The Passinwind Open Source Preamp

Discussion in 'Pickups & Electronics [BG]' started by Passinwind, Dec 25, 2016.


Tags:
  1. OJ Dorson

    OJ Dorson

    Jun 24, 2020
    Having read this thread many times and being super impressed and inspired, I wanted to share the filters I ended up going with for my Ibanez Musician. They are the Distillers listed here: JTEX Electronics | Jerry's Awesome Gadgets

    I went with one per pickup and was really happy with them. They're described as improved clones of the Alembic filters and with the Q switches I think he's really nailed it.
     
    Passinwind likes this.
  2. Passinwind

    Passinwind I know nothing. Commercial User

    Dec 3, 2003
    Columbia River Gorge, WA.
    Owner/Designer &Toaster Tech Passinwind Electronics
    Update:

    Looks like I'm finally more or less caught up on actual paying work for a week or three. I've built up a couple of things using the new DKRed boards, which went quite smoothly:

    NewPreamps.png

    So I think it's time to revisit the standalone LPF boards from earlier this year and see if I can find out what's creating the instability issues. I'm tempted to just roll an all SMD version, but somehow baby steps seem like a better call! :cool:

    I've also put up OSHpark shares for the mini mixer and the simple gain stage, which are both fully vetted now. I'll get to doing the documentation ASAP.
     
  3. OJ Dorson

    OJ Dorson

    Jun 24, 2020
    Okay... read through the whole thing again....

    Here's some questions...
    When there are two pickups which sound differently (neck/bridge) and have different resonant frequencies, wouldn't a freq/resonance filter for each pickup be the ideal setup?

    The resonance knob - seems to be a variable Q knob that changes the amplitude of the boost at the frequency set by the frequency knob. Is that right? What is the total boost when the resonance knob is "at 11"?

    If I were to describe the thing that is built using the board and components listed in post #1, would it be correct to call it a bass preamp with Bass and Mid cut/boost, and a LPF with adjustable Q with switchable Mids at *which frequencies?
    I've seen a dozen different boards in this thread... which board do we end up with? If I'm confident in my SMT soldering capabilities, is there an option for the "tiny board"?

    What frequency is the Bass knob centered at? (forgive me... this is probably listed but I didn't catch it)

    I'd like switchable mids at 350/800/1.2k... At least, I think I would... thoughts? I'd like to be able to boost the "low-mids", regular "mids", and "high-mids (before the LPF territory)" - what frequencies would I target?


    I think that I would love such a preamp... I'm anxious to order it and get started.
     
  4. Passinwind

    Passinwind I know nothing. Commercial User

    Dec 3, 2003
    Columbia River Gorge, WA.
    Owner/Designer &Toaster Tech Passinwind Electronics
    As always, that depends on your priorities. My goals were: DIY friendly, intuitive for as many players as possible, moderate current requirements, and no more than six knobs. And I purposely avoided cloning any of the classic existing filter preamps, because they tend to be pretty pickup-dependent and I was shooting for a more generic and lower cost approach. Beyond that, we generally want to neuter the difference between the pickups anyway for this exercise, the less character they bring to the party the better really. One more advantage of the approach I took is that it works in a pedal with no special wiring required on the bass.

    "At 11" in the very first graph = +11dB at the freuency of interest, the Y axis is gain in dB. Total gain can be up to ~12dB or so (there's a slight boost at the nominally flat setting), but I've found +10dB to be about the usable limit for the bass guitars I've tried with this simple LPF format.

    At this point there's still just the original through hole board as seen in the Wiki. If and when I get the standalone variable resonance LPF done to my satisfaction I'll work on more advanced, smaller versions of both that and the three bander. But it's always been my hope that others would jump in with some board designs, although after five years I'm not all that optimistic! ;)

    Shelving controls don't really have "centers" per se, the peak frequency changes with boost and cut settings and the filter slope does as well. IMO a reasonable approach might be to spec the halfway point along the slope at max boost, let's call it 180Hz. So here's a sweep of bass settings from max cut to max boost:

    OSBassSweeps.JPG

    As an aside, I find the trick to be to decide what boost level you are are actually likely to be using, and with that amount set the bass peak frequency at whatever the sweet spot is for your particular playing situation. For me as primarily a fretless four player looking for an upright-ish vibe, +6-8dB ~42-46Hz tends to work pretty well as the desgn target. If I want that to occur at a 3 oclock setting I need to put the peak at full boost (+12dB) at ~30Hz, as seen in the graph. And where the boost starts happening at an audible level matters too, in this case we could say that's something like +3dB at ~300Hz when bass is fully boosted. This all speaks to "play feel" and interactivity, and at some point that gets hard to talk about effectively.


    Those are decent choices and are certainly doable, but personally I often tend to prefer 400/700/1.6K as solid "money zones." 400 interacts mostly with the bass control, 700 with both bass and the LPF, and 1.6K mostly with the LPF.

    In all honesty, I'm just not sure how many of the more advanced designs will ever get to sharing status. It's a lot of work for something I don't even use or prefer for myself! So at this point I think trial pedal builds of the original design are probably still your best option.
     
