Why is my -12V from Voltage inverter becoming +12V after this transistor

[Skruffyhound]

See bottom right.


I have -11.81 from the inverter Vout, but the transistor has +11.95V on base and 11.96 at Vout (red type)

I don't understand this is directly from the app. note.

It's in this layout:

It's the transistor in the middle next to the letter I.

What is going on here?

[R.G.]

I think the problem is that you have +12 on the base. +12 on the base means you get the base conducting into the collector.

I would check both sides of the 10K resistor feeding the base and see if you don't have +12 on the "ground" side as well.

There are other possibilities, but that's where I'd start.

[Skruffyhound]

Thanks for the reply R.G. I checked "ground" to no avail, but something was wrong and for a while some chip voltages were off, I changed chips and transistors (no help) and a lot of scraping between traces and checking pinouts down the road
I'm back with the original chip and transistor. Chip voltages are spot on but I'm getting +11.97V out of the collector, 0V on base and -11.82V on emitter. Every datasheet I check shows the 2N2222 TO-18 package to be correctly placed.
What is going on here?

[Skruffyhound]

I'm almost there where I'm considering dumping this part of the circuit and taking my chances with pin 5 being dragged positive.

[Skruffyhound]

I should also say that this voltage inverter, the sub-mini circuit and the LM317 have all been tested on the breadboard.
I used the same transistor too.

[jdub]

If it worked on the breadboard, it's gotta be some sort of bridge or error on the perf- and the layout looks good as far as I can tell.  Any chance of pics?

[R.G.]

Thanks for the reply R.G. I checked "ground" to no avail, but something was wrong and for a while some chip voltages were off, I changed chips and transistors (no help) and a lot of scraping between traces and checking pinouts down the road
I'm back with the original chip and transistor. Chip voltages are spot on but I'm getting +11.97V out of the collector, 0V on base and -11.82V on emitter. Every datasheet I check shows the 2N2222 TO-18 package to be correctly placed.
What is going on here?

There are (at least) three possibilities here. (1) Measurement error is always possible. I do it all the time, sometimes making the same mistake a few times before I catch myself. (2) The base of the transistor could be open internally. (3) the emitter of the transistor could be open internally.

I pick those three out of the pot of possibilities because I trust that you've checked the datasheets and have the pins right, and it's not possible for a non-damaged NPN to have those voltages happen if the measurements are right. By the way, in the case that you're debugging something on a PCB, it's often good to take measurements on the pins of the device, as well as on the traces or solder joints the pins connect to. This helps in sorting out bad solder joints.

I suggest taking the transistor out and measuring the base-emitter and base-collector junctions for conduction the right way, no conduction the wrong way on the diode or resistance range of your meter. I'll have to think harder if the transistor shows good to the forward/reverse diode test.

The -11.x on the emitter side and 0 on the base side tell me that something is open from base to emitter.

[Skruffyhound]

Thank you guys.
I will make the tests R.G., I wish it had been mixing up the probes that was at fault, I'd have felt stupid but there would have been nothing to fix.
This unit is made up of a number of modules. I need to separate this module completely from everything else to be sure ...well, you know
John, I will return with pics, schematics and full layouts if I don't locate the problem very soon, there is a deadline on this as my mate needs to use it pretty soon
the end of next weekend would be a good point to be finally finished with this rather long project.

[Skruffyhound]

Well you were indeed correct R.G. as indeed was John and even I guessed right. 
The transistor was indeed open between base and emitter, this in turn was caused by a bridge between power rails on the second board, so after I disconnected the power connections to board two I was able to start putting things right.

I've now spent the day working on the pre-amp section on the same board but I'm banging my head on a brick wall again. After checking my layout thoroughly, I did find a small mistake near the 270 ohm resistor, but fixing it didn't help significantly.
After audio probing everything I'm really wondering where the gain is at. I changed the NE5332 for a TLO62 in case I'd fried the other chip with all the power problems.
If I plug my iphone through a little Carl Martin headphone amp and out the balanced out the circuit passes signal and the 10K pot produces a very little gain before cutting out altogether 2/3 of the rotation.
My bullet mic and another dynamic I have give nothing.

