18-channel DIY Stereo Analog Vocoder

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

Moderators: Kent, luketeaford, Joe.

Post Reply
User avatar
glubsch
Common Wiggler
Posts: 137
Joined: Sat Jul 28, 2018 4:10 am
Contact:

Post by glubsch »

Thalassa, what’s the temperature right now in Spain?
User avatar
julian
i like synthesisers...
Posts: 1579
Joined: Sat May 29, 2010 2:38 pm
Location: UK

Post by julian »

glubsch wrote:OK, I see more posts have come up in the meantime while I was preparing my previous post. To answer them all: yes, my current power supply concept uses a toroidal transformer inside the rack, followed by a linear PSU based on the MIC29300-12WT and Schottky diodes (combined with some beefy electrolytic caps). The toroidal is just 2x 12 V, therefore the power dissipation is quite acceptable. Hardly any ripple at the PSU output (I had to put major effort in a clean PCB design to ensure this). The caveat: There's mains power inside the rack.

BTW, here's a picture of the power board:

Image
This would, to me, really seem the best option - this psu design has stood the test of time, no 'special' components etc. and, as i say, anyone who can assemble the rest of the unit, can surely make sure the high voltages are correctly isolated.
.



_________

www.thebeast.co.uk - Bespoke silkscreened and engraved panel - devices and hardware for the DIY music community

Online shop for panels, boats, woodwork etc - https://thebeast.co.uk/?post_type=product

thebeast.co.uk is IOSS compliant. Read our statement here - https://thebeast.co.uk/?page_id=1931
User avatar
Thalassa
Super Deluxe Wiggler
Posts: 1055
Joined: Sun Feb 28, 2010 7:27 am
Location: Spain

Post by Thalassa »

glubsch wrote:Thalassa, what’s the temperature right now in Spain?
Where I live today will be 38 but during the weekend was 40 :deadbanana:

Coming back to the topic your PSU looks like a good solution for me. :tu:
User avatar
glubsch
Common Wiggler
Posts: 137
Joined: Sat Jul 28, 2018 4:10 am
Contact:

Post by glubsch »

A fellow wiggler gave me the hint that the vertical text on the left ("18-Channel Analog Stereo Vocoder and Formant System") may be in the way of the screw holes for those of you who want to place them at a different position. Who'd actually be interested in NOT using the suggested rack? I'm trying to get a feel for whether I need to offer the panel generally without the rack screw holes.

Personally, I'd prefer to have the front panel shipped completely finished if I started this DIY project. I don't have the precision tools to drill holes into the front panel myself but what about you?

Where else would it make sense to place the text? Underneath the handles? Suggestions are welcome.

Image
User avatar
paterursus
Wiggling with Experience
Posts: 406
Joined: Thu Feb 04, 2016 7:55 pm
Location: Virginia, USA

Post by paterursus »

I would prefer the rack/case solution you have shown so far. I personally have no need for the screw holes in a different position. BTW I also like the power supply configuration you have used.
Wiggling in the 16th century
User avatar
glubsch
Common Wiggler
Posts: 137
Joined: Sat Jul 28, 2018 4:10 am
Contact:

Post by glubsch »

OK, fellow wigglers, here's a new teaser. This is the complete list of semiconductor devices (except for the LEDs) used in the vocoder. It's a direct excerpt from the BOM.

Image

I've spent some time over the weekend to create a general overview (sort of a preamble) of the BOM. You can find it here. I hope this gives you a first glimpse of what to expect for this DIY project.

Some comments about the table shown above: The MC33079 are used for the 18 channel filters with the intent to reduce noise. If you prefer, you can just as well use the TL074 if you don't mind the extra noise. The THAT300 are used for the 1V/oct converters on the internal excitation boards. The MIC29300-12WT and the SB540 are used for the regulated power supply. The JFETs are used for analog switching purposes (such as in the voiced/unvoiced detector to switch between the excitation signals and the noise sources).
Last edited by glubsch on Tue Aug 07, 2018 4:13 pm, edited 1 time in total.
User avatar
ben_hex
Super Deluxe Wiggler
Posts: 6724
Joined: Sun May 12, 2013 12:58 pm
Location: West Yorkshire, England

Post by ben_hex »

This looks amazing!
All DivKid video series to be found here!
http://www.youtube.com/divkidvideo
User avatar
paterursus
Wiggling with Experience
Posts: 406
Joined: Thu Feb 04, 2016 7:55 pm
Location: Virginia, USA

Post by paterursus »

glubsch wrote:OK, fellow wigglers, here's a new teaser. This is the complete list of semiconductor devices (except for the LEDs) used in the vocoder. It's a direct excerpt from the BOM.

