DIY - Elektor Vocoder - General Build Thread

[Fitchie]

All of the listed mods are done, please see the attached photos and double check if they were done correctly

Looks OK.


Let's do some other checks:

• 1. Low pass filter
  
  Please only connect the power supply, input output module and the low pass filter. Add a 200 Hz carrier and scope on pin 6 IC 11 (LM 741, the one on your picture).
  What do you see on your scope ?
  
  See schematics below:
  
  
  
  2. High pass filter
  
  Please only connect the power supply, input output module and the high pass filter. Add a 3000 Hz carrier and scope on pin 6 IC 7 (LM 741, the one on your picture).
  What do you see on your scope ?
  
  See schematics below:
  
  

[Fitchie]

Scope here:

[MikeDB]

I understand why people like analogue, but isn't this one application when digital could do exactly the same job far simpler ?

[elmegil]

I understand why people like analogue, but isn't this one application when digital could do exactly the same job far simpler ?

Is a build thread for a specific project the best place to engage in philosophical discussions?

[yan6]

Working on the low pass first:

As far as I can tell in order to get signal to the area you circled in red I need to inject signal to the speech input and not the carrier. I jumped both so I had 200hz going into both the speech and carrier. I was following around the signals. From pin6 of A29 I dont get any input. I can see output as far as pin 7 of A26. It's really strange as per below:



I'm pretty sure I'm not seeing anything into pin 3 of A29

[Fitchie]

As far as I can tell in order to get signal to the area you circled in red I need to inject signal to the speech input and not the carrier.

Correct noticed:

• A = voice
  B = carrier

Please can you repost the scope screen dump ?
It looks like the picture is broken.

From pin6 of A29 I dont get any input. I can see output as far as pin 7 of A26. I'm pretty sure I'm not seeing anything into pin 3 of A29.

Based on this input I suspect the LM741 is over saturated.
Try to make very small changes with both trimmers (P5 on HPF and P9 on LPF).

[yan6]

Edit: looks like the pic upload is still broken but the original is showing again

[Boogie]

yan6

The signal at pin 6 A29 should look something like the yellow line o the plot below.
It's the envelope of the voice signal with for each filter a different time constant.
I should thus be a fluent line representing the particular filter's envelope response.
You should get likewise graphs at all filter boards.

The blue line in the plot below is the carrier amplitude modulated by the yellow line.
It's the signal at the output of the mixer with only 1 filter board active.

[Fitchie]

I'm pretty sure I'm not seeing anything into pin 3 of A29

Please can you scope on:

• - pin 1 A15 TL084
  - pin 7 A16 TL084

and

• - pin 1 A25 TL084
  - pin 7 A26 TL084

See above schematics.

[pixelforest]

did other people's heatsinks come with hardware for attaching them? mine from the mouser cart did not and there isn't hardware in the current BOM for this. the M3 x 4mm intended for the standoffs aren't long enough, so i'll have to source something longer as well as nuts.

[Fitchie]

did other people's heatsinks come with hardware for attaching them? mine from the mouser cart did not and there isn't hardware in the current BOM for this. the M3 x 4mm intended for the standoffs aren't long enough, so i'll have to source something longer as well as nuts.

The proposed heatsinks does not come with mounting hardware. Please use 4 ordinary nuts & bolts M3 x 8 / M3 x 10 or equivalent.
These nuts & bolts are not included in the BOM because you have to order them in quantities of 50, way to many.

[pixelforest]

did other people's heatsinks come with hardware for attaching them? mine from the mouser cart did not and there isn't hardware in the current BOM for this. the M3 x 4mm intended for the standoffs aren't long enough, so i'll have to source something longer as well as nuts.

The proposed heatsinks does not come with mounting hardware. Please use 4 ordinary nuts & bolts M3 x 8 / M3 x 10 or equivalent.
These nuts & bolts are not included in the BOM because you have to order them in quantities of 50, way to many.

can you add them to the BOM for those people who don't keep M3 hardware laying around? Would be very helpful for the BOM to have a complete listing of all the parts needed so that it is a complete reference document.

[Fitchie]

Here you are:

• Bolt (pack of 100, you need only 8):
  https://be.farnell.com/tr-fastenings/m3 ... ?st=M3%208
  
  Nut (pack of 100, you need only 8):
  https://be.farnell.com/tr-fastenings/m3 ... t=M3%20nut

Will add them to the BOM.

[pixelforest]

BOM on the build guide also diverges from the excel sheet - build guide lists P3 as a trimmer, should be P4. build guide says 100k trimmer in P1, excel says 10k.

