Author |
Message |
hoby
Joined: Jan 30, 2018 Posts: 21 Location: Copenhagen
|
Posted: Tue Jan 30, 2018 2:51 pm Post subject:
Yusynth VCO problem tracking at low frequencies Subject description: VCO tracking problem |
|
|
I have just finished building 3 yusynth VCO's using circuitboards from soundtronics.
Everything seems to work - different waveshapes etc, but i can not get the VCO to track at below 200 Hz - on all 3 modules.
I have calibrated the v/oct, and get good tracking for 6 octaves as long as I stick to frequncies above 200 Hz. When I go below 200 Hz the it goes out of tune, and, it also takes log before each note becomes stable - it glides before setteling. I am using a beatstep pro as input.
I have built the 12v version using the LM394s from ERicaSynth, (shifted so the pins match). I have also tried to replace the 394 with a matched pair of BC547, but same result.
Any ideas what I may have messed up? |
|
Back to top
|
|
|
gabbagabi
Joined: Nov 29, 2008 Posts: 651 Location: Berlin by n8
Audio files: 23
|
Posted: Thu Feb 01, 2018 10:57 am Post subject:
|
|
|
hi and to the forum
i would like to know if u can go below 200Hz without the beatstep only using the pots?
if so the question is what comes out of the beat step, also do they use the same GND?
if not you should search for unwanted offsets, i would start at T1 and R8
cheers
gabbagabi |
|
Back to top
|
|
|
hoby
Joined: Jan 30, 2018 Posts: 21 Location: Copenhagen
|
Posted: Thu Feb 01, 2018 11:49 am Post subject:
|
|
|
Thanks for our help Gabbagabi.
Yes, I can easlily make it go down below 200 with the pots. The out of tune "drift" happens below 200 Hz regardless of the pot setting, as long as the "offset" of the frequency knob is low enough, so that 0 volts CV will genereate a frequncy below 200. (I hope that sentence makes sense)
I have measured the voltage out of the beatstep, and it generates the right voltage when i measure the output with my multimeter. But I did that without the VCO connected, so perhaps I should double check that I also get the corect voltage at input terminals on the vco-board when the beatstep is connected.
I do use the beatstep on my TTSH (arp 2600 clone) and don't see the issues there. |
|
Back to top
|
|
|
hoby
Joined: Jan 30, 2018 Posts: 21 Location: Copenhagen
|
Posted: Thu Feb 01, 2018 1:36 pm Post subject:
|
|
|
Just noticed that i only have a 20k port for the T2 10 turn trimmer. I dont have a 22k/or25K so Ill have to get some before i can test, but perhaps that could be explanaiton? Last edited by hoby on Sat Feb 03, 2018 2:05 am; edited 1 time in total |
|
Back to top
|
|
|
gabbagabi
Joined: Nov 29, 2008 Posts: 651 Location: Berlin by n8
Audio files: 23
|
Posted: Fri Feb 02, 2018 12:14 pm Post subject:
|
|
|
ur welcome,
the value of the trimmer t1 should not be so crucial, but in the schematic the value is 220k --> do u have a 100K?
when u turn all pots down - the vco should produce the minimum frequency
this minimum frequency is set by the trimmer T1
please try to turn all pots to minimum and play with T1 and look if you can reach 20Hz
if not: (the following is regarding this picture: http://yusynth.net/Modular/Commun/VCO/VCO-wiring.jpg)
connect all red and purple wires from the midddle 5Port molex plug to ground and try again to reach 20Hz using T1
if so, something at the pots or 1V/Oct inputs is wrong
u have mounted the stuff already to a frontplate? Is the frontplate conducting?
so far,
happy troubleshooting and good luck
gabbagabi |
|
Back to top
|
|
|
hoby
Joined: Jan 30, 2018 Posts: 21 Location: Copenhagen
|
Posted: Sat Feb 03, 2018 2:25 am Post subject:
|
|
|
g.gabba wrote: |
the value of the trimmer t1 should not be so crucial, but in the schematic the value is 220k --> do u have a 100K? |
I have a 200K trimmer for T1. That should be alright according to the build instructions
g.gabba wrote: |
when u turn all pots down - the vco should produce the minimum frequency
this minimum frequency is set by the trimmer T1
please try to turn all pots to minimum and play with T1 and look if you can reach 20Hz
|
I can easily reach low frequencies, but the vco´s just become very unstable below 150 Hz. The lower the frequency the more unstable. There it really behaves as if it has glide or portamento. So when I change pitch below 150 Hz, the VCO glide or drift before setteling at a frequncy. Above 200 It does not.
