Author |
Message |
tommi
Joined: Dec 05, 2007 Posts: 247 Location: Italy
Audio files: 3
|
Posted: Fri Jan 08, 2010 11:36 am Post subject:
|
|
|
Here are two samples: in both i got a triangle from the xr vco and a saw from mfos vco.
In the first i got a sequence running. I start in both mode and i slowly turn up the resonance pot till fully cw. You can hear distortion growing.
Then i switch to lopass mode and i do the same with resonance.
Finally i switch to gate mode. Here the resonance seems to have no effect as it should, but the sum of the two inputs are overdriving more than in the other two modes.
In the second sample (no sequence) i start in both mode again, with the offset open and both volumes low, then i turn up the volume pots and here you can hear how they distort when at full volume. Then i turn up the resonance wich overdrives a lot...
then i do the same in the other two modes...
I think it's clipping distortion rather than noise. What you say?
Description: |
|
Download |
Filename: |
LPG1.mp3 |
Filesize: |
630.02 KB |
Downloaded: |
912 Time(s) |
Description: |
|
Download |
Filename: |
LPG2.mp3 |
Filesize: |
473.9 KB |
Downloaded: |
893 Time(s) |
_________________ http://soundcloud.com/mister-vommi
http://tideofsound.net |
|
Back to top
|
|
|
The Bad Producer
Joined: Mar 08, 2009 Posts: 282 Location: The Manhole
|
Posted: Fri Feb 19, 2010 5:10 pm Post subject:
|
|
|
Interestingly I've just completed another of these modules and it exhibits the same behaviour as Tommi's...
I've built a few already that worked well, but this latest is really weird...
I first used a TC4053BP (not sure where I got that from!) and as I turned the resonance knob up it would clip and then eventually it would just cut out - as if something was being fried, basically the module stopped working. I just swapped out the chip for a CD4053 and the problem is the same, clipping/distortion as the resonance gets turned up, it doesn't stop like the other chip, but it sounds pretty fruity - in fact if I thought it was supposed to do that I'd love it as a distortion effect!
The other modules I made use the MC14053, and they run at 15V. This one is for 12V (I changed R3 on the LPG PCB to 120K) but I'm testing at 15V, I didn't think this would matter...
Tommi, have you got any further in your investigations?
I'll test some more stuff tomorrow, but for now I'm off out!
Charlie _________________ http://loudestwarning.tumblr.com/ |
|
Back to top
|
|
|
tommi
Joined: Dec 05, 2007 Posts: 247 Location: Italy
Audio files: 3
|
Posted: Fri Feb 19, 2010 5:47 pm Post subject:
|
|
|
Hi Charlie,
I 've ended up leaving off the gate expantion board
But i still got another LPG module with the expantion board on the corner with the hope to fix it one day in the future...
If i could try one TC4053BP i would say if it works.
About the investigation, I am still on the path where i was at the time of my previous post, since then i 've built a new panel to leave off the exp. board and add the 3P3T switch. _________________ http://soundcloud.com/mister-vommi
http://tideofsound.net |
|
Back to top
|
|
|
fonik
Joined: Jun 07, 2006 Posts: 3950 Location: Germany
Audio files: 23
|
Posted: Sat Feb 20, 2010 1:30 am Post subject:
|
|
|
this is really annoying. i will take a look at the ICs i used.
however, the fact that you both run into problems as soon as powering from 12V seems suspicious to me...
anyways, i don't want to leave anybody out in the rain, so i could refund everybody who wants me to. though it would better to solve the problem, wouldn't it? _________________
cheers,
matthias
____________
Big Boss at fonitronik
Tech Buddy at Random*Source |
|
Back to top
|
|
|
The Bad Producer
Joined: Mar 08, 2009 Posts: 282 Location: The Manhole
|
Posted: Sat Feb 20, 2010 5:25 am Post subject:
|
|
|
Refund? WTF, I'd rather get it going!
However I'm not an EE so you'll have to bear with me! I tried it on 15V seems to be the same problem, I'll pull them both out and see if I can spot the difference, I have four good and one bad, and some more to make, so I'm sure it is solveable... _________________ http://loudestwarning.tumblr.com/ |
|
Back to top
|
|
|
The Bad Producer
Joined: Mar 08, 2009 Posts: 282 Location: The Manhole
|
Posted: Sat Feb 20, 2010 5:34 am Post subject:
|
|
|
Okay here are the differences on visual inspection between good and bad...
