Topic: Homemade Pre-amp Circuit Help

[joeldotc]

Hey folks. I have been looking into designing and building my own preamp which runs off of a 9 V battery, takes 3.5 mm stereo jacks, and has adjustable gain.

So far, I've come up with the attached circuit and the associated PCB layout which is centered around the TI OPA1612 op amp. It would have the input/output jacks attached at the input/output, and the LG1 and RG1 pads would lead to a two channel rotary switch providing 0, 5, 10, 20, 25, 30 dB of gain (the range could be adjusted to higher/different levels, but the corresponding capacitors would need to be adjusted to ensure a flat frequency response across the gain range).

I was wondering if anyone here who has experience with small circuit audio/preamp design would be able to chime in on the design before I send the drawings off to get printed. I've never designed a circuit before, and this is also my first time using circuit/PCB design software. I think it'll only cost a few bucks per board to print out and make, so if the easiest way to test it is by doing that then I don't mind at all.

I've ran the circuit through PSpice and the results seem promising, but I am in no way well educated in circuit design. I attached a couple screenshots of the PSpice simulation results for reference. I hope I'm not wayyy over my head and that this isn't just some pipe dream.

Any pointers, tips, or comments would be greatly appreciated!

Thanks!

[Sebastian]

It looks like it could work. However, I have a few suggestions.

• It looks like you're biasing the input signal via the R3/R4 voltage divider to bring it over 0V. After that voltage divider you have another pair of resistors (R5/R9) to keep R/L signals separated. Instead of this, I'd just run separate 100kΩ/100kΩ voltage dividers directly off the input voltage for both channels. It's the same amount of resistors and it's a much cleaner design.
• On the PCB, you have C13 as a bypass capacitor for the opamp (since you placed it very close to the IC). That's a 10uF as per your schematic. However, the datasheet of your opamp asks for a 0.1uF bypass cap. So you might want to keep C13, but place it somewhere near the power input, but add another 0.1uF ceramic cap (X5R or X7R) where C13 is currently placed on the PCB.
• Can you make C4/C9 a little larger? It looks like your 3 dB corner frequency is at around 25 Hz for 30 dB of gain. I'd want that to be below 20 Hz. But it probably won't matter in real-world scenarios...
• You can probably make C6/C12 way smaller. Your input bass cut filter has a higher 3 dB corner frequency anyway. And it's sometimes a good idea to cut some of the very low frequencies as they mostly contain mic handling/wind noise.
• You can completely remove R7/R11 and R8/R12. You probably have R11/R12 in there to limit current. But these resistors will also form a voltage divider with the recorder's input load, essentially decreasing your output voltage which you've just increased (and that was actually the whole point of making a pre-amp). Not by much, but it's still unnecessary. Also, your decoupling caps will block DC voltages, so the DC voltage on the opamp's output will always be 0VDC (or whatever output offset voltage it puts out).
• You don't need R7/R11 at all. Your recorder has an input impedance anyways, so just make an assumption about that (it's usually somewhere between 2kΩ and 10kΩ) and choose your decoupling caps accordingly. If you choose to keep R7/R11, at least increase their resistance to 10 kΩ. If you keep them at 3.3 kΩ and your recorder's input impedance is 10 kΩ, both of these resistances in parallel will result in a 2.8 kΩ impedance. And that would increase your output bass cut filter's 3 dB corner frequency unnecessarily.

You might want to take a look at this application note from Texas Instruments. It describes all the necessary steps and calculations to design a non-inverting mic preamp circuit: https://www.ti.com/lit/pdf/sboa290

I hope this didn't come off as too critical. You're almost there! However, be prepared that your final preamp will behave slightly different from your spice simulations. All the parts you're using have their tolerances - especially the capacitors. This is quite different from your ideal spice parts.

[joeldotc]

Sebastian, thank you! You weren't too critical at all, that is great help - thanks for the detailed reply. If you're ever in Ottawa I owe you a coffee or tea or something 

I followed all of your advice and used that TI application note for the required calculations. The response looks much flatter and better in PSpice now.

