[sdiy] Working with the ASM1 VCO.. weird MAT-02 behaviour...

[Spiros Makris]

It is true that a low charge current will be less ideal because then
leakage currents etc become significant. They usually cause a rise to the
minimum frequency the oscillator will still work at.
One can lower the frequency of any similar oscillator by increasing the
timing capacitor appropriately and that should not impact the tuning range
etc (referring to caps with similar specs and different nominal
capacitance).

Spiros


On Fri, 17 Jan 2020, 14:49 Magnus Danielson, <magnus at rubidium.se> wrote:

> Hi,
>
> Recall that as you go for very low frequencies, you get very small
> currents, as you have a Volt to exponential Current converter followed by a
> Current Controlled Oscillator. For the low frequencies to work, it needs to
> be very clean for instance. It's not scaled to be a LFO, you could do that,
> but care must be taken in such an effort.
>
> I should dig up my ASM-1 board and measure this as I once intended to do.
> I built the ASM-1, measured it with a scope to span a large range. I then
> got a rubidium clock, a counter and a couple of DMMs and intended to
> measure this more carefully, learned more and now have a excessive time and
> frequency lab, but still have not returned to the original project of
> measure the ASM-1 VCO. Also got to repair the original rubidium clock.
>
> Hmm, maybe I should resurrect the ASM-1 homepage I maintained, just as a
> reference.
>
> Cheers,
> Magnus
> On 2020-01-16 17:48, Jean-Pierre Desrochers wrote:
>
> Hi list,
>
>
>
> I'm working on the Elby-Design ASM-1 VCO
>
> https://www.elby-designs.com/webtek/asm-1/vco/vco-asm1-cct.pdf
>
> Im want to use it as a 1v/oct LFO (.05hz to 20hz)
>
> I first mounted the circuit on my protoboard with success.
>
> at first I used a matched pair of 2N3904 for Q1A & Q1B
>
> instead of the MAT02 shown in the schematic.
>
> I could reach very low frequencies (as low as .04hz).
>
> But to do so I needed to replace R9(56k) for a 33k temporarily. Fine.
>
> This morning I replaced both 2N3904 with a genuine MAT-02
>
> and started the same frequency measurements as yesterday.
>
> I bumped into something weird..
>
>
>
> Here are the actual measurements I first did using 2x2N3904:
>
>
>
> mVolt at 1k TEMPCO resistor vs Sawtooth period
>
> 158.0mV => 0.050sec
>
> 176.0mV => 0.100sec
>
> 193.8mV => 0.200sec
>
> 212.0mV => 0.400sec
>
> 229.8mV => 0.800sec
>
> 247.5mV => 1.6sec
>
> 265.5mV => 3.2sec
>
> 283.2mV => 6.4sec
>
> 301.7mV => 12.8sec
>
> 319.7mV => 25.6sec
>
> (18mV steps, nice exponential response!)
>
>
>
> Now here are the actual measurements with a MAT-02:
>
> mVolt at 1k TEMPCO resistor vs Sawtooth period
>
> 157.0mV => 0.050sec
>
> 175.0mV => 0.100sec
>
> 193.0mV => 0.200sec
>
> 211.0mV => 0.400sec
>
> 229.0mV => 0.800sec
>
> 247.0mV => 1.6sec
>
> 271.0mV => 3.2sec
>
> **** Now trying to go further increasing the TEMPCO mVolts*
>
> **** the oscillator starts backing up and re-increase its frequencies !!!*
>
> **** I could not go further lowering the frequency. *
>
>
>
> I tried 3 different MAT-02's and even LM394's
>
> and they all behave the same.
>
> Using different matched pairs of 2N3904 worked perfectly..
>
> I suspect  MAT-02’s minimum base to emitter current reached (???)
>
>
>
> What am-I missing here ??
>
> By the way, all the IC's I'm using here are genuine one's
>
> (TL082, MAT-02, CA3140, LM311).
>
>
>
> JP
>
> _______________________________________________
> Synth-diy mailing listSynth-diy at synth-diy.orghttp://synth-diy.org/mailman/listinfo/synth-diy
>
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at synth-diy.org
> http://synth-diy.org/mailman/listinfo/synth-diy
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20200117/ad62ccf8/attachment.htm>