Vox - Repeat Percussion (V 809)

poodlebra · Post by **poodlebra** » 03 Feb 2011, 09:59

not really a depth control, but the mix pot on my vero works well, giving the illusion of depth by throwing in a bunch of dry signal.

goes from total chop to dry.

Solidhex · Post by **Solidhex** » 06 Jul 2011, 19:48

R.G.'s suggestion about using the MPSA18 is spot on. I was using a 2N5088, switched it for a MPSA18 and the ticking greatly decreased.

ryanuk · Post by **ryanuk** » 12 Jul 2011, 08:27

dmk2113 wrote:Thanks RG - I don't know where we'd all be without you.

I'm happy to report that not only does the circuit work well, it worked well with all 11 of the UJTs -culled from 2 different sources - that I tried in it... all of them provided pretty much identical results. As for the NPNs, I tried BC108s 2N5088/5089, a bunch of Darlingtons, some SC1000s and a few others. They all actually sounded more or less the same, but the 2N5088 seemed to provide the best results.

dmk2113 - read the earlier posts but unsure what you've changed to eventually get all your UJTs working. RG suggested using PUTs; but you seem to have stuck with UJTs. How did you get em all to work?

Assume you used analogguru's correct schem?

Keen to know as I really dig the sound of this effect and might knock one up.

RyUK

RnFR · Post by **RnFR** » 12 Jul 2011, 12:48

going back to these PUTs, do they give the same waveform when oscillating? to me, that really seems like the key part of this effect. the UJT oscillator has a very particular waveform. RG, if you are around, did you happen to compare the two on a scope?

R.G. · Post by **R.G.** » 12 Jul 2011, 17:26

RnFR wrote:going back to these PUTs, do they give the same waveform when oscillating? to me, that really seems like the key part of this effect. the UJT oscillator has a very particular waveform. RG, if you are around, did you happen to compare the two on a scope?

I did. The PUT is the modern replacement for the UJT for a reason. The UJT does the waveforms it does because of the doping profiles, silicon resistivity, phase of the moon when it was diffused, and so on. The PUT can be programmed to cover the range of actions many UJTs will produce; not only that, once it's programmed by external resistors and caps, the PUT produces the waveforms and currents a UJT would. A PUT can be set up to mimic a number of UJTs. So PUTs effectively put UJTs out of business because they do the same thing, but controllably. They are not pin-for-pin replacements; you have to tinker the connections a bit.

More importantly, I built circuits with the UJT and the PUT. Both give the same sound. Back when I was looking at this, the 2N2646 was hard to find (for me at least) and cost from one to several dollars each. The 2N6027 is usually under a quarter.

You can do much the same thing as a PUT with a PNP/NPN setup and some resistors. Takes more tinkering with resistor values to get it functioning the same, but it's there as an option.

The real challenge with the repeat percussion is in isolating current flows and capacitive coupling so it doesn't tick, and keeping the current magnitudes on the timing cap so it will not stop oscillating because the current doesn't fall below the hold current level. These are issues that affect both UJT and PUT.

RnFR · Post by **RnFR** » 13 Jul 2011, 10:34

great, thanks for the reply RG! I'll have to pick up some 6027s next time I put in an order and finally give this a try. if I ever get around to it, I'll try and draw up a layout for the forum as well- making sure I of course try to follow the above mentioned directions for the least amount of ticking possible.

nice job on the Ludwig over at DIYS, btw! it looks like it's coming along nicely!

R.G. · Post by **R.G.** » 13 Jul 2011, 15:00

RnFR wrote:great, thanks for the reply RG! I'll have to pick up some 6027s next time I put in an order and finally give this a try. if I ever get around to it, I'll try and draw up a layout for the forum as well- making sure I of course try to follow the above mentioned directions for the least amount of ticking possible. nice job on the Ludwig over at DIYS, btw! it looks like it's coming along nicely!

Thanks!

There are a few places on the net that show the way to sub a PUT for a UJT. It's pretty simple when you think about how a UJT works - one terminal is connected to a ramping capacitor; when it hits X voltage, the thing conducts massively, dumping the timing cap, and stays conducting until the current drops below the holding current. Then it reverts to non-conducting and the ramping starts again.

