HFT

[Magnus Danielson]

From: "Tony Allgood" <oakley at techrepairs.freeserve.co.uk>
Subject: HFT
Date: Wed, 29 Dec 1999 20:56:25 -0000

> Hi all,
> 
> Take one sawtooth VCO and standard expo. Whats the best way to get high
> frequency trim? The Moog method and others uses a 10K trimmer in series
> with the timing cap, seems OK but changes the output waveform. Rene's
> and Ian Fritz's method has a 1M trimmer & diode attached from the input
> of the expo to the output of the servo op-amp. Seems to alter the pitch
> at any frequency. And then there's Oberheim SEM method which takes a
> sniff of the expo current and passes that back to the CV input. But
> which is best for results and ease of setting up.

Actually, the old de Franco compensation, namely a resistor in series with the
capacitor is very simple, easy to calculate and only effects the waveform as
the frequency gets high. This will surely change the overtone spectra, but
I don't think it will make a drastic change in the hearable range.

I think that the de Franco compensation is sufficient and good enougth
compensation for normal sawtooth useage. Having made measurements on my ASM-1
VCO it was not until I went into excess of say 40 kHz that there was a really
drastic change in waveform. Assuming the the largest overtone changes is a few
tones up from that we are high up. Personally I don't think it is a real
killer. Also, the compensation as done in ASM-1 and numbers of other synths
will not effect the PWM properties unless you are at the lower extreme, and
then we are talking low energy in the low range anyway, so the overtone
changes isn't great there either.

I still have to hear a valid reason for it to be really bad for normal synth
usage. Special applications is another thing.

Ah, btw, I finally got the feedback trick of the SEM. I thougth that only had
to do with the expo converter but now I understand it, it compensates for the
fix reset time... I have to take a look at the math... unless someone has done
that and can present it ;)

BTW. I have been considering methods for compensating for the drift in DC
offset, it's a simple trick thougth... just subtract half of the DC
contribution of the de Franco resistor.

Cheers,
Magnus