• 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

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Replaced the OPA2132PA with a fresh OPA2277P. No improvement, so I can rule out the opamp as at least the sole problem. I don't think there's anything wrong with my soldering nor my layout, unless I misread the schematic (which is possible). Right now, I think the main contender is the cartridges with fried electrets or preamps or whatever.

I hope I can find somebody with some conductive epoxy, because it sure is expensive.

EDIT: I'm not going to worry about the white stuff because fresh flux made it disappear.
 
You mentioned something previously about using a 1.5V battery after I wrote about measuring the signal levels, what were you doing with that, seeing if it would multiply that to 3 x 1.5 = 4.5V? The capacitor on the input would block DC voltage if the whole amp input was where it was connected instead of at the opamp input pin.

With the cartridge not connected, measure whether sufficient signal strength is coming from it. With it connected, measure whether opamp output pin AC voltage is 3X the input pin voltage. By comparing the voltage of each you should see where the problem lies. If you had to you could always temporarily hook a different source up like an mp3 player, might make it easier to have a constant tone at sufficient volume instead of playing it over speakers loud late at night for the mic to pick it up.
 
I was going to put the battery right on the differential input.

Or I could simply bypass the cap.

This would check that the circuit was laid out properly as well as if the opamp works, right?

Time to burn a 440Hz sine onto a CD, I guess.
 
You can bypass the cap, assuming you have the output hooked up to nothing just measuring the voltage, but what is the exact issue again, that you're sure it's supposed to be louder than it is, or you only expected it to be louder than it is? Do you still have the gain set at 3? Due to the time I'm already done for tonight, good luck.
 
No, I'm certain it's supposed to be louder. The cartridges, when hooked up, should be able to perform as recording microphones. I have to talk loudly near the cartridge when the circuit is plugged into my mic-in (not even line-in).

Somebody who built the same circuit confirmed for me that speaking loudly at arm's length is more than adequate in volume. The gain is as the resistors determine on the schematic.

Why is the gain 3 when Linkwitz says the circuit has a gain of 10dB?
 
Er... I hope my motherboard's line-out had output caps on it, because I just gave one channel the mic cartridge's DC bias (~6.4V) by feeding it to the regular mic-in on the circuit. Seems to be working ok now, though. God damn.

Anyway, at the given volume output, VAC of the line-out was about 81 mV. The VAC of the output from the circuit was 56 mV.

Wonder of wonders.
 
I'm not quite following what you mean, are you saying it's suddenly working properly now, without any changes?

Need more sleep, hope I get the following right-
As for a gain of 3 when the circuit adds 10dB, decibels are a logarithmic value, a ratio of output and input power. You can express the relationship like,
dB = 10 log (Output Power / Input Power)

Your opamp is a voltage gain stage, and we know power (changes) correspond to voltage squared (assuming matched impedance). Since it's a ratio, we don't need exact values for the voltage, can just use 1V for input and 3V for output when we set the circuit at 3X gain.

So a gain of 3X voltage means:
decibels = 10 log (3V^2/1V^2)
dB = 9.5

While I'm thinking about it, it's generally undesirable to pass an audio signal through a tantalum coupling cap. Most would say even an electrolytic is better if you don't have the budget or space for same value film cap. I was referring to the original project parts, but it looks like you did use a film cap for signal coupling?
http://www.maxim-ic.com/appnotes.cfm/an_pk/3171
 
Yes, I originally had a tantalum cap doing the DC blocking at the input, but I changed that to a film cap on somebody's advice.

I'm using a ceramic 2.2 uF on the output, as well. Hopefully I can find enough space to replace those with MKTs.

The output of the circuit is lower than the input, so no, it's not working.
 
Oh, now I understand what you were saying. Maybe you have done it already but I would retrace the circuit on the board, double-check the resistor values, check for cold solder joints and reflow any if they look suspect. You measured that 81mV at the opamp's input pin, not elsewhere, and 56mV on it's output pin?
 
The output was measured between the output after the output cap and 200 ohm resistor, and the virtual ground. The input was measured right from the output from the line-out after it was connected to the circuit (bypassing the cap) so yeah, it was at the input pin.
 
In case anything is lossy or has a bad solder joint, it would be good to measure at the opamp input and output pins, from the top side of the board. Beyond that I am running out of ideas except to double-check the other solder joints, component placement, etc. and measure various points in the circuit to see if anything looks like an unexpected value. Something must be going overlooked unless you are very unlucky and have 2 opamps in a row that are bad.
 
