Orange Squeezer +++ PCB

[nelson]

yeah post them in a .Gif file and I will update the project file.

[mojotron]

Thanks nelson for the layout to work with. I checked all of the schematic to the layout (from input to output, then output to input - pretty old-school but works..) and I found a few issues with the layout I posted my corrections to the parts layout as:

http://mojotronics.com/images/os+++lo_v3.GIF

Notes:
- most issues were around the FETs (the pinout) and the opamp
-red is when I added connections, dark blue is where I opened connections.
- I added some lables to the parts layout, these are needed for navigation

If someone could double check this work that would be great.

This is the corrected verson of the layout, with all of the issues fixed:



I made a board, and I'll start soldering in a little while.

On thing though, after looking around for info on how this was designed, I get the feeling that even this design has not been verified, much less the PCB layout. So, this might get interesting. As far as I can tell from looking over the design, I may indeed need to boost the signal - adding a small gain stage, and Mark's notes indicate some likely issues with high end... We'll see. I guess what I'm saying is that if anyone is really interested in this, you might want to wait until I get this build done, I may be adding things as needed. And, I post all the changes....

[nelson]

this was one of the first boards I designed, I am fairly certain i can make it neater and smaller, perhaps 1590b sized, I will wait till Mojotron submits his findings on whether a gain stage needs to be added and I will design from there.

can start with thinner traces and smaller pads

[mojotron]

this was one of the first boards I designed, I am fairly certain i can make it neater and smaller, perhaps 1590b sized, I will wait till Mojotron submits his findings on whether a gain stage needs to be added and I will design from there.

can start with thinner traces and smaller pads

Ok.. All done, I found that the layout needed to be flipped, which when you build a board with ICs and FETs... that is a photo-negative, you just need to bend all of the IC pins up and flip the IC and FETS. I corrected this and this is the final version of the layout (fixed link above too)



Gain was not an issue at all, the gain is just right - just enough to drive a tube amp a bit if you want. High end recovery is an issue - but that's a squeezer for ya...

The sound.. Well it's really awesome and very cool, if you like the orange squeeze. I would have to say, this was the first time where I liked the design without any mods, but I am actually thinking of mods already.

I think it sounds just right once I took C7 out, the sound (which is basically the same as the squeezer) is gently compressed and so natural it sounds very subtle when it's on, but as soon as you shut it off you hear a big difference. I would say it's "naturalness" is akin, if you can imagine, to using a boost all the time... you kind of feel naked without it.

The sound is actually hard to describe if you're used to harsh compressors - like my Fender COM-1. It's just like playing my Deluxe Reverb on 8 or 9, but with out the distortion. It's fat, and supper warm without being bassy or muddy. It's warmth is interesting, it kind of makes the middle pickup on a strat sound like a neck pickup, the neck pickup gets so fat that it sounds fatter than my Les Paul neck pickup. However, you do lose high-end, but it's a natural sound - not like the sound you get if you dump high-end with an RC-shunt circuit.

The controls are extreemly subtle. In fact, I was liking the sound so much that I spent most of the evening jamming with it (I don't have it in an enclosure yet), and at first I could not hear any changes that the attack and decay controls made. But, after 3 hours and some tweaking I could hear what these controls did. The attack control makes the compression effect max-out quickly or retards the compressed compensation a bit. The decay control seems to control the length of time until the compression starts to decay after reaching the max. The compression control is not that subtle, but has a big effect on the attack and decay controls. Gain, when maxed out, will drive an amps input fairly hard, and you get a hint of disortion - the gain is not subtle at all.

So, I'm really happy with this build. I actually really like the trace width and pad size. I think the only thing that I would change on this last version of the layout would be to move a few things around to make it a little smaller, only use 2 output pads on the controls where you can do the jumpers on the pots, and get rid of the jumpers. With 5 knobs, this would go into a big enclosure annyway... check out a full drive some time - that board is huge and loosly populated. I like the layout to the controls, it is very neat to wire up. On the jumpers, I agree with your earlier thoughts, if you insulate the jumpers they are fine, but I usually try to avoid them.

I'm woking on some mods to get better high end response, and to acentuate the effect of the attack and decay controls. But, I just ran out of solder braid.. so it may be a day or 2...

[nelson]

Try feeding the OS with a really hot signal, perhaps an active bass slapping/popping, there have been reports of the compressor completely dyign and fading back in. I believe then the internal gain will have to be lowered and a gain recovery stage added to end to rectify this.

I have updated the layout, could you check it over for me?