[yan6]

I'm finding that the leds for each of the filter cards is constantly half on, regardless of inputs. Any guesses what could be causing this.

[elmegil]

Ok, gonna try this again. It's been months with no update to the calibration procedures.

I've gone back to the original magazine articles, this is pretty much verbatim, and still is not making sense to me:

If professional measuring equipment is available, a more precise alignment procedure can be considered. Instead of blowing into a microphone, a test signal can be applied direct to the input of the module. A suitable test signal is a 500 Hz or less sinewave, superimposed on a fixed DC voltage.

What is the "input" that is being referenced here? The "voice" input of the entire unit? What is an appropriate fixed DC voltage?

Part of my problem is that the original assumes everything is already wired up to the front panel and everything. So I'm unclear as to how the original instructions map to the revised set up.

Has anyone actually calibrated their filters yet? If so can you explain to me like I'm five, cos I just don't grasp it.

[Fitchie]

Dear all,


I received feedback from 2 customers who had damaged their front panel while tightening the hex nuts (jacks, potentiometers and switches). They had used grip locking pliers for this, with the result that there were circular scratches on the aluminum.

So never use metal pliers or keys !
The anodised panel will be damaged in either way !


To assist them, I ordered 5 extra front panels from Schaeffer to replace the damaged ones. That means that another 3 spare front panels were available at the price of 60 euro plus shipping. If interested please send a private message.

Front Panel.png

I also designed a little tool to facilitate tightening the hex nuts (8 mm and 10 mm). If you can use a 3D printer somewhere, here are the .stl files to print your own set:

http://www.vocoder.eu/hex-fastener-tool.stl

Hex Fasteners.png

[Fitchie]

Ok, gonna try this again. It's been months with no update to the calibration procedures.

Hi Elmegil, I'm building a complete unit on demand for a customer who has no knowledge of electronics. So I've to have to adjust this vocoder as soon as completed (I'm now stuffing the last boards). All steps will be documented as published in the right order. If you have a little patience you can follow these instructions soon.

What is the "input" that is being referenced here? The "voice" input of the entire unit? What is an appropriate fixed DC voltage?

Both voice as well carrier. Forget about the DC voltage, just connect a sine wave (e.g. a 440 Hz 0,77 V RMS) to both inputs as documented here:
http://vocoder.eu/io-module-adjustment.html

Part of my problem is that the original assumes everything is already wired up to the front panel and everything. So I'm unclear as to how the original instructions map to the revised set up.

There is no need to connect all wiring to the front panel, temporary interconnect the output jack connection with the front panel potentiometer connection as advised here:
http://vocoder.eu/filter-adjustment.html

[Fitchie]

I'm finding that the leds for each of the filter cards is constantly half on, regardless of inputs. Any guesses what could be causing this.

That's correct. When a filter card is active, the LED is on half power. As soon as a signal passes through the filter card, the LED flashes fully.

[elmegil]

Ok, gonna try this again. It's been months with no update to the calibration procedures.

Hi Elmegil, I'm building a complete unit on demand for a customer who has no knowledge of electronics. So I've to have to adjust this vocoder as soon as completed (I'm now stuffing the last boards). All steps will be documented as published in the right order. If you have a little patience you can follow these instructions soon.

Great, thank you for the heads up.

[Fitchie]

Dear all,

A little bit off-topic, but given the time of year I think this is allowed for once...


Now that the 1st decade of 21st century is behind us, I wish you and your dear ones a fantastic 2020 where most of your dreams will come true and during which you'll be healthy enough to enjoy those wonderful moments all the way, both in your professional but most of all in your personal life.

fireworks.png

[nologin]

I just ordered all the components to build it. (I had bought the PCBs last year, built was pending). I should start the year well by building this vocoder i've been dreaming about for a long time.

[Croodey]

Hi,
i need help for the adjustment.
i have 2 vocoders here at the same satus (nearly finish ). one is built from myself. the other is built from a friend. i have on both machines the same problems and need help to understand the instructions - maybe german.
If i work how i understand it, there are many trimmers whitch changes nothing.

Thank you

[nologin]

Hi, so far i've only built that power supply, i've got a problem. When t ordered the BOM to reichelt, the 4700uF 40V capacitors were unavailable. I soldered 4700uF 25V instead i thought it would be OK, i must have been wrong.
IC19 is ok, +15V
IC21 is ok, +5V
IC22 is ok, -15V

IC20, is bad, and the output is -15V...
It should have a -5V output. At the input of IC20, there is -25V.

[elmegil]

At the *input* of IC20, that should be perfectly fine.

Verify that you actually have a 7905 regulator and not a 7915?