DIY - Elektor Vocoder - General Build Thread

From circuitbending to homebrew stompboxes & synths, keep the DIY spirit alive!

Moderators: lisa, luketeaford, Kent, Joe.

Post Reply
User avatar
Fitchie
Wiggling with Experience
Posts: 313
Joined: Sun Aug 30, 2015 1:49 pm

Post by Fitchie » Sun Feb 03, 2019 12:58 pm

Wolfo! wrote:IC3 pin 6 also flatlines ...
We are almost there !

It looks like after opamp IC3 there is no signal anymore.
What opamp did you use ? LM741 ? NE5534P ?

Wolfo!
Learning to Wiggle
Posts: 19
Joined: Wed Jun 03, 2015 4:29 am
Location: Belgium

Post by Wolfo! » Sun Feb 03, 2019 1:12 pm

Fitchie wrote: We are almost there !

It looks like after opamp IC3 there is no signal anymore.
What opamp did you use ? LM741 ? NE5534P ?
Hi Fitchie,

I used the parts specified in the BOM.

User avatar
Fitchie
Wiggling with Experience
Posts: 313
Joined: Sun Aug 30, 2015 1:49 pm

Post by Fitchie » Sun Feb 03, 2019 1:30 pm

Wolfo! wrote:I used the parts specified in the BOM.
That's the LM741CN/NOPB.

Strange, I did modify the offset compensation on the production boards, so the LM741 should work.
In the original Elektor design HA4741's were used (they have a smaller offset than the TL/LM series) with a greedy offset compensation, but the good thing was that 3 resistors were saved. :eek:
I'm going to check that I've overlooked something and Boogie may expect a phone call. 8-)

Feedback follows as fast as possible.

Perhaps superfluous, but please can you check R200 (180 K), R201 (180 K) and R202 (10 K) ?

User avatar
Fitchie
Wiggling with Experience
Posts: 313
Joined: Sun Aug 30, 2015 1:49 pm

Post by Fitchie » Mon Feb 04, 2019 5:33 am

Hi Wolfo!


Probably the offset correction causes the opamp (IC3) to go into saturation, especially in case trimmer P1 is not exactly in the correct position.
So remove IC3 and measure the voltage on pin 6. Micro turn P1 till you have (as good as you can approach) 0,0 V on this pin.
Put IC3 back into the socket and check again the signal on pin 6 with your oscilloscope.

Check also if you are using 1% resistors for R200, R201 and R202. If that is not the case, the offset control may be out of range as well.

Succes !

User avatar
elmegil
Super Deluxe Wiggler
Posts: 2888
Joined: Sun Apr 29, 2012 2:57 pm
Location: Chicago

Post by elmegil » Mon Feb 04, 2019 11:22 pm

I'm feeling pretty stupid. The calibration of the filters is not making sense to me.

I have the backplane jumpers over by the psu, and the out/pot jumper on the LPF card.

Step 1 produces no sound at all, without making any adjustments.

With no signal going into the carrier, however, it seems to me that it shouldn't be making any sound anyway.

All the pots (voice, carrier, out) are all the way up.

I'm using my Mutable Instruments module tester as a sound source -- I can generate white noise, so That's what I'm sending into the voice input. I've got it going through the same attenuator at the same setting that I used for I/O module adjustment to get a .77VRMS signal.

I'd assume I was doing something wrong, except then...

Step 2: I run the same white noise into the carrier input and I hear a low noise signal on the output. That's good, means my I/O module is actually passing audio. But I don't get how I'm supposed to apply this 200mV DC control signal when I have the output hard jumpered to the pot input. I haven't bothered to actually get a 200mV DC signal because that struck me almost immediately.

Step 3 I haven't tried yet, but am I correct to understand that the "output" mentioned here is the actual detection output, not the audio output??

I guess that's one of the sources of confusion. Terms "input" and "output" are mentioned, but it's not clear to me when that's referring to the control voltage ins & outs and when that's referring to one of the audio ins or the audio output of the vocoder.

User avatar
Fitchie
Wiggling with Experience
Posts: 313
Joined: Sun Aug 30, 2015 1:49 pm

Post by Fitchie » Tue Feb 05, 2019 12:47 am

elmegil wrote:The calibration of the filters is not making sense to me. Terms "input" and "output" are mentioned, but it's not clear to me when that's referring to the control voltage ins & outs and when that's referring to one of the audio ins or the audio output of the vocoder.
Fair enough, the instructions may not be very clear. I will describe these steps again and make them more understandable.
I do not speak native English for which I apologize, so please give me a few days time to perform this job.

