dual Vs single opamp for stereo mic preamp

[spaceludwig]

Thanks for the advice, PRR. I will follow your recommendation and add a stereo mini-plug beside the XLR output for added flexibility. I am trying to complete a functional  single channel version - which I will use with my shotgun mic - before I commit to a stereo version.

I've etched a PCB of the latest version and I am yet again having issues when I power it up. I end up with most of the voltage at the positive pin rather than evenly split (30V+ to 2V-). I don't understand why as the circuit is the same as the previous ones, which all worked, minus the output transformer. When I disconnect the circuit from the power supply I get +/- 16V. I have also removed the filter caps at the supply pins to make sure it is not an issue with them. I know I'm probably committing some very elementary error, but it is not obvious to me. Needless to say that it is extremely aggravating trying to find a problem that previously wasn't there.

 

[pstamler]

When I've had problems like that, it was usually because a ground wire wasn't connected properly.

Peace,
Paul

 

[spaceludwig]

When I've had problems like that, it was usually because a ground wire wasn't connected properly.

The PSU ground is connected to the circuit ground. The only other ground wire is from the XLR input and seems to be fine when I do a continuity test.  When I return perhaps I'll post a picture.

 

[spaceludwig]

OK, I've got to where I am getting the right voltage readings at the supply pins (4 & 7) when there is no opamp in the dip8 socket. However, the moment I drop an opamp in there the voltages start leaning towards the positive. I am including some pictures in the event the more observant posters reading this thread can detect the mistake(s) I am making.

Component side, supply caps removed ***NOTE: The dip8 socket is not oriented properly (tab indicating pin 1 should be facing other side) but I have been inserting the chips in the correct direction.

PCB side (negative rail on top)

Splitter circuit which is basically two 4.7k Resitors with two 220uF caps

The power supply that feeds the whole shebang

I am going crazy trying to locate the problem. I thought it might be the supply but I've used it as is successfully to power the various breadboard versions of the circuit and the two previous PCB's I made. Any help would be greatly appreciated.

 

[pstamler]

What's the "splitter circuit" about? Maybe a schematic of the whole thing would be useful.

Peace,
Paul

 

[spaceludwig]

What's the "splitter circuit" about?

Just to split 32 volts to dual rail, i.e. +/- 16V

 

[spaceludwig]

Here is the circuit to split the voltage from the PSU in half

And the preamp ciruit itself

I'm baffled by this. I just don't understand how the circuit can work one day and then the same circuit not work the next. There are no shorts anywhere that I can find using a continuity check on my multimeter.

 

[pstamler]

Have you tried pluging a different opamp into the socket?

Peace,
Paul

 

[spaceludwig]

Yes, same results. I recently received some tle2426 virtual ground rail splitter chips I ordered to make the power supply so I'll try to make another voltage divider circuit tomorrow and see if that changes anything  or inspires a fresh approach to solving this issue. I could always just etch another board but it irks me to no end not knowing why i'm having this problem. One of the aspects of electronics i like so much is that circuits are supposed to behave predictably so most problems can be traced logically. Making  a new board would feel like throwing out a computer that keeps crashing because a driver is not installed properly.

Anyhow, thanks for your help and interest. I know there are much more interesting threads in this forum, so any guidance is much appreciated.

 

[Andy Peters]

Yes, same results. I recently received some tle2426 virtual ground rail splitter chips I ordered to make the power supply so I'll try to make another voltage divider circuit tomorrow and see if that changes anything  or inspires a fresh approach to solving this issue. I could always just etch another board but it irks me to no end not knowing why i'm having this problem. One of the aspects of electronics i like so much is that circuits are supposed to behave predictably so most problems can be traced logically. Making  a new board would feel like throwing out a computer that keeps crashing because a driver is not installed properly.

Why not power the op-amp from the single rail you have (VCC (supply positive) to op-amp supply positive, and GND (supply return) to op-amp supply negative), and bias the inputs up to VCC/2?

-a

 

[robovox]

Is VGND of the breadboarded splitter circuit tied solidly to GND1 of the etched board?

 

[spaceludwig]

and bias the inputs up to VCC/2?

I didn't know you could do that and, therefore, don't know how to do that. If you could provide a quick schematic or link or elaborate I would be very happy to review the info.

Is VGND of the breadboarded splitter circuit tied solidly to GND1 of the etched board?

I don't know if it qualifies as "solid" but I used a jumper with alligator clips, one end clipped to the ground plane of the PCB and the other to the virtual ground on the breadboard. I measure different voltages when not in use (connection to ground) so I'm assuming it works. This worked fine when I powered the previous versions of the same circuit.

 

[Andy Peters]

and bias the inputs up to VCC/2?

I didn't know you could do that and, therefore, don't know how to do that. If you could provide a quick schematic or link or elaborate I would be very happy to review the info.

Use a simple resistive voltage divider to create VCC/2. Buffer it with a non-inverting op-amp. That op-amp's output is your half-rail bias.

Looking at your schematic, AC couple the input (say, a 47 uF cap) and instead of grounding the bottom of R2, connect it to the bias. Also, instead of grounding the bottom of R5, connect that to the bias.

That's it. Of course the op-amp output swings about VCC/2, which is fine, and your DC block at the output handles that.

-a

 

[spaceludwig]

Use a simple resistive voltage divider to create VCC/2. Buffer it with a non-inverting op-amp. That op-amp's output is your half-rail bias.

Looking at your schematic, AC couple the input (say, a 47 uF cap) and instead of grounding the bottom of R2, connect it to the bias. Also, instead of grounding the bottom of R5, connect that to the bias.

That's it. Of course the op-amp output swings about VCC/2, which is fine, and your DC block at the output handles that.

-a

Thanks, Andy. I'll try that this weekend. Perhaps I'll upload a schematic first to make sure I understood properly.

 

[ricardo]

and bias the inputs up to VCC/2?

I didn't know you could do that and, therefore, don't know how to do that. If you could provide a quick schematic or link or elaborate I would be very happy to review the info.

The circuit I posted does exactly this.  It is a very low noise amp to be powered from a single 9V battery.

 

[spaceludwig]

The circuit I posted does exactly this.  It is a very low noise amp to be powered from a single 9V battery.

If you have a link to your site or thread I'd be more than happy to read it and try it after I've gotten this circuit to work.

 

[ricardo]

If you have a link to your site or thread I'd be more than happy to read it and try it after I've gotten this circuit to work.

http://tech.groups.yahoo.com/group/micbuilders/
Look in my Files directory.  You have to join.

Maybe read it before further work on your circuit cos it addresses many of your problems.

 

[spaceludwig]

Thanks, Ricardo.

 

[spaceludwig]

I finally seem to have resolved the issue with the preamp. Seems to have been a combination of both the hookup of the gain pot and perhaps some leaks in the traces things seemed to have gone to normal after I ran a nail between the traces to cut through any potential shorts.

I will post them shortly.

Cheers

 

[spaceludwig]

Hey everyone,

So unit now works. I've been working on the power supply, the schematic of which I'll upload later tonight.

Question: Is there any advantage/disadvantage to using a lower value for the gain pot? i.e. 10K or even 1K rather than 20K? Would it impact the output impedance?