g.gabba wrote: |
u have mounted the stuff already to a frontplate? Is the frontplate conducting?
|
Yes, I have made my own aluminium front plates, so they are indeed conducting. I will try to remove the pots from the plate to make sure nothing strange is going on there.
g.gabba wrote: |
happy troubleshooting and good luck
|
Thanks - hopefully Ill have some time to look into it today |
|
Back to top
|
|
|
gabbagabi
Joined: Nov 29, 2008 Posts: 651 Location: Berlin by n8
Audio files: 23
|
Posted: Sat Feb 03, 2018 6:32 am Post subject:
|
|
|
conducting panels are good, just make shure that it is somehow connectet to GND
glhf
gg |
|
Back to top
|
|
|
hoby
Joined: Jan 30, 2018 Posts: 21 Location: Copenhagen
|
Posted: Sun Feb 04, 2018 10:42 am Post subject:
|
|
|
OK - so I had a further look, but still not found the solution. Also tried he boards/pots when they were not connected to the panel, but it made no difference. Ive put on the oscilliscope on the sine output, and it is acting really strange at low frequencies
first- here is a screenshot at 3.75 kHz - where it is acting quite normal - the sine is a bit sharp at the top end but I think this is ok
at 300 Hz it is more like a shark shape
And at 100 Hz it quite far from a sine wave.
I have put the scope on the output of "U6b" from the scematic. Again, at higher frequencies it look like a saw wave, an at lower frequencies it gets an exponential decay.
Could it just be that the 12V version is simply underpowered or something? I feel tempted to change one of the boards to the 15 version. Right now I am running the from a dual power supply each deliver 2 A, so they should definitely be powerful enough |
|
Back to top
|
|
|
gabbagabi
Joined: Nov 29, 2008 Posts: 651 Location: Berlin by n8
Audio files: 23
|
Posted: Mon Feb 05, 2018 8:26 am Post subject:
|
|
|
wow thats strange!
i would conentrate on the raw-saw at U6B cos this has to be right first.
1. is it the same behavior on all 3 VCO?
2. please double check the Caps C8 C11 C10 if they have the right value
3. please make shure that the FET is soldered in in the rigth direction, so that source and gate are not swopped
otherwise iam running a bit out of ideas
gl, gabbagabi |
|
Back to top
|
|
|
hoby
Joined: Jan 30, 2018 Posts: 21 Location: Copenhagen
|
Posted: Mon Feb 05, 2018 1:06 pm Post subject:
|
|
|
Yeah - All three boards act the same way, and I have double checked the caps so many times I have lost count
The J112 is also placed correct according to the PCB layout. I have tried replacing the TL074 with another brand, so yes, I am also running out of good ideas.
But thanks for your help anyway |
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
Posted: Thu Feb 15, 2018 12:38 pm Post subject:
|
|
|
can you post a good photo of one of your module viewed from component side ? _________________ Yves |
|
Back to top
|
|
|
hoby
Joined: Jan 30, 2018 Posts: 21 Location: Copenhagen
|
Posted: Fri Feb 16, 2018 1:24 am Post subject:
|
|
|
picture hosting
Here is a link to a highres photo to the board.
https://www.dropbox.com/s/rxikjpsl20f8hdy/DSCF1050.JPG?dl=0
I just grabbed one of the boards before going to work this morning, and did not realize if was one where I had taken the LM394 & tempco off the board.
I used the LM394 from Erica Synth, shifted one position to the left on the board, and 3300ppm 1K tempco from Thonk.
I will post a new photo when I get home, where they are also installed
Here is a link where the LM394 is installed. The tempco is just mounted for illustrative purpose. I will use heat transfer compound etc. as per the description on your website.
https://www.dropbox.com/s/whe2rp7q9fsg9ra/DSCF1056.JPG?dl=0
url for an image |
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
Posted: Fri Feb 16, 2018 4:15 am Post subject:
|
|
|
OK I have found a mistake : I noted that you soldered a 1K resistor on the PCB at the print of R12. R12 is the tempco resistor therefore either you connect it on the PCB at R12 print OR (it is an exclusive OR) you put it on pins 4 and 5 of the LM394 socket, but do not a connect a tempco on the socket and a 1k res on R12 print at the same time.