Image

I've spent some time over the weekend to create a general overview (sort of a preamble) of the BOM. You can find it here. I hope this gives you a first glimpse of what to expect for this DIY project.

Some comments about the table shown above: The MC33079 are used for the 18 channel filters with the intent to reduce noise. If you prefer, you can just as well use the TL074 if you don't mind the extra noise. The THAT300 are used for the 1V/oct converters on the internal excitation boards. The MIC29300-12WT and the SB540 are used for the regulated power supply. The JFETs are used for analog switching purposes (such as in the voiced/unvoiced detector to switch between the excitation signals and the noise sources).
This looks really doable! All the ICs seem pretty generally accessible. Nice job on the start to the BOM! I'm more and more excited for this.
Wiggling in the 16th century
Ayab
Wiggling with Experience
Posts: 371
Joined: Wed Dec 18, 2013 7:19 am
Location: London, UK

Post by Ayab »

Excellent project. Count me in for 1 x set of pcb's and a panel
Dr Gris
Wiggling with Experience
Posts: 415
Joined: Tue Feb 14, 2017 5:14 pm
Location: Sweden

Post by Dr Gris »

Maybe I'm confused...
I see talk about rack ears and handles.
I don't get the point of handles.
What the front panel really need are holes for mounting it in a 19" rack.
In the last pic I don't see any. Just the holes for the case (Hifi2000) and the not needed handles.
Or am I wrong??

//M
User avatar
paterursus
Wiggling with Experience
Posts: 406
Joined: Thu Feb 04, 2016 7:55 pm
Location: Virginia, USA

Post by paterursus »

Dr Gris wrote:Maybe I'm confused...
I see talk about rack ears and handles.
I don't get the point of handles.
What the front panel really need are holes for mounting it in a 19" rack.
In the last pic I don't see any. Just the holes for the case (Hifi2000) and the not needed handles.
Or am I wrong??

//M
If you look at glubsch's post above on this page, the panel drawing is wider than the one in the photo on the first page. The two additional holes on both outer edges appear to be the rack holes on ears. The inner screw holes look to be the ones for the case.
At least that's how it looks to me. Would be good to get it clarified.
Wiggling in the 16th century
Dr Gris
Wiggling with Experience
Posts: 415
Joined: Tue Feb 14, 2017 5:14 pm
Location: Sweden

Post by Dr Gris »

paterursus wrote:
Dr Gris wrote:Maybe I'm confused...
I see talk about rack ears and handles.
I don't get the point of handles.
What the front panel really need are holes for mounting it in a 19" rack.
In the last pic I don't see any. Just the holes for the case (Hifi2000) and the not needed handles.
Or am I wrong??

//M
If you look at glubsch's post above on this page, the panel drawing is wider than the one in the photo on the first page. The two additional holes on both outer edges appear to be the rack holes on ears. The inner screw holes look to be the ones for the case.
At least that's how it looks to me. Would be good to get it clarified.
I guess you're right, the talk of handles made me think the outer holes was for those.
And by looking at the pic the outer holes looks to be too far from the edges to fit standard 19" mounting, but that could be me being fooled...

//M
User avatar
paterursus
Wiggling with Experience
Posts: 406
Joined: Thu Feb 04, 2016 7:55 pm
Location: Virginia, USA

Post by paterursus »

Now that you mention it, they do look a bit too far from the edges. Hopefully glubsch can clear it up.
glubsch, the panel will have ears for mounting in a 19-inch rack, right?
Wiggling in the 16th century
User avatar
dingebre
Wiggling with Experience
Posts: 323
Joined: Tue Mar 24, 2009 1:28 pm
Contact:

Post by dingebre »

Here are some screen captures from a Front Panel Designer (Front Panel Express/Schaffer) .FPD file. The FPD file is also attached. It is a standard Schroff 19 inch wide panel with standard holes. I arbitrarily selected a 2.0mm thickness and 3U high. The rack holes should at least be elongated to allow for tolerances in rack widths. The information box in the top left corner of each image has the dimensions and locations of the individual holes on the left side. The right side would be a mirror of the left. Obviously, other mounting holes may or may not be needed depending on the enclosure and other constraints. Not everyone will want a rack mount, either (like me :) )

The software is a free download and will at least give the proper dimensions and locations for the mounting holes for the rack and for handles if those are wanted. These files are only examples which I hope will help.

