CV influencer! (broken NPN?)

[antonis]

Sorry Matthew but are you pretty sure that oscillations (if they indeed are) come from Emitter followers..??

[Matthew Sanford]

Nope, not sure. But I think my lack of base current limiting may have caused it. I replaced the first stage npn and put Rb, all good. Second shows voltage at all pins even with the base disconnected. I'll change it after work and put a 100r Rb

[Matthew Sanford]

Leave Rb there..
(it prevents Emitter follower oscillation..)

Thank you Antonis for leading me through this. I'd got excited when DiffEQ advised making it a current source, threw it on the breadboard (with loads of other circuits getting power from the 9v 0.3A wall wart) with no Rb and it worked like a charm! Put it in things like that but not sharing power, no Rb, and it worked...till it didn't. New npns and 100R Rb and it behaves, though I'm going to try 1K and in between to get a sweet spot (or trim pot...).

But I've found solder&heat shrinking wires to the (l)Dr LED with a couple daughter boards, 4 flying pots and switches, is a great way to break the LED legs off to 0.1mm, so I'm working cut up solderable bb for stability and learning to cut out LEDs to put new ones and still get it light tight with the ldr soldered in.

Just that'll take me a bit, but wanted to report back on my current learning (get it? Current?! Ok...takes the fun outta the pun)

[Matthew Sanford]

Ok just to finish up. Thank you again Antonis, gotta start there. 100R for the base worked ok, but I went with 1ks though probably could've gone higher like 10k; figuring I get 5mA tops base current, noticed Druid does 10k on the 3904 so 500uA to a min 50nA. The 2222 is min 20nA so base current could reduce.

One bothersome thing, the Cf on the LPF (1st section) accounts for the npn voltage drop, the inverter Rf did not. Rather than riddle I changed the 10k:10k voltage divider from 5v to 5k6:10k to give inverter bias about 3.2v rather than 2.5v, therefore offsetting it higher by about the drop, so they seem even-ish.

Tacked on to the FilterFX it makes fun by changing the rate while the depth goes inverse, so less on faster, or cycling through different waveforms, etc...haven't tried much as I get stuck in seeing how to play with the current weird.

I'm going to get sound samples up, but want er boxed first - as you know, she looks like ultraviolet catastrophe! So for my vein of abuse I need to get to making the extras on one board and manage/route wires sensibly. Little by little

[antonis]

One bothersome thing, the Cf on the LPF (1st section) accounts for the npn voltage drop, the inverter Rf did not.

That shouldn't happen..
As long as BJT's B-E junction is set inside NFB loop, op-amps outputs should account for those V_BEs..
(even for voltage drop inequalities..)

[Matthew Sanford]

I figured, but for this one I want it in the box already, so the higher bias also works though may not hit the lows, and I don't have the time.

I have another to fix (add Rb, replace npns) which had worked fine, so I'm guessing it is proper and should act so

[Matthew Sanford]

So wasn't just base resistor. After another gone on the lpf side even with 10k base resistor, but after reading on emitter w/reactive load (Cf qualifies, yes?) I've put 47R on the collectors for both sides, 10k Rb on LPF & 1K got inverter in range. Probably don't need the collector resistor on the inverter, but it works. Once I change the Frankenstein look to a pirate and get it in the box I'll share it.

Thanks again!

[Matthew Sanford]

Finally worked on cd4007 to replace this Lm358 inv/buf with this from this edumacational page



Which led me to breadboard this


Except I switched it so all four npns have bases tied, otherwise a slight BE drop from the led npn's voltage. But with the CMOS low output current no base resistors seem necessary, and I felt best to have one NPN for each out - this is to send the StompLFO through, to power many vactrols from each output.

I will keep the LM358 triangle set up with a switch jack to replace the Schmitt feeding the pot for tap tempo (small triangles or flat tops and bottoms without proper pot adjustment I'm guessing).