User avatar
elmegil
Super Deluxe Wiggler
Posts: 2888
Joined: Sun Apr 29, 2012 2:57 pm
Location: Chicago

Post by elmegil » Tue Feb 05, 2019 1:16 am

Fitchie wrote:
elmegil wrote:The calibration of the filters is not making sense to me. Terms "input" and "output" are mentioned, but it's not clear to me when that's referring to the control voltage ins & outs and when that's referring to one of the audio ins or the audio output of the vocoder.
Fair enough, the instructions may not be very clear. I will describe these steps again and make them more understandable.
I do not speak native English for which I apologize, so please give me a few days time to perform this job.
I mean no criticism of you :) Thank you for your effort.

User avatar
Boogie
Veteran Wiggler
Posts: 507
Joined: Thu Nov 10, 2016 12:56 pm
Location: Belgium

Post by Boogie » Wed Feb 06, 2019 9:51 am

Fitchie
Boogie may expect a phone call.
Yip, I got your phonecall.

So for initial calibration, just do not populate the P9 on your filterboards.
When using modern opamps, the input bias current is small enough to get a functional system without trimming this opamp's bias.

When the filter board is functional, one can add the trimmer P9 (latest modif on this offset topic) and trim till edited: pin 6 of the opamp reads about 0.0xx mV. Best to start with the trimmer in the middle position and then trim very gentle up or down.

For those using NOS, one may have the larger offsets as in 'the good old days'.

Before mounting P9, and not applying any input signal, measure the output voltage at edited: pin 6 of the 471 with a voltmeter.
If this is in the 0.something mV, do not populate P9 yet, and follow the above procedure when everything is working.
Else, place P9 (one may replace it with a 10k or 20k for finer trimming control) and trim the voltage on edited: pin 6 to a number as low as possible, e.g. 0.0xx mV.

Good luck.
-------------------------------------------------------
“Digital generated ‘analog’, is missing its analog soul and therefore isn’t analog any more.”
Jupiter 8 clone: https://www.muffwiggler.com/forum/viewt ... p?t=209322
Minipops 7 clone: https://www.muffwiggler.com/forum/viewt ... p?t=199360
ClapTrap clone: https://www.muffwiggler.com/forum/viewt ... p?t=202248
www.nabla-instruments.com

User avatar
Fitchie
Wiggling with Experience
Posts: 313
Joined: Sun Aug 30, 2015 1:49 pm

Post by Fitchie » Wed Feb 06, 2019 4:33 pm

Boogie wrote:Else, place P9 - one may replace it with a 10k or 20k for finer trimming control.
Thanks @Boogie for this feedback !

The 100 K Ohm potentiometer P1/P5/P9 in series with the 180 K Ohm resistors R200 and R201 gives a control range of -4 .. +4 V on the inverting input of the LM741 opamp and this is a bit too much to be able to fine-tune the offset in an easy way. The result is that (in case of an inaccurate setting of P1/P5/P9) the opamp immediately saturates.

The hard way is to trim P1/P5/P9 until you measure approximately 0.0 mV on pin 6 of IC3.
But there are 2 much simpler solutions to refine this adjustment:
  • - replace the 100 K potentiometer P1/P5/P9 on all filter boards by a 10 K Ohm or 20 K Ohm equivalent (10 trimmers in total)
    or
    - replace both resistors R200 and R201 on all filter boards by a 1 M Ohm equivalent (20 resistors in total)
When adjusting the filters, this is the first step you have to take !


@Wolfo!
Please can you confirm this modification solves the problem you had to adjust the LM741 offset ?
.

User avatar
elmegil
Super Deluxe Wiggler
Posts: 2888
Joined: Sun Apr 29, 2012 2:57 pm
Location: Chicago

Post by elmegil » Wed Feb 06, 2019 5:17 pm

Boogie wrote:Fitchie
Boogie may expect a phone call.
Yip, I got your phonecall.

So for initial calibration, just do not populate the P9 on your filterboards.
That's kind of problematic since I've already built all the boards :)
Boogie wrote: When using modern opamps, the input bias current is small enough to get a functional system without trimming this opamp's bias.

When the filter board is functional, one can add the trimmer P9 (latest modif on this offset topic) and trim till edited: pin 6 of the opamp reads about 0.0xx mV. Best to start with the trimmer in the middle position and then trim very gentle up or down.

For those using NOS, one may have the larger offsets as in 'the good old days'.

Before mounting P9, and not applying any input signal, measure the output voltage at edited: pin 6 of the 471 with a voltmeter.
If this is in the 0.something mV, do not populate P9 yet, and follow the above procedure when everything is working.
Else, place P9 (one may replace it with a 10k or 20k for finer trimming control) and trim the voltage on edited: pin 6 to a number as low as possible, e.g. 0.0xx mV.

Good luck.

So I fully expect pulling P9 off the PCB to destroy at least one of the 10 trimmers, and I don't have any spares.

Any alternative suggestion?

Will substituting the larger resistors as Fitchie mentions do the job?

