Ian Fritz Sawtooth VCO not resetting

synchroma

Hi there,

I'm trying to base a module on the Ian Fritz sawtooth VCO, and I've run into trouble. I've made some small adjustments, such as using a 5.6V Zener diode-based reference voltage.

From my scope I can see that the integrator is charging to 5.38V (the actual zener voltage) and then staying there.

You would think that the comparator (LM319) is not doing its job, and I can reset the integrator by shorting out two of the pins.

It's been made using DesignSpark from RS components, can anyone think of a reason, or has anyone encountered this before?

Scratching my head (no emoticon for this Smile

Ed

Fritzy_Saw.jpg

Description:

Filesize:

102.51 KB

Viewed:

291 Time(s)

This image has been reduced to fit the page. Click on it to enlarge.

synchroma · Posted: Mon Sep 02, 2013 4:16 am Post subject:

Oh yes, forgot to mention:

Using a 2N4859 instead of the 2N4393 (can't seem to find a 2N4391 from my suppliers)

The op-amp AD711 is actually an LF351 in the circuit.

CR1 is the zener - the voltage reference is working properly - +5.38V

Exponential converters are on another board.

frijitz

synchroma

OK. It is based on:
http://home.comcast.net/~ijfritz/sy_cir2.htm

I've made adjustments like, the 2N4859 - looks like the KE4859 in some reference notes, and I can't get hold of a 2N4391 easily in the UK. Low Rds and fast switching time.

LF351 has the same slew (20v/usec) as the Burr-brown op amp which is also hard to source in the UK.

Apart from that, I'm not sure what changes you are referring to, but I'm open to suggestions.

Cheers,
Ed

frijitz · Posted: Mon Sep 02, 2013 3:09 pm Post subject:

According to the data sheet I'm looking at, the LF351 SR is 13V/us. Additionally, you are using a 5.5V reference vs the 4.0V of my circuit. You could try increasing the size of C4 to compensate for your changes.

It's almost impossible to read you schematic, so I can't comment further.

Ian

synchroma · Posted: Mon Sep 02, 2013 7:07 pm Post subject:

I stand corrected. And I'm feeling pretty humbled by your comments. I don't quite understand what you're saying about the schematic, because if I click on it it will enlarge and I can read it OK in the browser. DesignSpark doesn't make perfectly readable schematics but it does work OK for making boards and doesn't cost $0000s like Eagle.

I'll try what you said, and post again if I have any success.

Best wishes, and thanks for the circuits.

frijitz · Posted: Tue Sep 03, 2013 9:51 am Post subject:

synchroma · Posted: Tue Sep 03, 2013 10:11 am Post subject:

Hey man, that's cool. Yes you're right about the schematics that the software makes - they are a pain in the ass in terms of readability, but it's the only software I can get that does autorouting without paying a small fortune for it.

I'll try what you suggested once I've put the kids to bed tonight and post back. You've already taught me a lot here Smile

gdavis · Posted: Tue Sep 03, 2013 1:31 pm Post subject:

synchroma · Posted: Tue Sep 03, 2013 2:13 pm Post subject:

Hello gdavis,

Sadly no, I can't live with the tiny board sizes of free Eagle!
This VCO is just a halfway stage to the waveshape modulation oscillator I'm building - a bit like the Ian Fritz SNICster, but using a different technique. I already modelled it in DSP software and have been looking for an original way to do it for some time, and my boards are 160x100mm and 203x114mm.

Also, I want to sell some boards eventually, so Eagle prohibits this.

Yes, DesignSpark schematics suck! And the downloadable components usually don't have the correct symbols (you can see two ugly rectangular ICs where there should be an op-amp symbol and another two for the LM319) - but if you can live with these then the autorouting is good after a bit of practice.

OK, I'm off to try a 100pF cap now Wink

synchroma · Posted: Sun Sep 08, 2013 12:43 pm Post subject:

Hello Ian,

I'm realising now that a very small tweak to my waveshaper will allow me to implement the exact circuit you describe. However, I can't get all the parts for it, so I'm going back to the Bernie Hutchins design of 1976. I'll take it from there. All I need is a good ramp or sawtooth, and then I can make the audio frequency triangle -> sawtooth circuit I've designed work I think. This is a commercial project, so I don't want to rip off your SNICster - although that is significantly less complicated than my design in terms of BOM / component count.

However, if I can make it work I will publish designs for the waveshaper on my blog for the DIY community. But I want to make some new features of analog synthesis available to the masses.

I also have a new take on the envelope generator - attack, decay, sustain, swell, release. I'll post when I can test and refine my design.

Best wishes,
Ed (synchroma audio engineering, UK)

synchroma · Posted: Tue Sep 17, 2013 5:06 pm Post subject:

Hello Ian,

I've just ordered an AD817ANZ op-amp, with a very fast slew rate and settle time. I've got some 3.9V zener diodes and a pack of PN4391 FETs coming from Ray Wilson, so I should be able to make the circuit I built work. However, I'm hungry for knowledge so I really want to know how you would calculate the reset time of the comparator+integrator+FET combination.

Sorry for naive questions - I studied electronics years ago, but have only come back to it with any passion recently. Is the time-constant derived from the voltage divider (R23 and R24) and C4 in your schematic? And how does the slew rate and settle time affect this?

Thanks for your help and advice,
Ed

synchroma · Posted: Tue Sep 17, 2013 5:10 pm Post subject:

Oh yeah, I remember I was trying to cut corners with the LF351. Wrong slew - my bad Surprised