Author |
Message |
synchroma
Joined: Oct 18, 2011 Posts: 72 Location: London
Audio files: 2
|
Posted: Sun Sep 29, 2013 6:35 am Post subject:
Sawtooth VCO - help!!! Subject description: I built a Bernie Hutchins sawtooth VCO and can't get it to work properly |
|
|
Can anyone help me out here? I'm losing the will to live.
My 3rd attempt at building an integrator + reset sawtooth VCO.
I'm following the schematics for the ENS 76 VCO option 1:
http://www.synthsource.com/ens76/vco2.htm
But for the exponential converter I'm following Ian Fritz' Sawtooth VCO schematics:
http://home.comcast.net/~ijfritz/sy_cir2.htm
It doesn't make any sense: all I get is a vaguely trapiezoidal wave at ~200KHz. Panel controls don't make any difference.
Sorry about the DesignSpark schematics - I've tried to make them as readable as possible. Panel controls are on a separate board and are not shown, but they closely follow the Ian Fritz schematic with the addition of an extra exponential FM control+socket.
?????
Description: |
And this is the exponential converter |
|
Filesize: |
78.08 KB |
Viewed: |
261 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
Description: |
This is the ENS-76 VCO option 1 board |
|
Filesize: |
96.89 KB |
Viewed: |
234 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
Description: |
This is the output of the oscillator |
|
Filesize: |
177.96 KB |
Viewed: |
255 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
|
|
Back to top
|
|
|
synchroma
Joined: Oct 18, 2011 Posts: 72 Location: London
Audio files: 2
|
Posted: Sun Sep 29, 2013 6:36 am Post subject:
|
|
|
Oh, by the way, the 1.5K resistor in the sawtooth schematic is actually a 15K resistor. I already spotted and corrected this error. |
|
Back to top
|
|
|
elmegil
Joined: Mar 20, 2012 Posts: 2177 Location: Chicago
Audio files: 16
|
Posted: Sun Sep 29, 2013 3:06 pm Post subject:
|
|
|
Very high frequency oscillation sometimes means your op amps need compensating capacitors, usually in the 10pF to 100pF range (I'd start smaller and work up).
Dunno if that's much help, but probably worth a shot |
|
Back to top
|
|
|
synchroma
Joined: Oct 18, 2011 Posts: 72 Location: London
Audio files: 2
|
Posted: Sun Sep 29, 2013 3:25 pm Post subject:
|
|
|
There was one on the original Electronotes schematic. Bernie Hutchins hinted that it might no be necessary if you used an LF351 which is what I actually used, but I think you might be right. The first schematic has a 39pF cap as compensation.
I'll try it, and let you know how I get on.
Thanks,
Ed |
|
Back to top
|
|
|
piedwagtail
Joined: Apr 15, 2006 Posts: 297 Location: shoreditch
Audio files: 3
|
Posted: Sun Sep 29, 2013 10:31 pm Post subject:
|
|
|
Best to help yourself by reducing the risk of it not working.
Junk the expo circuit and mock up a simple linear one instead to first get the integrator working, then re-introduce the expo. Don't try and solve too many variables.Keep looking.Use a continuity tester to check joints.
The elephant in the room is the the jfet
http://www.geofex.com/article_folders/fetmatch/fetmatch.htm
how many have you tried?
R |
|
Back to top
|
|
|
synchroma
Joined: Oct 18, 2011 Posts: 72 Location: London
Audio files: 2
|
Posted: Mon Sep 30, 2013 2:19 am Post subject:
|
|
|
Ah. Ouch!
Too many variables, but I'm determined to get a reliable design together. I'll try a linear current source with it. I've been going down the road of gradually splitting everything into it's component parts, which is why I'm only building the core of the VCO here.
JFETS - they do seem like a poisoned chalice in terms of reliability and specification. I am thinking that maybe a fast analogue switch like 1/4 of a CD4066 would be better. |
|
Back to top
|
|
|
synchroma
Joined: Oct 18, 2011 Posts: 72 Location: London
Audio files: 2
|
Posted: Mon Sep 30, 2013 3:13 am Post subject:
|
|
|
Oh yes, only tried the 2N4391 so far. I have a 2N4859and a 2N4856 somewhere, so I think I'm going to build another with a transistor socket to test them. |
|
Back to top
|
|
|
jackdamery
Joined: Apr 26, 2010 Posts: 75 Location: UK
|
Posted: Mon Sep 30, 2013 4:20 am Post subject:
|
|
|
Are you building this particular VCO for any reason? If not, I can suggest an easier one.
