Author |
Message |
kx
Joined: Dec 05, 2009 Posts: 17 Location: New Zealand
|
Posted: Sun Dec 06, 2009 3:05 am Post subject:
|
|
|
Thanks so much for your reply
Sorry about the pictures before, it looks unbalanced in them, but in reality it is balanced (DC mode, the highest peaks go to +4v and the lowest valleys to -4v).
Can't figure out why it's clipping the tops in the output. If I turn the shape trimmers the other way, the waveform gradually turns into a small square wave. |
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
Posted: Sun Dec 06, 2009 3:43 am Post subject:
|
|
|
I think this is due to the fact one saw (the large one) triggers the convesion but the small one don't, therefore the signal is half converted.
There is nothing to do except submitting a perfectly symmetrical 10Vpp sawtooth. _________________ Yves |
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
Posted: Sun Dec 06, 2009 3:51 am Post subject:
|
|
|
Check very thoroughly the value of the resistors, R8,R9,R11 and R12 : 100K1%, R14,R15,R16 : 47K, T1 250K, is D2 oriented correctly ? _________________ Yves |
|
Back to top
|
|
|
kx
Joined: Dec 05, 2009 Posts: 17 Location: New Zealand
|
Posted: Mon Dec 07, 2009 3:12 am Post subject:
|
|
|
Ok, to try and pinpoint the problem I've tried these things.
First - instead of using that imperfect saw wave, I've got a saw wave coming from the PC and breadboarded a simple op-amp to amplify it to +/- 5v. - Still the same 'chop off' of the top.
Then I checked the 100k 1% resistors. Took them out and checked their values. In fact, (despite being advertised as 1%) I got some values like 98k and 102k. So I replaced them with much more accurate 100k's with values like 99.9k and 100.0k. Unfortunately this also didn't change the problem.
I couldn't find 250k trimmers, so I had to settle on 200k's. On the forum earlier it was said that 200k would be OK so I tried them. To check, I also tried putting a 47k resistor in series to make a trimpot from 47k - 247k, but it didn't really make any difference either.
the 47k's on the board are also double checked and 1% metal film resistors.
I have the 1N4148 Diodes in their specified positions with the black lines nearest the outer edges of the board...
I'm not sure what to try next, maybe it calls for a replacement TL074? Maybe I should try a LM324 or TL084? I'm running out of things to replace
Thanks again for all your help |
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
Posted: Mon Dec 07, 2009 3:28 am Post subject:
|
|
|
Hi Kx
I asked for a check of these components because in 90% of the cases it comes from an error of this kind.
200K trimmers is fine too.
Did you check that the diodes are sane ?
Can you post pictures with the PC enerated sawtooth and for various settings of the trimmer : (five positions : 0%, 25 % 50%, 75%, 100%) ?
For the oscillograms measure them at pin 7 of U3b. _________________ Yves |
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
Posted: Mon Dec 07, 2009 3:33 am Post subject:
|
|
|
And also make yourself sure that the MOD IN #1 input is connected to the 0V when setting up T1 _________________ Yves |
|
Back to top
|
|
|
kx
Joined: Dec 05, 2009 Posts: 17 Location: New Zealand
|
Posted: Tue Dec 08, 2009 3:35 am Post subject:
|
|
|
I think I have found the root of the problem. While doing the latest set of tests you suggested, I realized that I wasn't doing the adjustments while looking at the correct output. I was monitoring it from the final output (after the addon mixer).
Here is the input wave : (1 square = 5V)
After monitoring from pin 7 on U3b, and then pin 1 (U3a) for the balance, I could get the correct swaying waveform when un-shorting mod 1 from ground.
Pin 7 output :
Pin 1 output :
I also noticed that the output from pin 1 is inverted from the ouput from pin 7. The problem is, when this shifted, inverted wave is mixed with the original input wave (as is hardwired in my version - I wasn't too worried about quadrature LFO usage) the resultant wave is a square wave, and the effect is the same as PWM on a pulse output, as shown in these pictures :
=
Just wondering, is this the correct usage? Maybe I've wired it wrong, since I was expecting to be mixing 2 non-inverted sawtooths. |
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
Posted: Tue Dec 08, 2009 5:35 am Post subject:
|
|
|
I'll do some checking tonight and let you know. _________________ Yves |
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
Posted: Tue Dec 08, 2009 5:49 am Post subject:
|
|
|
OK I did some simulations and you are right, I did not realize that the waves were in opposite. I will let you know how to modify the add-on board to solve this. In fact I will propose a new add-on board !!! _________________ Yves |
|
Back to top
|
|
|
nerdware
Joined: Jul 11, 2009 Posts: 91 Location: UK
|
Posted: Tue Dec 08, 2009 9:20 am Post subject:
|
|
|
Could this be why I hear a beating effect in your demos whenever both animators are used? _________________ http://soundcloud.com/nerdware/
"render unto digital what is due to digital, render unto analogue what is due to analogue" |
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
Posted: Tue Dec 08, 2009 11:28 am Post subject:
|
|
|
Hi Kx
Here is how to fix the add-on board to obtain the right mixture.
