SN Voice troubleshooting: EG working

[nutoniom]

Edited: see post on 02/20/11

Hello,

I'm finally getting around to finish my SNV and I'm having a couple of problems.

But first off, I'm using Bill's SmS pcb. And I am triggering using seq-303 feeding a doepfer MCV-4

What works: The pulse sound 1V/oct , with coarse and fine tuning operating as expected (althought I find the coarse range to be limited:it goes very high very quick)

Linear and Ex FM both work feeding them the MCV-4 gate ( fun ! : ) )

LFO works as expected.


What doesn't work:

Tri out: no clue why. I can see the triangle waveform nicely at pin 2 when the molex connector is disconected. It disapears when I connect, weird...

Noise: flipping switch 2, shuts the output at pin 13 of SN chip.

Finally, and most importantly to me: Gate. The env is constantly locked. But I don't think it's my switch. When I flip the switch I can hear the enveloppe fade out in relation with the release pot. If my release pot is fully CCW, flipping the switch halts the envellope instantly, set fully CW, it fades slowly, halfway CW, fades half the time. So my reasonning is my gate signal gets knackered somewhere along the way.

Tracing it from the jack, I get a healthy 5Vp-p square wave signal which gets attenuated to 0.5 v by R39 and enters the base of Q5, comes at at collector at 1v p-p inverted square. I can trace this to C3, after wich it becomes a straight line on the scope with sharp peaks. This is what I get at pin 3 of the 4046. The vco output of the 4046, pin 4, looks nothing like a squarewave to me: it's like a pulsing beam that seems to be almos 4 volts of amplitude.

pins 9, 14 and 16 of the 74HC4046 all receive 5V.

I have soldered the 4046 directly to the board could have I fried it (I solder pretty quick and alternating)

Any help would be greatly apreciated.

thanks, jf

[magman]

jf,

One thing that sounds a bit strange to me is that you have used a 74HC4046, which is effectively a TTL device. The design calls for a CMOS CD4046, which is a slightly different beast.

As this chip provides the external clock for noise generator for the 76477, this could explain at least one of your problems.

I'm not sure what your electronics experience is, so a few comments if you decide to replace this chip. First, cut the legs off the chip body itself (the chip is worth nothing compared to getting another PCB and wasting your time), then desolder each pin individually.

I would suggest using a socket when you replace the IC (but I suspect you've probably decided to do this anyway).

Hope this helps

Regards

Magman

[nutoniom]

Hello Magman,

i have not seen any mention other than 4046 pll in the docs i got for the pcb, so i figured it would work as they seemed pretty much the same, except for pin 15 which is a zener 5.2 on the cd4046 vs a pc3 out for the 74HC4046.

Anyhow, i have snipped the chip out (you only try to unsoldered a 16 pin IC once in a lifetime: been there, done that, got the t-shirt...) and soldered a socket. Replaced with another 74HC4046 and I get the same exact results, so I guess I'll be ordering some Cd4046 unless I get some other suggestion ?

thanks,

jf

[nutoniom]

[State Machine]

[nutoniom]

nope, J3-2 not shorted: i would have been an extremely happy camper if it had been the case : )

Thanks for the generous offer on the cd4046 Bill, but I won't need them as i ordered from mouser already.

Will keep you posted on my progress

thanks !,
jf

[nutoniom]

Well I got my cD4046, popped it in and tadam! i've got noise, beautiful noise: it's Defender all over again : )

Alas, still no gate, just the fixed enveloppe.

And the triangle output is still not working, which is strange because I can see the triangle wave on the scope on the foil side pad, but not in the molex connector's groove (where the little pin locks). But the triangle I can live without, it's the gate thing that drives me mad, any pointers will be more than appreciated

thanks,
jf

[nutoniom]

Ok, have a bit more time for this and would need someone to validate or invalidate my theory.

Shouldn't I see 0.1 V p-p after R40, on it's way to pin 23 of the SN76477? And I mean clearly see, with the volts/div setting on 0.1 v, a 0.1 v p-p square wave? As in 100000/104700. Because i don't see that at all, it's a flat liner. Please let me know before i unsolder things (I hate unsoldering things just as much as i love soldering them, I'm weird like that).

thanks in advance,

jf

[nutoniom]

mmm, ok, I guess it's more like 4700/104700, so 0.044 v, which seems consistent with what i see on the scope. Darn!

[nutoniom]

I know this is probably heresy and risk burning at the stake for it, but i'm very tempted to bypass R39 and live with A/D pots' very limited range.

[nutoniom]

Replaced Q5 and it is now working.

thanks for watching : )

[State Machine]

[nutoniom]

No problem Bill, I was just being my funny self (I'm an acquired taste I guesss...)

But I shout victory too quick, as it still not functionning quite right. More rounds of fun : )

The LFo, which was working, isn't anymore and I can't switch from fixed to enveloppe any more. I guess all that manipulation as knackered the wiring.

I'll keep you posted : D

[nutoniom]

Yup, knackered wiring: it's working fine now.

On to the Mankato: more rounds of fun : )