On LPG board itself:
good uses VTL5C3
bad uses VTL5C3/2
good R3 is 150K
bad R3 is 120K
On the exp board:
good uses HCF4053
bad has CD4053 (tried TC4053, but much worse)
tested both at 15V...
I'll plug both in and see if I can get some readings... maybe get out the o'scope... _________________ http://loudestwarning.tumblr.com/ |
|
Back to top
|
|
|
The Bad Producer
Joined: Mar 08, 2009 Posts: 282 Location: The Manhole
|
Posted: Sat Feb 20, 2010 5:56 am Post subject:
|
|
|
Okay, 1st test, I put the HCF4053 in the bad module, same problem distortion on 'both' and 'lopass' modes when resonance turned up, the offset pot has little (no) effect, I wonder if that is the problem (the offset trim does have some effect...)
I also tried the CD4053 in the good module and no problems at all, so I think we have discounted the chip IC as being the problem...
Basically the good module is good, no matter what 4053 is used, and the bad is bad no matter what 4053 is used... _________________ http://loudestwarning.tumblr.com/ |
|
Back to top
|
|
|
fonik
Joined: Jun 07, 2006 Posts: 3950 Location: Germany
Audio files: 23
|
Posted: Sat Feb 20, 2010 6:24 am Post subject:
|
|
|
The Bad Producer wrote: | However I'm not an EE so you'll have to bear with me! |
i fear you'll have to bear with me first. i am not an EE too
anyways, thank you for your hard work so far.
concerning the differences: R3 is on the 'diode side' of the LPGs circuit, so i cannot see where/how it should interfere with the gate board. same for the vactrols. AFAIK hte only difference between single and dual vactrols is the diode count.
just a wild guess: if it is not the IC and it is not the 12V (less headroom, signal width)... the LPG is only connected to the 4053. all other components of the gate board only control the 4053. so what about the bufferred 7.5V (6V) on the gate board? maybe you could try another opamp or try filter caps between pins 5/3 and GND?
if you eventually get out the scope you may want to measure the +/-7.5V rail? _________________
cheers,
matthias
____________
Big Boss at fonitronik
Tech Buddy at Random*Source |
|
Back to top
|
|
|
The Bad Producer
Joined: Mar 08, 2009 Posts: 282 Location: The Manhole
|
Posted: Sat Feb 20, 2010 6:35 am Post subject:
|
|
|
Hi Fonik, I've got a rocksteady 7.5V rail, no problems there, I've just been poking around with my DMM, and I found something very interesting out.
I put one probe to GND and one to pin S3 on the mode switch, I was just randomly measuring voltages and comparing between the two, and when I probe that point the 'clipping' stops pretty much at most knob settings, I need to fiddle around more maybe...
Anyway, maybe this is a big clue! I have no idea what the DMM is doing when I do this, is it acting like a resistor? _________________ http://loudestwarning.tumblr.com/ |
|
Back to top
|
|
|
The Bad Producer
Joined: Mar 08, 2009 Posts: 282 Location: The Manhole
|
|
Back to top
|
|
|
The Bad Producer
Joined: Mar 08, 2009 Posts: 282 Location: The Manhole
|
Posted: Sat Feb 20, 2010 6:47 am Post subject:
|
|
|
Okay, I've just been looking at the schematic, and S3 should connect to GND (S4) when MODE is activated to lopass or both, I just stuck a cable to test the result and it got rid of the clipping on both modes when connected to ground... that seems to be the problem, but I'm not sure why it only manifested itself on the fifth build?
weird, I'll double check my wiring on the 'good'ones, I made four at the same time, so if I made a mistake I will have made it on all four! _________________ http://loudestwarning.tumblr.com/ |
|
Back to top
|
|
|
The Bad Producer
Joined: Mar 08, 2009 Posts: 282 Location: The Manhole
|
Posted: Sat Feb 20, 2010 7:04 am Post subject:
|
|
|
hmmm, that's rubbish, it seems to work better in lopass mode like this, but it messes up the other modes...
