• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Quick question about a microphone pre-amp

What exactly will buffering the ground do?

(mechanical background 🙁)

Also, I don't know how much current the op-amp will need. If it's low, I could use a TLE2426 according to that site, I think.

The microphone cartridge itself needs I = V/R => 9/~1000 ~= 9mA... or is that wrong?

I can't remember any circuit math for the life of me.
 
Mic cartridge spreadsheet says the max current consumption is 0.5 mA.

I wonder what the op-amp needs, then.
 
Well, I know I should stay under 20mA total according to the tangentsoft page. But I don't know what the op-amp needs (the OPA2134, not the TLE2426).
 
Originally posted by: Howard
http://www.linkwitzlab.com/sys_test.htm#Mic
http://www.linkwitzlab.com/images/photos/mic-amp.jpg

Can the virtual ground created by the 2 9V batteries be shared by multiple channels without any adverse effects? What's shown is one channel only.
Click to expand...

Yes, the virtual ground can be shared. However, you will have to make sure they each have their own dedicated ground trace connected at a single point on the PCB to minimize crosstalk. IE, don't daisy-chain the common grounds like what you do with SATA drives.
 
traces.jpg

Will it suffice if I jump the middle two grounds at several locations? Each one is the ground for its own channel, so if they can be shared, I'm going to jump them.

I can't fathom putting multiple leads into the same hole.
 
Originally posted by: Howard
traces.jpg

Will it suffice if I jump the middle two grounds at several locations? Each one is the ground for its own channel, so if they can be shared, I'm going to jump them.

I can't fathom putting multiple leads into the same hole.
Click to expand...

blahblah is talking about using a star ground as opposed to a daisy chain. A star ground is where you run a lead to each of your ground originating from the same terminal (node) on the ground. Daisy chain is where you hook up the ground to the ground of the previous item. Daisy chaining can introduce noise and interference, one that has caused many problems in sensitive equipment that I have worked on. If you are taking the ground off of a bus, then this will be essentially the same as a star ground since you are taking the ground point all from the same node. If you want to be pendantic, you could try to tak the leads from holes as close to each other as possible but I am of the thought that you should have as short of wires as possible and as tidy of a board as possible.
 
Well, I'll do a star ground if I end up getting noise, but this way will be faster for me to set up.

EDIT: Would it be a good idea to flood the ground traces with solder to decrease the resistance?

EDIT2: Heck, I can solder solid wires on top of each trace and solder it to the traces along the length. Should drop the resistance by a heck of a lot.
 
Almost there... Just need the TLE2426, two caps, a few wires, two 9V batteries, the RCA jacks, the mic cartridges (prepped but not yet soldered) and that should be it.

before_TLE.jpg
 
Still need to attach the input/output jacks and the battery +/-, but the power supply section works.

top view

The big caps are Rubycon 50V 220uF, paralleled. The small medium-blue ceramic caps in the middle are 50V 0.47uF, also paralleled.
 
OPA2134 uses 4mA/channel, so 8mA total. Plus 0.2mA/TLE2426

There is no need to flood the copper tracks with solder or solder wires on, their resistance is not significant at this low current level. However, they are fairly fragile over time, oxidation as tiny bare copper tracks so ideally you would put a coat of lacquer over them when the entire thing is finished. With thicker tracks I might've suggested tinning them with solder to prevent oxidation but on boards like that the heat can cause them to delaminate, start peeling off.

I try to avoid looking at point to point wiring so this is just a passing comment, I didn't trace your layout to see if it applies. The two ground paths should not have multiple jumpers all along them necessarily, jumper them so the TLE2426 pin connected to them, is in the middle of the path between them (no other jumpers).

You didn't mention, but I assume this is going into a grounded metal box... it ought to since it doesn't have a ground plane.
 
Rebuilt the thing so it doesn't look like a piece of crap. Only problem is, I can't get enough gain out of the thing, but that's way better than the previous builds where nothing worked at all. I think I'm destroying the preamp circuits inside the cartridges when I'm soldering to them, so I'm going to try using cartridges that have been connected with conductive epoxy.

Might also need to install a DIP socket for the opamp since I'm a little paranoid about that thing being defective... multiple times (don't ask). I didn't want to use a socket because I was planning on using decoupling caps soldered directly to the voltage rail legs of the opamp but now I'm just going to try without them because I'm not using an AC-derived supply anyway.

