Author Topic: Low output and distortion at lower gain levels

March 20, 2022, 04:12:04 PM
Read 1761 times

ouchmouse

  • Newbie
  • *
  • Posts: 7
  • Karma: +0/-0
    • View Profile
So I finished building the MP573 and did all the tests, I thought it was working fine, then I realized the gain wasn't too high.

Checking the voltage of each pin (I didn't write them down), but they were all exact or within a very small margin of the correct figures.
The transformers resistances all seem correct.

There's some heavy saturation/distortion in the signal way before the LED turns red. I put some sine waves through it and they're about 40dB quieter than they should be. To achieve unity on the input I need about 40dB of gain and at that point the top half of the waveform is squared off.

One other thing is that when I use the rotary encoder at 30, 40, 50 and 60dB of gain there's a fairly big pop sound.

I couldn't seem to follow the input with the multimeter, I could only get one value at the input transformer pin 7 and and TP1 and it was the same with or without any input.

Any idea what could be wrong?

March 21, 2022, 09:02:48 AM
Reply #1

JPK

  • Administrator
  • Hero Member
  • *****
  • Posts: 998
  • Karma: +29/-0
    • View Profile
One other thing is that when I use the rotary encoder at 30, 40, 50 and 60dB of gain there's a fairly big pop sound.
I would start here. You probably have a reversed cap.
Check the direction of C8, C14, C19, C20, C22, C23, C24, C27, C29.
JPK

March 24, 2022, 02:08:04 PM
Reply #2

ouchmouse

  • Newbie
  • *
  • Posts: 7
  • Karma: +0/-0
    • View Profile
All the capacitors are in the correct direction and correct location. Solder seems fine.

I worry I burnt the outside of the relays a little as I put the wrong diodes in at first and had to swap them. But I guess it’s only cosmetic.

March 25, 2022, 10:59:12 AM
Reply #3

JPK

  • Administrator
  • Hero Member
  • *****
  • Posts: 998
  • Karma: +29/-0
    • View Profile
Check if you can "see" the input transformer primary by measuring its resistance with your ohmmeter between pins 2 & 3 of the input XLR or pins 8 & 10 on the extender.
You need to be powered on to do this.
With the impedance switch on 1200 you should read 48 ohm,
on 300 you should read 12 ohm.
JPK

March 30, 2022, 10:03:19 AM
Reply #4

ouchmouse

  • Newbie
  • *
  • Posts: 7
  • Karma: +0/-0
    • View Profile
I realized the output transformer connector had a bad solder. After I fixed it the headroom is fine, there's more gain and less distortion, working as usual and sounding great!

There were two different PCB parts for the output transformer connector, at first I figured I used the wrong one, so I tried to unsolder, but then figured both would work, but I guess I forgot to re-solder one pin.

There is still a click at each gain stage above 20 on the rotary encoder, I can mostly deal with this for now, but it would be nice to fix it in the future.

March 30, 2022, 10:14:18 AM
Reply #5

JPK

  • Administrator
  • Hero Member
  • *****
  • Posts: 998
  • Karma: +29/-0
    • View Profile
There is still a click at each gain stage above 20
Check if you can measure any DC voltage against TP2 on 
- transformer T1 pin 7, 
- R38 top pin,
- R39 low pin.
Check at different positions of the gain switch. Only a few millivolts can produce a click.
JPK

April 06, 2022, 09:48:20 AM
Reply #6

ouchmouse

  • Newbie
  • *
  • Posts: 7
  • Karma: +0/-0
    • View Profile
Sorry for the late responses.

Checking T1 Pin 7 I measure:
10 - 0v
20 - 0v
30 - 0.01v
40 - 0.05v
50 - 0.07v
60 - 0.07v
There are spikes up to 0.14 when changing the gain.

R39 I measure 0 on each gain, but again some small spikes up to .1v
R38 measures the same.

Not really sure what that voltage measurements on the input transformer means, how can I fix this?

April 06, 2022, 12:19:58 PM
Reply #7

JPK

  • Administrator
  • Hero Member
  • *****
  • Posts: 998
  • Karma: +29/-0
    • View Profile
You may have a leaking C14. Double check its direction.
You can try to replace it.
JPK

April 08, 2022, 04:20:19 PM
Reply #8

ouchmouse

  • Newbie
  • *
  • Posts: 7
  • Karma: +0/-0
    • View Profile
I realized that C20 was actually the wrong way around, I swapped it, now I get zero input level.
There are clicks when changing the gain, but a lot smaller than before though.

I guess I should replace C20?

The bias wouldn't need to be adjusted would it?

Edit: I think the output transformer wiring was bad so I ordered another header pin and I’ll try soldering the cables again, hopefully they won’t be too short now.
« Last Edit: April 09, 2022, 05:21:18 AM by ouchmouse »

April 09, 2022, 03:53:40 PM
Reply #9

JPK

  • Administrator
  • Hero Member
  • *****
  • Posts: 998
  • Karma: +29/-0
    • View Profile
now I get zero input level
You may have disturbed something else. The cap alone should not cut the signal completely.

I guess I should replace C20?
Very probable. The tantalum caps don't like being reverse polarized.

The bias wouldn't need to be adjusted would it?
No.

hopefully they won’t be too short now.
There is a good margin.
JPK

April 10, 2022, 10:42:31 AM
Reply #10

ouchmouse

  • Newbie
  • *
  • Posts: 7
  • Karma: +0/-0
    • View Profile
(This post was edited a few times.)

I replaced the capacitor at C20, for some reason there's louder pops now than I had the first time and still no output after rewiring the output transformer.

Then I put another capacitor in and it's better, there are still pops, but -18dBu or less, the LED turns green, but not red and not peaking at 0dB like before.

With JMP3 in place it does work, so I think I may need to replace the 2N7000.

Thank you for the continued support and apologies.

I rechecked all the DC voltages, (before changing the capacitor).

A - 2.3
B - 3.7
C - 22.4
D - 11.9
E - 1.8
F - 11.4
G - 0.9
H - 4.1
I - 0.3
J - 3.5
K - 22.4
L - 3.0
« Last Edit: April 10, 2022, 03:50:03 PM by ouchmouse »

April 11, 2022, 11:41:17 AM
Reply #11

JPK

  • Administrator
  • Hero Member
  • *****
  • Posts: 998
  • Karma: +29/-0
    • View Profile
The voltages look good.

With JMP3 in place it does work, so I think I may need to replace the 2N7000.
Yes very possible. Be careful keeping your body grounded when handling the 2N7000. This can be done by working without your shoes on.

for some reason there's louder pops now
You could try a simple good quality electrolytic 22uF-35V in C20, in case the new tantalum is also leaking.
JPK

April 12, 2022, 04:42:23 AM
Reply #12

ouchmouse

  • Newbie
  • *
  • Posts: 7
  • Karma: +0/-0
    • View Profile
I replaced the tantalum twice, so the pops are way quieter now.

I always built with no shoes, but I have vinyl flooring. I’ll take care to ground myself then try it next time, ordered a few 2N7000s in case.

All should go well.