By the way you did not have to solder straps since Soundtronics PCBs are double sided and the straps are replaced by tracks. _________________ Yves |
|
Back to top
|
|
|
hoby
Joined: Jan 30, 2018 Posts: 21 Location: Copenhagen
|
Posted: Fri Feb 16, 2018 5:22 am Post subject:
|
|
|
You are absolutely right - I did remove the R12 on my other boards, when I installed the tempo. It will not work with both as you say, but I the odd behaviour pictured above with the scopes was with the standard "R12" removed, and only the tempco.
I did not know about the "vias" as you mentioned. But I guess it is ok to have them there. |
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
Posted: Fri Feb 16, 2018 5:26 am Post subject:
|
|
|
hoby wrote: | I did not know about the "vias" as you mentioned. But I guess it is ok to have them there. |
Yes it doesn't make a difference.
OK I continue to check your photo. _________________ Yves |
|
Back to top
|
|
|
hoby
Joined: Jan 30, 2018 Posts: 21 Location: Copenhagen
|
Posted: Fri Feb 16, 2018 5:34 am Post subject:
|
|
|
Thanks - I really appreciate it - let me know if you need a closeup photo of anything. |
|
Back to top
|
|
|
hoby
Joined: Jan 30, 2018 Posts: 21 Location: Copenhagen
|
Posted: Wed Feb 28, 2018 3:41 pm Post subject:
|
|
|
OK - so last weekend i tried to convert one of the VCO's to running +/- 15 volts, (changed the 78l15 and 79l15, and the three resistors) to see if that would make a difference. But unfortunately everything is the same.
I am starting to run out og ideas - Could this issue have something to do with the type of J112, TL072 or LM311 I have used?
And if so, it would be great to get a link to where I could buy some, that have worked on other modules. |
|
Back to top
|
|
|
gabbagabi
Joined: Nov 29, 2008 Posts: 651 Location: Berlin by n8
Audio files: 23
|
Posted: Sat Mar 03, 2018 9:17 am Post subject:
|
|
|
if iam looking to the saw-wave it looks like it is running through a kind of filter
my suspect is the JFET,
my foggy memory say that i saw that effect already once
i would try to test substituting him with a PNP Bipolar transistor
u have nothing to loose
good luck mate |
|
Back to top
|
|
|
gdavis
Joined: Feb 27, 2013 Posts: 359 Location: San Diego
Audio files: 1
|
Posted: Sun Mar 04, 2018 11:27 am Post subject:
|
|
|
At this point, I would personally stop taking shots in the dark and start troubleshooting specific points in the circuit to narrow down the culprit.
The decay you're seeing indicates that the expo converter isn't acting as a constant current source like it should, so the problem is probably around there somewhere.
I would suggest starting by checking the following:
1. 5V reference : check at U7a pin 1, C8, Q1 drain and R13 to make sure they all have the same constant voltage of approximately 5V.
2. Virtual ground: U3b pin 6, the other side of R13 and pin 1 of the LM394 (or the collector of the BC547 connected to R13). These should all be around 0V.
3. U3b pin 7 should be a constant negative voltage (for a given frequency setting) but not too close to the negative rail voltage.
Let me know what you find and we'll go from there. _________________ My synth build blog: http://gndsynth.blogspot.com/ |
|
Back to top
|
|
|
hoby
Joined: Jan 30, 2018 Posts: 21 Location: Copenhagen
|
Posted: Sun Mar 04, 2018 1:28 pm Post subject:
|
|
|
Thanks the input. I’ll try it out and let you know. I am away for a few days but will take a look as soon as I get home |
|
Back to top
|
|
|
hoby
Joined: Jan 30, 2018 Posts: 21 Location: Copenhagen
|
Posted: Thu Mar 08, 2018 3:00 pm Post subject:
|
|
|
gdavis wrote: |
I would suggest starting by checking the following:
1. 5V reference : check at U7a pin 1, C8, Q1 drain and R13 to make sure they all have the same constant voltage of approximately 5V.
