The Passinwind Open Source Preamp

[Jeff Siddall]

I see there is an OPA2206 in SOIC8 for around $3. Looks to be pretty much identical...

Hmmm... I haven't even looked at that one, but literally just yesterday I put in an order for a few OPA2205ADR. They have dropped to about half their original price ($6.35 CAD) and I needed some other stuff from Mouser anyway so pulled the trigger. They are backordered, so not sure when I will actually get them.

 

[Dadum]

Has anyone had success using ferrite beads at the input of a stomp box preamp? (I think it’s a different situation than the in-guitar preamp boards). Mouser sells them and you can buy just one or two of them at a time. They can be somewhat long so some rearranging of stuff may be necessary.

I any case, if someone did have success, I would appreciate knowing the model number. At the moment, the Fair-Rite type 31 ferrites seem interesting.


Sorry if this was discussed before in this thread.

 

[dcoyle]

This is a link that Charlie sent me in March, c/o Talkbass, that I didn't see till now.

It describes an overview of the circuit and has graphics of what it does. Might be useful if you've just joined the thread.

Invalid Link Removed

 

[Jeff Siddall]

I see there is an OPA2206 in SOIC8 for around $3. Looks to be pretty much identical...

Interestingly, for Mouser at least, the price of the 2205 and 2206 are within about $0.10 of each other

For anyone observing, the difference between the two amps is:

The OPA206 and OPA2206 are related devices with the same op-amp core, but with the added feature of input overvoltage protection ±40 V above the supplies​

 

[Sanj]

@Passinwind , thank you for contributing this project. I haven't built any musical electronic gear in at least twenty years, but the ready availability of your PCB inspired me to dive back in. My goal was to have a standalone HPF more for speaker protection than anything else, hopefully this is a valid use case for your HPF design.

I ended up repurposing an old MXR stomp box (for the second time!) to use as the enclosure. The PCB is attached to a piece of thin styrene with double-sided foam tape, which is in turn attached to the back of the frequency pot. I designed a cover sheet in Illustrator and printed it on a photo printer, with a thin plastic laminate covering. For a first attempt I'm pretty pleased with the results.

I was hoping to convert one of the remaining boards for use with a true bipolar power supply for use inside my rack-mounted preamp by eliminating the Vref circuitry, and substituting preamp V- for ground, and GND for Vref. Your thoughts?

 

[Passinwind]

@Passinwind , thank you for contributing this project. I haven't built any musical electronic gear in at least twenty years, but the ready availability of your PCB inspired me to dive back in. My goal was to have a standalone HPF more for speaker protection than anything else, hopefully this is a valid use case for your HPF design.

I ended up repurposing an old MXR stomp box (for the second time!) to use as the enclosure. The PCB is attached to a piece of thin styrene with double-sided foam tape, which is in turn attached to the back of the frequency pot. I designed a cover sheet in Illustrator and printed it on a photo printer, with a thin plastic laminate covering. For a first attempt I'm pretty pleased with the results.

I was hoping to convert one of the remaining boards for use with a true bipolar power supply for use inside my rack-mounted preamp by eliminating the Vref circuitry, and substituting preamp V- for ground, and GND for Vref. Your thoughts?
View attachment 5245875 View attachment 5245876 View attachment 5245877 View attachment 5245878

Cool, good job! I do simple laminates like that a lot, especially for prototypes.

I'm just starting to play catchup after a fairly long computer replacement/migration adventure and will get back to you on the bipolar conversion idea ASAP. Looks like I'll need to sort through the last few pages later today as well.

 

[bowthing]

Interestingly, for Mouser at least, the price of the 2205 and 2206 are within about $0.10 of each other

If I remember right, that's comparing the VSSOP 2205 to the SOIC 2206. The 2205 in SOIC is the expensive one.

 

[Passinwind]

@Sanj : your idea looks doable, depending of course on how sharp your board rework skills are. Probably the most challenging part would be floating the pin 4 ground on both opamps (top left pin in screen grab) and then hardwiring in the V- rail:

Those are 20mil traces, so not too hard to cut pretty cleanly with a sharp X-acto knife. I'd probably install the opamps first and then cut the traces, otherwise the pin 4 pads may want to lift during opamp installation. You could kludge the primary Vref filter cap to serve V- instead, and lose the supply divider resistors. I can walk you through the rest in PM if you like.

I do also have a bipolar HPF board design, but it's currently attached to a one band parametric EQ as well. My tester didn't show the increased headroom I hoped for though, and the second build test has been languishing at the back of my work queue for over a year now, I think. I did a matching power supply daughter board to derive +/- 15V from a standard 9V power supply, which seems to work just fine. The HPF/PEQ board is a lot bigger than the standalone HPF board, naturally. If and when I get this working to my satisfaction it would be freely shared too, FWIW.

 

[Passinwind]

I would love to get my hands on the graphic files to get the V4 boxes drilled and painted, it might be a bit expensive to send them to the US and then to Australia to get them done.

Happy to accommodate you if I can Mick, as I will need to work up Tayda format files anyway at some point. Not sure if V4 fits comfortably in a 1590BB2 enclosure, and IIRC the next size jump at Tayda is all the way up to 1590XX, which is definitely bigger than needed. The V4 graphics I did for @dcoyle are meant for always-on applications, would you prefer something with a footswitch though?

 

[Sanj]

@Sanj : your idea looks doable, depending of course on how sharp your board rework skills are. Probably the most challenging part would be floating the pin 4 ground on both opamps (top left pin in screen grab) and then hardwiring in the V- rail:

That's a method I hadn't considered, though I'm sure the rework is doable.