I've attached new screenshots of the circuit and PCB (which is practically the same size as the 9V battery now!). I changed some of the reference names of the components, but the schematic is more or less exactly as you described it. Let me know if you see anything else I've done wrong.

Thanks again!

[Sebastian]

Yes, it looks a lot cleaner now. Just one more thing: What do you need C1/C6 for? Are you trying to smooth your audio signal?

[joeldotc]

Yes, it looks a lot cleaner now. Just one more thing: What do you need C1/C6 for? Are you trying to smooth your audio signal?

I think you're right that C1 and C6 are a bit redundant. They could help filter out the DC from the input signal, but I don't think they're required since C2/C7 are also doing that job and can probably do it alone. I will remove those as well.

[Sebastian]

I think you're right that C1 and C6 are a bit redundant. They could help filter out the DC from the input signal, but I don't think they're required since C2/C7 are also doing that job and can probably do it alone. I will remove those as well.

You're deliberately applying DC voltages to both sides of  C2/C7, so there's no need to "filter out" anything.

At this point, I thing you're good to get your PCBs made and to test your circuit with real-world parts. Have fun sodlering!

[joeldotc]

You're deliberately applying DC voltages to both sides of  C2/C7, so there's no need to "filter out" anything.

At this point, I thing you're good to get your PCBs made and to test your circuit with real-world parts. Have fun sodlering!

Ah you’re right.

Again, thank you for the help! I really appreciate it. Gonna get the boards printed and order some components now

[weroflu]

i use tle2426 rail splitter for  virtual ground.

what are you using for the gain switch?

i have lots of completed schems/ layouts if you want.

[joeldotc]

i use tle2426 rail splitter for  virtual ground.

what are you using for the gain switch?

i have lots of completed schems/ layouts if you want.

The rail splitter is a good idea!

For the gain switch I was planning on just using a small rotary switch with a nice low profile knob.

I would love to see your completed schems/layouts!

[Chilly Brioschi]

Do I see pads for a soic-8 package for the op amp?
Are you using a socket to permit rolling, at least at the onset and during testing?
Or is this much smaller than that?

Kudos for home brewing/experimenting!

Any suggestions for breadboarders or lesser hackers?
That pwb (PCB) looks like it will be field-ready in short time!

[joeldotc]

Do I see pads for a soic-8 package for the op amp?
Are you using a socket to permit rolling, at least at the onset and during testing?
Or is this much smaller than that?

Kudos for home brewing/experimenting!

Any suggestions for breadboarders or lesser hackers?
That pwb (PCB) looks like it will be field-ready in short time!

Thanks 

It is an soic-8 op amp! I'm not using pins/sockets, so if I wanted to swap around different op amps, they would just need to be soldered in/out which isn't too big of a deal for me. I really like how the OPA1612 looks from its spec sheet, but it would be quite interesting to see how other opas would work too.

I've only ever done minor circuit design before. This was really my first time dipping my toes into any significant electrical design. I found that KiCAD is amazing to use and really recommend it. It only took me about an hour of youtube videos and a few hours of messing around to understand how pcb design works. I really recommend just downloading the software and to try copying an existing Texas Instruments design (like the one Sebastian sent) on it - that'd probably be the fastest way to learn.

I have an engineering degree (mechanical, not electrical), but I'm now a high school teacher on summer break so I've just been spending my free time scratching my engineering itch with this project!

I also chatted with weroflu and decided to change my design - I made it more similar to weroflu's, and now it'll fit inside an altoids which solves the need of making a custom enclosure for what's really a prototype board that I'm sure I'll tweak and change lots after I build and test the first version.

Once it's built I'll post updates here!

[Chilly Brioschi]

Browsing around the 'Net, I found this:

https://www.thatcorp.com/datashts/AES129_Designing_Mic_Preamps.pdf