I will also update the drawings above with an update version of the add-on PCB.
Description: |
The red arrows show which tracks must be cut and the blue arrows show the connections to be created with solder bridges |
|
Filesize: |
26.12 KB |
Viewed: |
18402 Time(s) |
|
Description: |
Once the wrong tracks are cut and the solder bridges (in red) made, solder two 150K resistors as shown |
|
Filesize: |
27.9 KB |
Viewed: |
18399 Time(s) |
|
_________________ Yves |
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
Posted: Tue Dec 08, 2009 12:03 pm Post subject:
|
|
|
OK I have changed the drawings at the beginning of this thread : new corrected addon-pcb _________________ Yves |
|
Back to top
|
|
|
kx
Joined: Dec 05, 2009 Posts: 17 Location: New Zealand
|
Posted: Tue Dec 08, 2009 3:22 pm Post subject:
|
|
|
Hi Yves,
Thanks so much for adding those alterations, I applied them to my board this morning, and it works great. Sounds great too! My original 'imperfect waveform' does really well also, and in particular very interesting results come when using input from the 'skew' output of the TH XR-VCO.
Cheers |
|
Back to top
|
|
|
Sebo
Joined: Apr 27, 2007 Posts: 564 Location: Argentina
|
Posted: Wed Dec 23, 2009 4:46 pm Post subject:
|
|
|
Hi Yves:
I'm about to start building the Sawanimator module, I etched the PCB and
the add-on PCB. I want to mod it (adding a switch) to have individual otputs
for SAW1 and SAW2, as the add-on is intended for, but also all three signals
in one out (original SAW+SAW1+SAW2), this is handy when you only use
input 1 (a single SAW goes to both processors).
In the new circuit add-on you insert the processed SAW1 in the + input of the
summing OpAmp, and if I want to add SAW2, I have to do it to that node,
I don't know how to do it that way (summing non inverter OpAmp), is there an
easy way to modify the add-on PCB to achieve that?, if not, I will build a little
circuit on stripboard.
Thank you. _________________ Sebo
---------------------------------------
My Music:
https://www.facebook.com/cosaquitos/ |
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
Posted: Thu Dec 24, 2009 1:14 am Post subject:
|
|
|
Sebo wrote: |
In the new circuit add-on you insert the processed SAW1 in the + input of the
summing OpAmp, and if I want to add SAW2, I have to do it to that node,
I don't know how to do it that way (summing non inverter OpAmp), is there an
easy way to modify the add-on PCB to achieve that?, if not, I will build a little
circuit on stripboard.
Thank you. |
You do not need to modify the circuit, send the SAW 1 signal on one lug of the switch, on the other lug of the switch solder a 150K resistor and connect a wire between the resistor and the pad on the add-on board corresponding to the non inverting input of the Opamp. _________________ Yves |
|
Back to top
|
|
|
Sebo
Joined: Apr 27, 2007 Posts: 564 Location: Argentina
|
Posted: Thu Dec 24, 2009 7:13 am Post subject:
|
|
|
Thanks Yves!!!
One last question:
All three inputs will have a gain of 1 (right?), if all signals are 10Vpp I think
the signal could become too high and clip, should I reduce the gain of inputs?
If so, I did the math using the substractor opamp equation, and I think this
values could work:
R5=200K, R7=R6=R6'(the new input)=100K, and the resistor between the
non inverter input and ground=50K
Thanks a lot for all you help.
Merry Christmas, and the best 2010 to you and your family.
Sebo _________________ Sebo
---------------------------------------
My Music:
https://www.facebook.com/cosaquitos/ |
|
Back to top
|
|
|
Appliancide*
Joined: Jul 04, 2007 Posts: 126 Location: Paul lives in a 1920’s film
|
Posted: Sun Dec 27, 2009 1:22 pm Post subject:
|
|
|
If I were going to add a switch for the mixer function, would I just have it disconnect the saw in from the mixer? Or would it be better to bypass the whole mixer circuit?
