Deluxe Micro Man

[rustypinto]

Hello All,
In spirit of all the cool projects being created, i thought i would share my Deluxe Memory Man work-alike. The motivation behind this project was to use the MN3008 BBD, which is easier to get than the MN3005 (although this won't be true soon enough). Also, SmallBear is selling counterfeit MN3008 pairs (pictured below) at a reasonable price. Its a tough project, so good luck!

Here is the project file : Deluxe Micro Man Rev A

Pics:






Short clip: http://soundcloud.com/youngpedals/deluxe-micro-man

[oldschoolanalog]

Thanks for sharing this nice project "over here".
Your efforts are very much appreciated!

[Processaurus]

You do nice mechanical design!

[Ummagumma]

I would love to see this reworked for 3208!  

[rustypinto]

Thanks for sharing this nice project "over here".
Your efforts are very much appreciated!

You do nice mechanical design!

Thanks!

I would love to see this reworked for 3208!  

It would be quite easy since i used a positive supply. At a minimum you would simply replace the 7815 with a 7809. Looks like you would have to lower R1/R2 since they set the DC bias of the compressor portion of the compander. 47k/47k would probably work (the datasheet has the equation).

EDIT: you would have to swap the supply pins in the layout. I always forget that MN320X/MN300X have opposite supply pins!

[maarten]

Nice project! As I have some MN3008's still lying around, I might give this a try....
What is C46 (DNI ??) for? I did not see it in the picture...
Thanks for contributing this.
Maarten

[Scruffie]

Thanks for sharing this nice project "over here".
Your efforts are very much appreciated!

You do nice mechanical design!

Thanks!

I would love to see this reworked for 3208!  

It would be quite easy since i used a positive supply. At a minimum you would simply replace the 7815 with a 7809. Looks like you would have to lower R1/R2 since they set the DC bias of the compressor portion of the compander. 47k/47k would probably work (the datasheet has the equation).

EDIT: you would have to swap the supply pins in the layout. I always forget that MN320X/MN300X have opposite supply pins!

I never noticed you swapped the power arrangement... not a bad idea... I see you simplified a few parts aswell.

I recon with 2 x 3205s (instead of 4 x 208s) I might be able to shoe horn it into a 1590BB with onboard pots and no stand up resistors... running at 9V the bridge rectifier and regulator could happily be dropped and a zener used in its place to give the BBDs that slightly lower voltage they prefer (and allow for people to run the opamps at 12V should they prefer while protecting the BBDs) then as you say, just a case of swapping the pin arrangements and the compander resistors.

Cheers for the inspiration! 3205 Memory man here I come  

One slight edit... to that, VGG should be 14/15 so to save on having a voltage reference i'd add a diode to each to drop VGG by 0.6V... close enough to 14/15th of 9Vs.

What about the outputs of the BBDs though... would they go to ground now instead or am I going insane.

[Govmnt_Lacky]

Bookmarked! 

[Taylor]

Very generous contribution and great PCB design. I would not be man enough to tackle that on a single-sided board.

[Scruffie]

Nice project! As I have some MN3008's still lying around, I might give this a try....
What is C46 (DNI ??) for? I did not see it in the picture...
Thanks for contributing this.
Maarten

I think that stands for Do not install, the cap's in parallel with the clock cap so it's probably for experimental purposes.

[aron]

Beautiful!

[damianlothers]

you're my hero and muse

[Brymus]

Heck Yeah
Excellent!
Such a well put together project,thanks for sharing.

[rustypinto]

I never noticed you swapped the power arrangement... not a bad idea... I see you simplified a few parts aswell.

I recon with 2 x 3205s (instead of 4 x 208s) I might be able to shoe horn it into a 1590BB with onboard pots and no stand up resistors... running at 9V the bridge rectifier and regulator could happily be dropped and a zener used in its place to give the BBDs that slightly lower voltage they prefer (and allow for people to run the opamps at 12V should they prefer while protecting the BBDs) then as you say, just a case of swapping the pin arrangements and the compander resistors.

Cheers for the inspiration! 3205 Memory man here I come 

One slight edit... to that, VGG should be 14/15 so to save on having a voltage reference i'd add a diode to each to drop VGG by 0.6V... close enough to 14/15th of 9Vs.

What about the outputs of the BBDs though... would they go to ground now instead or am I going insane.

I was never a fan of shunt regulators for analog delays, but it is certainly possible to save space. The problem is making the supply pins happy while dealing with the current requirements of the clock/CD4047, which changes over the full range of 100kHz-10kHz. The regulator method is more stable. Even if you use 12VDC on the op-amps, you can get away with a 78L09 powering the BBD's/clock, which is not terribly space constraining.

I tried to speak a little bit about VGG in my design notes. The datasheet specifies it to be 14/15V away from Vdd, but it still works if Vgg = Vdd. If you use the bias scheme i did, you can't get away with a simple diode or resistor drop because the voltage is in the wrong direction. You have to use a resistor divider to create ~1V with this scheme. I did breadboard it with both the resistor divider and where Vgg = Vdd, and i could not hear any difference at all. I remember reading an old BBD app note (Panasonic or Reticon?) that explained this pin in more detail, and mentioned that Vgg could be the same as Vdd for convenience (and it sure was convenient for the density of this layout), but said ideally it should be ~1V away from the supply.

[rustypinto]

What is C46 (DNI ??) for? I did not see it in the picture...

I think that stands for Do not install

Correct, it stands for Do Not Install.

