----------
The DC offset is low enough that I would not worry about it.
That said, I have also used the same parts and boards as you and the measured DC is essentially 0 (my meter does not have precision better than 0.1 mV). I am not sure what is happening in your case.
Alvin,
the Pass B1 Buffer is AC coupled at both input and output.
The DC blocking capacitors at In/Out prevent the DC bias inside the amp/buffer from escaping through the interconnects.
If you are reading 6-9mVdc of offset at the output, then you have severe leakage through the output capacitor. That means your DC blocking capacitor is broken.
Latest measurements: DC offset 1.0mV - 1.2mV on both channels. 0.0 mV AC ripple on both channels.
Capacitors are not perfect, but I guess they are good to go. Worse case of 50x gain down the chain will give me 60mV DC at speakers.
Lessons in measurements - the power supply leads was distorting the numbers yesterday as they were close to the output leads. Moving them apart gave better numbers.
Thanks for all your help
Capacitors are not perfect, but I guess they are good to go. Worse case of 50x gain down the chain will give me 60mV DC at speakers.
Lessons in measurements - the power supply leads was distorting the numbers yesterday as they were close to the output leads. Moving them apart gave better numbers.
Thanks for all your help
With a good plastic film capacitor for C101 and using R105 as the grounding resistor should should not be able to measure any output offset.
Something is broken !
Either the capacitor or the voltmeter or your measurement method.
If you have an electrolytic for the C101 you may be able to measure that voltage, but that would imply an leakage of ~ 1.1mV / 221k = ~ 5nA
Change to a Plastic film type. MKT would do, MKP is better, but audibly they probably perform the same.
Something is broken !
Either the capacitor or the voltmeter or your measurement method.
If you have an electrolytic for the C101 you may be able to measure that voltage, but that would imply an leakage of ~ 1.1mV / 221k = ~ 5nA
Change to a Plastic film type. MKT would do, MKP is better, but audibly they probably perform the same.
Right now, I am using a metallised polypropylene (Gold Obligatos). I will look for a replacement to test to make sure it is not the capacitor.
Measurement method: Short the input and measure the voltage across the outputs (unloaded). Did I get this wrong? I don't think it is the voltmeter as I got similar results using another unit.
Last edited:
I read plenty of good results from basic power supplies, even SMPS, and that if doing a linear PSU, it should be single rail. My question regards doing it as a balanced setup. From the linked thread I see that one could use two B1 boards to do an effective yet still pretty simple balanced config.
For the power supply, would there be benefit in doing a separate rail for each board, with its own winding, diodes and bridge? (crosstalk or other) Or is that a waste of time, and I should just take the DC output of a single rail and split it to the two boards?
For the power supply, would there be benefit in doing a separate rail for each board, with its own winding, diodes and bridge? (crosstalk or other) Or is that a waste of time, and I should just take the DC output of a single rail and split it to the two boards?
Switched caps - similar results. Repeated the measurement with a other Voltmeter - similar results. 8.7 VDC across the caps.
Checked the soldering points on the board. Looks ok. Puzzled.
Sent from my iPad using Tapatalk
Distortion from Pass B1
I have sorted out the DC offset issue - just leaving it on for a period of time will bring it down to zero offset.
I continued on with the build, and I am now hitting a strange problem. The sound is slightly distorted, and 'buzzy'. Norah Jones sounds like she is having a cold, and there's a 'bzz' sound at the edge of every tone. It's like she's having a lisp in the lower registers. This is a standard B1 build based on PassDIY PCB and matched JFETS (370s) provided by PassDIY. Power supply is a regulated LM317 supply at 18V DC. The distortion (and noise?) is on both channels. I have tried the following trouble-shooting steps:
1. I suspected a ground loop, so I measured the signal ground with reference to the mains Earth. Open circuit. The signal ground is a 'floating ground'. Not likely to be a ground loop, but to be sure, i revised the grounding scheme to a 'ground lift' with 10 ohm between the signal ground and the main's earth. No impact on the sound.
2. Replace the input and output capacitors. No effect
I am thinking of the following next steps, and would appreciate any input:
1. Replacing the stereo attenuator.
2. Shortening the wires to the input selector and volume control (it's long now - 4 inches) but I am skeptical that this will do anythng
3. Replacing JFETS - again, skeptical on this as unlikely for both channels to be affected.
Appreciate any input. Thanks!
I have sorted out the DC offset issue - just leaving it on for a period of time will bring it down to zero offset.
I continued on with the build, and I am now hitting a strange problem. The sound is slightly distorted, and 'buzzy'. Norah Jones sounds like she is having a cold, and there's a 'bzz' sound at the edge of every tone. It's like she's having a lisp in the lower registers. This is a standard B1 build based on PassDIY PCB and matched JFETS (370s) provided by PassDIY. Power supply is a regulated LM317 supply at 18V DC. The distortion (and noise?) is on both channels. I have tried the following trouble-shooting steps:
1. I suspected a ground loop, so I measured the signal ground with reference to the mains Earth. Open circuit. The signal ground is a 'floating ground'. Not likely to be a ground loop, but to be sure, i revised the grounding scheme to a 'ground lift' with 10 ohm between the signal ground and the main's earth. No impact on the sound.
2. Replace the input and output capacitors. No effect
I am thinking of the following next steps, and would appreciate any input:
1. Replacing the stereo attenuator.
2. Shortening the wires to the input selector and volume control (it's long now - 4 inches) but I am skeptical that this will do anythng
3. Replacing JFETS - again, skeptical on this as unlikely for both channels to be affected.
Appreciate any input. Thanks!
Probably RFI. Make sure the distance of the pot to the input of the transistors is short. Try also to install a 270pF capacitor at pot wiper to gnd.
Sent from my Nexus 4 using Tapatalk
Thanks. I will try that over the weekend if I can find the time.
Sent from my iPad using Tapatalk
I reduced the distance of the pot by more than half, and added a 100pf capacitor (as i had those on hand) at the pot wiper to gnd. Better, but I am still having the 'bzz' sound at the edge of the tone. It sounds like a 'crack' in the woofer, but the speakers are perfectly fine with my integrated amp. I can sense the potential as the rest are excellent - just the 'bzz' 'bzz' sound. Frustrated.
Sent from my iPad using Tapatalk