Neovibe Debugging Help Please

[Jaicen_solo]

Ok, after many years, i've finally soldered up a neovibe. I got all the board components in tonight, and I decided to just wire up the Depth and Speed controls, and see the bulb do it's thing.
Unfortunately, it doesn't do anything at all, and Q13 (bulb driver) gets really hot.
I noticed while debugging that C24 and C25 were only 16v caps, so I pulled them out. Voltages are as follows. Any suggestions, other than replacing the caps?

Q13
E:1.9
B:2.5
C:15

Q12 & Q13
E:0
B:0
C:15

[R.G.]

Ok, after many years, i've finally soldered up a neovibe. I got all the board components in tonight, and I decided to just wire up the Depth and Speed controls, and see the bulb do it's thing.
Unfortunately, it doesn't do anything at all, and Q13 (bulb driver) gets really hot.

Q13 E voltage looks high. What resistance do you read between Q13E and ground with the power off?

Have you double checked the pinout and insertion direction?

What voltage appears across the bulb?

I noticed while debugging that C24 and C25 were only 16v caps, so I pulled them out. Voltages are as follows. Any suggestions, other than replacing the caps?

16V caps for C24 and C25 are distinctly marginal. If you put more than about 18V on them, they will probably die. Capacitors are rated for about 10% more than their working voltage, but this is a surge rating, not something to be applied all the time. I would replace them if it were mine. Leaking hot capacitor goo smells terrible. 

Q12 & Q13
E:0
B:0
C:15

If you mean Q11 and Q12, there is a problem, probably bases shorted to ground.

[Jaicen_solo]

Thanks for the reply R.G.

I did indeed mean Q11 and Q12.

On Q13 I got 22v at the collector before I pulled the caps, now i'm getting 15v at the collector, and around 200OHM short to ground (dependant on trimmer setting) on the base with the power off.
With power, Q13 gets hot and puts 0.2v across the bulb.

I'll get the correct caps in later today, any suggestions as to where I should start looking?

[R.G.]

Pull out C23 temporarily.

With power off, measure and report the values of R47-R50 and TR1 as measured at the leads of the resistors themselves, and measure the range of resistance that Tr1 will adjust to.

If Q13 gets hot, that is another way of saying that the product of the voltage across it and the current through it is too high. The testing is to evaluate whether this is caused by the resistors telling the transistor to pull too much current, or by the power supply being too high, or the bulb not sagging enough as more current goes through it.

If you have a bulb rated for too much current, this could also cause the same thing. What is your bulb rating?

What happens if you replace the bulb temporarily with a 1K resistor?

[Jaicen_solo]

The bulb is supposedly a 12v /60mA, though it measures about 20 Ohm.
I'll try all that, and get back to you. What do you expect to happen if I swap out the bulb for a 1K resisitor?

[Jaicen_solo]

Ok, so I pulled out C23 and the bulb which stopped Q13 from getting hot.

Voltages are as follows:

Q11
E: 23V
B: 4V
C: 12-14V (cycles, so suggests that the LFO is working perhaps?)

Q12
E: 10-13V
B: 12-14V
C: 23V

Resistors measure as follows:

TR1 adjusts from 0-200 Ohms.
R47 = 100K 
R48 = 68ohms
R49 = 4k7
R50 = 46.7K

[R.G.]

The bulb is supposedly a 12v /60mA, though it measures about 20 Ohm.
I'll try all that, and get back to you. What do you expect to happen if I swap out the bulb for a 1K resistor?

A bulb is much lower resistance when it's cold. You would expect a 12V/60ma bulb to be 12/0.06 = 200 ohms but it is very common for a bulb to be 10 times less when cold.

The 1K resistor can be thought of as "one volt per milliampere". 20ma is 20V, and it doesn't change with heating up. So a 1K resistor lets you actually measure the current, where a bulb doesn't.

[Jaicen_solo]

Thanks once again R.G, I should have known that but I guess this thing has frazzled my brain.
Anyway, I happened to have a bulb from a christmas light lying around, so I thought i'd try it just in case, and guess what, the damn thing works perfectly! Q13 still gets warm, but not overly so. I might swap it out for a darlington just to be safe.

