Messages

Wow that's really cool, and there are so many sounds with such a simple (dare I say elegant!) design!
I'd love to hear more vibrato (full wet, minimum delay) samples with various depth settings!
Amazing job, thanks for sharing!

And yeah, it does make a difference! Thanks a lot!

What surprises me is that the Deluxe Memory Man actually goes to 100% dry or wet.

Funny enough, this seems to be due to the large electrolytic caps used in the DMM on the dry/wet outputs. I remember posting this finding here a long long time ago (after I had tried 1uF film caps in the Dirtbag project) and never quite knowing exactly why (maybe due to the cap's reactance/ESR, or something else).

You could try swapping C4 for 47uF and C38 for 22uF like the DMM and see if you get more range on the blend pot (with the negative side going toward the pot in each case).

Wow, really... I dont know much but I would never have imagined it would have an impact! I'll definitely try it out soon, and I'll report here
Thanks a lot!

To get fully wet or dry I think you'd need to offboard the blend pot and convert it to be something similar to the Pork Barrel blend with the middle lug connected to ground or VB. I think it was mentioned in one of the development threads about blend circuits best working this way (?) as it'll shunt all of the signal you don't want somewhere, rather than being passive.

I wouldn't have a suggestion as to the accompanying resistor values best suited. Nor do I really have any idea what I'm talking about but as the saying goes, the best way to get the right answer is to post the wrong one 

Thanks! Yes I thought about doing that, a properly done mix section, but I like the idea of having the real DMM output section...

Not sure that the buffers would help, but worth a try.

I think the reason you are going to get some signal is that the blend pot is 10k, so at most you are adding a 10k resistor in the signal path.

First thing that popped into my head, If you really want to kill the signal from one or the other, maybe expose T1 which will let you completely turn down the wet, and do a similar pot to ground for the dry signal, which would let you completely turn down the dry...and sum those together with an op amp.  It means you have an extra control to fiddle with, but it would definitely let you control the mix all the way to silence from either side.

Thank you for your answer!

What surprises me is that the Deluxe Memory Man actually goes to 100% dry or wet. At least in my memories of it. I remember that I could get a pure vibrato sound.
I've thought about external pots for dry and wet levels but honestly it's just too many pots! I like the simplicity of a blend pot.

Also, I've seen 10k used as a blend pot, but only with properly buffered outputs before the pot. That's why I'm asking the question!

Maybe I should just try a 50k blend pot in my Dirtbaby, and that'd be enough...

Hi everyone

I've been using a Dirtbaby for years now, and Im wondering about something. The blend pot won't go to 100% dry or 100% wet, it's always a mix (with really low level of wet signal when fully CCW, but still you can hear it a bit, and same thing for full CW)

Would it help to add an output buffer to the dry path and the wet path?
Im thinking simple BJT buffers after C4 and C38, so right before the blend pot.

What to you guys think?

I found the simplified Diamond Compressor schem. It was actually in my Diamond folder, d'oh. It doesn't show the EQ part but it does show one approach to simplifying it. Courtesy of Fred Briggs.

Unfortunately, my 4:2 project seems to be a bust. Just can't figure out what's wrong. I breadboarded the envelope section and it works as it should. On two built PCBs, no envelope response. Possibly something could have gone with the boards themselves but I don't see any defects. Last thing to do is compare voltages on the PCB and breadboard to see if it has any clues. If not, I'll just set it aside for another time.

Sorry about the 4:2 project!

Yes I remember that simplified Diamond comp.
I actually prefer the Afterlife/4:1 by a large margin, because it can get subtler levels of compression (the Diamond, even at minimum, still compresses quite a bit) and to my ears it has more headroom (I powered the Diamond at 18v, but the 4:1 at 9v).

So, from what I have been able to grok the 470k resistor connected to vRef is either about reducing current consumption or possibly minimizing current offset between the inverting and non-inverting inputs of the op-amp. Whether that came directly from experimentation or as a "design practice" I don't know. What I do know is that if you are buffering the Vref supply is that you don't need it. Just connect the non-inverting pin to vRef directly. It's already a much lower impedance source than the typical voltage divider via two resistors.

On the buffers - IIRC this is something I looked into when I started on the project but I don't remember exactly what I arrived at. Certainly you can simplify those parts to a single emitter follower and I think some worked out a project that did that (although I don't remember who).

My advice is to breadboard the circuit and really thoroughly investigate these issues to arrive at the solution that has the fewest parts for the best results. I don't see any reason you cannot couple the ICI.2 output directly to the Tilt EQ without the buffer stage but I'm also not smart enough to say that without actually testing it, haha.

One other piece of advice I can give: play with C7 (your schem). Using a higher value there can increase the overall sustain but the biggest impact it has is how quickly the compression turns on. When you put in larger values the opto LED starts lighting up a bit slower. This can actually be a good thing because the response of the compressor delays slightly and makes it more musical, IMO. IOW, it's not quite as immediately squishy but has more "bloom", if that makes sense.