I found a crocodile clip and messed around a bit more, effect goes away when S6 is connected to ground too, and this seems to work better on both and lopass modes, this is odd, as S6 isn't supposed to go to ground!
Anyway, any more thoughts on what it could be? _________________ http://loudestwarning.tumblr.com/ |
|
Back to top
|
|
|
The Bad Producer
Joined: Mar 08, 2009 Posts: 282 Location: The Manhole
|
Posted: Sun Feb 21, 2010 5:23 am Post subject:
|
|
|
So I took some scope shots, this is in lopass mode with a sine wave modulating a sine wave, so you have clean, then clipped pic.... I won't posit an explanation!
I'm still wondering if it is more to do with my build of the LPG itself? Also to note, the offset trim has as much of an affect on clipping as the resonance control...
I did read this:
http://www.geofex.com/article_folders/cd4053/cd4053.htm
Does the 4053 need biasing to avoid this distortion? Does the exp. board do this? I'll take some more measurements of voltages etc... I am now running it on 12V with the same distortion...
Charlie _________________ http://loudestwarning.tumblr.com/ |
|
Back to top
|
|
|
The Bad Producer
Joined: Mar 08, 2009 Posts: 282 Location: The Manhole
|
Posted: Sun Feb 21, 2010 9:23 am Post subject:
|
|
|
hmmmmmmm, interesting progress, I took a look at the other LPGs I've built, and I mixed up the resonance and offset controls (10k and 50k pots) so they obviously didn't work as expected - sounded good to me - I mean now I've put them to how they should be they distort and don't sound good... maybe I should swap them back...
The problem may then be to do with the Resonance mod of the 292? Fonik did you build it as per the docs? 50K pot, R18 is 50K to ground? I've just noticed on the board this has a * next to it (as does C10 - 330f - optional?)
I just swapped out R18 from 50K to 51K, and it has made a noticeable difference, the resonance is pretty fucking fruity, and just clips at the end of its range, but not so bad (it was from 12 'o' clock on that it sounded crap) I might try a slightly higher value like 56K and see what happens... _________________ http://loudestwarning.tumblr.com/ |
|
Back to top
|
|
|
The Bad Producer
Joined: Mar 08, 2009 Posts: 282 Location: The Manhole
|
Posted: Sun Feb 21, 2010 10:10 am Post subject:
|
|
|
Okay, as far as I'm concerned the problem is solved, I replaced R18 with 56K on the 15V version of the LPG, and 82K on the 12V version (could probably go a little higher as there is occasionally a bit of clipping at full offset and resonance, but I can live with that, I reckon 100K would get rid of it totally though)
This is all on the LPG board itself, so I can say that I don't see any problems with Fonik's expansion board now that I've made these modifications...
I have no idea how or why these changes got rid of the clipping (I had no clipping without the expansion board) but I'm glad that it has gone!
I guess I can stop with my constant updates!
I have even thought about adding a little switch to choose between two resistor values for R18, so that I can switch in overdriven resonance (distortion) at will... _________________ http://loudestwarning.tumblr.com/ |
|
Back to top
|
|
|
fonik
Joined: Jun 07, 2006 Posts: 3950 Location: Germany
Audio files: 23
|
Posted: Sun Feb 21, 2010 10:32 am Post subject:
|
|
|
that's great news, charlie. thank you for your hard work! i will add your findings to the documentation on monday. and i will take a closer look on the changes you made. (and your distortion switch would change a bug to a feature!)
and guess what? i allready wrote a reply to your last post, but you've been faster! (see below).
again, great work, dude. you saved my day. thanks.
Quote: | The Bad Producer wrote: | I'm still wondering if it is more to do with my build of the LPG itself? Also to note, the offset trim has as much of an affect on clipping as the resonance control...
