DIY - Elektor Vocoder - General Build Thread

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

Moderators: lisa, Kent, Joe.

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

Post by Fitchie » Sat Jan 26, 2019 2:42 am

elmegil wrote:Input calibration....
"Now your speech input has been calibrated for using line levels."
What if I have a mic I want to use? :D
It seems to me I can always turn down the input pot to use line input but I can't turn up to use the mic if I've calibrated for line levels....
You have 2 options:
  • Using a microphone for live speech performance

    The speech input volume potentiometer is followed by a preamplifier with an adjustable gain between x1 till x1000 for any input sensitivity between 7,7 mV and 7,7 V.
    The input impedance is 10 kOhm and in practice microphones with almost any impedance can be used. Depending on the chosen microphone adjust P16 accordingly to get a sufficientl loud signal without distortion.

    Using a line output with recorded speech

    To use a line output as speech input apply the same signal as used for the carrier to the speech input and turn P16 until both signals are almost equal (peak-to-peak + 15 V RMS).

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

Post by elmegil » Sat Jan 26, 2019 3:34 am

Right, but to calibrate the gain so I'm not clipping the mic.... ?

"Depending on the chosen microphone adjust P16 accordingly to get a sufficiently loud signal without distortion."

So basically I have to adjust this by plugging the A connection into some kind of amplifier?



Also, with the line level input to the voice I see that I'm basically half-wave rectifying the voice signal, that doesn't seem right. That's when the gain is dialed back to the point where I'm not distorting horribly....

Image

(and yes, I know that has the input voltage pp lower than it should be; I have a mutable instruments module tester generating the sine wave, through an attenuator, measured as .77VRMS on my meter (the RMS on the scope is baaad), and then put it through a mult to send the same input to both inputs. For whatever reason, the multing of the signal is cutting it in half; regardless this shows the not-quite half-wave signal I'm getting on A on the backplane)


Also ignore the offset of "2>" it's another quirk of this DSO Quad scope.

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

Post by Fitchie » Sat Jan 26, 2019 4:51 am

elmegil wrote:Image
Please confirm?
  • Scope channel A (light blue) = carrier input measured on pin B
    Sope channel B (yellow) = voice input measured on pin A

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

Post by Fitchie » Sat Jan 26, 2019 1:25 pm

And please can you post a detailed picture of the input/output board?

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

Post by elmegil » Sat Jan 26, 2019 6:14 pm

Fitchie wrote: Please confirm?
  • Scope channel A (light blue) = carrier input measured on pin B
    Sope channel B (yellow) = voice input measured on pin A
Yes.


Further thinking about it though, I realized that the same thing I have to do to make the two traces line up means that the bottom half of the curve is hitting outside the range of the scope itself. When I moved the yellow trace up, the rectification went away.

Sorry to alarm anyone :)

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

Post by Fitchie » Sun Jan 27, 2019 1:10 am

No sweat !
:tu:

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

Post by Wolfo! » Sun Jan 27, 2019 10:33 am

Houston, we have a problem...

I passed the calibration of the i/o board. Now I'm trying to calibrate the filter boards, but I'm encountering problems.
- Bridges between G-H and I-J are set
- Only psu, io board and filter board inserted
- Filter board has the bridge.

When following the calibration procedure (blowing in the mic) I have absolutely no output from the device, measuring on the out jack. Turning the volume up/down briefly distorts the scope "flatline". Mic volume is at max.
Inserting a sine on the IN pin of the filter board has no result (no output).

I've installed the vocoder in it's basic config (all the filter boards including lpf and hpf, io board and psu), G-H and I-J still bridged.
I introduce a speech signal (literally a speech on youtube) on the voice input, no output. I can see the leds above the filter controls react on the input signal.
When I also introduce a sine on the carrier input, still no output. When I cranck the carrier pot to max, it breaks trough to the output signal but distorted (the sine is a square).

One more thing: The leds are on when nothing is connected (about 50% of the brightness of the power led). On both lpf and hpf I've followed the instructions and connected all the grounds together: these leds are the brightest (about 50% of the power led). On the other filter boards the ground of the led is connected separately to one of the grounding connections of its filter board. (I use connectors, it was easier to apply a separate connector on the led ground) These are less bright, again about half as bright as the two boards with grounds grouped.

I guess something is wrong with the io module, but I have not enough knowledge of the thing to test.
Fitchie, can you help me out?

