Updating my program-suite to version 11 (from a few years old ver9) so I can use up to date components (I must admit I feel pretty handicapped using multisim/ultiboard when everyone else seems to go LTspice, hehe)
I think I'm going to stick with my design with dual 6171's, and stick a charge pump voltage inverter in there to feed their -input.
Modified design with added power supply bypass capacitors and feedback caps and resistors as well.
I think a ICL7662CPA invterter will do fine for the negative supply (-4,5 - -20v, and it can get it's input from the same +12v as the rest of the circuit) and they will not burn a hole in my wallet at 5 bucks/piece.
Youd' think I would have learned by now that my "bright" ideas I get in the middle of the night arent always as bright as they seem at the moment .
The additional caps and resistors I placed on my "bright idea" schematic are specific for the ideal 6171-setup... That's why. I figured I'd start it safely with recommended components and then strip it down as I go along. Looks like my little driver is going a way of it's own, so I'll keep working on that one as a separate project, replacing the 317 with something that can handle the inputs, and copying your design for a second testing on my end with the (mostly) same setup.
edit: offtopic:
I found an old box with a few thousand random components in it, leftovers from our certification-classes =)
Loads of 1n4007, 47nF 100V tantcaps, 4116r lf (isolated resistor ic's, 1kOhm, 2,5W), hef4049 hex inverters, 470mF 40V elec caps, 100uF 25V elec caps, 150nF 250V tant caps, 220nF 250V tant, minimal surfacemount stuff (am p73), topline plcc20, plcc32, qfp-100-de (qfp = dummy components), surfacemount diodes, zeners, caps, resistors, unmarked coils measuring 3x3x3mm, lucbdg2207bdg, a few lm324n, and other random odds and ends... might find something useful in there =)
Personally, I'm doing it the wrong way on purpose with my weirdo design, to learn more =) -And, at the same time, going to do one or a few of yours as well, not just to learn, but to also get something that works
Very nice to hear that it works offscreen! =) I have the same problem with multisim. It's simulations are not 100% reliable, and sometimes gives the same kind of bugs you described, without protesting at all. I cant tell you how many times I have invented the infamous perpetuum mobile in that damn program simulation. My fav was when i got pretty much what I'd call a fusion generator from a simple setup of amps and caps, with 9Vdc input. For some reason it gave me 96TV output . -Normally that would generate an error-message, lol.
Anyway, I'm gonna order some LM359N's and NE5230NG's tomorrow, and try to make them order a few other random 50Mhz amps and switching regulators as well, for testing.
I managed to get a pcb layout done in school, I'll post it in a bit
Edit: Added pic.
Pcb size: 1,5 x 1,6" single layer
Last edited by Occularis; 04-06-2010 at 12:24.
Reason: Added pic