That's why UJT oscillators refuse to work at too-high speeds. If you decrease the timing resistor enough to speed up the timing cap a lot, it may provide more current than the holding current and the UJT locks on and never falls out of conduction, so the cap can't charge up again.

Renegadrian · Post by **Renegadrian** » 30 Jul 2011, 22:53

Built one using the vero layout - only differences, I didn't wire the wet pot and tried 2n3904 or bc548 as trannies - I got 5 motorola 2n2646 locally, 0.90 € each. no "chopping" tremolo but almost direct sound and a persistent ticking...grr...suggestions!^?

R.G. · Post by **R.G.** » 05 Aug 2011, 23:18

The ticking is a result of the layout.

A UJT works by storing up a charge in that timing capacitor then dumping the whole cap semi-instantly into the local ground. Since all wires are actually low value resistors, the wires from the UJT to the timing cap pulse with the passage of the discharge currents. That's your tick. If you cannot have that current pass through conductors that lead only back to the capacitor being dumped, then the sudden change in voltage on the "ground" will be carried off as a tick. It takes some work to make sure that does not happen. You have to think of every wire as a resistor, and then think about how to make the "tick" not flow through wire/resistors that the output will see as a moving ground.

Star grounding is not the only way to do this, but it's one of the few ways you can realistically say before hand that will work. Vero makes this much more difficult. I have not looked at the vero layout, but at a minimum, the timing cap needs to be on the opposite side of the UJT from the side that handles signal. It also helps if there is enough local power supply decoupling that all spike currents of frequencies above perhaps 100Hz can be satisfied locally from the decoupling capacitors and not from a remote power supply/battery that's hidden behind 100mOhm + 400uH of resistive/inductive wire lead.

The PCB layout I did shows a tick of under 100uV, which is sufficiently down in the mud that it effectively can't be heard.

The chopping should not be affected by this. That's another problem.

poodlebra · Post by **poodlebra** » 06 Aug 2011, 09:08

am i missing something? i can't see a PCB layout by RG here...?

thought i might remake this now i can etch PCBs - never been 100% happy with the one i made.

R.G. · Post by **R.G.** » 06 Aug 2011, 19:05

poodlebra wrote:am i missing something? i can't see a PCB layout by RG here...?
thought i might remake this now i can etch PCBs - never been 100% happy with the one i made.

Sorry - there isn't one available publicly. I don't release everything I do to the internet, for a number of reasons. Quite a bit, but not everything. This is one that's not public.

I didn't mean to be saying "I did something and you can't have it!". But people doing their own layouts needed to know that the ticking was a layout issue and that quiet *is* possible.

poodlebra · Post by **poodlebra** » 07 Aug 2011, 11:33

damn. ah well...

Renegadrian · Post by **Renegadrian** » 08 Aug 2011, 22:12

ok I'll try to get it silent!!! Thx for the precious info as always!!!

lolbou · Post by **lolbou** » 10 Jan 2012, 21:05

Can anyone with a genuine unit measure the minimum and maximum rate frequencies please (fairly hard to scope out of a youtube video

)? Though I know frequencies can be computed out of circuitry values, I do measure wide variations with the 16 devices I have here (which include 4 2N1671). While you're at it, Vv and Vp of the sawtooth at the UJT's emitter can be interesting for me too... Though I don't think they care that much with tight specs back when the V809 was made...

I get the 16 to oscillate, and 4 of these stuck at high rate setting (some late 80's Motorola ones from the same batch). When you lower the rate (ie: increase timing R), you get the oscillation back, but the lower side of the sawtooth is flat (at Vv : valley voltage). This flat portion disappear with even more increase. But again, increasing timing R means lowers min and max frequencies, so you have to adjust timing C accordingly.

Anyway, I'm collecting as much data as I can on these 16 UJTs because just to help sorting these out for best results along with some measurable parameters. I'm putting all of these in an excel sheet to cross these parameters with the waveform observed...

One thing that bugs me though is that Vv (valley voltage = bottom of the sawtooth) and Vp (peak voltage) does vary too from batch to batch (though being nearly identical in the same batch).