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

Post by Fitchie » Mon Jan 28, 2019 12:17 am

Wolfo! wrote:Fitchie, can you help me out?
Off course I'll help you out !
Wolfo! wrote:I introduce a speech signal (literally a speech on youtube) on the voice input, no output. I can see the leds above the filter controls react on the input signal.
When I also introduce a sine on the carrier input, still no output.
I'll answer in more detail tonight, but basically I think 2 things are causing the described problems:
  • - your filter boards are not are not sufficiently adjusted
    - do not introduce a single frequency sine on the carrier, but apply a full audio spectrum signal
Nothing to worry about, at the beginning I also thought that the device did not work at all.
:oops:

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

Post by Fitchie » Thu Jan 31, 2019 1:46 am

Wolfo! wrote:I passed the calibration of the i/o board. Now I'm trying to calibrate the filter boards, but I'm encountering problems.
- Bridges between G-H and I-J are set
- Only psu, io board and filter board inserted
- Filter board has the bridge.
Correct, this is the vocoder in the most simple configuration:
- bridged filter board (input connected to output)
- bridged backplane (G-H and I-J connected)
- power supply (+/- 15V and +/- 5V)
- input/output module

Wolfo! wrote:When following the calibration procedure (blowing in the mic) I have absolutely no output from the device, measuring on the out jack. Turning the volume up/down briefly distorts the scope "flatline". Mic volume is at max.
Inserting a sine on the IN pin of the filter board has no result (no output).
As discussed in a private mail, this is not OK.
So please see below the schematics of a filter board:

Image

Please can you put an oscilloscope on IC2 pin 14 and IC2 pin 8 and post some screen dumps? Do this also for IC1 pin 8 and IC1 pin 7.
If you have decent signals here, the filter boards are fine.

Due to elimination now only the OTA's (CA3080) remain. Unfortunately there are fake copies on the market.
But please post the requested screen dumps before we draw conclusions.

Wolfo! wrote:One more thing: The leds are on when nothing is connected (about 50% of the brightness of the power led). On both lpf and hpf I've followed the instructions and connected all the grounds together: these leds are the brightest (about 50% of the power led). On the other filter boards the ground of the led is connected separately to one of the grounding connections of its filter board. (I use connectors, it was easier to apply a separate connector on the led ground) These are less bright, again about half as bright as the two boards with grounds grouped.
Looks correct, on both vocoders I did make, the LEDs are slightly on when there is no input and blink depending on the signal level. I do not notice intensity differences at the different LEDs. And the connection of the ground seems to have nothing to do with it. Please check R33 (22 K), R34 (120 K) and R35 (1 K).

Wolfo! wrote:I guess something is wrong with the io module, but I have not enough knowledge of the thing to test.
OK, let's do some checks on the ipunt/output module.
Please can you put an oscilloscope on IC16 pin 6 and IC18 pin 6 and post some screen dumps?

User avatar
Ricochet
Learning to Wiggle
Posts: 48
Joined: Sat May 06, 2017 3:16 am
Location: Germany, Muenster

Post by Ricochet » Thu Jan 31, 2019 7:17 am

After I had unfortunately ordered the wrong CONEC sockets (with solder lugs) :doh:
I see, that REICHELT didn't have the right ones (with solder pins) in their assortment :bang:

Image

I have managed to squeeze the solder lugs so that they fit into the pads :yay:

Image

Image

But.... :woah:

there is now a small distance of 2mm between socket and PCB ...
Image

My question:
is it so to use or not :hmm:
You do not have the required permissions to view the files attached to this post.
sailing away with the first winds of the new day !

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

Post by elmegil » Thu Jan 31, 2019 12:10 pm

An opinion only:

As long as your connectors are square, and consistent height, I would expect it to be fine. If they're crooked, or some are higher and others lower, then I would not want to do it that way.

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

Post by Fitchie » Thu Jan 31, 2019 2:00 pm

Ricochet wrote:My question:
is it so to use or not :hmm:
Go for it !
Ordering new connectors is quite expensive and your solution works as well, as long as all pins come neatly through the PCB with sufficient space to solder.
And in case you mess it up, there is still the possibility to order new connectors. In that case I'll send you a replacement prototype backplane for free.

Suggestion:
Drill the PCB and screw the connectors to the backplane in the foreseen holes with 3 mm bolts and nuts. As there are no traces behind the mounting holes, it's safe to do this.
Apply washers or bushings between the connector and the PCB to align them exactly on the same height.

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

Post by Wolfo! » Fri Feb 01, 2019 11:30 am

Fitchie wrote: Please can you put an oscilloscope on IC2 pin 14 and IC2 pin 8 and post some screen dumps? Do this also for IC1 pin 8 and IC1 pin 7.
If you have decent signals here, the filter boards are fine.
This flatlines, only noise of a few mV
Fitchie wrote: OK, let's do some checks on the ipunt/output module.
Please can you put an oscilloscope on IC16 pin 6 and IC18 pin 6 and post some screen dumps?
Also flatline ...

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

Post by Fitchie » Fri Feb 01, 2019 12:49 pm

Wolfo! wrote:This flatlines, only noise of a few mV
Please can you check these points on 1 of the filter boards ?

Image

A1 / IC1 / Pin 14
A3 / IC1 / Pin 8

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

Post by Wolfo! » Fri Feb 01, 2019 2:00 pm

Fitchie wrote: A1 / IC1 / Pin 14
A3 / IC1 / Pin 8
Nothing :(

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

Post by Boogie » Fri Feb 01, 2019 4:01 pm

hey Wolfo!
can you check the power supply ?
pls check voltages at pins 4 to GND and 11 to GND for IC 1 and IC 2
and check at pins 7 to GND and 4 to GND for IC 3 and IC 4
-------------------------------------------------------
“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 » Fri Feb 01, 2019 4:53 pm

Please can you check the bridges on the backplane ?