|
Definately something wrong here. I get around 4.4V - (also on Q1, source though)
The voltage after the 78L12 drops to about 10.6V, so I guess I have to play around with the voltage division between R49 and R50. So perhaps 47K for R49 and 51K for R50 should get me close.
gdavis wrote: |
2. Virtual ground: U3b pin 6, the other side of R13 and pin 1 of the LM394 (or the collector of the BC547 connected to R13). These should all be around 0V.
|
These seems allright, so Ill start changing the resistors and take see if that helps. |
|
Back to top
|
|
|
hoby
Joined: Jan 30, 2018 Posts: 21 Location: Copenhagen
|
Posted: Thu Mar 08, 2018 3:35 pm Post subject:
|
|
|
gdavis wrote: |
3. U3b pin 7 should be a constant negative voltage (for a given frequency setting) but not too close to the negative rail voltage.
Let me know what you find and we'll go from there. |
Changing the resistors did unfortunately not help much.
The voltage on pin7 goes from -0.51 at the highest frequencies, to -0.38 at low frequencies. I guess this is pretty close to the rail voltage, so there might be something here. |
|
Back to top
|
|
|
gdavis
Joined: Feb 27, 2013 Posts: 359 Location: San Diego
Audio files: 1
|
Posted: Thu Mar 08, 2018 8:48 pm Post subject:
|
|
|
hoby wrote: | gdavis wrote: |
I would suggest starting by checking the following:
1. 5V reference : check at U7a pin 1, C8, Q1 drain and R13 to make sure they all have the same constant voltage of approximately 5V.
|
Definately something wrong here. I get around 4.4V - (also on Q1, source though)
The voltage after the 78L12 drops to about 10.6V, so I guess I have to play around with the voltage division between R49 and R50. So perhaps 47K for R49 and 51K for R50 should get me close.
|
This is normal. It doesn't have to be exactly 5V, just close enough that you can compensate with the trims. As long as the voltage is stable the tuning should be stable.
hoby wrote: | gdavis wrote: |
3. U3b pin 7 should be a constant negative voltage (for a given frequency setting) but not too close to the negative rail voltage.
Let me know what you find and we'll go from there. |
Changing the resistors did unfortunately not help much.
The voltage on pin7 goes from -0.51 at the highest frequencies, to -0.38 at low frequencies. I guess this is pretty close to the rail voltage, so there might be something here. |
The rail voltages are the approx. +/- 10.6V you see out of the 7xL12, you're near 0V which is in the middle. So that's fine. I'd expect it to be more negative for higher frequencies though (I may need to double check that).
Based on that, off hand I'd say double check C7 (100pF), that both emitters of J4 are connected to R15, R15 is not shorted and that the base of J4 that should be connected to ground actually is. _________________ My synth build blog: http://gndsynth.blogspot.com/ |
|
Back to top
|
|
|
gdavis
Joined: Feb 27, 2013 Posts: 359 Location: San Diego
Audio files: 1
|
Posted: Thu Mar 08, 2018 9:48 pm Post subject:
|
|
|
gdavis wrote: |
hoby wrote: | gdavis wrote: |
3. U3b pin 7 should be a constant negative voltage (for a given frequency setting) but not too close to the negative rail voltage.
|
The voltage on pin7 goes from -0.51 at the highest frequencies, to -0.38 at low frequencies. I guess this is pretty close to the rail voltage, so there might be something here. |
The rail voltages are the approx. +/- 10.6V you see out of the 7xL12, you're near 0V which is in the middle. So that's fine. I'd expect it to be more negative for higher frequencies though (I may need to double check that).
|
Oops, I mixed up my nanos and picos. That voltage range is where it should be. _________________ My synth build blog: http://gndsynth.blogspot.com/ |
|
Back to top
|
|
|
hoby
Joined: Jan 30, 2018 Posts: 21 Location: Copenhagen
|
Posted: Fri Mar 09, 2018 12:31 am Post subject:
|
|
|
gdavis wrote: |
The rail voltages are the approx. +/- 10.6V you see out of the 7xL12, you're near 0V which is in the middle. . |
Ah yes, my mistake. I misread you and thought it was not supposed to be close to GND, but, you clearly wrote negative rail.
I'll look up on your point this afternoon evening. |
|
Back to top
|
|
|
|