I did read this:
http://www.geofex.com/article_folders/cd4053/cd4053.htm
Does the 4053 need biasing to avoid this distortion? Does the exp. board do this? I'll take some more measurements of voltages etc... I am now running it on 12V with the same distortion... |
good find, charlie.
so according this article a bias/offset could possibly cause distortion. the input caps together with the resistor to GND form a filter that removes possible offset.
so far so good (or not so good in our case). how to proceed? just add three caps and resistors to remove the offset, or find the source of the offset.
the latter would be interesting, just because some units work, and some not (i built two by myself plus the breadboard circuit, all all worked).
the first would possibly be the easiest solution for you in this particular case, however it would be not that convenient to add three caps and resistors to the three inputs... |
_________________
cheers,
matthias
____________
Big Boss at fonitronik
Tech Buddy at Random*Source Last edited by fonik on Sun Feb 21, 2010 12:48 pm; edited 1 time in total |
|
Back to top
|
|
|
The Bad Producer
Joined: Mar 08, 2009 Posts: 282 Location: The Manhole
|
Posted: Sun Feb 21, 2010 10:41 am Post subject:
|
|
|
I'm quite happy that it works too! I should be clear - no changes seem to be needed to your board, it is the LPG board which I guess is supplying too much 'something' (current? offset?) to your board, unless the resonance is tamed... Do you think it could be to do with nolinearities in the vactrols? I'm building a few more this week, I'll try some more experiments out (try two vactrols on 12V etc)...
I am surprised that the offset control on the LPG board has so much affect on the distortion, being, as you said, on the LED side of the circuit...
What does changing R18 to a higher value do? I have no idea, but you may!
cheers! _________________ http://loudestwarning.tumblr.com/ |
|
Back to top
|
|
|
tommi
Joined: Dec 05, 2007 Posts: 247 Location: Italy
Audio files: 3
|
Posted: Sun Feb 21, 2010 1:25 pm Post subject:
|
|
|
Charlie,
It seems that you solved the problem! That's GREAT! soon I will try to change r18 on my LPG+exp. board and see what happen (and, of course, post here my result). Thanks for your precious investigation!
You know, when i was desperately trying to fix the problem, at a point i was tempted to change the value of r18 to a bigger one for lowering the range of the resonance, but the prejudice that the problem were on the exp. board stopped me to do that!
Charlie, i have a question for you: why you have some module working and others not?
Wich value for R18 on the modules that are working? _________________ http://soundcloud.com/mister-vommi
http://tideofsound.net |
|
Back to top
|
|
|
The Bad Producer
Joined: Mar 08, 2009 Posts: 282 Location: The Manhole
|
Posted: Sun Feb 21, 2010 1:35 pm Post subject:
|
|
|
The modules I thought were working had the wrong value for the resonance knob, it was 10K, so when I changed them to 50K I got distortion as you did...
But, I know that for months now I've had no problems with the expansion board, I just had LPG's with less resonance!
I'm hoping that someone will chime in as to WHY R18 is causing problems! I've just soldered a pot in place of R18 to get an absolute value, as with 82K (12V version) I still occasionally get clipping at certain settings of the offset pot and trim...
I'm gonna eat dinner and then check it out in a bit.... _________________ http://loudestwarning.tumblr.com/ |
|
Back to top
|
|
|
fonik
Joined: Jun 07, 2006 Posts: 3950 Location: Germany
Audio files: 23
|
Posted: Sun Feb 21, 2010 3:26 pm Post subject:
|
|
|
BTW what quad opamp did u use for the LPG? TL084 or TL074? which version? AC, BC or C? _________________
cheers,
matthias
____________
Big Boss at fonitronik
Tech Buddy at Random*Source |
|
Back to top
|
|
|
The Bad Producer
Joined: Mar 08, 2009 Posts: 282 Location: The Manhole
|
Posted: Sun Feb 21, 2010 3:39 pm Post subject:
|
|
|
I used TL084 CN on both 12V and 15V versions...
I've just spent a bit more time with this and I feel I'm back at square one.
I thought I'd fixed the 15V version, so I started on the 12V, I replaced R18 with a 1M pot to give myself a reading when the clipping stopped, seemed to be roughly 120K or something, then I replaced that with a 100K pot and a 100K resistor to ground, didn't do anything, 100K pot and 27K resistor to ground = nothing. I thought i was getting somewhere but as soon as I play around a bit it comes back, if you sum two audio inputs they need to be on half volume (input attenuator) or it goes really clippy... I also changed C10 to 22pf (would this make sense?)
