Grounding micpre

[pazduha]

This is my first project attempting to make a Hamptone JFET preamp.
Included below is a layout of a wiring. Purple lines are a shield of a shielded cable.
There is a bit of a problem with a hum. The level of noise with power off is -90db on a converter.
With all the pots to the max I get -50dB. The gain of single stage (lets say hjfp1) is 24dB.
I would be glad if u can help me with the grounding.
Included are the pictures of preamp and a freq graph of the noise.

Thank you,
Miha

 

[Khron]

Is this with nothing plugged into the XLR input? And/or which input (XLR or instrument) is selected? Or do you get the same noise level on either one?

Try connecting, for testing, a 150 ohm resistor between pins 2 and 3 - that's how noise specs of mic preamps are measured.

 

[pazduha]

I have a jack plug for di that disconects the mic side of the wiring.

I did some tests:

mod 1: Nothing pluged in with thick wires connecting all the pin1 to the earth nut at the back.
This seems to make the amp 10dB more quiet. (I have to rerun tests again i might have messed up the measurement, coz of the small box, putting wires back on is a nightmare)

mod 2: Pins 1 only connected to the metal near the plug as in layout picture
-No Di nothing pluged in : peak -33.7 dB (strong 100hz peak with harmonics)
-150 ohm pins 2,3: peak -33.4 dB peak same as before.
- guitar on 0 pluged in DI: -29.5 dB peak same as before.

I would say the noise is the same on all plugs.

 

[ruffrecords]

Where the mic signal cable reaches the board, what is the screen connected to?

Most of your noise seems to be 100Hz and its harmonics which suggests insufficient smoothing or poor
or ground layout. Can you post your PSU schematic?

To eliminate external noise getting into the mic XLR you should run tests with an XLR with pins 2 and 3 shorted together (nothing on pin 1) plugged into the mic input.
Cheers

Ian

 

[pazduha]

Mic cable screen from pin1 is not connected. Only pins 2&3 go to input transformer.

This is the PSU:

 

[JohnRoberts]

There should be lots of discussion referring to "pin 1 problem" (perhaps do a search).

XLR pin 1 typically gets hard bonded to chassis ground and if not careful shield ground currents can contaminate signal grounds.

JR

 

[pazduha]

To clarify it more:
1. mic in screen is connected to pin1 of XLR mic input, pin 1 is connected to metal near the XLR mic input.
2. it goes to pad where screen is connected to screen going to input transformer.
3. screen is not connected at input transformer
4. input transformer output cable: screen connected to 0 (black), one side of transformer secondary and shield
5. this goes to DI jack plug, screen is not connected there
6. DI jack cable goes to the hjfp module 1 input. Screen connected on one side to the sleeve and on the other side to 0 (black).
7. hjfp module 1 after cap1. Screen to 0 (black)
8. Pot1 both screens together at pin1
9. same thing for module 2 and pot 2
10. After pot2 screen is connected to XLR output pin 2, signal to pin 3. XLR output pin 1 is connected to metal near this pin.

 

[pazduha]

Does this change matter? :


(I suggest downloading both layouts and scroll between them to see differences)

Cheers, ty

 

[Voyager10]

Looking at the pictures, the power transformer is pretty close to the high gain circuitry, and I wouldn't be surprised if it's just picking up the magnetic field.

Can you (safely) move the transformer out of the box, on some long wire? If the hum level changes with transformer position, it's not the ground layout which is at fault.

 

[pazduha]

I had a single stage of hjfp made on a protoboard without box, runing 24V from a power supply from this veroboard that is now in the box ( with both stages of hjfp on the veroboard completly disconnected). The only thing connected to protoboard was 6 wires:
24 V +-
input +-
output +-

input was connected only to DI jack without any mic wires
output was connected only to output XLR pin 2&3, with pin1 to metal. No pots.

I was using a different transformer on a long wire. Veroboard had no grounds connected just two wires for 22V AC and 24 V +- DC to the protoboard ofc.

Noise inside daw was -85dB.
---
This new transformer has extra winding for phantom.
N turns is:
P1 - 2500
S1 - 456
S2 -239

Its dead quiet to the ear. Old transformer was buzzing a bit. Also I didnt notice any significant level in change of noise when it was close. Only if I lifted it close to mic XLR area.

 

[ruffrecords]

Mic cable screen from pin1 is not connected. Only pins 2&3 go to input transformer.

OK, that is good. What is the secondary voltage and current rating of the transformer?

Edit: You are feeding the phantom power to the mic connector via 6K8 resistors?