Does this schematic look right? Because I have checked pretty carefully through my layout but I didn't breadboard it before building, kicking myself now
http://sound.westhost.com/project122.htm

[PRR]

> Does this schematic look right?

Rod's plans *are* correct as-drawn.

(I did check, because his "conventional" topology was not clear at first glance, and I had to sorth through it.)

The question is: what is your as-built?

Are there any DC voltages anywhere?

> cutting out altogether 2/3 of the rotation

Sure sounds like bad DC voltages.

[Skruffyhound]

Thanks for taking a look Paul.
TL062
Pin     Voltage
1      -10.66
2      -2.62
3      -5.32
4      -11.67
5      0.0
6      0.0
7      5.37
8      11.96

Pins 5 and 6 look a bit shifty to me, but I can't find anything wrong as yet.

The pot has got -2.48V on terminals 1 and 2(they are connected) and 0V on terminal 3
I can't get to it again tonight but I'll have another look tomorrow evening.               

[PRR]

Why does pin 3 have -5V on it??

It is only connected to zero V via 10K. It ought to be at zero.

(The other voltages appear to be sane, IF pin 3 is at -5V. The symmetry with +5V at pin 7 is notable.)

> TL062

That is a terrible low-Z mike preamp. If I may believe my own fingers, really terrible hiss relative to low-Z mike.

Or... if 'forced' to use TL062, you cudda saved a lot of supply debugging by using two 9V batteries. Battery life with the cheapest 9Vs is 2,000 hours. Or a couple years gigging 4 hours a night 5 nights a week.

Yeah, the DC voltages should be correct with TL062, so fix that while you order a 5532 or LM833.

[Skruffyhound]

Thanks Paul,
                  I will try with the 5532 I took out, or the one from the other board, after I have a look at pin 3. Lets hope I get some mic signal through and some gain.
Then it'll be on to board 2 ... ha ha.
                  I didn't realize the TL062 was so rubbish here, I have them as a cure for ticking LFO's

[Skruffyhound]

Right, so still hacking away at this, although it's now officially the biggest PITA I've encountered in any debug.

In my build pin 3 just is -5V, ground is zeroV and there are no solder bridges or other explanations that I can find.
I have been over my layout many times now, pin by pin, crossing off components on the schematic as I go. Nothing.

To add insult to injury, I breadboarded it last Sunday and using my mate's mic managed to get it to pass signal, just.
Voltages are laughable though, just the negative power rail on pin 4 and the positive power rail on pin 8, everything else is Zero  

So I breadboarded it again with a different topology.
And again  

I found a stash of 5532's in the parts bin and have tried a few, without effect.
It's not a complicated circuit, in fact I pretty much have it memorized.
What is going on here.

It would be a shame to scrap this project over one circuit, but pretty soon I'm going to have to tell my mate to buy something if I can't get to the bottom of this.
I've got the etch design, switching design and box layout printed out and waiting and the box all shined up and I've built the two boards with heat sinks and stand-offs. Even used the teflon wire  .
One lousy opamp stands in my way. Wonder what I'm missing.

[Skruffyhound]

The pot is just as useless on the breadboard, needs to be full up before anything happens.

[Skruffyhound]

Ok I'm back. Now I have this working pretty well on the breadboard
http://sound.westhost.com/project66.htm

I don't understand the gain/attenuation control here.
Whichever way round I try, whatever value pot I try,whatever taper I try, I get no signal until resistance is almost zero.
So close to zero that it could be a switch in fact.
I've tried changing the 100K out for lover values across the pot, no change.

Gain for the circuit seems pretty respectable.
The only place where I vary from the project is that my power supply is +/- 9V because that's what I have in my breakout box.

Ideas would be greatly appreciated.

[PRR]

EDIT--- duplicate question.

http://www.diystompboxes.com/smfforum/index.php?topic=102046.msg902857#msg902857