[moosapotamus]

Thanks for that detailed report, mojo. This was originally intended to be more bass-friendly than the stock squeezer. I'll be interested to see what other mods you might come up with. And as nelson said, I'd also be interested to hear how your unit responds to a really hot input signal. Maybe just put a boost in front of it if you don't have a bass.
Here's one of the previous threads nelson was referring to...
http://www.diystompboxes.com/sboxforum/viewtopic.php?t=34590

Thanks
~ Charlie

[mojotron]

Thanks for that detailed report, mojo. This was originally intended to be more bass-friendly than the stock squeezer. I'll be interested to see what other mods you might come up with. And as nelson said, I'd also be interested to hear how your unit responds to a really hot input signal. Maybe just put a boost in front of it if you don't have a bass.
Here's one of the previous threads nelson was referring to...
http://www.diystompboxes.com/sboxforum/viewtopic.php?t=34590

Thanks
~ Charlie

Hey Charlie, I was hoping you would chime in - this is a cool circuit I want to thank everyone for putting this out there.

I don't have a bass around, but I could see how it is bass friendly - the EQ is right in the range where I imagine it would do more heavy lifting then what I put it through with my strat, 335 and Les Paul. I guess it's been a while since I played through an original OS, but if my memory is correct the sound is pretty much the same as the OS - with more controls. When I compared it with my Fender Com-1, which has an attack control.., I really notice the mid boost. It's not a bad boost at all - with a strat - it's the sort of clean/yet-not-squeeky-clean sound that I could play all day. The sound is really smooth compared to the Fender Com-1.

I can re-create that output drop (then re-apearrence) issue by lowering the trimmer connected to the compression control and then whacking one of the middle strings really hard - at least just after I power it on. So I can see Mark Hammer's first point in the other thread. But, that seemed a little too simple to correct with the trimmer, so I assume there is another issue out there when shoving a huge signal into the input. And, the trimmer pot might only address the power-up part of this. I'll play around with this some more tonight.

With 1" of solder braid left, I'll either have to get more braid tonight or be really judicious on my next few moves here.  

[mojotron]

I'm wondering if anyone else has built one of these?

As far as OS-type compressors go - this one is my favorite - I'd like to compare notes if you have built one of these.

Mostly, I'm wondering if someone else has tried to use smaller caps in place of the 4.7uF caps and if this would enable the attack/decay controls to have more of an effect on the compression....

[Mark Hammer]

Re: Treble loss.

Simplest thing to do is to make a crude bright switch arrangement.  Stick a 10k fixed resistor (and a .01uf cap in parallel) just ahead of the output volume pot.  It *will* reduce overall output volume, but there was generally enough to start with anyways, so no great harm done.  This will give a bit o a boost above 1600hz.

[mojotron]

Re: Treble loss.

Simplest thing to do is to make a crude bright switch arrangement.  Stick a 10k fixed resistor (and a .01uf cap in parallel) just ahead of the output volume pot.  It *will* reduce overall output volume, but there was generally enough to start with anyways, so no great harm done.  This will give a bit o a boost above 1600hz.

Hmmm... I will try that - treble loss is a bit of an issue. Funny thing is that I had grown to like it - it's sort of twisted how the mind works...

So, Mark does this just address treble loss or does this make the decay and attack controls seem a bit more responsive?

My thought was that if I changed C5, C10 and C11 to 2.2uF or 1uF then the charge time ramp on these caps would allow the decay (and perhaps attack) of the effect to be a bit more responsive to their respective controls. Right now the decay time seems pretty long anyway. 

I'll have to play around with this, simulation is a bit difficult with this as what I really need is a piece-wise input and that can not be done on any "free" sim-tools I have seen.. 

Thanks!
Mike

[nelson]

I never did get round to building this 

I could probably make the board ALOT smaller too.

[Ben N]

Just on a quick glance, there a couple of things that I see that could be tightened on your next revision, and maybe those will open other possibilities. But I'm no expert, so take it FWIW, and if I am wrong about these, I'll be happy to learn about it..

(Referring to the component view in the v2 pdf): 
1) The 4u7 coming off the 10k pot has long traces that can easily be eliminated--that cap could easily fit right there by the pot pads. 
2) The 150k R off the wiper of the gain pot can be flipped around 180 deg to shorten that trace. 
3) Pin 5 connects to a stack of two resistors and a cap in parallel.  Again, referring to the drawing, note that the 470k and the cap both connect to ground, while the 390k connects to the other (right) side of the layout.  So it makes sense to move the 390k to the top of the stack and flip it 180 deg, thereby significantly shortening the trace on the other side of that resistor.  At the same time, you will only need a short bridge between the ground sides of the 470k and the cap and a single trace to ground.
4)  Also, on the lower left of the pic, I would take that 10k that connects to the junction of the Ge diode and the 4u7 cap, and flip it, that is, looking at the trace from the R to the diode, put a pad at the elbow for the flipped resistor, then use what was the other resistor pad for the cap instead, eliminating the remaining trace to the cap.  Now you can also shorten the trace on the other side of the cap by putting the second pad for the cap closer to the 4k7.
5) Move the attack pot to the south end of the board.
I haven't come up with anything for that power jumper, but if it doesn't cause any problems, who cares.

More generally, I wonder if turning the 4558 90 deg, so that pins 5-8 are facing "up" wouldn't save some real estate. Maybe not.
HTH,
Ben