Image

You have to connect G & H and I & J.

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

Post by Wolfo! » Fri Feb 01, 2019 5:09 pm

Fitchie wrote:Please can you check the bridges on the backplane ?

Image

You have to connect G & H and I & J.
They are connected. Will check the voltages tomorrow ...

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

Post by Wolfo! » Sun Feb 03, 2019 10:15 am

Boogie wrote:hey Wolfo!
can you check the power supply ?
pls check voltages at pins 4 to GND and 11 to GND for IC 1 and IC 2
and check at pins 7 to GND and 4 to GND for IC 3 and IC 4
Hi mr Boogie,
The voltages seem to be correct:
IC1/2: pin4 14,77V pin 11 -15,07V
IC3/4: pin4 -15,07V pin 7 14,73V
That is with power, io board and ons filter board installed.

I'm thinking about 2 things, but more on a feeling base:
- the ca3080s are bad (but I don't know this can explain everything)
- I messed up the panel wiring of the io section (albeit I've checked this a hundred times). I used other jacks (pj301bm from thonk), but the wiring is correct.

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

Post by Fitchie » Sun Feb 03, 2019 10:25 am

Wolfo! wrote:I'm thinking about 2 things, but more on a feeling base:
- the ca3080s are bad (but I don't know this can explain everything)
- I messed up the panel wiring of the io section (albeit I've checked this a hundred times). I used other jacks (pj301bm from thonk), but the wiring is correct.
If there is only a 'flat line' on IC1 pin 14 and IC1 pin 8, there is in all probability no signal from the input/output interface on both the carrier (pin B) as the voice input (pin A).

In the described setup as you have now, add a sine wave around 440 Hz on both the carrier & voice input and put your scope on the A & B pins at the right side of the backplane.
Please post a screen dump of both signals.

Image

Not only the speech signal but also the carrier signal does not arrive at the CA3080's, so they remain outside of suspicion at this time.

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

Post by Wolfo! » Sun Feb 03, 2019 10:51 am

Fitchie wrote: If there is only a 'flat line' on IC1 pin 14 and IC1 pin 8, there is in all probability no signal from the input/output interface on both the carrier (pin B) as the voice input (pin A).

In the described setup as you have now, add a sine wave around 440 Hz on both the carrier & voice input and put your scope on the A & B pins at the right side of the backplane.
Please post a screen dump of both signals.
This is the setup. Sine is inserted trough the front panel jacks, both voice and carrier pots are at about 50%.
Yellow is pin A, green is B. Ground is taken from the io board.

Image

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

Post by Fitchie » Sun Feb 03, 2019 11:30 am

Wolfo! wrote: This is the setup. Sine is inserted trough the front panel jacks, both voice and carrier pots are at about 50%.
Yellow is pin A, green is B. Ground is taken from the io board.
Excellent ! The bus print does feed the signals to the filter boards on the correct levels and that's exactly what we want, so let's check the opamps on the filters again.

Please can you insert one of the filters (wherever you want, except on the 3 connectors at the left side).
Put your scope again on IC1 pin 14 and check the signal. Do the same for IC1 pin 8 and post a screendump of your scope.

Image

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

Post by Wolfo! » Sun Feb 03, 2019 11:52 am

Fitchie wrote: Excellent ! The bus print does feed the signals to the filter boards on the correct levels and that's exactly what we want, so let's check the opamps on the filters again.
Please can you insert one of the filters (wherever you want, except on the 3 connectors at the left side).
Put your scope again on IC1 pin 14 and check the signal. Do the same for IC1 pin 8 and post a screendump of your scope.
Hi Fitchie,

Thanks for the help, and sorry for the image quality: the Bitscope app crashes when trying to save the wave, so I have to use screenshots ...

Filter card (3900hz) is in the lpf slot, in-out jumper is put in place. Inserting sine in both voice and carrier on the front panel (same as above). Ground taken from io board, yellow is pin8 and green is pin14. Sine 500Hz:
Image

with higher freq (5KHz):
Image

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

Post by Fitchie » Sun Feb 03, 2019 12:15 pm

Good news !
The signals from the input/output module are clearly available for both speech as carrier.

Let's check the complete filter board:

Image

Legenda:
  • Green : already checked & OK
    Red : speech signal, check all outputs from the consecutive opamps (A2, A5, A7, A9, A6 and A8)
    Blue : carrier signal, check output from the opamp (A4)
    Orange : VCA, unable to check with oscilloscope
Can you check also transistor T2?
This should be a PNP transistor type BC 557.

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

Post by Wolfo! » Sun Feb 03, 2019 12:44 pm

IC1 pin 1 (yellow) and pin 2 (green)

Image

IC2 pin 1 (yellow) and 7 (green)

Image

IC2 pin 8 (yellow) and 14 (green)


Image

IC3 pin 6 also flatlines ...

Post Reply

Return to “Music Tech DIY”