MIDI-CV converter and CV Recorder / Sequencer

wmonk · Posted: Thu Jan 28, 2010 4:49 pm Post subject:

Cool to see so much interest in the boards. The first will be the MIDI2CV converter. Are there any connector types I should include on the board? (I'll add the 6-pin MOTM for myself)
Should I include an onboard 7805 regulator? Some other things I must add?

Tell me! Very Happy

I'll work on the PCB design next weekend.

jmejia · Joined: Mar 12, 2009 Posts: 114 Location: portland

I'm in for the sequencer! maybe 2!

diablojoy · Posted: Thu Jan 28, 2010 11:25 pm Post subject:

b-funk · Joined: Jul 17, 2007 Posts: 193 Location: Berlin

Hi,

I'm interested in 2 P-Sequencer boards.

As I have a PIC burner I don't need the PICs, but would offer help if anyone needs to get PICs burned.

+1 for the onboard 5v reg
+1 for the 4 pin .156 power connector
+1 for .1 connectors for all other connections

Best wishes,

Tobias

Luka · Joined: Jun 29, 2007 Posts: 1003 Location: Melb.

that site has some awesome projects
i might try the 2181 vca

xjscott · Joined: Apr 25, 2007 Posts: 232 Location: Appalachia

I'd be interested in 2 of the MIDI2CV PCBs.

mattias · Joined: Jan 26, 2010 Posts: 28 Location: Cologne

I'm in for 2 MIDI2CV PCBs with PIC and DAC.

ejr27233 · Joined: Feb 08, 2010 Posts: 52 Location: UK

I'd like 2 P-Sequencer and 1 Midi>CV boards.

wmonk · Posted: Wed Mar 10, 2010 5:57 pm Post subject:

If you thought this project was dead, I have good news: it isn't! Smile

Routed the PCB for the MIDI2CV today. For those who happen to have Eagle on the PC, please check my layout.

Cheers,
Wmonk

midicv.brd

Description:

MIDI2CV pcb layout

Download (listen)

Filename:

midicv.brd

Filesize:

34.98 KB

Downloaded:

167 Time(s)

midicv.sch

Description:

MIDI2CV schematic. Please don't distribute this. You can get the original schematic at www.acoupel.com

Download (listen)

Filename:

midicv.sch

Filesize:

295.45 KB

Downloaded:

141 Time(s)

xjscott · Joined: Apr 25, 2007 Posts: 232 Location: Appalachia

BTW, Eagle's available for Mac also, and I think Linux as well.

Are there specific changes to the schematic we should be taking a look at here?

Opening the board I got a "no forward - back annotation will be performed" warning, not sure what that means or if is to be expected.

Traces are real curvy, I didn't even realize that was an option, interesting.

Note to others: rename midicv_114.sch to midicv_655.sch if you want to be able to do error checks between the two.

wmonk · Posted: Wed Mar 10, 2010 6:25 pm Post subject:

I used Eagle on linux, and yes, it's available for OSX too.
Think the main changes are the power connector, the 5V regulator and the use of a double sided board instead of single sided. There still is one SMD capacitor, I'll provide these for those who want that. I had a few problems with the power connector side of the board. It's all properly connected, but I think it can be better. So if you have some hints and suggestions, I would like to hear them Smile

scriptstyle · Joined: Jan 22, 2008 Posts: 250 Location: nj

i think that your board is cramped in areas, there is open real estate so you should be able to even it out a bit. i would also say that it would be nice that all the through holes are the same size for continuity(at least among ICs). my last gripe would be that your traces seem a bit sloppy.

xjscott · Joined: Apr 25, 2007 Posts: 232 Location: Appalachia

OK here's some comments on the layout. I started to make changes and was thinking I was going to be able to do them and just post something showing an example, but this is my first time with Eagle and I am really slow at it, so I thought I'd get you some feedback rather than wait.

* The two DACs I recommend rotating 90 degrees clockwise so pin 1 is downwards. DACs have an analog side and a digital side, and it is mirrored in the silicon. You want the digital side to be facing your MCU and the analog side to be facing away, and towards your op amps. This will result in less digital noise bleed.

* And of course all analog on one side of the board all digital on another. You have sort of followed that, with left being analog, right being digital.

* The noise filtering caps on the power pins should be placed as close to the power pins as possible. I wasn't sure this was the case with all of them.

* I recommend maintaining a separate analog and digital ground. They can be connected back at the power bus, but it generally helps to have separate ones. I realize the original schematic doesn't do this, but it was for a one sided board and so that would have probably been a problem to route.

* The power and ground busses should be physically larger than the signal traces if possible.

