News:

Forum may be experiencing issues.

Main Menu

First SMD build!: Dirtbag (Deluxe Memory Man) with V3205 Mods

Started by thomasha, October 22, 2019, 10:03:59 AM

Previous topic - Next topic

thomasha

Hi,
this project took a long time to be finished, but was totally worth it!
I started from the Dirtbag schematic and made my first attempt with SMD parts.

Many thanks for the Total recall and Dirtbag PDFs, they have a lot of info, and are really helpfull! The mods for the v3205 worked like a charm.

Got some tweezers and started this one. Well, the tweezers were too big for the 603 resistors,
but luckily my wife has some really sharp tweezers for eyebrows, and they were really good for the job.

I added all the v3205 mods to the layout, because for a first try I would not use expensive MN parts...
Let's see some pictures!
I used a double sided PCB, so the thermal transfer was interesting, but etching was OK. Here I already added the connections from one side to the other.



Then some components:


I wish it looked better. My early mistakes are highlighted with green wire. Maybe in my next layout I do a better job designing both layers, so that 2 layers are enough.
The small board is for the SA570 and the first TL072. It fits over the pots, between the enclosure wall and the input jack.

A view from inside the box:

The input jack moves over the pots when the cable is connected. The gap is really small.

Then a crappy pic of the build:

I soldered everything like a lid, where most of the wires to the pots work like a hinge, with the overload circuit holding the other side of the board down.

Another bottom view:


And the final result:


and a sample:


Now, some questions...
1- I did the calibration with a 250 Hz 0.5v p-p sine wave, and noticed that the signal distorts at the compander when the LEVEL pot is above 9 o'clock (1.8v p-p at pin 7 of the first BBD). In other words, the compander distorts the signal, visible with an oscilloscope, before the BBD does. The compander gives a fairly high gain (0.2v p-p clean to 1.3v p-p clean). Is that right? Vref of the compander is 1.77V, what is well within spec, I guess.

2- I wanted to reduce the gain of the second stage, before the compander, so that both compander and BBD would distort at the same time.
Unity with true bypass is achieved with the LEVEL pot at 12 o'clock, what means distortion when using humbuckers. Would it make sence to reduce the gain pre-compander to avoid this kind of distortion?

Cheers
Thomas

alanp

"A man is not dead while his name is still spoken."
- Terry Pratchett
My OSHpark shared projects
My website

madbean

I think that's the big compromise with running it @ 9v when using v3205 (meaning distortion). There's just less headroom so it starts clipping early on, and the compander does not respond that well to it.

Things I found that decreases distortion on v3205 version:
Put a 1uF between pins 6 and 5 (in series with the 68k resistor) of the compressor side. I found this on the Memory Boy when I was tracing it out.
You can also try lowering the rectifier cap (pin1) to 470n or 330n.
Increasing the 100k series input resistor to 220k will help some, too.
Series resistance at the input of each BBD - like 100k or higher.

This is an insane creation and kudos to you for attempting it! My advice would be if you have this much gumption then do another version in a 1590A or B that's 100% DMM (18v or 24v power with MN3005). I think you will be much more satisfied with it than the v3205 version.

Invertiguy

 :o Holy crap! If your builds get any more dense they just might collapse into a black hole and swallow the earth! Seriously awesome work, man!
Doomsday Devices

Scruffie

The internal bias of the compander is 1.8V but did you adjust the external biasing resistors for the output (Pin 7)? It is offset a bit and compander distortion is subjectively nicer than BBD.

The reason there's the extra gain stage up front of the compander is the feedback and dry signal form a divider, so it needs some compensation to get infinite repeats. You can probably afford to cut it a bit though.
Quote from: madbean on October 22, 2019, 11:21:57 AM
Put a 1uF between pins 6 and 5 (in series with the 68k resistor) of the compressor side. I found this on the Memory Boy when I was tracing it out.
That should reduce the companders gain and cuts a tiny bit of bass, probably a good option, if you can actually get it in there! If not you could just try removing that 68k.

This was designed at the time of weak single coils, you're cutting the headroom already by lowering the voltage so the input not offering a huge range isn't too surprising. Plus it wasn't designed to be true bypass so unity gain would originally be wherever the level pot was set.

It's a bit of an irritating design regarding level as it's cut and boosted five times before it even gets to the BBD, plus bias shifted.

But besides all that, that's nuts  8)
Works at Lectric-FX

jimilee

5 points for using smd and 5 points for etching that MF! 10 points for Gryffindor!


Sent from my iPhone using Tapatalk
Pedal building is like the opposite of sex.  All the fun stuff happens before you get in the box.

madbean

I was so inspired by your build that this morning I decided to give it a try. Here's a pic (I had to blow it up a bit). It's not working, though.


jimilee

Pedal building is like the opposite of sex.  All the fun stuff happens before you get in the box.

nzCdog

looks pretty good bean, gl with troubleshooting ;D

Wow Thomas!  This made my morning... inspiring work 8)

thomasha

Thanks guys, I will try some different configurations, but to me it is already a good analog delay. The modulation is really good!

QuotePut a 1uF between pins 6 and 5 (in series with the 68k resistor) of the compressor side. I found this on the Memory Boy when I was tracing it out.
I added the 1uF and a 6.8k resistor, as in the PDF

QuoteYou can also try lowering the rectifier cap (pin1) to 470n or 330n.
I tried to understand how the compander works, and one of the questions I had was how the control scales for lower voltage (the rectifier and variable gain). I can try this.

QuoteThis was designed at the time of weak single coils, you're cutting the headroom already by lowering the voltage so the input not offering a huge range isn't too surprising. Plus it wasn't designed to be true bypass so unity gain would originally be wherever the level pot was set.
That makes a lot of sense. The first thing I noticed when I finished it is that it sholdn't be true bypass. The level drop to achieve the clean repeats would never be a problem, plus it can work as a booster. In the send and return section of the amp sometimes it is possible to adjust the volume too, in that case it is just a matter of adjusting the signal going to the pedal and the signal coming backl. Not that it matters when everything is already distorted...


m-Kresol

holy crap. this takes it to another level. 5 pots in a 1590LB...
Chapeau good sir, chapeau
I build pedals to hide my lousy playing.

My projects are labeled Quantum Effects. My shared OSH park projects: https://oshpark.com/profiles/m-Kresol
My build docs and tutorials

oip

you damn maniac!  amazing.

i guess this is blasphemy but.. PCB fab houses are pretty good

dawson

Criticism is encouraged: constructive, or otherwise.

benny_profane

That's astonishing. Well done! I'm trying to see if I can center a SPST switch for a 1590lb tap tempo...

Also...

ATTN: pickdropper

thomasha

Well,
I tried it again. This time with the MN3005 at 15v and a jack for tap tempo.
First some guts.



I almost hat to add a 2 mm spacer between lid and enclosure, but I could reduce the height of the voltage doubler and switched jack.

I don't think the tap is something usable with a 600 ms delay, but it was worth the try. I guess the next one will be a "normal" one.