Living VCOs

[camelneck]

In the BOM, Juergen says that we could substitute other matched transistor pairs for the LM3046's. I was wondering if anybody had
compared the tracking between a VCO that uses the LM3046 versus
a matched pair with a bit tighter specs such as the 2SC1583.
(It's been my experience that using a transistor pair with tighter matched specs will improve tracking unless there are other factors that limit tracking within the VCO.)

I was thinking about using 2SC1583's if I knew it would improve the tracking. Otherwise, I'll probably just use LM3046's since I have a lot more of them and they are significantly cheaper.

[rutgerv]

[rutgerv]

In addition to my previous post I would appreciate some comments/suggestion on an idea to add a linear FM input to the living VCO. The attached schematic shows part of the expo convertor with an additional inverting op-amp and resistor connected to the shared emitters of the matched NPN pair (LM3046).

I have a hard time understanding the exact operation of the BC560Cs (Q1 and Q2 in the original living VCO schematic). Is the shared emitter pin of the LM3046 the appropriate place to inject the linear FM signal?

And is there a risk of making the VCO less stable (or too much "living") with such a modification?

Rutger

---

spice1.jpg
Description:
Filesize:	50.47 KB
Viewed:	48939 Time(s)

---

[/mr]

Thanks for an interesting post.

[rutgerv]

I believe this is due to the time needed for the capacitor in the pulse-shaping circuit to charge and discharge. The spike could be made more narrow with different capacitor values, but as a consequence the pulse height will decrease. At some point it may not trigger the transistor anymore. This could probably be fixed with an op-amp in the sync circuit, but I didn't want to take the route of adding that many components.

[synthetic]

Parts ordered, getting started on this next week. Has anyone found a good source for vernier pots/knobs, or is ebay the best bet?

[LFLab]

Last time I checked Reichelt.DE had good prices on multiturn spectrol pots and vernier dials, but they are on this side of the pond, not sure where to get them on your end, they can be pretty pricey, so surplus (or eBay) might be your best bet.

[emdot_ambient]

I just got 3 from this seller: http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=280504171334&ssPageName=STRK:MEWNX:IT

Good price, brand new item in box with all the mounting stuff, quick delivery.

Big, but really good quality.

[synthetic]

Thanks, emdot. How big is it, what's the diameter? The auction doesn't specify.

Also, never having used this type of control, is there a course control? Is that what the little tab does? Or do you have to turn over and over to reach the end?

[emdot_ambient]

They're something like 46mm wide, maybe a bit more (46.4mm or so).

[LFLab]

[emdot_ambient]

JarnoBassplayer's correct: that "tab" as you put it, is actually called the brake. It locks the dial so you won't accidentally knock it once you've set it to where you want.

Also something to note about buying multi-turn pots...I've read complaints that a lot of the multiturn pots are too loose--they turn REALLY easily.

Well, I've found some that to me feel perfect when used with regular knobs. They feel a little looser than alphas, which actually I like.

HOWEVER, this particular dial, which is made for up to 14-turn pots, has its own inherit resistance. It feels perfect with no pot to offer resistance. When I use one of my "perfect" feeling 10-turns......suddenly the dial has too much resistance, feels downright stiff.

When I connect it to a very loose Vishay I have, viola! perfect again.

So, now I know why many multi-turn pots are as loose as they are: used with the correct dial they're marvelous!

[synthetic]

The more I think about it and look around, I'm starting to think that something like the 921b tune control is what I want to do. This system has a fine tune +/-12 semitones, then it has an octave switch. I think I would use a 3-position toggle instead of the six-position rotary, but put this on each of the oscillators.

Now I just gotta figure out how to do that. Does this seem possible?

[synthetic]

I plugged it in tonight and my whole modular went nuts, so there seems to be something wrong with the power supply. I measured between the different legs and there is about 3k ohms resistance between -15V, ground, and +15V. What should I be seeing here? Thanks.

[Jaba]

Hi,

I am very interested by this VCO design, mostly because of this "linear detune" feature;

I would like to know if there are easy ways to add this feature to existing VCOs too.

If I got it right, in any VCO that has this same architecture, with expo converter and OpAmp-buffered timing cap, it should be possible to just add the three resistors (R14, R15, R16) and a pot and voila, is it really that simple ?

