neo-vibe doesn't oscillate at slow speed

[alderbody]

i just built my neo-vibe and it only oscillates at full speed!
as soon as i turn the pot just slightly down, the oscillation stops and the bulb doesn't flash...

A strange thing happens when i touch some leads with the tip of the multimeter. When i release it, let's say from the base of the darlington, the circuit oscillates instantly and then it dies out...
This happens whenever i touch some part of the feedback stage.

It also happens at power up. An instant oscillation and then nothing.

Below, i attach the voltage readings.
My supply is at 23.5V, the caps are 1uF 63V electros,
and the resistors are carbon film.
The diodes are the recommended ones and the transistors are 2N5089.



what do you think is wrong with it?
Should i put closer tolerance resistors? Another set of transistors? Non polarised caps?

please help!...

[R.G.]

Your particular Q11 and Q12 are not giving you enough current gain. This happens in real Univibes as well as clones. I've repaired several with this flaw.

The clue is that it oscillates a little bit at any disturbance, and at top speed. Below that it rings for a few cycles and then fades out.

Before you start swapping transistors, verify the value of every resistor. I'm forever mixing up 4.7K and 47K resistor because the shades of red and orange on the resistors are not at all standard. Try a 2N5089 for Q11 if you can.

[alderbody]

Thanks R.G.

I'll try these and report my results

btw, all transistors except for the bulb driver are 2N5089.

[R.G.]

Then replace Q12 witn a 2N3904. The Q11/Q12 darlington suffers from all the normal problems of a discrete darlington. The first transistor must operate at quite low currents, so unless it's designed for high gain at lower currrents, the gain isn't what you'd expect from the datasheets. The second transistor does all the current, so it needs to have good gain at high currents. The 2N5089 is a higher gain version of the 2N5088, and may not have good high current performance - hence my guess that a 2N3904 will work better.

This is not a cut and dried situation. Low gain in the LFO on the vibe is a fairly persistent problem and may be due to a fault in the components that feed back to keep oscillation going as well. The transistor gain problem is just more common.

[Eb7+9]

... two ways of increasing the gain of that oscillator at low frequencies is by replacing the first transistor in the Darlington pair by an n-channel jFET - 2n5484 will work ... another way is to replace the whole pair by a jFET op-amp wired as a follower ... basically decreasing the loading at the high-Z node ...

~jc

[R.G.]

I'm sure both those would work, but it's much simper to get the transistor pair right if  you've already built the thing and it's not working.

You can also replace both Q11 and Q12 with an integrated darlington like the MPSA13, which *has* been designed internally for the current level differences between the first and second transistors and so provides high current gain.

[alderbody]

thanks,

i'll try these, too...

for the time, i replaced the relative resistors with metal film 1%, but still nothing.

i put bases in the darlington traces in order to be able to test many devices.

but a nasty thing happened. now, it doesn't oscillate at all!...

but there has to be a problem with the diodes because the votage reading is 2.0V instead of 12.0V that it used to be...

i'll replace them to see what happens

would another type of diode work in there?

anyway, i'll report my progress soon

thanks again.

[dr.angus]

Hi!

Did you check the wiring of your speed pot? It should be connected according to the schematic (in R.G.s Neovibe article) not the parts layout (otherwise it will be like an on/off switch).


Have fun,

matze

[alderbody]

no matter what i tried, the damn thing won't oscillate.

most of the time it won't oscillate at all...

i'll check what dr.angus suggested to see if i got it the wrong way.

i replaced every part in the osc. circuit, tried combinations of transistors, but still nothing!


i think i'll grab a hammer and smash the sH__t out of it!....

[R.G.]

no matter what i tried, the damn thing won't oscillate.
...i think i'll grab a hammer and smash the sH__t out of it!....

Mother Nature is trying to teach you something. Usually when she gives me this kind of lesson, I really feel silly once I find out what it was that she was trying to teach me.

Pinout, soldering, wiring, circuit board flaws, that kind of thing can all conspire.

You know, I was about to ask you to measure all the voltages, but I see that you did. Unless there is a silly mistake in measuring voltages, something is seriously wrong with Q11's base. It's shown at 4.3 V, and it ought to be more like 13.7. Can you measure your current voltages and post them back?

[puretube]

caps...

(esp. the upper one won`t like being mis-polarized!)

[R.G.]

I didn't weigh in on that because the cap is not what's causing the problem if the base is that low. The normal voltage on the base would make the cap properly polarized.

[alderbody]

i changed the wiring of the speed pot and re-wired it according to the schematic and not the layout.

the "on-off" opreation stopped but i still can't make it oscillate at slow speed.

R.G.  i think the base voltage is around 11+volts now. 4.3V was very low indeed.

i'll keep on trying whatever you suggest and i'll report my progress again.

thanks for the tips

[alderbody]

i measured the voltages around the darlington.

now i have 6.5V on the base, 24.1V on the collectors, 13.5V on the emmiter and 14V on the link between Q11's E and Q12's B.

the funny thing is when i somehow touched the E-B link with the tip of the mulimeter the LFO started working...  I checked just in case the soldering had a flaw, but everything was fine.
Then i connected one side of a 2.2M resistor i had lying around to the ground and with the other i started touching the leads of the transistors. Again, sometimes when touching the E-B link the LFO worked fine, although it looked more like an "on-off flash" and not this wavy "dark-bright" transition...
When that happened, the voltage reading on the "C" lug of the depth pot (negative lead of C22)was at -3V... ()

then i tried it without the resistor and... the second transistor exploded!    hehe...


all this time, the voltage readings on the bulb driver were:

B:4.7V, C:19.9V and E:4.1V


well, how about that?....

[alderbody]

well?...

any opinions?

....

[R.G.]

There is something wrong with the biasing setup or the connections to Q11. Since you're replacing Q12 anyway, remove Q11 and measure the voltage at the holes where Q11 is going to go. Then pop back here.

[A.S.P.]

information lost...

[R.G.]

Good points!!

[alderbody]

hmmm, i'll check these too.


do you think that non-polarised caps would have closer tolerances?


i'll get back with my results later.

thanks

[alderbody]

good news!!!!

this morning i bought some NP 1uF caps and i installed them just a while ago.
i only put two because the third turned to be a 3.3 (by mistake of the shop... )

...and the oscillation just started!  ...at any speed!!!!!!!....


just by replacing two bloody caps!...

i think i'll replace the third one and maybe the two 10uF to see what happens.


i guess i found the weak link in the chain...


so many thanks!!!