Buchla 'Timbre' [Easel/259]

[Photon]

I haven't looked too close at this, but it looks like a different trace was connected rather than the one suggested. I'm not sure which is correct.

---

fix.png
Description:
Filesize:	286.26 KB
Viewed:	336 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

[TekniK]

[ultrashock]

Dear guys!
Have you experimented with op-amps other than discounted CA3160?
Since the D. Bucha used 3160 in 259's Timbre, there were RC4136 (which are basically the uA741 in origin) in the Easel, and thus Lanterman supposed to substitute them by the "nearest" RC4558.
I made some Timbre model in ISIS Proteus. Nothing really special, just a Timbre-circuit with the scheme aforesaid and I tried LM358, TL072/074 and other ones ^_^ (unfortunately I did not managed to find the RC4558 or RC1558 spice models) and even so I observated some sine "shaping" into square wave with enriched harmonics in virtual oscilloscope (CPU activity was almost 100% ). That is why I made the assumption for myself that it could be almost any op-amp (even better hi-fi OPA2137 or OPA4134), the difference could be only in the different sonic possibilities.
What do you think about this?

p.s. : I also found some substitutes for J201: sst201, bf512, mmb4393.. yes, they are smd ones.

p.s. 2: thank you for the trick of having the 75K and 49K9 resistors, but I can not find those you told 72K and 43K in the circuit

[janvanvolt]

A short debug/wiring question (and verification):

- i am using R4558 and the respective resistors
- i am _currently_ using 1N4148, the 1N45 are on order.
- i build the PCB with "buchla_timbre_164.pdf", not with the recent one.
- if i apply the named kludges, e.g. 2 wires plus one +V connection, all should be fine ?
- i am NOT applying the small red dot on the lower part of the PCB

correct ?

1) Regarding wiring, do i need to connect some of the PRE Gain/Post to something (except the switch) ?
2) Audio level seems a _bit_ low ( need to crank up my VCA full to get a signal), so do i need to tweak there something? ( Resistor X?)
3) XFADE itself does not seem to work at all.
4) Literally the question stands: how to verify proper operation ?

For the curious: Panel attached.

---

panel.png
Description:	MOTM Panel
Filesize:	18.95 KB
Viewed:	379 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

[forbin]

Hi Ultrashock

I used TL702's for ALL of the opamps and it worked fine -- it isn't really critical. Probably a bit of a waste using OPA's though... The FET probably doesn't need to be a J201 -- really any linear JFET will work it is more where the Vgs threshold is that is important. It is really just being used as a sort of voltage controlled resistor to attenuate the input signal. The 72K and 43K resistors are not direct values in the circuit but formed with two resistors in parallel. Don Buchla used a 49.9K//390K to get 43K -- i found that it was a lot easier to get that value using a 470K//47K as they are easier to get than the 49.9K. Ditto for 75K//1.5M using a 180K//120K.

Hi Janvanvolt
If you have the OLD PCB then you need TWO +V connections plus the other two jumpers so FOUR bits of wire/bridges in total. I still havn't got back to the XFADE section -- try putting an LED in circuit first as it gets rather expensive if you toast a VTL52!

By stepping the control voltage and feeding in a sinewave in simulated this. Which was pretty close to what I measured. The problem that you get in reality is making the control voltage adjust the amount of folding evenly. Ideally the gate of the FET should be driven by the control voltage through an opamp with gain and offset control to maximize the usable range... fairly messy to configure though...

---

wavefolder_sim.jpg
Description:
Filesize:	284.54 KB
Viewed:	482 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

[janvanvolt]

Just to be sure (again):

- left side of "fix.png" (previous posts) is the right one, so :

1) two RED connections to V+ on the left side,
2) orange wire
3) left wire on of the upper right, as the right wire already is going thru on my PCB.

please confirm shortly.

[forbin]

Hi Janvanvolt

yes -- the red +V connections, the orange wire and the one left wire; i moved the existing upper-track a bit to show it could be done easily on the PCB...

[mono-poly]

I still think this is awesome

[Dego]

Me too!

With odd and even you got yourself a nice vco...

http://hem.bredband.net/bersyn/VCO/vco_complex.htm

[janvanvolt]

Any news on the XFADE Section ?

[forbin]

Xfade is crawling towards completion...

I have put in an LED and can certainly modulate it's brightness so the voltage control section seems good. Having said that the control range does seem pretty small. After about 1.7V the current limit protection circuit has kicked in -- this is equivalent to about 7mA through the LED in the vactrol. I have a suspicion that using a 1N4148 diode instead of a 1N457 may be the cause... Fixes are probably lowering the series resistor for the Vactrol LED; the data sheet suggests that the maximum current is 40mA and 7mA is well shy of that.... My Vactrol has turned up so I will try it as it certainly should be OK!

Will let you know how go tomorrow!

[forbin]

XFade -- well fitted the vactrol and all seems well! The vactrol is horribly slow but i suppose that is the effect that we are after... The worry about the current through the LED seems somewhat misplaced and if anything the 7mA is more than enough and will certainly prolong it's life. Here is a CRO screen capture to display effect of a 3 second sawtooth modulating a sinewave.

---

dsc03920.jpg
Description:
Filesize:	459.94 KB
Viewed:	1872 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

[janvanvolt]

I was using 1N4148 (and exchanged it now to 1N467) but i have issues with the XFADE. It does not seem to work at all ( or very quiet)

1)
Do i need to connect or use the Pre Gain/Post Gain Timbre Out ?
And what about XFADE.IN 1 Buffer out and XFADE.IN 2 Buffer out ?
(i did connect the Switch up/down)

2)
By watching Aarons Video ( http://il.youtube.com/watch?v=Pi3H1M555-M&feature=related) i saw that he swapped some PINs on the OPamps.
Is this fixed in the current design ?

[janvanvolt]

anybody able to help me troubleshooting ?

[TekniK]

[janvanvolt]

[TekniK]

I can't help you with that as i did not built the pcb yet,and i think am not gonna built the panner section anyway.

[janvanvolt]

[forbin]

Hi Janvanvolt

I will take another look at it in the morning and see if I can give you some more help... It does work but the scaling of the control and audio signals is a bit unusual for most modern synths. This part of the design was originally designed for providing a sort of morph between different waveforms from a VCO and that isn't really how it is being used here... I have sort of ended up using it just as a VCA (or gate to use Buchla parlance).

[TekniK]

I built the timbre circuit but it does not sound like it should be,did drive it with a proper sine but its to fuzzy,as if there is to much distortion at input.

The other pcb in the module you can see on the picture is the original serge version.

---

Timbre.jpg
Description:
Filesize:	852.28 KB
Viewed:	1886 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

[janvanvolt]

I am likely parting the module. Too much to hassle with. I have a dozen other open projects still pending and it does not bring me anything.

[vtl5c3]

only a dozen?

[forbin]

Hi Janvanvolt; sorry it has taken so long to reply! I think that the problem you are seeing is the perennial one with a lot of Buchla designs of that era in that the FET biasing of the Vgs is pretty critical. It tends to vary with FETs of the same type and even manufacturer. I had a lot of grief with the one in the waveshaper of the 258 VCO. It almost needs the gate to be driven with an op-amp with adjustable gain and offset to optimize the range of operation. All I can really suggest is experiment with other FET's and see... not very helpful really...