If yes,
How many volts are you supplying it?
First thing, remove the expo converter circuit so you can just troubleshoot the oscillator. Connect a pot biased between +V and -V to the input. Tweak the pot until you get the slowest possible oscillation.
Try increasing the value of the capacitor (C2) between the comparators input and output to ~100pf. If this makes no difference then...
Try using a larger cap in the integrator. increase 2.2n to 10n. If this makes no difference...
Try a Tl071 instead of a LM311 for the comparator. You will definitely need C2 to be 100pf for this. If its still too fast increase C2 to a higher value.
Replace the voltage divider on pin 3 of the comparator with a pot so you can adjust the level at which the oscillator resets.
Re: the JFET. You can also try a BF245. Last edited by jackdamery on Mon Sep 30, 2013 4:51 am; edited 3 times in total |
|
Back to top
|
|
|
Jerry_100
Joined: Nov 18, 2011 Posts: 6 Location: Bavaria
|
Posted: Mon Sep 30, 2013 4:33 am Post subject:
|
|
|
Is U4 meant to be a dual power supply since the original circuit IC2 runs single power supply ? |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Mon Sep 30, 2013 7:46 pm Post subject:
|
|
|
It probably won't work with a TL071 for the integrator. It's a tight design, and you cannot just make arbitrary component substitutions.
Ian |
|
Back to top
|
|
|
jackdamery
Joined: Apr 26, 2010 Posts: 75 Location: UK
|
Posted: Tue Oct 01, 2013 12:42 am Post subject:
|
|
|
frijitz wrote: | It probably won't work with a TL071 for the integrator. It's a tight design, and you cannot just make arbitrary component substitutions.
Ian |
Do you mean the comparator? I subbed the TL080 in the Jupiter 8 - a similar circuit for a TL071 and it performed just fine.
synchroma can you point your oscilloscope at pin 7 of the LM311 so we can see what the comparator is doing. I suspect the pulse width is too long and we are seeing the integrator charge to full (the ramp) then wait to be reset (the flat line) then discharge slowly (the falling ramp). |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Tue Oct 01, 2013 9:53 am Post subject:
|
|
|
jackdamery wrote: | Do you mean the comparator? I subbed the TL080 in the Jupiter 8 - a similar circuit for a TL071 and it performed just fine. |
No, I meant what I said. And I don't see how your Jupiter 8 comparison is relevant.
Ian |
|
Back to top
|
|
|
synchroma
Joined: Oct 18, 2011 Posts: 72 Location: London
Audio files: 2
|
Posted: Fri Oct 04, 2013 12:57 pm Post subject:
|
|
|
OK. I'm beginning to realise how critical the choice of op amp is for many of the designs I find out there! Apologies for using the "TL071" symbol in my schematic - it was a means to an end to get a working symbol/PCB component. I realised I had a fairly bogus Eagle op-amp library I imported into my CAD package with the wrong pinouts, and had to go with the crappy downloadable part spec from the software makers :-/
So, the TL071 is actually an LF351 in the circuit. Forgot to mention it, sorry! I was going with what Bernie Hutchins wrote that "You can also use an LF351 for the CA3140. You can also use an LF351 for the CA3130, but it would be run on the bipolar supply, and without the 39p compensating capacitor." (source: http://www.synthsource.com/ens76/vco4.htm )
It seems that this is incorrect. It didn't work.
So...I've bought a load of components: CA3130 and CA3140, OPA2132 (can't seem to get the OPA132 in the UK) LM311 and LM319 comparators, LM329DZ precision regulator and a bunch of other tightly spec-ed components (2N4391, 2N4859, 60K4 resistor etc). And, I've made a board with three different saw cores on it which (if my kids let me) I'll solder and test over the weekend.
Hutchins / Mikulic (ENS 76)
Rene Schmitz
Ian Fritz (with the other half of the LM319 used for PWM)
I still have the expos I made from Ian Fritz's design so I'll plug them in once I know the saw cores work (I'll knock up a linear current source to test them first).
Cheers
Description: |
Only top copper is shown. If they all work as intended, and if there is some interest, I'll make the design available here. |
|
Filesize: |
712.11 KB |
Viewed: |
197 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
Last edited by synchroma on Sun Oct 06, 2013 11:26 am; edited 1 time in total |
|
Back to top
|
|
|
synchroma
Joined: Oct 18, 2011 Posts: 72 Location: London
Audio files: 2
|
Posted: Fri Oct 04, 2013 5:37 pm Post subject:
|
|
|
PS jackdamery, thanks for your tips. The most useful here. I need to understand the relationships between reset time, slew rate and capacitor value. I'll experiment with what I've got.