* Consider filling in empty spaces on the board with ground planes. This will improve noise characteristics and also reduce the amount of etching chemicals needed.

* I am not sure about that C7 under the IC - is that a sole SMD capacitor that is on the second side? Usually one wants to avoid having parts on the other side, but maybe it was unavoidable.

* I agree with the other comment that things are really crowded in areas and empty in others. I don't think this could be robotically populated because of some of the spacings. Not that we are doing that, but some things seem crammed in there.

On the 5V regulator I had a question. I had thought that the 5V power bus on your 6 pin bus is already regulated to exactly 5.0V from your chassis power supply, is that not the case? If it is so, would this have less than 5V after dropping across the regulator. But there must be something I am not grokking here.

Floppy · Posted: Fri Mar 12, 2010 1:12 am Post subject:

i would take 2 p seqs as well.
do they offer the board only or also the frontplate ?

wmonk · Posted: Fri Mar 12, 2010 9:05 am Post subject:

Thanks for the comments xjscott! This is really helping me! Smile

I can answer one question immediately, about the 5V regulation.

skylab76 · Joined: Dec 30, 2007 Posts: 38 Location: Austin

Oh man , I said no more projects, but I'll take two P-sequencers, please.
Very Happy

loydb · Joined: Feb 04, 2010 Posts: 393 Location: Providence, RI

I'd grab a couple of each.

mono-poly · Posted: Sun Mar 14, 2010 1:56 pm Post subject:

I would love a midi to cv board.

droffset · Posted: Sun Mar 14, 2010 5:10 pm Post subject:

I'm in !

skylab76 · Joined: Dec 30, 2007 Posts: 38 Location: Austin

wmonk · Posted: Mon Mar 15, 2010 12:10 pm Post subject:

I changed some things on the PCB, and have some questions. Almost done routing.

xjscott · Joined: Apr 25, 2007 Posts: 232 Location: Appalachia

I think SMD caps and resistors are not a problem in a project. It's just when you get to the SMD ICs that you need superhuman skills and dexterity. They are not really intended for hand soldering of course.

I was hoping you wouldn't ask about the DAC power! When I've done this in the past the DACs and ADCs had separate analog and digital power pins and I was confounded about this one with only the one.

Let's look at the MCP4822's data sheet.

http://ww1.microchip.com/downloads/en/devicedoc/21953a.pdf

Page 14
7 - AVss is the analog ground pin

So analog it is.

With the LEDs that mirror the gate outputs that's a good question. If they were PWM signals from the PIC modulating the brightness of the LEDs they'd be digital grounds. But these are just solidly lit and tapping the gate signal, which is an analog control signal so analog ground, but comes from the PIC and so has a real nice sharp and harmonically rich edge. The diodes mean current is only flowing one way so any noise there is cut in half. I'm not sure. If you choose analog it is a path for some PIC noise to come into the analog ground on the board, if you choose digital it is a path for some PIC noise to come into the synthesizer being driven. Neither is a perfect choice but we have to pick one. Maybe either way would be ok. I'd be surprised if it would be a noticeable issue. Let's say analog, unless someone has another thought on this.

Oh, and analog ground traces should not run under digital components, particularly the high clock speed components like the MCU.

blue hell · Posted: Tue Mar 16, 2010 8:37 am Post subject:

Yesterday in the chat room I suggested to woody to use a digital ground with the LED grounds from the following reasoning: The the gate outs are digital in nature, i.e. on or off, so a bit of noise on them is of no concern. At the same time they will be switched and you would not want that to end up as clicks in the CV signals.

wmonk · Posted: Wed Mar 17, 2010 3:24 pm Post subject:

Thanks for the help. Came up with this. Actually more messy then the previous one, but it has some improvements. Have to add the ground planes, and remove the ground routing on a few sides. I'm not an expert after all. Wink

midicv2.brd

Description:

The pcb layout

Download (listen)

Filename:

midicv2.brd

Filesize:

32.62 KB

Downloaded:

112 Time(s)

midicv2.sch

Description:

Download (listen)

Filename:

midicv2.sch

Filesize:

296.15 KB

Downloaded:

115 Time(s)

wmonk · Posted: Mon Mar 22, 2010 8:47 am Post subject:

And updated the board. Please check for errors, then I can order a prototype Smile

Cheers,
Woody

midicv2.sch

Description:

Download (listen)

Filename:

midicv2.sch

Filesize:

296.13 KB

Downloaded:

108 Time(s)

midicv2.brd

Description:

Download (listen)

Filename:

midicv2.brd

Filesize:

34.49 KB

Downloaded:

115 Time(s)