out the my spade mount lug thermistor IS NOT equivalent to the the one on the BOM. Changed out thermistors, let it fully warm up and biased both channels to 62mV and less than 1mV offset. Let is sit for a while and turned it off. Let it cool down to room temperature and turned it back on.
Results 61 to 62mV on each channel, 2.8mV offset on right, 0.5mV on left. Time for some listening.
Thanks,
Steve
This is good news
Thanks for reporting. This is strange if the datasheet is showing same resistor/temperature curve for both parts...
Fab
.