Thanks for the explanation about the 470k to vref!
Yeah, I dont see any reason to use a buffer before the tilt EQ, but Im not smart enough either haha

About the 100uf cap, I remember reading a lot about it on freestompboxes (where people believed it was 100n at first) and midwayfair recommending to socket it and maybe use a switch for different decay settings (22uf, 47uf, 100uf etc.)
But after trying different options I'm happy with 100uf there usually.

Oh by the way, I'm pretty sure you can remove R22 and R27 on the Rough Cut as they're bypass resistors (and see how you can bypass the tilt eq and keep just the inverting opamp) : http://revolutiondeux.blogspot.com/2012/06/diamond-compressor.html
(pic attached)

Anyways, I'll let you know if I breadboard it! Thanks a lot for all the info!

I think it would work. I'd stick the other half of the OP275 as a Vref buffer like the RoughCut. R28 may need to be lowered in value - not sure about that. The one drawback I can see is that the 4:1 is very bright as is and the tilt eq may end up boosting those highs into unpleasant territory. Also, with the 4:1 having separate clean and comp controls you'd have to kind of reset both for volume output depending on how bright the EQ is.

I have a second version (the 4:2 for 1590B instead of 1590G) that I've been working on that reduces the clean and comp to a single blend knob, but it does not have any eq. I'd show the schematic for that but it turns out I'm having a problem with it that I haven't solved yet.

Quick drawing of a possible evolution for the 4:1
Clean & comp = MIX
Master Volume
Tilt EQ at the output like the Diamond
Sensitivity trimpot (thanks midwayfair)
And buffered vref 

I never understand when the + pin of an inverting opamp needs to be connected to vref through a resistor (470k in the EQ section here). And why did Diamond use two transistors instead of one as a buffer... no idea haha! Oh well!

No, you're right. I was misremembering the buffer thing. You do have one half an opamp leftover so maybe there's something else to use it for. On the 4:2 I used an extra op-amp as an output gain control.

The 4:2 the problem I am having is that the LED is not responding to the envelope detector. It just stays at full brightness (I tested this part out before soldering in any opto). I need to look at it again - I think something was throwing voltages off. It's been a week and that feels like last month already.

Oh yeah that's odd. Is it the same envelope circuit as the Afterlife/4:1?
I think a mix section followed by a gain stage, similar to the VFE Bluepring or your Florist, would work well for a comp. Maybe I should try that and incorporate a tilt EQ the somewhere on the signal path (but would it affect only the compressed signal, or the general output, that's my question of the day haha)

I think it would work. I'd stick the other half of the OP275 as a Vref buffer like the RoughCut. R28 may need to be lowered in value - not sure about that. The one drawback I can see is that the 4:1 is very bright as is and the tilt eq may end up boosting those highs into unpleasant territory. Also, with the 4:1 having separate clean and comp controls you'd have to kind of reset both for volume output depending on how bright the EQ is.

I have a second version (the 4:2 for 1590B instead of 1590G) that I've been working on that reduces the clean and comp to a single blend knob, but it does not have any eq. I'd show the schematic for that but it turns out I'm having a problem with it that I haven't solved yet.

Thanks for the quick reply!
I dont understand the part about buffering vref, I dont see it on the Rough cut schematic.
I also dont understand why R28 should be lowered, but that's quite easy to try!

Usually I lower the impedance of the 4:1 to a pretty standard impedance, and I dont find it overly bright! But I see what you mean about the volume thing. Though I do like to have two seperate volumes now! But yeah maybe a simple clean blend would work well.

Whats the problem you're having with the 4:2? Sure I cant help you but you never know!

Hey,

So the 4:1 is by far my favorite compressor. I prefer it to the Diamond, especially for acoustic guitar.
But I remember enjoying the tilt EQ for electric guitars, and I got an idea: why not add the tilt EQ to the 4:1 ?

So here's my idea (sorry for the poorly photoshop schematic)
And for reference, the schematics for the 4:1 and the Rough Cut (Diamond compressor)

What do you guys think? That'd work right? Added bonus: no more phase reversal, since the tilt EQ reverses phase, as does the 4:1. Now maybe I could put the EQ on a switch, and when off, add an inverting buffer, so the signal is alway in phase whether the EQ is on or off. We'll see...

Thanks for all the info, yeah I noticed there was different smd sizes!
Between all the madbean delays and the deadastronaut ones, I think I will come up with something eventually
Thanks!

Wow that's so beautiful, amazing job! I would love to hear it too!
What modulation circuit did you use? Lately I've been thinking of building a Deathlaw with a little less filtering on the repeats and adding the modulation section from the Dirtbaby.

I've seen an old MXR unit that delivered 10v instead of 9. And Ive read somewhere about some units producing random voltage spikes.
Anyways, I'd use tc1044s (max input voltage 12v)  or lt1054 (15v) instead if it happens again!