David
Attachments
panel 04.jpg
panel 04.jpg (246.16 KiB) Viewed 837 times
panel 03.jpg
panel 03.jpg (254.79 KiB) Viewed 839 times
panel 02.jpg
panel 02.jpg (257.91 KiB) Viewed 840 times
panel 01.jpg
panel 01.jpg (249.5 KiB) Viewed 851 times
Example 19 inch.fpd
(854 Bytes) Downloaded 30 times
David M. Ingebretsen
AnalogueRealities.com

dingebre@3dphysics.net
dingebre@CFandE.com
User avatar
glubsch
Common Wiggler
Posts: 137
Joined: Sat Jul 28, 2018 4:10 am
Contact:

Post by glubsch »

I'm currently on travel, hence my late reply. I mentioned in an earlier post that I needed some help with the dimensions. dingebre, your FPD is of great help. Thanks a lot.

The panel I prepared was with the holes for the Schroff handles but not with the elongated holes for rack mounting purposes. I'll be updating the front panel over the weekend.
User avatar
glubsch
Common Wiggler
Posts: 137
Joined: Sat Jul 28, 2018 4:10 am
Contact:

Post by glubsch »

I guess the handles need the size that match the holes for enclosure. There's no need for extra screw holes.
User avatar
dingebre
Wiggling with Experience
Posts: 323
Joined: Tue Mar 24, 2009 1:28 pm
Contact:

Post by dingebre »

glubsch wrote:I'm currently on travel, hence my late reply. I mentioned in an earlier post that I needed some help with the dimensions. dingebre, your FPD is of great help. Thanks a lot.

The panel I prepared was with the holes for the Schroff handles but not with the elongated holes for rack mounting purposes. I'll be updating the front panel over the weekend.
You're very welcome. The images are only a "preliminary" guide, but I have found the software to be reliable and accurate.

Thank you for making the project available. I am following with great interest.

David
David M. Ingebretsen
AnalogueRealities.com

dingebre@3dphysics.net
dingebre@CFandE.com
User avatar
guest
Super Deluxe Wiggler
Posts: 7119
Joined: Mon Aug 19, 2013 11:49 am

Post by guest »

glubsch - first off, great work on this. it looks like years have gone into this. ive been following the thread, and noticed that there were not enough standard VCA chips in your BOM for 18 channels. it looks like youre doing a PWM multiplier for the VCA? if you have time/inclination, id love to hear about the design tradeoffs you considered when coming to the decision about how to implement your VCAs.
openmusiclabs.com
User avatar
glubsch
Common Wiggler
Posts: 137
Joined: Sat Jul 28, 2018 4:10 am
Contact:

Post by glubsch »

guest wrote:glubsch - first off, great work on this. it looks like years have gone into this. ive been following the thread, and noticed that there were not enough standard VCA chips in your BOM for 18 channels. it looks like youre doing a PWM multiplier for the VCA? if you have time/inclination, id love to hear about the design tradeoffs you considered when coming to the decision about how to implement your VCAs.
Oh, you are the great dude who wrote the paper "Minimizing Distortion in Operational Transconductance Amplifiers", aren't you!? I read it with great interest, probably more than a year ago, after some time of learning and designing my own modulars using the LM13700, looking for solutions to minimize noise. I used the LM13700 for my first incarnation of the vocoder but looked for an alternative after having been unhappy with the noise adding to the audio and the bleedthrough. When I looked for an alternative for the VCA, I came across Elliott's page. So, yes, it is a PWM-based VCA, based on the principles shown in Elliott's Figure 11.

The 18x CD4053 on the list of ICs should have given it away :hihi:
User avatar
glubsch
Common Wiggler
Posts: 137
Joined: Sat Jul 28, 2018 4:10 am
Contact:

Post by glubsch »

Below is the updated front panel design based on dingebre's FPD template.