Found the problem. Your advice to check directly at the pins was sound (no pun intended). VAC at input pin was fine, and the VAC at the output pin was about 3* that (around 83mV and 250mV). The VAC at the output RCA jack was anywhere from 0 to 17mV.

It was the stupid ceramic capacitor. For some reason, it just wasn't passing AC. I took it out and moved the signal wire to the RCA jack to the end of the 200 ohm resistor (just as in the schematic) and the output was 3* input.

Except now I need more gain. It's too quiet right now. If I replace the 5K resistor with a 100 ohm resistor, I'll get about 40dB, right?

In the future I think I might put in a SPDT switch to control the gain, because anything good for recording won't be good for near-field speaker testing.
 
If your target is around 4V, that's what you'd have to use. That's a lot of gain for audio with a slow opamp, at this stage some people might use a faster one on a PCB or use two opamps so they're not cutting the resolution so much from the high gain. If the OPA2132 still works, you might want to use it instead.
 
Yeah, I agree with mindless, 40 dB is a lot of gain for this kind of circuit, it is a pre-amp after all. Why don't you get a pot and put that in the feedback loop and adjust the gain on the fly to what you want. Then you can measure the pot's resistance and replace it with the appropriate resistor.

Going back to the fact that this is a pre-amp, aren't you going to be feeding this into an amplifier or gain stage?
 
Originally posted by: Born2bwire
Yeah, I agree with mindless, 40 dB is a lot of gain for this kind of circuit, it is a pre-amp after all. Why don't you get a pot and put that in the feedback loop and adjust the gain on the fly to what you want. Then you can measure the pot's resistance and replace it with the appropriate resistor.

Going back to the fact that this is a pre-amp, aren't you going to be feeding this into an amplifier or gain stage?
Click to expand...
Nope. Output goes to line-in and gets recorded.

EDIT: I think I will try using a pot, though. Unfortunately, I only have log pots on hand so I'll have to pick up a linear pot. I could do it with a log pot, but a linear pot would be nice to have anyway.
 
Originally posted by: Howard
What's the difference between the 2277 and the 2132?
Click to expand...

Higher gain bandwidth product and slew rate among others, it's a better opamp for audio and the higher your gain the more the gain bandwidth product difference helps. OPA2277 is really below the grade of opamps most would want to use for audio. At first I'd thought it was OPA2227 which is closer to 2132.
 
Oh. Right.

Bah. Looks like I'm gonna put in a 2132, then.

EDIT: I might not need as much as 40dB of gain, but let's say I want to move to the OPA2228 in the future. I'd want to keep the gain above 5, which is definitely doable; the "higher" resistor would just be 1/4 the value of the feedback resistor, while the "lower" resistor would be whatever it takes for recording (in my application) gain. I read on the tangentsoft site that high-frequency oscillation (from the finicky behaviour of faster op-amps) can sometimes be remedied by increasing the gain. This is perfect for the higher-frequency 2228, right?

I'm going to add the two decoupling caps and two ferrite beads on the outputs just in case, once I get everything sorted out. Do I simply thread the wire through this, or do I need to wrap it around?

EDIT2: Turns out I need about 53dB of gain.
 
I don't know what effective frequency range you'll need for the beads IF they're needed, but that one you linked seems quite large. You probably would need at least one tight wrap around it otherwise the wire isn't going to be in close contact with something having such a difference between wire diameter and ID. I'd look for something closer to 5mm OD and ID only enough to pass the wire through.
 
How are they most effective? Inside the feedback loop, or just inside a few loops of the wire that goes to the signal conductor of the jack?
 
They're usually on the output, when it comes to the feedback loop you want as short and few components as possible.

I suggest trying whatever you end up doing without these beads, adding them only if needed. Also, if RF pickup seems to be a problem, consider putting a small, roughly 200pF capacitor (film or mica) between the noninverting and inverting opamp inputs. However, as always if we want to keep drifiting down the path of what is ideal vs reasonable per the budget or goal, the sky is the limit. It's really your call, when you have a result that is acceptable per the use.
 
A ferrite bead in this application would be more for prevention of oscillation rather than RF pickup or whatnot, right?

Let me show you what it looks like right now. The black circle to the left of the left battery at the end of the green wire is the cartridge.

IMG_0714.JPG
IMG_0715.JPG

I'm not dissatisfied, that's for sure.
 
Yes, I just can't predict when a board like that will encounter problems, at what opamp frequency it would start.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top