0
1
2

EDIT: I'm putting it into a plastic box. Everything that needs to be connected to ground can be; why do I need a conductive enclosure? Isn't that going to worsen ground loops?

EDIT2: I think my rubbing alcohol is extremely hard, as in "hard water". I used it to clean off the flux and I got something that looks like hard water deposits. Time for some vinegar, perhaps? Damn stuff is 99% pure, too.

5
IMG_0700.jpg
IMG_0702.jpg
 
According to the schematic drawing posted earlier, http://www.linkwitzlab.com/images/photos/mic-amp.jpg gain is increased by using different values for the 10K resistor in the feedback loop (between opamp output and inverting input), and the 5K pull-down to ground (being a voltage divider). Change their values so instead of a gain of [(10K/5K) + 1] = 3, you end up with what you want instead. Might take a few tries to reach what you want, I'd make the 5K smaller instead of 10K larger if you have the parts to do so.

Decoupling caps are important to use for good sound quality, it is not depending on whether it is a battery or AC source. It is even more important since you have no ground plane as this shunts HF noise. At least think about putting a 0.1uF film or ceramic cap between the opamp +- power pins, if it's not easy to do so with short leads to +/Gnd and -/Gnd (again, I am not looking at the board and traces, just generalizing about the need for cap leads to not be very long). It may not even be stable, may oscillate without the decoupling cap(s) in certain situations.

Your enclosure won't be a ground loop if it is only connected at one point. It is to keep noise out, your amp will amplify everything, including RF noise, even radio stations received over the air, if you don't have shielding or a very good ground plane. When you used the rubbing alcohol did you just pour it or scrub with a toothbrush? It usually gets off pure rosin core flux pretty good unless the flux is cooked too long with the iron, which is much easier to do on a board like you're using because there is so little copper and it's got so many holes it doesn't solder nearly as easily as a normal production PCB with plated holes. If all else fails, you could scrub the areas with a mild abrasive like toothpaste, then scrub with rinsewater and be sure it's dry before powering it again.
 
Originally posted by: mindless1
According to the schematic drawing posted earlier, gain is increased by using different values for the 10K resistor in the feedback loop, and the 5K pull-down to ground (being a voltage divider). Change their values so instead of a gain of [(10K/5K) + 1] = 3, you end up with what you want instead. Might take a few tries to reach what you want, I'd make the 5K smaller instead of 10K larger if you have the parts to do so.
Click to expand...
Yeah, but I think something is defective. When the circuit is plugged into my motherboard's mic-in, I get barely higher output than from my headset which, to the best of my knowledge, does not have any preamplification whatsoever.

Decoupling caps are important to use for good sound quality, it is not depending on whether it is a battery or AC source. It is even more important since you have no ground plane as this shunts HF noise. At least think about putting a 0.1uF film or ceramic cap between the opamp +- power pins, if it's not easy to do so with short leads to +/Gnd and -/Gnd (again, I am not looking at the board and traces, just generalizing about the need for cap leads to not be very long). It may not even be stable, may oscillate without the decoupling cap(s) in certain situations.
Click to expand...
It could be possible that there's noise on the rails, but at this point I'm not worried about it. However, how would I tell if the opamp starts to oscillate?

Your enclosure won't be a ground loop if it is only connected at one point. It is to keep noise out, your amp will amplify everything, including RF noise, even radio stations received over the air, if you don't have shielding or a very good ground plane.
Click to expand...
Fair enough. Right now I don't have any problems that I can see besides the fact that it's just too damned quiet. It's basically worthless right now. If I get noise, I'll add the decoupling caps and the metal enclosure.

When you used the rubbing alcohol did you just pour it or scrub with a toothbrush? It usually gets off pure rosin core solder pretty good unless the flux is cooked too long with the iron, which is much easier to do on a board like you're using because there is so little copper and it's got so many holes it doesn't solder nearly as easily as a normal production PCB with plated holes.
Click to expand...
I scrubbed it with an old toothbrush. There isn't any flux left as far as I can tell, and this is the first time I've encountered this problem. Come to think of it, last time I used the same rubbing alcohol and I didn't notice any deposits.

Could flux be plain white? I'm too lazy to turn on the iron to tell by heating up a joint.
 
Originally posted by: Howard

Yeah, but I think something is defective. When the circuit is plugged into my motherboard's mic-in, I get barely higher output than from my headset which, to the best of my knowledge, does not have any preamplification whatsoever.
Click to expand...