So I thought I'd have a play around with the 15V version, and the clipping is still there... or rather it came back... I'm at a bit of a loss really! _________________ http://loudestwarning.tumblr.com/ |
|
Back to top
|
|
|
fonik
Joined: Jun 07, 2006 Posts: 3950 Location: Germany
Audio files: 23
|
Posted: Mon Feb 22, 2010 5:29 am Post subject:
|
|
|
unfortunately i don't have a non working unit here. so i can't do any measurements.
in regards to the multiple inputs: if you mix enough sources, you will end up with clipping - as soon as the sum comes near the rails. since the gate board is powered by +/-7.5(resp. +/-6V) this limit will be reached earlier.
in regards to C10: it is there to prevent the opamp from selfoscillating(?), so i don't think you will gain something from your change.
should we assume that offset is the problem? and you don't want to add those filters to the switch boards inputs? then we should still ask where the offset comes from.
just to recall: in which mode(s) did you experience the clipping? LoPass only? _________________
cheers,
matthias
____________
Big Boss at fonitronik
Tech Buddy at Random*Source |
|
Back to top
|
|
|
The Bad Producer
Joined: Mar 08, 2009 Posts: 282 Location: The Manhole
|
Posted: Mon Feb 22, 2010 6:20 am Post subject:
|
|
|
My solution was to remove C10 entirely and put a 100K trim in line with a 100K resistor in place of R18. It seems to be a trade off in that you lose the deeper resonance but gain a cleaner signal. The clipping only happened in Both and Lopasss mode and with R18 moved up to 82K only in Lopass mode...
I can't be bothered to add the filters to the board, as that would mean desoldering and generally making a mess, at least for this 12V module...
My other 15V modules on the other hand I am tempted to experiment more with, maybe try biasing the 4053 or putting a 100K pot on panel, plus maybe a switch to choose between 2 values of R18 to allow the crazy distortion - after playing around with it all weekend, I've found at some settings it sound pretty good, rather like Mike Ruberto's version of the Synthacon filter which has a 'Turbo' switch...
In short the distortion can be removed without biasing the 4053, and with the added benefit of a distortion 'feature'...
I'm guessing that another way of getting rid of this distortion is to lower the gain on the input audio summing amplifier when using the expansion board... _________________ http://loudestwarning.tumblr.com/ |
|
Back to top
|
|
|
fonik
Joined: Jun 07, 2006 Posts: 3950 Location: Germany
Audio files: 23
|
Posted: Mon Feb 22, 2010 6:36 am Post subject:
|
|
|
The Bad Producer wrote: | The clipping only happened in Both and Lopasss mode and with R18 moved up to 82K only in Lopass mode... |
both mode! that's funny, just because the switch should be all open then.
i was thinking of the resonance circuitry being a possible source of offset, but then it should have no effect in both mode. however, in gate mode S1 and S3 is tight to GND. maybe this has an effect on U1As and U1Cs offset? anyone?
Quote: | In short the distortion can be removed without biasing the 4053, and with the added benefit of a distortion 'feature'... |
could you put your solution into one short phrase/sentence? i am a little bit confused right now and am short on time... _________________
cheers,
matthias
____________
Big Boss at fonitronik
Tech Buddy at Random*Source |
|
Back to top
|
|
|
tommi
Joined: Dec 05, 2007 Posts: 247 Location: Italy
Audio files: 3
|
Posted: Mon Feb 22, 2010 8:08 am Post subject:
|
|
|
Quote: | could you put your solution into one short phrase/sentence? i am a little bit confused right now and am short on time... |
I hope not to add confusion: I think he means that his solution for the clipping is to increase R18 to 82k. But there could be the feature of a distortion switch (SPDT to choose between 50k and 82k for R18).
Quote: | I'm guessing that another way of getting rid of this distortion is to lower the gain on the input audio summing amplifier when using the expansion board... |
Yes, it should be! And it can be done by lowering the value of R27, 28 and 29 (the sum resistors), let's say... from 100k to 50k isn't it? _________________ http://soundcloud.com/mister-vommi
http://tideofsound.net |
|
Back to top
|
|
|
|