  5. Glad I held off a wee bit on putting in my Oshpark order!

    Thanks for the update and the new kit.

    Lots to digest in your latest post above, I'll do my best to do so.
    Mm goi sai, sifu.
     
    Passinwind likes this.
  6. Passinwind

    Passinwind I know nothing. Commercial User

    Dec 3, 2003
    Columbia River Gorge, WA.
    Owner/Designer &Toaster Tech Passinwind Electronics
    Lots of typos too, even though I let the post simmer overnight before putting it up here. Typing has been the hardest part of this whole adventure, truthfully.
     
  7. I used to get paid to look for typos, when I first joined TB it was still so ingrained, but TB was/is rife with 'm, what with ESL etc...
    That's a while ago now (getting paid and joining TB);
    I didn't notice any typos, prolly will now that yerv menshunned it. :laugh:
     
    Passinwind likes this.
  8. OJ Dorson

    OJ Dorson

    Jun 24, 2020
    That's understandable - thank you for sharing that.

    Reading this it seems as there are still all these possibilities... If I put your board together with the parts list you supplied, do I get +10dB of boost with the Response knob at Maximum?
    What parts would I change if I want, say, 18dB of boost with the response knob at Maximum?

    How do you feel about this description - would you say it's accurate?

    That graph was really helpful - Thanks. I think this gives me the information I was looking for to understand this.

    If I build the board with the parts as supplied, will the switch use the 400/700/1.6k centers for the Mid control?

    Lastly... can I assume that the "mini mixer" you mentioned about is similar to the Noll Mixpot, or an active pickup blend? I recently purchased a Noll Mixpot and it's good, but I'd much rather build my own. Would you mind detailing what your "mini mixer" is and pointing me to the board and components?
     
  9. Passinwind

    Passinwind I know nothing. Commercial User

    Dec 3, 2003
    Columbia River Gorge, WA.
    Owner/Designer &Toaster Tech Passinwind Electronics
    A little over ten dB as spec'ed. 18, no can do, we're already pretty near the instability zone at +12dB or so. That's a disadvantage of using the simpler one opamp filter format and a number of parts without super tight tolerance specs. I'm wrestling with this today actually, hopefully I'll have better insight shortly. The original version oscillated like crazy, its a very tweaky circuit.


    You can call the midrange a boost/cut bandpass filter, and yeah to the rest of that.

    Mid switching is an option beyond the basic BOM and there's no set default tuning really. There's room there for a better switching arrangment though, one of many rabbit holes perpetually in the queue.

    Yep, similar to the Noll, but as basic as can be. One opamp, two inputs, one master gain trimpot, one output. Documentation pending, as I only built the first one a couple of days ago. The various public board shares I've done, including the mini mixer, are here: OSH Park ~ Shared Projects by PW3B-LPF-OS
     
  10. OJ Dorson

    OJ Dorson

    Jun 24, 2020
    Awesome... that's really all very helpful. I should have some spare money to order everything soon.

    The mixer - so, no adjusting for pickups with different output levels, then?
     
  11. Passinwind

    Passinwind I know nothing. Commercial User

    Dec 3, 2003
    Columbia River Gorge, WA.
    Owner/Designer &Toaster Tech Passinwind Electronics
    Its a dead simple cookbook inverting amplifier/mixer circuit, you can either run volume pots into it or change gain per channel via the input resistors. Or you can use one or two of the matching simple gain boards as front ends, or a FET buffer I'll be sharing soon as well. Kind of like a modular version of what the Noll does. I'd not seen that one, but I going to ask my luthier friend to order one the next time he gets a parts shipment in from Germany so I can check it out. Klaus Noll makes and sells nice stuff, I actually use his parts quite a bit.
     
  12. Passinwind

    Passinwind I know nothing. Commercial User

    Dec 3, 2003
    Columbia River Gorge, WA.
    Owner/Designer &Toaster Tech Passinwind Electronics
    Update:

    Revisited the two problematic standalone LPF prototype boards today and eventually decided to just try building up the third (last) one from scratch:

    PW23B_v1.png

    And oddly enough, third time's the charm. This one worked flawessly the very first time...yay. I already have a smaller version modeled up with only two through-hole parts, ready to order as soon as I put this one through its paces in listening tests in the next few days.Tentative sizing for v2 is 1.55 x 1.08 inches, which could stack pretty much perfectly with my standalone Bass/Mids two bander. Or for those of you who want one filter per pickup, two of the LPF boards plus the mini mixer are at least on the table now as a sooner than later option. Better a year or three late than never, I reckon. ;)

    There are a variety of ways I can go with dimensioning and easier or more advanced build formats, by all means hit me with your thoughts on that. Personally, I am really liking using mostly 1206 SMDs with a few 0805s and 1210s in strategic locations.
     
    Jeff Siddall and Andyman001 like this.
  13. OJ Dorson

    OJ Dorson

    Jun 24, 2020
    I would definitely build a couple of those...
    Do I remember seeing that stacked pots for Freq/Response will never work?
     