Cheers

Ian

 

[pazduha]

I know nothing about winding tranformers and electronics. But as I see it, my reasoning was:

Number of turns:
P1 - 2500 0.15mm wire
S1 - 456 0.15mm wire
S2 -239 0.3mm wire

U:
P1 230V
S1 44.5V (measured)
S2 23.3V (measured)

I again assumed:
0.3mm wire Imax = cca 200mA
0.1mm wire Imax = cca 50mA

 

[ccaudle]

mic in screen is connected to pin1 of XLR mic input, pin 1 is connected to metal near the XLR mic input.

I see black paint on the front and back panels. Is the area where the front and back panels touch the side/bottom piece free of paint?

The only thing connected to protoboard was 6 wires:
24 V +-
input +-
output +-

So single 24V supply, +24V/0V, or +24V and -24V? I assume the first, but the way it is written could be ambiguous.
Assuming the first, is the power supply 0V connected to chassis at some point?

 

[pazduha]

I see black paint on the front and back panels. Is the area where the front and back panels touch the side/bottom piece free of paint?

No, both front and back panels are painted. Both have 2 bolts at the bottom and two screws at the top for connection. Front panel front side is sanded where the bolts meet the panel. I will completely sand the back sides of both panels where they meet the bottom panel tomorrow.

So single 24V supply, +24V/0V, or +24V and -24V? I assume the first, but the way it is written could be ambiguous.

Yes, its +24V/0V.

Assuming the first, is the power supply 0V connected to chassis at some point?

If you mean protoboard test circuit then no.
If you mean "finished" veroboard boxed version, then yes. At one point right after the power supply. If you look at the PSU pic right at the - sign (also on the layout).

Ty.

 

[Voyager10]

Is this the "Hamptone" preamp mentioned in this thread?

Several posts note that it takes a lot of power supply current. It's quite possible that the power supply isn't adequate, and as a result you're getting ripple on the supply which the 7824 can't filter out.

Simple steps to test this would be

• Measure the supply current to each HJFP module. If it's over 40mA it's too much.
• Increase the smoothing capacitor (470uF) between the bridge rectifier and the 7824.

 

[ruffrecords]

I know nothing about winding tranformers and electronics. But as I see it, my reasoning was:

Number of turns:
P1 - 2500 0.15mm wire
S1 - 456 0.15mm wire
S2 -239 0.3mm wire

U:
P1 230V
S1 44.5V (measured)
S2 23.3V (measured)

I again assumed:
0.3mm wire Imax = cca 200mA
0.1mm wire Imax = cca 50mA

Just to be clear, are you saying you wound this mains transformer yourself?

Cheers

Ian

 

[pazduha]

Is this the "Hamptone" preamp mentioned in this thread?

Yes, this is the preamp. But my version is a bit different. Its done with 2SK170BL JFETs and resistor values are different.
I measured the U drop over a 33Ohm resistor before each stage and I get:
886mV for m1, I= 26.8mA
968mV for m2, I= 29.3mA

Just to be clear, are you saying you wound this mains transformer yourself?

Yes. Specs as writen above.

 

[My3gger]

I'm telling Miha for months to only use select 2N5457, or 2SK170 he picked instead after turning trimmers a bit, then do the rest to make the circuit work properly... We talked about the need for chasis panels conducting, signal wires shielding, using known good power trafo i gave him for testing, taking power trafo out of chasis for now, grounding (a few Hamptones i made 15-20 years ago work fine with phantom an all, so do jfet inputs on DIs), pin 1 problem, Ian's Grounding 101 and other reading, phantom wiring, ergonomics, using LM317 with more filtering because scope showed 7mV of ripple, operating with hard data not guessing, using good test setup at my place pretty much anytime because DVM isn't enough, start testing with dynamic mic, etc, etc.
Layout is good, so is material, the second module and coupling caps could be positioned elsewhere to shorten signal wire runs, although i don't think this is a problem if signal wires are shielded where needed and power trafo works properly.

 

[pazduha]

This is the Linear Voltage Regulator I have:
https://si.farnell.com/stmicroelect...13?CMP=e-email-sys-orderconfirmed-GLB-Product
My DVM is probably not accurate to measure the ripple i get? But the U AC from transformer should be enough to power it?
Its strange to me, that one stage works without noise. Could this doubling of current demand ond the PSU be the couse of the noise?

 

[pazduha]

Edit: You are feeding the phantom power to the mic connector via 6K8 resistors?

Yes, It was done according to this schematic:
https://www.jensen-transformers.com/wp-content/uploads/2014/08/as016.pdf