Geofex cheap good polarity protector transistors

[GGBB]

http://www.geofex.com/article_folders/cheapgoodprot.htm

I have put together one of these but I'm getting about -1.5 volts out when I reverse the battery instead of 0.  When hooked up properly I get Vin=Vout.  I am wondering if the transistors matter.  Instead of 2N3904/6, I am using the only PNP I had laying around - 2N4126 and also a 2N4124 NPN.  Also - don't know if this matters but I have no circuit connected, just the DVM.

Here's the details:

battery 8.91V (no load)

Positive
VIN +8.69V
VOUT +8.69
NPN
C=0.02V
B=0.73V
E=0V
PNP
C=8.69V
B=7.92V
E=8.69V

Negative
IN -8.78V
OUT -1.44V
NPN
C=-1.20V
B=-0.59V
E=0V
PNP
C=-1.44V
B=-1.83V
E=-8.78V

[Canucker]

Did you see the one I posted in another threat on the polarity protection topic? http://farm5.static.flickr.com/4034/4635974624_e8f3894336_o.jpg   I'm totally obsessed with these these days. Has anyone out there actually messed up a pedal cus they didn't have polarity protection???

[slacker]

Try putting a 1k resistor across the output to simulate the load from a pedal, you'll hopefully find the voltage drops to a very small negative amount.

[GGBB]

Try putting a 1k resistor across the output to simulate the load from a pedal, you'll hopefully find the voltage drops to a very small negative amount.

No luck - it went down(up) but still getting about -1.2V out.  Thanks though.

[R.G.]

I would have said that the 1K would drop the voltage to nil too. Let me go look at the original files a bit. It's been some years since I did that, but as I remember it did have a bit of leakage unloaded, but a load dropped it to insignificant. It'll take just a bit to refresh me.

[GGBB]

Did you see the one I posted in another threat on the polarity protection topic? http://farm5.static.flickr.com/4034/4635974624_e8f3894336_o.jpg   I'm totally obsessed with these these days. Has anyone out there actually messed up a pedal cus they didn't have polarity protection???

Yes - I did see that and used it to verify my own layout.  Thanks.  I also put together a perf board layout:



Edit:  Maybe this is the problem?  Is it okay?  When the perfboard didn't work I breadboarded it with the same result (values above are from the breadboard).

[GGBB]

http://www.geofex.com/article_folders/cheapgoodprot.htm

I have put together one of these but I'm getting about -1.5 volts out when I reverse the battery instead of 0.  When hooked up properly I get Vin=Vout.  I am wondering if the transistors matter.  Instead of 2N3904/6, I am using the only PNP I had laying around - 2N4126 and also a 2N4124 NPN.  Also - don't know if this matters but I have no circuit connected, just the DVM.

Here's the details:

battery 8.91V (no load)

Positive
VIN +8.69V
VOUT +8.69
NPN
C=0.02V
B=0.73V
E=0V
PNP
C=8.69V
B=7.92V
E=8.69V

Negative
IN -8.78V
OUT -1.44V
NPN
C=-1.20V
B=-0.59V
E=0V
PNP
C=-1.44V
B=-1.83V
E=-8.78V

I looked at the datasheets for the transistors and noticed that the maximum E-B voltage for the 2N4126 is 4.0V and I measured 6.95V.  I'm assuming that is a problem, but the question is whether this is a cause or a symptom.

http://www.fairchildsemi.com/ds/2N/2N4126.pdf

[R.G.]

OK, did some looking.

I think base-emitter breakdown might be what is getting you, OK.

Try this: put a silicon diode, any kind, 1N4148, 1N914, or 1N4000 series would be good, in series with the base of the NPN. If that doesn't help, put another one in series with its collector, both with cathode toward the NPN transistor. I think that will help.

[GGBB]

OK, did some looking.

I think base-emitter breakdown might be what is getting you, OK.

Try this: put a silicon diode, any kind, 1N4148, 1N914, or 1N4000 series would be good, in series with the base of the NPN. If that doesn't help, put another one in series with its collector, both with cathode toward the NPN transistor. I think that will help.

Thanks R.G.  I tried the diodes as you suggested, but no luck.  This must be something related to the 4124/4126 transistors - I picked up some 2N3906 today and tried one with a 2N3904 and this gets me down to about -0.2V without the extra diodes - I assume that won't cause any problems.  If anyone is interested, the numbers with the two extra 1N4001 diodes (and 1K load resistor) are:

Negative
IN -8.69V
OUT -1.12V
NPN
C: -0.02V
B: -0.03V
E: 0V
PNP
C: -1.11V
B: -1.75V
E: -8.68V

[R.G.]

Thanks R.G.  I tried the diodes as you suggested, but no luck.  This must be something related to the 4124/4126 transistors - I picked up some 2N3906 today and tried one with a 2N3904 and this gets me down to about -0.2V without the extra diodes - I assume that won't cause any problems. 

Hmm. Yeah, if the 3904/3906 work OK, must be something unusual about the 4124/4126. My first suspicion was the Vbd breakdown, as it was yours. The diodes would have fixed that.  You're right, -0.2V won't cause any problems, as it won't turn on even a silicon Schottky junction in reverse.

If anyone is interested, the numbers with the two extra 1N4001 diodes (and 1K load resistor) are:
Negative
IN -8.69V
OUT -1.12V
NPN
C: -0.02V
B: -0.03V
E: 0V
PNP
C: -1.11V
B: -1.75V
E: -8.68V

I'll think about that some more.

[GGBB]

Thanks again R.G.