Thanks,
Paul _________________ http://appliancide/blogspot.com |
|
Back to top
|
|
|
Sebo
Joined: Apr 27, 2007 Posts: 564 Location: Argentina
|
Posted: Wed Dec 30, 2009 9:01 pm Post subject:
|
|
|
Hi Yves:
Do you read my last post?
I will try the circuit with unity gain and see what happens, if the signal is too
hot I will try to half the gain of the inputs (are my maths OK?).
Thank you again.
> *Appliancide* : I can't understand what you mean. _________________ Sebo
---------------------------------------
My Music:
https://www.facebook.com/cosaquitos/ |
|
Back to top
|
|
|
Appliancide*
Joined: Jul 04, 2007 Posts: 126 Location: Paul lives in a 1920’s film
|
Posted: Wed Dec 30, 2009 9:53 pm Post subject:
|
|
|
Sebo wrote: | > *Appliancide* : I can't understand what you mean. |
If I am adding the additional circuitry that mixes the original saw with the shifted saw, where should I add a switch to remove the original saw from the output? Or would a wet/dry pot be better?
Cheers,
Paul _________________ http://appliancide/blogspot.com |
|
Back to top
|
|
|
Appliancide*
Joined: Jul 04, 2007 Posts: 126 Location: Paul lives in a 1920’s film
|
Posted: Tue Jan 05, 2010 1:53 pm Post subject:
|
|
|
*Appliancide* wrote: | Sebo wrote: | > *Appliancide* : I can't understand what you mean. |
If I am adding the additional circuitry that mixes the original saw with the shifted saw, where should I add a switch to remove the original saw from the output? Or would a wet/dry pot be better?
Cheers,
Paul |
Has anyone done this? _________________ http://appliancide/blogspot.com |
|
Back to top
|
|
|
Sebo
Joined: Apr 27, 2007 Posts: 564 Location: Argentina
|
Posted: Tue Jan 05, 2010 5:16 pm Post subject:
|
|
|
Hi:
You can add a switch cutting the SAW1 and SAW2 inputs in the Add-On PCB.
I don't see much improvement in a wet/dry pot as the most useful mix between
both signals is 50/50 (at least to me).
I just will add a switch to add Processed SAW2 to Output 1, this is useful when
only one signal is connected as you can get the mix of the SAW1 (input)
plus both Processed SAWs at one output.
I will be working in this module next few days. _________________ Sebo
---------------------------------------
My Music:
https://www.facebook.com/cosaquitos/ |
|
Back to top
|
|
|
ultrashock
Joined: Dec 10, 2009 Posts: 40 Location: Vienna.AT
|
Posted: Tue Jun 29, 2010 2:40 am Post subject:
|
|
|
Dear Yves!
Kindly ask you to answer the question concerning your saw animator!
When the RATE pot is at the ZERO level - do we have the CLEAR saw at the output? I mean could we use the ZERO level as the "crossfader" beetween pure saw and "animated"? |
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
Posted: Tue Jun 29, 2010 2:54 am Post subject:
|
|
|
ultrashock wrote: | Dear Yves!
Kindly ask you to answer the question concerning your saw animator!
When the RATE pot is at the ZERO level - do we have the CLEAR saw at the output? I mean could we use the ZERO level as the "crossfader" beetween pure saw and "animated"? |
No you misunderstood, the Rate is setting the speed of the internal LFO.
It is not a mixing pot. If you use the add-on board shown above you will have a fixed mix of the raw sawtooth and of the modulated (shifted) ones. If you don't use this board you need an external mixer. _________________ Yves |
|
Back to top
|
|
|
ultrashock
Joined: Dec 10, 2009 Posts: 40 Location: Vienna.AT
|
Posted: Tue Jun 29, 2010 4:09 am Post subject:
|
|
|
Thank you very much for such fast responce!
When I answered about the "rate" knob I meant if the modulation at ZERO rate-level will be so minimal as indistinguishable on ear that we can use is as almost the clear saw |
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
Posted: Tue Jun 29, 2010 4:28 am Post subject:
|
|
|
OK I get your idea, at zero rate the period of modulation is something 5s. Therefore it souns like a sawtooth that would slightly go out of tune. _________________ Yves |
|
Back to top
|
|
|
|