My original thought was to connect preamp V+ to board V+, preamp GND to board Vref (using the + pad for C4), and preamp V- to board GND. R11, R12, C3 and C4 would be omitted, as the preamp supply should already have good regulation, and the hardest part would be attaching the negative side of C11 and C12 to Vref instead of GND. The biggest concern would be that the ground plane is now at V- instead of ground. Proposed changes in red below.

 

[Passinwind]

That's a method I hadn't considered, though I'm sure the rework is doable.

My original thought was to connect preamp V+ to board V+, preamp GND to board Vref (using the + pad for C4), and preamp V- to board GND. R11, R12, C3 and C4 would be omitted, as the preamp supply should already have good regulation, and the hardest part would be attaching the negative side of C11 and C12 to Vref instead of GND. The biggest concern would be that the ground plane is now at V- instead of ground. Proposed changes in red below.
View attachment 5246912

Please don't post my schematics in this thread in the future, especially ones with your own mods, as it makes version control completely impossible for me. I can put up verifiably workable mods on the GitHub repository and that way I can always edit or delete them as needed. No worries this time though, let's get this figured out for you as best we can.

Running the ground plane at V- instead may work OK, in all honesty I have no idea which of our proposed methods might be easier for you. But as always, I'll be happy to help out where I can!

 

[Sanj]

My apologies, I'll avoid that in the future. I appreciate the feedback.

One further question -- C2 is attached to GND in the original, if running bipolar supplies would it be OK to leave it as is, now at V-?

 

[Passinwind]

My apologies, I'll avoid that in the future. I appreciate the feedback.

One further question -- C2 is attached to GND in the original, if running bipolar supplies would it be OK to leave it as is, now at V-?

That's a very good question. In some cases I've seen problems from running that node at Vref rather than ground, and I found a number of contradictory online posts by actual EEs about which way is correct. But if necessary you could just omit C2 and run the top of R3 to ground, I reckon. I'll play around a bit with my model and see what I can suss.

 

[Jeff Siddall]

If I remember right, that's comparing the VSSOP 2205 to the SOIC 2206. The 2205 in SOIC is the expensive one.

No, actually, both were the SOIC8 version:

Link Removed

 

[Passinwind]

That's a very good question. In some cases I've seen problems from running that node at Vref rather than ground, and I found a number of contradictory online posts by actual EEs about which way is correct. But if necessary you could just omit C2 and run the top of R3 to ground, I reckon. I'll play around a bit with my model and see what I can suss.

@Sanj : my model thinks terminating C2 to V- will still work fine, so it's probably at least worth a try if you decide your proposed mods are preferable for whatever reason.

 

[Sanj]

Great, thank you! I'll keep you posted.

 

[Jeff Siddall]

That's a very good question. In some cases I've seen problems from running that node at Vref rather than ground, and I found a number of contradictory online posts by actual EEs about which way is correct. But if necessary you could just omit C2 and run the top of R3 to ground, I reckon. I'll play around a bit with my model and see what I can suss.

I would guess the reason to have C2 go to GND rather than Vref is that in a non-inverting configuration like the first stage the op-amp inputs and R3 are going to be moving by the same amount as the input signal. R3 is only 30K so relatively low impedance compared to Vref itself, which means C2 will fairly strongly couple the input signal to Vref. Moving it to V- gives a low impedance path to a rail rather than Vref, which should make Vref more stable.

The obvious downside to having C2 go to GND is that it needs to be charged to Vref voltage. If it is discharged and power is applied, the inverting input is going to be pulled down to the GND rail and thus the op-amp output will be pulled to the V+ rail. That would normally show up on the output as a loud "pop" but given the rest of the circuit is an LPF it should damp that somewhat.

When power is removed from the circuit, C2 will then need to be discharged through either the ESD protection diodes on the op-amp (if there are any) or through the feedback resistor R4 depending on how the output of the op-amp swings when powering down. Either way it shouldn't cause any issues.

With a dual supply Vref becomes a low impedance GND so you could probably move R3 there.

Does that make sense?

 

[Tassieviking]

The V4 graphics I did for @dcoyle are meant for always-on

I would be happy with the same graphics you have used so far, I don't think there is a need for a stompswitch for most people.
I think I used a switch in the V2 out of habit since it is a stompbox.
Thanks Charlie

 

[Passinwind]

I would be happy with the same graphics you have used so far, I don't think there is a need for a stompswitch for most people.
I think I used a switch in the V2 out of habit since it is a stompbox.
Thanks Charlie

Cool, that's my personal preference as well. There's still the option of using the second toggle for a bypass if desired, among other possible functions.

@dcoyle: I put up a share on OSHpark for the simple gain board, you can readily omit the trimpot and just run it as a unity gain buffer if you like. This is rev4, so it's pretty well vetted by now.

Most of these boards shares can be downloaded as standard Gerbers and ordered at other fab houses, but some of older ones for daughter cards may need a re-do since OSHpark used to accept native KiCAD .brd files in lieu of Gerbers.

 

[Jim C]

I would be happy with the same graphics you have used so far, I don't think there is a need for a stompswitch for most people.
I think I used a switch in the V2 out of habit since it is a stompbox.
Thanks Charlie

I have found that I really didn't need a stomp switch as it is almost always on. I do like the mini in/out switch so you can quickly determine the amp EQ vs. the pedal.