User avatar
Fitchie
Wiggling with Experience
Posts: 313
Joined: Sun Aug 30, 2015 1:49 pm

Post by Fitchie » Wed Feb 06, 2019 5:25 pm

elmegil wrote:Will substituting the larger resistors as Fitchie mentions do the job?
Yes, this solution is also fine and costs less than swapping the trimmers. Preferably try it out on 1 filter print before desoldering the others.
I plan to perform/document some extra tests next weekend and will update the adjustment procedures on the vocoder website.

Smart advice: carefully cut the resistors in 2 and desolder pin after pin from the PCB.

User avatar
elmegil
Super Deluxe Wiggler
Posts: 2888
Joined: Sun Apr 29, 2012 2:57 pm
Location: Chicago

Post by elmegil » Wed Feb 06, 2019 9:45 pm

Great thank you. Yeah, desoldering resistors is a walk in the park compared to trimmers. I can do IC's without sockets better than I can do trimmers :D

User avatar
Fitchie
Wiggling with Experience
Posts: 313
Joined: Sun Aug 30, 2015 1:49 pm

Post by Fitchie » Thu Feb 07, 2019 12:49 am

elmegil wrote:Great thank you. Yeah, desoldering resistors is a walk in the park compared to trimmers. I can do IC's without sockets better than I can do trimmers :D
That is completely true ! :nod:

it is very important to make use of 1 M Ohm 1% resistors !
The deviation of the resistors may otherwise cause the potentiometer to fall out of range, and then the adjustment fails again.

This is a good candidate from the same MF50 series as used for the other resistors:

Image

Wolfo!
Learning to Wiggle
Posts: 19
Joined: Wed Jun 03, 2015 4:29 am
Location: Belgium

Post by Wolfo! » Thu Feb 07, 2019 2:28 am

@Fitchie,

I replaced P1 with a 20k multiturn pot, and can trim the offset to 0,000V!
Have to do some further testing on the other boards, but this looks good!

User avatar
elmegil
Super Deluxe Wiggler
Posts: 2888
Joined: Sun Apr 29, 2012 2:57 pm
Location: Chicago

Post by elmegil » Thu Feb 07, 2019 8:33 am

Fitchie wrote:
elmegil wrote:Great thank you. Yeah, desoldering resistors is a walk in the park compared to trimmers. I can do IC's without sockets better than I can do trimmers :D
That is completely true ! :nod:

it is very important to make use of 1 M Ohm 1% resistors !
Perfect, I don't know if I even have any 5% resistors around any more, I've been buying only 1% for whatever application for so long now :)

User avatar
Boogie
Veteran Wiggler
Posts: 507
Joined: Thu Nov 10, 2016 12:56 pm
Location: Belgium

Post by Boogie » Thu Feb 07, 2019 12:51 pm

Hello,
it is very important to make use of 1 M Ohm 1% resistors !
Perfect, I don't know if I even have any 5% resistors around any more
The absolute value is of not so much importance.
It's the two resistors having approx the same value which is of importance.
Anything between 750k and 1M should be fine.
So for those who have a bunch of resistors in stock,
just make sure you cherry-pick them comparing the values.
An absolute spread of +2% to -2% is an absolute maximum if one wants to cover the maximum input bias current of 100 nA positive and negative.

Thus using 1.020 M and 980k is the max spread if one wants to be sure to cover the whole range of bias currents when using NOS.

Using 990k and 980k is thus a perfect approach as well.

Below a simulation schematic I've used to calculate some values. (TINA simulator)

Image
-------------------------------------------------------
“Digital generated ‘analog’, is missing its analog soul and therefore isn’t analog any more.”
Jupiter 8 clone: https://www.muffwiggler.com/forum/viewt ... p?t=209322
Minipops 7 clone: https://www.muffwiggler.com/forum/viewt ... p?t=199360
ClapTrap clone: https://www.muffwiggler.com/forum/viewt ... p?t=202248
www.nabla-instruments.com

User avatar
Fitchie
Wiggling with Experience
Posts: 313
Joined: Sun Aug 30, 2015 1:49 pm

Post by Fitchie » Sat Feb 09, 2019 1:55 am

Fitchie wrote:There are 2 much simpler solutions to refine this adjustment:
  • - replace the 100 K potentiometer P1/P5/P9 on all filter boards by a 10 K Ohm or 20 K Ohm equivalent (10 trimmers in total)
    or
    - replace both resistors R200 and R201 on all filter boards by a 1 M Ohm equivalent (20 resistors in total)
And there even is a third solution without desoldering any component !
Simply solder an additional 12 K Ohm resistor over the trimmer as suggested below:

Image

Perform this patch at the solder side from the PCB.
If you use a 1/8 watt resistor, you can solder the resistor straight between the legs of the trimmer.
Last edited by Fitchie on Sat Feb 09, 2019 1:59 am, edited 1 time in total.

