A (not so new) Shruthi-1 filter board is coming
  • 6 months ago I set myself to rework “from scratch” the SMR-4 circuit.

    Some of my motivations:

    • Fix some so-so design choices in the SMR-4 that were on the “cargo cult” side (eg: I did the expo converter with NPNs because it was what was used in the material I studied).
    • Simplify the trimming/calibration procedure, source of many support nightmares.
    • Take advantage on some tricks I learned while working with the IR3109 (“sweet resonance”).

    Another motivation:

    • Get rid of the expensive / hard to find LT1054 ; and go single supply.

    This last point took me in many different directions and made me face many unseen problems (power supply noise coupling through the mid-rail, power supply noise coupling through inputs used to offset/null the VCA CV, dependency on Vregs actual voltage, impossibility to use the Shruthi from a 9V battery). In the end, solving all these problems meant more parts (be they additional op-amps, caps, Vrefs) or more expensive parts (better Vregs), which kind of defeated the original plot.

    So I’ve decided 2 weeks ago to get back to the good old “battle-tested” LT1054-based split supply. Noise levels are back to that of the current design (= typical of LM13700s). The new design is simpler and only needs V/Oct trimming. The board will have patch points for BP and HP modes (the same way the current board has a 2-pole and 4-pole jumper) ; though it is not planned to add digital control of the routing at the moment.

    I’ll have prototypes ready in 10 days, I’d love to get your feedback about it since I’m considering making it the new “default” filter board with the kits.

  • nice. I really like your LT1054 design, elegant and simple. I’ve done a lot of +5/-x/+x psus on various projects and the Shruthi approach is the easiest..

  • I still think you should have the filters and the controller board as different kits… other than that I’m fine with the LT1054. Perhaps you should consider switching to 9V or even 12V AC supply and have “proper” voltages inside, but we’ve done that discussion before.

  • @Lindeborg: anything new a 12V AC supply could bring (besides removing the ability to power the unit from a 9V battery)? The filter board size is a more limiting factor than the 100mA provided by the LT1054 for the negative rail… Requiring a proper AC transformer would make sense for a larger project, though…

  • Not really no. As you write; not for this application, some headroom that is it. I was just thinking out loud, sorry bout that. I think, as I wrote, that the LT1054 solution is just fine and quite elegant.

  • great! i surely take one if it sound like the old one, just to be sure of the trimmings!!! it's my favorite!
  • how difficult would it be to add firmware/menu control over lfp/hp/bf filter modes? cause that would be extremely cool, obviously.

  • I think the SMR-4 is/was a great filter. One of my favourites actually. But my number 1 has to be the IR3109 filter. If the SMR-4 MK II has some added IR3109 influences and inherits some of its sonic characteristics, then I certainly wouldn’t complain :)

    No real suggestions from me though.. Other than to say that im sure the simplified trimming will be a good thing! Many of us have built several Shruthi’s now, and trimming becomes pretty straightforward. It’s another matter for those part time hobbyists just making the one however. It took me a few goes to get my first one right..

    On a related note.. And maybe this is a small suggestion after all.. Trimmer location. Having built a few Shruthi’s now, I noticed the IR3109 variant was damn nearly the most convenient to trim out of them all, simply because the more critical trimmers are all located along the front of the unit. Meaning you can assemble the entire thing, including enclosure (minus the smaller front vertical panel) and then get to those 3 trimmers quite easily. Which is much easier than pulling it all apart and linking the 2 boards together with jumper wires.
    Perhaps a switch to side (rather than top) accessed trimmers would make it easier and more convenient for builders too?
    Like this..

    If possible, you might then also be able to get away with no longer having to ship the temporary jumper cables in the kits.

  • @mic.w: it would take 100 bytes of code and two chips on the board + making the project much harder to troubleshoot -> not worth it I guess, but maybe I could do an “XT” version with this later.

  • @Luap: arghh, I thought I could have keet this part of the plan a secret :)

  • Oh, and it’s using the “grid-like” groundplane of the most recent boards, for easier soldering. Another change is the use of large holes for the DC power connector, allowing the use of the most common kind of DC connector (something coming with the Polivoks filter board too).

  • I might be a total ass but I have spent a few more hours with it and I was still not happy with the way it sounds. It lacked the “strength” and “solidity” of the SMR4, I’m sure you know what I mean ; and the way the resonance appeared was just weird (at 45 it started self-oscillating but onyl for high cutoff values, and by 63 it was still not oscillating with the lowest cutoff values – and no dub basses is a no-no).

    I traced it to another major change it had compared to the original SMR-4: I used the Darlington buffers integrated to the LM13700 instead of a TL074 as buffer between filter stages. It took me a while to realize that it was not a benign choice. So I tore down half the breadboard, put the TL074 back in and it instantly rocked.

    So now I’ll have to reroute the whole board, probably the 4th or 5th time :) And I wouldn’t call this a sequel or a reboot, it’s more a remastered classics edition…

    Pure labor of love :)

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

In this Discussion