With this JFET bias, i found 1nF to give me a usable clock range of 9.5kHz - 80kHz. If you've ever worked with JFETs, you will find that the channel resistance (which is being used to modulate the clock) is not terribly consistent. You may need to slightly shift the clock frequency if your JFET happens to be different from the one i used, and having two capacitor footprints can get you strange values like 330pF + 470pF = 800pF.

[Scruffie]

I never noticed you swapped the power arrangement... not a bad idea... I see you simplified a few parts aswell.

I recon with 2 x 3205s (instead of 4 x 208s) I might be able to shoe horn it into a 1590BB with onboard pots and no stand up resistors... running at 9V the bridge rectifier and regulator could happily be dropped and a zener used in its place to give the BBDs that slightly lower voltage they prefer (and allow for people to run the opamps at 12V should they prefer while protecting the BBDs) then as you say, just a case of swapping the pin arrangements and the compander resistors.

Cheers for the inspiration! 3205 Memory man here I come 

One slight edit... to that, VGG should be 14/15 so to save on having a voltage reference i'd add a diode to each to drop VGG by 0.6V... close enough to 14/15th of 9Vs.

What about the outputs of the BBDs though... would they go to ground now instead or am I going insane.

I was never a fan of shunt regulators for analog delays, but it is certainly possible to save space. The problem is making the supply pins happy while dealing with the current requirements of the clock/CD4047, which changes over the full range of 100kHz-10kHz. The regulator method is more stable. Even if you use 12VDC on the op-amps, you can get away with a 78L09 powering the BBD's/clock, which is not terribly space constraining.

I tried to speak a little bit about VGG in my design notes. The datasheet specifies it to be 14/15V away from Vdd, but it still works if Vgg = Vdd. If you use the bias scheme i did, you can't get away with a simple diode or resistor drop because the voltage is in the wrong direction. You have to use a resistor divider to create ~1V with this scheme. I did breadboard it with both the resistor divider and where Vgg = Vdd, and i could not hear any difference at all. I remember reading an old BBD app note (Panasonic or Reticon?) that explained this pin in more detail, and mentioned that Vgg could be the same as Vdd for convenience (and it sure was convenient for the density of this layout), but said ideally it should be ~1V away from the supply.

(Moved our discussion from FSB over... no point in having a conversation across two forums  )

I quite like shunt regulators not just for there small space but also because the 320X Series of BBDs seem to like a bit less than 9V so it's easier to get a Zener in 8v2 than a regulator... just a fairly neat and tidy simple solution.

Hmm... I think i'll risk it on the 4047, if it was a problem, hardly an issue to wire up a little daughter board 9V or 12V Regulator and the BBD is still going to be regulated by that 8V2 regardless... I guess you could even run it higher than 12V although i'm not sure how much good it'd do.

Ahh of course yeah... it would be a bit of a pain to magic up -0.6V but as i'm giving them positive voltage, should be all good... I doubt there is much different in it really but, good practice 'n' all for the cost of 2 diodes.

One other question... not that I intend to do it right now but would using quads for the 8 Opamps in the audio section introduce any problems/hiss?

It will if its a LM324  :mrgreen:

I seriously doubt it though, but i've never tried it so i can't say for sure. I used SIPs because they lend themselves extremely well to dense designs. With this type of pedal in general, its the clock noise you have to worry about. You notice that i pushed the 4047 as far off to the corner as possible, and made sure to use shielded hook-up wire. I've done other delay designs where the clock just bleeds into the bypass or other parts of the signal path. You may not face this with board mounted pots, but then if you do, its much harder to fix it.

Yeah I don't think a quad would add noise either but thought it best to ask... SIPs i'm imagine do really help with this layout but if I can avoid ordering special parts, I will. The Quad idea was less about this layout anyway and actually leaning more towards a stripped down version of the Memory Man... perhaps using the Stereo Memory Man as a base which is already +V so adapting to the 3205 would be a breeze, dropping the stereo bit, and taking it back to just 4 OpAmps in the Audio Path then adding the LFO of the Deluxe Memory Man so it can do Chorus & Delay at the same time... possibly with 2 x 3205 I haven't decided on that.

Indeed, i'm very aware of clock noise... yeah board mounters do normally help a fair amount and on my current PCB layout (which i'm probably going to start again as I ended up leaving a huge chunk of unused board space which I could really do with!) it's up at the top corner with the BBDs and useing as short as possible clock pin tracks.

[karol3188]

beautiful project!!!

I've only MN3005 now...can I substitute MN3008 with MN3005?

What should I modify?


Thanks!

[Govmnt_Lacky]

I believe

2 x MN3008 = 1 x MN3005

[rustypinto]

I believe

2 x MN3008 = 1 x MN3005

Correct.

beautiful project!!!

I've only MN3005 now...can I substitute MN3008 with MN3005?

What should I modify?


Thanks!

To use MN3005, the mod isn't really that clean, but it is possible. Note that if you stuff 4xMN3005's, the CD4047 will not be able to drive that much capacitive load at higher clock speeds (eg. ~20-100kHz).

To use 2xMN3005, place them at IC2 and IC3. Everything between C30 and C32 is not populated (including C30), and then everything between the output of IC8B up to R41 is not populated. Jumper R16/C17 to C32, and then jumper the output of IC8B to R25.

[Govmnt_Lacky]

I believe

2 x MN3008 = 1 x MN3005

Correct.

Pinto,

Can I assume that you used the MN3005s from Steve at Smallbear? The ones that are ACTUALLY 3008's 

Seeing as you have 4 of them in your lovely circuit above