Using a constant tone from the source, measure AC voltage at the opamp positive input pin. Compare to output pin. Is it 3X gain? Note that you may have just made an antenna when you put the probe on the input pin. Remember you have a 10K pulldown at the positive input and something, 200 or 800 ohms (can't read it) between output and jack.

It could be possible that there's noise on the rails, but at this point I'm not worried about it. However, how would I tell if the opamp starts to oscillate?
Click to expand...
You might hear motorboating (sometimes, or always), other HF noise slightly or moreso, it might get hot, or use more current than expected, generally a scope measurement is the way to be sure, or use a time-tested design on quality board, decoupling, and not too fast an opamp with a fair amount of gain and assume it's ok with high probability of being right.

I scrubbed it with an old toothbrush. There isn't any flux left as far as I can tell, and this is the first time I've encountered this problem. Come to think of it, last time I used the same rubbing alcohol and I didn't notice any deposits.

Could flux be plain white? I'm too lazy to turn on the iron to tell by heating up a joint.
Click to expand...

Normally only water based flux leaves white residue, but perhaps if you've used the same rubbing alcohol a lot it is contaminated now, but either way it is only a cosmetic issue. See my prior remark about a gentle abrasive like toothpaste (it seems I never really finish any post, edit and re-editing a lot when I think of more related things to add).

 
Originally posted by: mindless1
Originally posted by: Howard

Yeah, but I think something is defective. When the circuit is plugged into my motherboard's mic-in, I get barely higher output than from my headset which, to the best of my knowledge, does not have any preamplification whatsoever.
Click to expand...

Using a constant tone from the source, measure AC voltage at the opamp positive input pin. Compare to output pin. Is it 3X gain? Note that you may have just made an antenna when you put the probe on the input pin. Remember you have a 10K pulldown at the positive input and something, 200 or 800 ohms (can't read it) between output and jack.
Click to expand...
It's 200 ohms, as per the schematic. I think I'll try using a 1.5V battery for the input, since I only have a +/- 6V supply to play with right now (two 6V batteries in series).

It could be possible that there's noise on the rails, but at this point I'm not worried about it. However, how would I tell if the opamp starts to oscillate?
Click to expand...
You might hear motorboating (sometimes, or always), other HF noise slightly or moreso, it might get hot, or use more current than expected, generally a scope measurement is the way to be sure, or use a time-tested design on quality board, decoupling, and not too fast an opamp with a fair amount of gain and assume it's ok with high probability of being right.
Click to expand...
I see. Unfortunately, I can't check for any of that right now. I'm using the OPA2132PA now which should be a little more resistant to oscillation, apparently.

I scrubbed it with an old toothbrush. There isn't any flux left as far as I can tell, and this is the first time I've encountered this problem. Come to think of it, last time I used the same rubbing alcohol and I didn't notice any deposits.

Could flux be plain white? I'm too lazy to turn on the iron to tell by heating up a joint.
Click to expand...

Normally only water based flux leaves white residue, but perhaps if you've used the same rubbing alcohol a lot it is contaminated now, but either way it is only a cosmetic issue. See my prior remark about a gentle abrasive like toothpaste (it seems I never really finish any post, edit and re-editing a lot when I think of more related things to add).
Click to expand...
Well, my solder is definitely not the water-soluble type. It's RA rosin.

more pics

IMG_0700.jpg
IMG_0702.jpg
 
Not sure what that residue is, I'm guessing either there was a coating on the board that is heat degraded, small amount of ROHS plating on leads was eaten off by the flux, or you have been snorting powdered sugar while you work. Probably just some additive in the flux to make it more active, what brand and part # is it?
 
There's food at my computer desk but not at my bench, hehe.

AIM Solder, SnPb 63/37, Lot# C-44271, RA 3%

Honestly I've never seen this stuff before. Looking back at my old pics, there might have been a hint of it... but nowhere as much as on the board after the last cleaning.
 
Maybe brushing it with alcohol again to dissolve whatever that is, then rinsing with water and shaking excess water off would help, or just ignore it since it shouldn't effect anything except possibly getting lacquer to adhere well if you were going to paint the tracks on the back, but there is no need to do that since those are plated now. Hmm, 11:50 here, in 10 minutes I'll be turning into a pumpkin, good night.
 
Thanks for the help.

EDIT: Replaced the 0.47 uF noise reduction cap on the TLE rail splitter with the recommended 1.0 uF. Installed a DIP socket for the op-amp. Will try an OPA2277. Pics to come.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top