Thanks,
Ed |
|
Back to top
|
|
|
synchroma
Joined: Oct 18, 2011 Posts: 72 Location: London
Audio files: 2
|
Posted: Sun Oct 06, 2013 11:28 am Post subject:
|
|
|
I notice that threads go cold on this forum after the first post from everyone. So, I'll start a new thread with the results I have from my build, which is less than encouraging as far as me working this out is concerned!
If only Thomas Henry had an integrator+reset design. His designs actually work.
I have now etched an built 5 sawtooth VCO cores without a single working oscillator. Mighty miffed. |
|
Back to top
|
|
|
elmegil
Joined: Mar 20, 2012 Posts: 2177 Location: Chicago
Audio files: 16
|
|
Back to top
|
|
|
synchroma
Joined: Oct 18, 2011 Posts: 72 Location: London
Audio files: 2
|
|
Back to top
|
|
|
elmegil
Joined: Mar 20, 2012 Posts: 2177 Location: Chicago
Audio files: 16
|
Posted: Sun Oct 06, 2013 3:22 pm Post subject:
|
|
|
What's the other side of the board look like? |
|
Back to top
|
|
|
synchroma
Joined: Oct 18, 2011 Posts: 72 Location: London
Audio files: 2
|
Posted: Mon Oct 07, 2013 4:34 am Post subject:
|
|
|
Here is the board, both the artwork and a superimposed component view. It's not very well labelled I'm afraid, but I've taken as much care as possible to follow the schematics I found online. You can view the full 600dpi from the zip file.
A few changes - the ENS76 Option 1 VCO has two voltage followers providing 2 outputs (LF353 to the right of the board) and I've put together a PWM section from the other half of the LM319 in Ian Fritz's schematic.
Description: |
|
Filesize: |
433.47 KB |
Viewed: |
161 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
Description: |
|
Filesize: |
291.16 KB |
Viewed: |
201 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
Description: |
|
Download |
Filename: |
SawCores.zip |
Filesize: |
1.78 MB |
Downloaded: |
148 Time(s) |
|
|
Back to top
|
|
|
synchroma
Joined: Oct 18, 2011 Posts: 72 Location: London
Audio files: 2
|
Posted: Mon Oct 07, 2013 4:39 am Post subject:
|
|
|
Oh yes I forgot to mention - the first version I made that started this thread off - the JFET connections were wrong. I've been double-careful to get them right this time. |
|
Back to top
|
|
|
elmegil
Joined: Mar 20, 2012 Posts: 2177 Location: Chicago
Audio files: 16
|
Posted: Mon Oct 07, 2013 5:35 am Post subject:
|
|
|
That helps, but I was actually more interested in a photo of the back of the board |
|
Back to top
|
|
|
synchroma
Joined: Oct 18, 2011 Posts: 72 Location: London
Audio files: 2
|
|
Back to top
|
|
|
synchroma
Joined: Oct 18, 2011 Posts: 72 Location: London
Audio files: 2
|
Posted: Mon Oct 07, 2013 2:07 pm Post subject:
|
|
|
By the way all, the waveform at the top of this post was gone after I tested it again. It went back to what I've got each time - the capacitor charges and keeps on charging - no reset.
I think I'm going to build some test circuits to see if my chips are fried. |
|
Back to top
|
|
|
elmegil
Joined: Mar 20, 2012 Posts: 2177 Location: Chicago
Audio files: 16
|
Posted: Mon Oct 07, 2013 6:07 pm Post subject:
|
|
|
Your 3130 doesn't have the compensating capacitor between pins 1 & 8 that is in the original.
In the original circuit, the 3130 is powered single supply, so pin 4 is ground, not Vss.
Here's the schematic I'm reading that from: http://www.synthsource.com/ens76/vco2.htm (2 refers to page number, not option number).
Does anyone else know if there's a decent way to test an expo converter by itself? That's the other place I'd be looking, but I'm not sure whether running a resistor to ground would let you check the output or would just load it down. |
|
Back to top
|
|
|
synchroma
Joined: Oct 18, 2011 Posts: 72 Location: London
Audio files: 2
|
|
Back to top
|
|
|
|