I still need some help though. The screw holes are placed for the panel to fit the Slim Line 3U Case. They have a center-to-center spacing of 105 mm. The handles that are supposed to fit are available here. Can someone tell me how they are supposed to be attached to the front panel? Are they supposed to sit over top of the holes meant for the panel to be screwed onto the enclosure? :hmm: :help:

Image.
Last edited by glubsch on Fri Aug 10, 2018 3:52 pm, edited 1 time in total.
User avatar
glubsch
Common Wiggler
Posts: 137
Joined: Sat Jul 28, 2018 4:10 am
Contact:

Post by glubsch »

D33K wrote:Awesome work OP. Did you consider making a modular design where individual channels could be added/removed? It may be have been feasible without the the controls on the bottom
I thought about it at first but found it too cumbersome to come up with a reasonable design. That was more than 2 years ago when I thought about it so I don't remember specifically what made me decide against it. I think it was mostly the fact that the 18 channel filter boards need to be interconnected. That is actually accomplished via a 16-wire ribbon cable.
User avatar
truman_k
Common Wiggler
Posts: 93
Joined: Sun Jun 26, 2016 10:29 am
Location: Hong Kong

Post by truman_k »

I always dreamed of getting a fully blown analogue vocoder. Here is the one that is do-able. Not spending a huge amt of money in getting the expensive THAT chips.
User avatar
guest
Super Deluxe Wiggler
Posts: 7119
Joined: Mon Aug 19, 2013 11:49 am

Post by guest »

glubsch wrote:The 18x CD4053 on the list of ICs should have given it away
yeah, that and the large "PWM" on the silkscreen!! ive been looking into PWM multipliers for a while now, and am suprised i dont see more of them in synth projects. what was the noise floor on the PWM version versus the LM13700? and did you consider the SSM2164 at all?

im glad you read the paper. a fair bit of work went into that, and its always nice to hear that others have gotten some use out of it.
openmusiclabs.com
User avatar
glubsch
Common Wiggler
Posts: 137
Joined: Sat Jul 28, 2018 4:10 am
Contact:

Post by glubsch »

truman_k wrote:I always dreamed of getting a fully blown analogue vocoder. Here is the one that is do-able. Not spending a huge amt of money in getting the expensive THAT chips.
Yes, these are the only ICs that are a bit more expensive than the others but there are only two of them and they are easy to find. In fact, as I generally was looking for a design based on easily procurable parts, I even sent an email to THAT Corp and asked about whether I need to worry about the THAT320 (not used in the vocoder) going end-of-life at some point. The email reply was prompt and I felt assured there are no plans to take the transistor arrays off the market.

Having made the statement about easy procurement though, I've been caught by surprise that the LM3916 is no more available. My intention though is to sell the first PCBs based on my design using the LM3916 and replace it with a different design later.
User avatar
glubsch
Common Wiggler
Posts: 137
Joined: Sat Jul 28, 2018 4:10 am
Contact:

Post by glubsch »

guest wrote:
glubsch wrote:The 18x CD4053 on the list of ICs should have given it away
yeah, that and the large "PWM" on the silkscreen!! ive been looking into PWM multipliers for a while now, and am suprised i dont see more of them in synth projects. what was the noise floor on the PWM version versus the LM13700? and did you consider the SSM2164 at all?

im glad you read the paper. a fair bit of work went into that, and its always nice to hear that others have gotten some use out of it.
I used a fairly non-scientific way to judge the noise: I simply listened to the results and thought I can do better :hihi:. I am quite pleased with the PWM-based solution. However, I don't want to leave a false impression and make it very clear to the community: this design is analog and therefore there's still noise. The VU's are there to help set the input signals (Voice and Carrier) high enough for the output signal to be as far away from the noise floor as possible. I've been quite content with the noise levels and therefore decided not to include a compandor. 8-)

I dismissed using the SSM2164 in general for a number of simple reasons: they are hard to come by and expensive, and if you can get them, they are only available as SMD (the only SMD parts used in this vocoder are the decoupling capacitors and those for the 8-pole filters). Besides, they'd be an overkill for this design.
Post Reply

Return to “Music Tech DIY”