The drive signal from the emitter is divided and filtered by the 333k/82k/68n network but would it give the same depth or effect on the input signal? (I haven't breadborded this part of the circuit yet, but any input is appreciated

).

ps: and what about the Rate pot? if it's really 100k, it divides the min frequency by two...

acidfuzz · Post by **acidfuzz** » 11 Jan 2012, 18:33

Don't have time to measure the min./max. rate freq. right now, but after trying to build one of these off the wrong schematic floating aground years ago I finally popped open an original + tweaked the crap out of this circuit. I can confirm that the rate pot in both the plug in units and the Vox Ultrasonic guitar I have are 100k reverse audio taper. Actually they measure a bit higher around 125K, but probably just drifted over time. If it's getting all your UJT's to work in this circuit that bugs you, just replace the 4.7K resistor between the UJT's emitter (center leg) and rate pot with a ~20K trim pot. Using the 100K C taper rate pot in conjunction with a 20K trimmer has allowed me to use ALL 2n2646 UJT's I've tried without sticking open and with pretty consistent results in terms of the min./max. rate using the stock 10uf timing cap. FYI, I've also tried using more modern PUT's such as the 2n6027 as someone suggested earlier in this thread. They work, but don't sound the same to my ear, and it adds a few more parts to the circuit to get them ticking right. I know on paper all the 2n2646's are supposed to be ~ the same, but I've also found they all sound slightly different from one another. It's very subtle, but the old '60's -'70's production GE brand sounds the best to my ear, with the Motorolla's a close second. For some reason the newer brands, I think Multicomp?, just don't sound the same. I haven't scoped this, nd it doesn't make sense in theory, but to my ear they sound like they're shaving a bit off the low end and don't pass the full signal. Just a bit stale compared to the GE's.

The 333k/82k/68n network does work best as far as dividing and FILTERING the drive signal + I wouldn't touch that section or you just loose volume or get a very pronounced ticking. The rest of the circuit is up for grabs + I have a hand drawn schematic with lots of mods including added cut/chop, depth, volume, and tone controls + an led that flashes with the LFO. Those mods allow you to dial in subtle amp like bias vary/opto tremolos as well. I use that modded version for my space percussion pedal + the "repeater" section of the "sonic boom" pedal I build for Pete Kember. I'll post the schematic.

RnFR · Post by **RnFR** » 13 Jan 2012, 03:15

looking forward to seeing it!

disorder · Post by **disorder** » 14 Jan 2012, 03:40

I haven't built one but I'm helping a friend mod his... does increasing the resistance of the rate pot increase or decrease the rate?

lolbou · Post by **lolbou** » 14 Jan 2012, 21:24

disorder wrote:does increasing the resistance of the rate pot increase or decrease the rate?

The faster rate remains the same since you get this rate when the pot resistance is 0. The slower rate is slower with a higher pot value.

Here's the formula, with R being 4k7, C being 10µF and eta being the intrinsec standoff ratio (somewhere inbetween 0.5 and 0.

.

Code: Select all

f = 1/[RCxln(1/1-eta)]

disorder · Post by **disorder** » 14 Jan 2012, 23:40

Nice thank you! Theres a quite a bit of his in this build.. Enough so that the hiss gets "repeated" and you can hear it distinctly through the amp when not playing anything. How can I cut down on this? This build has a MPSA18 replacing the BC108 and inplace of the BC109 is the BC108. Could these transistor substitutions have any cause for the hiss/noise?

lolbou · Post by **lolbou** » 16 Jan 2012, 08:49

Thanks for your input acidfuzz.

It seems that "eta" is important (see formula above), but Vp and Vv (highest and lowest voltages of the sawtooth), and, even more Ip and Iv (currents at these emitter voltages) are crucial to get the UJT to work.

All are parameters that depends on the UJT itself. Only "eta" is constant, while the others are slightly varying with Vbb (interbase voltage).

I'm still working on it, but if you do get the scope to have min/max frequencies and Vv and Vp at the emitter, it would give a sorta target to aim at...

disorder wrote:Nice thank you! Theres a quite a bit of his in this build.. Enough so that the hiss gets "repeated" and you can hear it distinctly through the amp when not playing anything. How can I cut down on this? This build has a MPSA18 replacing the BC108 and inplace of the BC109 is the BC108. Could these transistor substitutions have any cause for the hiss/noise?

Try MPSA18 all the way? Did you use metal film resistors? If your hiss is modulated ( = "repeated"), it may be hiss coming from your input (or from the unit before?). How did you power it?

Vox - Repeat Percussion (V 809) [schematic]