I can't really explain why the original bulb doesn't work, it doesn't even light up in the circuit.
Anyway, i'll wire up the audio section tomorrow and see if I need to do any debugging there. Fingers crossed!
Thanks for your assistance R.G.

[Jaicen_solo]

Just a quick question, the bulb driver transistor is still getting pretty hot using the new bulbs I just got.
Is there any reason why I shouldn't use the two sets of pads for Q13 to double up the transistors, (i.e, use two MPSA14's)?
It seems to work just as before, but neither get as hot as just one trannie.

I'm using a 12v 60mA bulb:

http://www.maplin.co.uk/Module.aspx?ModuleNo=1955&OrderCode=BT43W

Should I just look for a lower current bulb?

[R.G.]

If it works for you, use two. The engineers credo is "If it works, do it. We use equations, but we get it done even without them."

The smaller a bulb's power rating (volts*amps), the faster its time response to a changing voltage across it. So bigger power rated bulbs will respond slower and have a lower top speed response to the LFO. Smaller power bulbs respond faster.

Notice that it's OK to put a resistor or zener in series with the bulb to eat some of the excess voltage. This moves the power wastage out of the transistor and into the resistor or zener. It's the same power, just moved into something other than the transistor.

[Jaicen_solo]

That's the sort of info I was hoping for. I thought about a series resistor, perhaps 100 Ohm or something similar, but then I thought that it was effectively the same as TR1. Am I right in thinking a fixed resistor wouldn't affect the speed of response because it's resistance doesn't alter with current like a bulb? The cold resistance of the bulb is 20R.
I think this would be a more satisfactory workaround than two transistors, though that does seem to work.

[Jaicen_solo]

Just to bring some closure to this thread, I have finally got my Neovibe to work!
I had a minor hiccup with a broken (as in, one leg broke off!) voltage regulator which wasn't giving me +15v. I've replaced that and i'm now getting a swooshy univibe sound!
It's going to need some dialing in in terms of the LDR positioning, as I don't think the throb is quite there yet, but overall i'm really pleased. It's very quiet as well, even though I have the power transformer inches from the board and it's all unshielded.
In retrospect, I don't think that the wet/dry mix trimmer was strictly necessary, as I have it pretty much 50/50 anyway, so if I build another that will get left out.
Anyway, thanks for all your help, and a great layout R.G, even if it was all just stupid mistakes on my part!

[yeeshkul]

The wet/dry mixer is supposed to be 50/50 for the maximum effect. The trimmer is fun, but a good DMM to pick the 2 very same resistors will do.

EDIT: Also, for maximum throb you want to use the JC's mod - 250k trimpot that makes the voltage divider at the lamp driver base - this does miracles. In my first build i just put the trimpot "leg to leg" on RG's layout.

[R.G.]

In retrospect, I don't think that the wet/dry mix trimmer was strictly necessary, as I have it pretty much 50/50 anyway, so if I build another that will get left out.

The manufacturer sure felt that way. It all depends on how lucky or unlucky you get with both the gain through the audio phase path and the exact resistors.

If you have to choose skill or luck, take luck! 

Anyway, thanks for all your help, and a great layout R.G, even if it was all just stupid mistakes on my part!

You're welcome, and remember - all mistakes look stupid after you find out about them - at least all of mine do. 

[Jaicen_solo]

On the subject of stupid mistakes, last night my speed pot touched the input jack. I think it shorted it to ground, because there was a spark and my LFO stopped LFO'ing. The bulb is stuck on a dim glow.
Guess I fried Q11 and Q12.

Also, now I have to crack out the soldering iron to fix this thing again, (though my Iron is dead now too  ), can anyone suggest the best way to add a front panel control for the bulb bias? I get an impression that it would be a nice addition from some samples i've heard of the Foxrox vibe, though I can't find any proper demos of the control on its own. Ideally, I want the control to be useful across its whole range, so perhaps a 100K pot, with two 50K trimmers on the outside lugs to ground and +V? That way I can pre-set the minimum and maximum width.