  14. Passinwind

    Passinwind I know nothing. Commercial User

    Dec 3, 2003
    Columbia River Gorge, WA.
    Owner/Designer &Toaster Tech Passinwind Electronics
    Never say never, but it's really all about economy of scale. Find me ten thousand users and I'll find you a solution. :cool:

    It's a big ask, you want closely a matched reverse audio taper dual pot for the frequency sweep and that's already pretty much unobtainium even without adding a second dual pot on a concentric shaft arrangement. The Alembic Q switch or Wal switched pot solutions are sensible workarounds IMO. But I really like what @Jeff Siddall is working on with digital pots and hope he manages to bring that to fruition.

    BTW, I will be revisiting that max gain brick wall I mentioned, since I'm using different opamps there may be a little more margin to play with now. Self oscillation is really no bueno though...:beaver:
     
    Andyman001 and Jeff Siddall like this.
  15. One of the issues you run into with digipots is that, in general, all the pots in a package are the same size, they tend to be the lower resistance values (ex: 1K, 5K, 10K, 50K) and most are linear taper.

    But if the design allows for some flexibility, digipots allow for some very nice capabilities. For example, reverse tapers are trivial, as are partial ranges (ex: using 10% to 90% of the range). This also allows for some flexibility with pot values, especially if you just need a variable resistor instead of potentiometer. Ex: a 10K pot programmed for 0-50% range effectively allows a 0-5K variable resistor. If you have spare pots free on a device you can also parallel two digipots to make the equivalent of a single pot with half the resistance.

    Regarding control, even some of the smallest controllers should be able to handle digipot duties. Ex:
    https://www.mouser.ca/ProductDetail...el/ATTINY85V-10SH?qs=rBGENRD8NwIeGVZl36XmsA==

    Those are a buck or two in a SOIC8 package with power consumption in the hundreds of microamp range at low clock frequencies with no external components. They have enough I/O to drive a SPI or I2C pot chip plus plus 3 inputs to control the digipots.

    It's an interesting rabbit hole for sure.
     
    delta7fred and Passinwind like this.
  16. I forgot to ask originally: how closely do the pots need to match? My experience is that there is significant variance between individual devices on the same chip -- still within stated tolerance of +/-30% for the device I chose, but more than I would have thought would have been typical. Linearity within a pot is very good, just not that closely matched between pots.
     
  17. Passinwind

    Passinwind I know nothing. Commercial User

    Dec 3, 2003
    Columbia River Gorge, WA.
    Owner/Designer &Toaster Tech Passinwind Electronics
    Raw value tolerance isn't a big deal but section matching matters a lot. I ideally need 5% or better, it's a pretty big PITA actually.

    This morning I ordered some v3 test boards, this is just the standalone LPF:

    PW23B_3_0.JPG


    One notable tweak is that I've added a parallel cap to allow much better granularity in filter alignment tuning. Just being limited to vanilla ISO values (4700pF, 5600pF, 6800pF, 8200pF, and 10,000pF, for example) is very limiting for this particular circuit and hopefully the inevitable compromises can move out to some less painful aspects now. There are six 0805 parts if we use the parallel cap, five otherwise.
     
    Jeff Siddall likes this.
  18. Hmmm... 5% is awfully tight. A device like this 20% tolerance dual pot part that costs about a buck could work otherwise:
    MCP4651 - Mixed Signal - Digital Potentiometers

    Well, 3 devices actually (a dual 5K, a dual 10K and a dual 50K)

    There are some tighter tolerance 8% dual pot parts, or even 1% single pots, but they are more expensive and come in a much more restricted set of packages and values -- notably none are available in 5K.

    Six individual 10-pin packages is still a lot of board real estate, and things like MSOP might be too hard to solder.

    So yeah, no ideal options.
     
    Passinwind likes this.
  19. Passinwind

    Passinwind I know nothing. Commercial User

    Dec 3, 2003
    Columbia River Gorge, WA.
    Owner/Designer &Toaster Tech Passinwind Electronics
    5K is not an important value really, it just happened to dovetail with one reasonably workable tradeoff set. I've started to play around with the state variable format again in modeling, as it doesn't necessarily require any dual pots at all. But as usual, that just trades for added size, complexity, and cost elsewhere. ;)
     
  20. Passinwind

    Passinwind I know nothing. Commercial User

    Dec 3, 2003
    Columbia River Gorge, WA.
    Owner/Designer &Toaster Tech Passinwind Electronics
    Unexpected update:

    I ordered a test set of SMD boards for the standalone LPF on Tuesday. OSHpark comped me a service upgrade to their fast track and somehow the boards arrived in my mail box this morning. I went big on the tuning range at ~275Hz-10KHz for my first build, added a few dB of extra boost capability, and just finished initial bench testing. Looks great so far, totally stable and the dual audio taper pot in reverse trick seems to work well. I can't wait to have a listen tomorrow.

    .................................... :bassist:

    PW23B_v3.JPG
     
  21. Primary

    Primary TB Assistant

    Here are some related products that TB members are talking about. Clicking on a product will take you to TB’s partner, Primary, where you can find links to TB discussions about these products.

     
    Sep 28, 2021

Share This Page

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.