odd(?) transistor switch behaviour

My best guess is that the entire circuit jumps to +15V when you kill the ground to it. It is still connected to the amp's ground and when the coupling caps charge to +15 the thing pops. Try using a high side switch and see if that helps (a PNP switching the +15V).

Jim

Try a much higher value base resistor. Just a hunch.

Graham

--
                                +15v
                                 |
                                 +----+
                                 |    |
                                 | [100nF]
  control voltage       ___    |/     |
            o--------o-|___|- -|     GND
                         3k    |>
                                 |
                                 o
                                 |
                                 |
                                 |
                                 o
                                GND

transistor to

function

in

on

switch

to a

I'm gonna check and see if I have any PNPs sitting around here and give that a try. I wish I'd have thought of that before I built the circuit into my pedalboard - I wouldn't have needed to use an inverter to get the control voltage right.

Thanks

transistor to

function

in

on

switch

to a

Hi Graham;

Thanks again for the insight. I'll try a higher resistor while it's hooked up to the scope, but I can't actually use a higher resistor on the pedalboard. The 15v comes from the amp through a 100k resistor, and the Ron of the transistor is too high to actually fully switch functions if I use more than 3k (actually it still doesn't quite fully switch with 3K, but I cant go any lower because it's being driven by a 4049.).

>

to

function

in

on

switch

to a

Thanks for the tip John. I'll scrounge up a 100nF cap and see if that fixes things. Also, is the value critical? I'm assuming not, but I'm just wondering how you decided on 100nF.

Thanks again

--
WAG, basically. ;) I figured the 15V was coming from a pullup, and I
thought it likely to be somewhere between 10k and 100k, so 100nF
would give a time constant of 1ms at 10k and 10ms at 100k.  What it
does is increase the time it takes for the transistor end of the
pullup to get back to 15V, so that should also "soften up" the
notch.  The value isn\'t critical except that it might be too small.
It occurs to me that the notch might also be in the switching signal
on the base, so you might want to look at that line and filter it if
you find the notch there.  Which transistor are you using?

fixes

Thanks again John. The transistor is a generic 2222.

So that's 150 uA.

With a typical small signal NPN device you'll need no more than 3uA of base current to switch that !

Eh ?

That makes no sense at all..

Graham

He sucked his finger.

It's amazing how often 0.1 uF comes up as a result of that.

Graham

--
Awww... Poor baby\'s upset because she couldn\'t come up with it no
matter what she sucked. :-(

hooked

base

functions

OK I'm glad this came up because it's been puzzling me for a while. The amp has an LED indicator for each channel, so whichever one is on has its LED lit. If I ground the connector (i.e., 15v to ground) the channels switch and

1 LED is on. However, if I connect the 15v to ground using a 15ohm resistor, one LED will light fully, and the other will still be partially (or fully) on. The same goes when using a transistor as a switch. I originally used a 10k resistor for the base, but 1 LED remained partially on. Using a lower value resistor (like 3k) turns the LED almost completely off. I figured from this that driving the base with more current reduced the Ron to a low enough value that it switched completely. It does seem odd, though, with a 100k resistor in there that an extra few ohms would make that much of a difference. You may remember from my previous posts on this topic that I was going to try a low Ron N-channel MOSFET to get away from this problem. Also, I've tried using an optoisolator to do the switching, but it also leaves one LED partially or fully lit.

Can you shed some light on this (no pun intended).

Thanks

Oh, one other quick question - I'm using a 4049 inverter set up to be a latching flipflop and it seems to have died on me. When the input of any of the inverters is grounded, its output should be a logic high shouldn't it? I've been testing the various inverters in the package to see if there is only one damaged or if the whole thing is shot, and none of them give a logic high with the input grounded, although they do show 0VDC when a high is applied to the input.