User avatar
Fitchie
Wiggling with Experience
Posts: 313
Joined: Sun Aug 30, 2015 1:49 pm

Post by Fitchie » Sat Feb 09, 2019 1:58 am

Cost/skil comparison:
  • Solution 1
    10 x trimmer 10 K Ohm
    Desoldering necessary

    Solution 2
    20 x resistor 1 M Ohm 1 %
    Desoldering necessary

    Solution 3
    10 x resistor 12 K Ohm 1/8 watt
    No desoldering necessary
:tu:

User avatar
jbdiver
Wiggling with Experience
Posts: 287
Joined: Tue Aug 31, 2010 11:39 pm
Location: Minneapolis, USA

Post by jbdiver » Sun Feb 10, 2019 11:03 pm

I finished soldering all the boards today. It took me 8 days to complete the boards. It's not a difficult project, but the part count is fairly high. If you used the Mouser BOM in this thread, you will be missing a few parts (the quantities in the official BOM for these parts is also incorrect). I'll try to update that BOM for others in the future. Here are the parts missing:

8 - 10K resistors
1 - 22K resistor
4 - 0.1uF film capacitors

Here are the additional unused parts which can be removed:

1 - 22K resistor
4 - 47K resistors
2 - 4148 diodes

The 100K resistors in the BOM have some sort of plastic coating on one of the leads. That coating is a pain to scrape off, so that part should be swapped for a regular resistor.

Also, I'm going with the 1M resistor fix for the filter boards (see Solution 2 above) which means that 20 1M resistors need to be added to the BOM and 20 180K resistors removed.

BTW, I can't seem to find the panel wiring diagram on the vocoder.eu website. Where is it located?

User avatar
Fitchie
Wiggling with Experience
Posts: 313
Joined: Sun Aug 30, 2015 1:49 pm

Post by Fitchie » Mon Feb 11, 2019 1:07 am

jbdiver wrote:BTW, I can't seem to find the panel wiring diagram on the vocoder.eu website. Where is it located?
Here you are:
http://vocoder.eu/bom.html

3 downloads are available:
  • BOM
    PCB layouts
    Panel wiring instructions

User avatar
Fitchie
Wiggling with Experience
Posts: 313
Joined: Sun Aug 30, 2015 1:49 pm

Post by Fitchie » Mon Feb 11, 2019 1:10 am

jbdiver wrote:The quantities in the official BOM for these parts is also incorrect.
Did you use a recent version of the BOM?
And please let me know what quantities are not correct to update the BOM where necessary?

Thanks beforehand.

User avatar
jersupereq
Common Wiggler
Posts: 223
Joined: Sat Aug 16, 2008 9:24 pm
Location: SF

Post by jersupereq » Mon Feb 11, 2019 2:15 am

Fitchie wrote:Cost/skil comparison:
  • Solution 1
    10 x trimmer 10 K Ohm
    Desoldering necessary

    Solution 2
    20 x resistor 1 M Ohm 1 %
    Desoldering necessary

    Solution 3
    10 x resistor 12 K Ohm 1/8 watt
    No desoldering necessary
:tu:
Would 1M2 or 910K resistors work? I don't currently have enough 1M laying around

User avatar
Fitchie
Wiggling with Experience
Posts: 313
Joined: Sun Aug 30, 2015 1:49 pm

Post by Fitchie » Mon Feb 11, 2019 2:32 am

jersupereq wrote:Would 1M2 or 910K resistors work? I don't currently have enough 1M laying around
910 K Ohm is fine, be sure they are 1 %.
Don't use resistors greater than 1 M Ohm, small deviations can lead to big differences.

User avatar
elmegil
Super Deluxe Wiggler
Posts: 2888
Joined: Sun Apr 29, 2012 2:57 pm
Location: Chicago

Post by elmegil » Mon Feb 11, 2019 8:45 am

jbdiver wrote:Here are the parts missing:

8 - 10K resistors
1 - 22K resistor
4 - 0.1uF film capacitors

Here are the additional unused parts which can be removed:

1 - 22K resistor
4 - 47K resistors
2 - 4148 diodes
Um. How does that 22K not cancel out? :-D

Aside from that thank you. Let me know if you have issues updating the BOMs.

User avatar
jbdiver
Wiggling with Experience
Posts: 287
Joined: Tue Aug 31, 2010 11:39 pm
Location: Minneapolis, USA

Post by jbdiver » Mon Feb 11, 2019 11:10 am

Fitchie wrote:
jbdiver wrote:The quantities in the official BOM for these parts is also incorrect.
Did you use a recent version of the BOM?
And please let me know what quantities are not correct to update the BOM where necessary?

Thanks beforehand.
Line 16 in Noise Generator: quantity should be 3 - 0.1uF film capacitors

This is the only error I could remember offhand. I'll check again later.

Post Reply

Return to “Music Tech DIY”