For VCOs with a passive integrator it may be more complicated, I guess.
I read on JH's site that he published something about this, is it available ? can I read it ? would someone point me to it please ?

I would even try to apply this linear detuning to other VCOs such as the Thomas Henry xr2206 based and 4046 based designs, or the OTA based ones, but I have no idea what the rules are: can I just add a small current with a high impedance, or there is more that I should be aware of ?

one last questionmark: if I use a "normal" VCO (i.e. stable and well tuned) to put a saw wave into one of the sawshifters such as Yusynth's SawAnimator, or CGS38 SawPitchShifter, ETI '83 "Waveform Multiplier", ... etc., isn't it more or less the same as Linear Detuning ? sounds similar, and in theory too ...

I also want to experiment with HardSync following rutgerv's suggestions, as soon as I have the LivingVCOs PCB in my hands, but this is another story, I'll report the results, if any.

Paolo

[oldcrow]

Hey folks, can someone tell me where these resistors connect on their board? It is not in JH's schematic. One end is taken to ground, and I imagine the other end is attached to something like the pulse output as part of a level shift. It may be he omitted these resistors in the end and used the diode clamps only.

Crow
/**/

[The Bad Producer]

I have a board in front of me, as it happens! But which resistors do you mean?

Edit: Oops, it was in the subject!

R6 connects from junction R31/32 to GND...

[oldcrow]

Thanks, it was attached how I thought. --Crow

[prgdeltablues]

I'm planning on building a Living VCO as the first module in my next cabinet, by way of my tribute to Juergen. It looks as if it should run off 12V (which I use) rather than 15V - there's a comment early on about changing some resistor values: does anyone know which? I'm guessing possible candidates are R7, R12, R13, R19, R29, R30, R31, R28, R34, R37 (and the equivalents on the second and third VCOs). I also suspect not all would actually need changing.

Peter

[prgdeltablues]

Can anyone explain how Q1 and Q2 work in the expo converter?

It looks as if they provide the current reference that is more usually done by an opamp, but - unless there is an error on the schematic, the collector of Q2 is not connected, and the b-e junction looks to be reverse biased. Is this a noise source then ? - the 'noisy servo' the late and sadly missed Juergen refers to?

I've tried to model one section of the VCO in LSpice, but so far without luck - I get a low amplitude oscillation at about 1 MHz at the saw output, almost independent of the CV input. Not sure what is wrong...

Peter

[prgdeltablues]

Answering my own post (how sad is that?) in case anyone else is interested. I've got the Living VCO core running fine off +/- 12V, with the following resistor changes (referring to page 1 of the schematics - obviously the same changes to the other two cores as well).
R5 - 120k
R7 - 390k
R12 - 1k8
R22 - 47k
R24 - 4k7
R33 - 33k
R34 - 10k
R36 - 560k

Not yet built the driver section.

I also used two hand matched BC549C's instead of the LM3046 - seems to track fine, and used a 47k for R10 - with a 10k trimmer (the target resistance is just under 56k, so this is still within range).

[jumunius]

[prgdeltablues]

Hi

Just built the driver section of my Living VCO. It works fine, except for the portamento. When I turned that down to 0. the signal out was distorted: eventually I identified it as the same sort of warbling effect you get if you plug a lead into a V/Oct Cv input and just hold the end, which makes me suspect interference of some sort. Shorting out the 1MEG portamento pot solved the problem - nice clean signal.

Before I decide just to do without the portamento, I wondered if anyone had any clues. I thought maybe having unshielded leads to a 1MEG resistance could be the source of the problem?

Peter

[prgdeltablues]

Problem solved!

I'd read the schematic as showing the 1k and 1u cap to earth as hanging off the centre tap of the 1MEG pot, such that there was always 1MEG resistance between the two opamps. But in fact it ought to be read as a variable resistor, with the 1k and cap to earth off one end. Now working fine. Great vibrato.

To complete the info on running the Living VCO on +/-12V: for the driver, I changed R160 to 75k (and also added a 10k to ground off the Depth pot, ass I'm using banana jacks), R150 to 10k, and R140 to 150k.

Now just got to make the front panel.

Peter

[kvitekp]

Hi,

Is the pcb for JH Living VCOs still available for sale anywhere?

Thanks,
/Peter