"squeeling" noise in mic preamp

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

aviel

Well-known member
Joined
Feb 6, 2018
Messages
96
Hey folks,
i need your help here!
i am building a mic preamp (based on hamptone FET).  my preamp is 2 "gain stages":
Tranny->Stage1->Pot (Voltage divider) - > Stage 2 -> Pot (voltage divider) -> Line driver (IC)

1. when no mic is conneted i get crazy oscilations.
2. when mic is connectedm i get high pitch "squeeling" background noise. that "squeeling" noise changes pitch and level when i touch the body of the knobs, as well as changing pitch when rotating the knob.
i used a star ground (See pictures). and placed 470uF and 1000uF capacitors between the +16 and gnd.. grounding the body of the pot's isnt helping. it sounds like some ground loop, but how is that possible?

34y7er6.jpg

rkwr9s.jpg


 
In a high gain circuit if output leaks into the input it can oscillate... layout matters and circuit design matters.

There is a wealth of experience on this forum, but our guesswork will not help you.

JR 
 
ruffrecords said:
Please post a scematic then we can halp you.

Cheers

Ian

Hey,
Attached the schematic.

Cheers!
 

Attachments

  • fet-preamp (1).png
    fet-preamp (1).png
    186.9 KB · Views: 69
You almost certainly have far to much overall gain. You can probably dispense with one of the FET preamps. You also have the output op amp right next to the input transformer; possibly the worst place for it.

Cheers

Ian
 
ruffrecords said:
You almost certainly have far to much overall gain. You can probably dispense with one of the FET preamps. You also have the output op amp right next to the input transformer; possibly the worst place for it.

Cheers

Ian

Where should i place it?
Also, why is that so bad.? Could there actually be leakage from the opamp to the transformer? The circuit does have a lot of gain, assuming i want to stick with 2 stages, how would you reduce the gain of a stage? I could remove that bypass capacitor.
But seems that this design is what people like about the hamptone.
 
Hi Aviel
The basic spacing for wideband amplifiers should be 20dB per inch to maintain stability. The spacing can be closer if you use metal shields to make the SIGNAL PATHS over the shields to meet needed spacing.
I would suggest that the POTS be moved to the UPPER EDGE above each amplifier stage and the output stage placed where the pots are now after extending that edge of the PCB. This means the signal path is a straight line and not circle back, as the phase shift and the coupling input to output will create an OSCILLATOR STAGE.
Duke
 
aviel said:
Where should i place it?
Also, why is that so bad.? Could there actually be leakage from the opamp to the transformer?
Absolutely, just stray capacitance from the op amp output to the input of the transformer could be enough. A better place for it physically is after the second FET stage
The circuit does have a lot of gain, assuming i want to stick with 2 stages, how would you reduce the gain of a stage? I could remove that bypass capacitor.

Yes, that would reduce it to a more sensible amount. Exactly how much gain do you need?

Cheers

Ian
 
Audio1Man said:
Hi Aviel
The basic spacing for wideband amplifiers should be 20dB per inch to maintain stability. The spacing can be closer if you use metal shields to make the SIGNAL PATHS over the shields to meet needed spacing.
I would suggest that the POTS be moved to the UPPER EDGE above each amplifier stage and the output stage placed where the pots are now after extending that edge of the PCB. This means the signal path is a straight line and not circle back, as the phase shift and the coupling input to output will create an OSCILLATOR STAGE.
Duke

Thanks for your inputs!
It will be hard to do i assume if i want it to fit into a 500 series module like now...
 
Hey!
i redid the layout - see attached picture.
overall it's much quieter. i STILL get some high pitch background noise (around 6Khz, sounds like  a sine wave).
it also doesn't increase with gain at all, and not very loud, but still there.

How would you tackle/debug this?
 

Attachments

  • InkedWhatsApp Image 2018-12-30 at 14.50.57_LI.jpg
    InkedWhatsApp Image 2018-12-30 at 14.50.57_LI.jpg
    1.2 MB · Views: 37
ruffrecords said:
Start my screening the audio connections e.g. to and from the pots, inputs and outputs.

Cheers

Ian

how do i do that? do i need to get a special screened wire? 
arent the screened wires much bigger in diameter?
 
aviel said:
how do i do that? do i need to get a special screened wire? 

Yes you need some audio coax to do this - the kind of thing commonly connected to RCA jacks
arent the screened wires much bigger in diameter?
Overall they are. The point is they help stop unwanted interactions between different parts of the circuit.

The key thing you need to understand is that in a high gain circuit such as this, the performance owes as much to the construction as it does to the design.

Cheers

Ian
 
I remember building a VCA limiter and getting the same sort of sine wave oscillation (very low but audible)

My best guess is it had something to do with the transistor. I checked some tech papers and the transistor in question was designed to function below 100° C voltages.

In a small active speaker amplifier I saw a copper heatsink shape of an arch above an IC, soldered to the board.
 
efinque said:
I remember building a VCA limiter and getting the same sort of sine wave oscillation (very low but audible)

My best guess is it had something to do with the transistor. I checked some tech papers and the transistor in question was designed to function below 100° C voltages.

In a small active speaker amplifier I saw a copper heatsink shape of an arch above an IC, soldered to the board.

Interesting.. It gets a bit hot indeed, but far from 100C. I would also assume that if it indeed thermal noise, it will not be there the moment i turn on the unit, what do you think?
 
aviel said:
Interesting.. It gets a bit hot indeed, but far from 100C. I would also assume that if it indeed thermal noise, it will not be there the moment i turn on the unit, what do you think?

Heat can do a lot of damage to components, and DIL IC's are no different, actually I think they can be even worse due to the small size and pin spacing etc.

As the dielectric/insulation/complementary metal oxide layers wear or melt due to heat the thermal characteristics change in relation, ie. it doesn't matter because the change from room temperature to operating temperature is/can be nearly instant.. many manufacturers know this but lack the proper (or ideal) materials.

One trick would be to pull/source components from devices that are operated in unusual conditions, high temperatures, moisture etc.. they're rarely hi-fi (unless you work in NASA or something) but worth trying.

For example an amplifier I'm making uses a 0,1uF/230V cap I pulled from a hair dryer because a) I happened to have one at hand and b) it most likely withstands heat (and high voltages) very well.

EDIT : also I think a DIL is somewhat tricky to cool with a heatsink as it has next to nothing you can attach it to.. and to use a fan in a mic pre isn't really practical unless you have to. I don't have much experience with IC's but I'd guess a heatsink in the NC pins could work in theory, as to how much it dissipates heat is another story.. many recommend copper or similar wishalloy because it has excellent properties for this but in general is considered expensive.
 
aviel said:
1. when no mic is conneted i get crazy oscilations.
2. when mic is connectedm i get high pitch "squeeling" background noise.
Is the input xfmr case grounded?

BTW, I see you use what seems to be a THAT 1646; do you know this circuit needs to be driven by a very low impedance circuit, which the wiper of a potentiometer is not?
I also see you run your stages at 25+mA; is it really necessary?
 
abbey road d enfer said:
Is the input xfmr case grounded?

BTW, I see you use what seems to be a THAT 1646; do you know this circuit needs to be driven by a very low impedance circuit, which the wiper of a potentiometer is not?
I also see you run your stages at 25+mA; is it really necessary?

Case is not but the screen pin is. I can try to ground the case also.
Regarding the That 1646, i didn't actually know that. What are my options in that case? I could use a lower pot, 5k for example..
 
Back
Top