Questions Orange Squeezer with mods.....R.G. and Mark Hammer...works good now

[KMS]

The OS from GGG.


I have read a lot about this little stompbox called the OS.

I notice that Mark Hammer has drawn up some mods for it on GGG located at this link
http://www.generalguitargadgets.com/diagrams/orangesqueezer_marks_variable.gif

I am also aware of Mark's "Tangerine Peeler" but I don't really feel like making a PCB for it right now (I will definitely do that sometime in the future if Mark does not beat me to it.)

One question;

The Mark Hammer mod schemo shows the opamp with pins 5,6,and 7 connected to the higher output mod and on t JD's original schemo pins 5,6, and 7 are used in a different manner (I think for the gain).

Does it make any difference (phase maybe) if I use pins 1,2,and 3 for Mark's mod instead of pins 5,6,and 7?

The reason why I ask is because using pins 5,6,and 7 for the higher output mod would mean a change in wiring and I already made the GGG PCB.

I'm trying to cut down on PCB designs, it takes me a long time to draw them up.  I guess I could make an extension mod PCB (very easy board) and route shielded leads from the opamp section of the GGG PCB over to the mod board.  I just don't know enough about opamps to draw my own conclusion here.

Another question;

I could only get JFET 5485(s) and I have heard nothing about whether those will work in the OS or not.

Will the 5485 work?

I have the 1N34 diode and the 4558 IC.

I am going to make some other mods too based on R.G.s comments at GGG and some of the build reports I read at Tonepad......

R.G. mentioned that two resistors dominate the attack and decay functions of the envelope follower so I am assuming that pots could be used in those locations to get some extra time variance on those functions.

Will I get anything from such a mod (Pots at R12 and R11) or will that just allow me to adjust for the JFET variables?

I'm going to try some 10turn trim pots on this project, one 10K for the bias pot and another 500K in place of the 100K at R12 (if that has potential).  Also plan to use a 2K trim pot in place of the 1.5K R11.  and 200K in place of R9 220K and 10uf for C7 and 1uf NP for C6. And a dual gain 10K audio for the brightness and level. Any other suggestions are welcome as I am wiring this up on my experimenter board now.

I don't want true bypass, so I will be looking for a pot connection that I can operate with a SPDT (and maybe a resistor on the SPDT) that will defeat the circuit and yield a non-envelope follower function with the output level remaining the same with no compression, or no attack, or an instant release/decay.  Any ideas on that?

I'm not sure how I will place it my chain(s) yet but I will start with in the No. 1 position in my box just before my GEO A/B/Y splitter. 

Thanks in advance,

You guys have help me become "an obsessed stompboxer" !  I could never repay you.

Don't forget to check out the box...............

http://www.elixant.com/~stompbox/smfforum/index.php?topic=39106.msg292389#msg292389

[alextdel]

I have built this with an extention board for the second half of the op amp. Works now. The schematic is missing a coupling cap between the two stages. If you don't put this in the biasing on the amp is all over the place. Put a 4.7uF between the level pot and the pin 3. I also played with the value of the C9 as I could hear no difference with the brightness pot at either end.

Its a nice compressor. Good luck

[KMS]

So if I can assume......you wired the opamp with pin 1,2, and 3 to the output as you said to put the 4.7 uf between the level pot and pin "3". 

Thanks for the tips.  I'll look into the Brightness issue over at GEO as R.G. has some EQ info over there.

[alextdel]

All correct - make sure you earth the non inverting input and short the output to the inverting input of the op amp for the NON-used channel. I found the circuit was very noisey with this channel not connected to anything.

[Mark Hammer]

I have built this with an extention board for the second half of the op amp. Works now. The schematic is missing a coupling cap between the two stages. If you don't put this in the biasing on the amp is all over the place. Put a 4.7uF between the level pot and the pin 3. I also played with the value of the C9 as I could hear no difference with the brightness pot at either end.

Its a nice compressor. Good luck

Nota bene: There are two OS derivatives mentioned in KMS' note.  The 4.7uf cap described by Alex is for the Tangerine Peeler.  The biasing of the modified OS posted at GGG is fine from all reports.

I just added on the pinouts of the two op-amps in the GGG posting after I drew it up, without reference to anyone's particular layout.  Use of the left side or right of the dual op-amp you happen to use for IC1a and IC1b will make no difference if you are perfing/vero-ing it, although clearly if you are adapting someone else's existing layout, you will need to pay attention to what they have used for the firststage in their layout.

FETs....

The 2N5457 and the 82k fixed resistor at its drain act like the two halves of a volume pot.  As the FET resistance drops in response to envelope signal at its gate, this mimics the ground leg of that "pot" getting smaller, which is the same as turning down the volume of the input to the op-amp.  You CAN use other FETs, since they will all work as voltage-controlled resistors in one manner or another.  But I have no idea if the value of that 82k resistor and the 10k trimpot would be appropriate for the FET you mention.  They might be, but it is not guaranteed.  Be prepared to experiment a bit until you get it right.

I don't want true bypass, so I will be looking for a pot connection that I can operate with a SPDT (and maybe a resistor on the SPDT) that will defeat the circuit and yield a non-envelope follower function with the output level remaining the same with no compression, or no attack, or an instant release/decay.  Any ideas on that?

Scrap this idea.  Nice out of the box thinking, but it will unfortunately get you a huge unwanted volume boost.  The gain of the op-amp is preset in the OS.  Unlike other types of compressors where the gain is reduced to produce compression (see the Dynacomp for an example), the OS prefers to keep the gain fixed and turn down the input to achieve compression.  If you cancel that "turning down" (by disabling the envelope signal feeding the FET), you revert to full gain with no attenuation = big loud.

I suppose you *could* consider using one switch to both disable the sidechain AND add another resistor in parallel across the 220k feedback resistor to drop the gain, but my hunch is that would require a lot of experimentation to get it right, and would still end up being cumbersome.

If you are attempting this sort of "nonbypass" to preserve the output buffering of the OS, then what I would suggest is using a 3PDT to select between either the OS or a separate onboard FET-based or op-amp-based buffer/gain stage with its own output level control.

If your goal is to switch between light and heavy compression, then consider using a footswitch to add another fixed resistor in parallel with the 82k input resistor.  Dropping that combined resistance will reduce the compression amount but limiting how much attenuation is produced when the FET resistance goes lower.

[KMS]

You gave me some good pointers.

The FET info helps me a lot....to understand the circuit and what I can experiment with to tweak it too.


I don't want true bypass because it always has some very small pop sound.  I know how to get rid of the pop another way using led/ldr and a capacitor all on a separate power source but that is also cumbersome to do and it adds about 100 to 300 ohms in line with my signal (not really a big deal adding 300 ohms?).

My supplier has plenty of ten turn trim pots all for 50cent to $100 each so I might still give the defeat mode a whirl with a trim pot connected to the 220K R, but first things first.....hoping to fire it up yet tonight.  If it sounds really good........maybe I don't need to bypass it at all.


Thanks a million for all you do.

[KMS]

The OS does work with the 2N5485

I got it to sound like the sample with no mods.

I have since made some of my own mods.....sort of just experimenting and I don't know what I'm doing.

I got 2 more seconds of "apparent sustain" from the original on the 5th string doing a 15th fret bend.

[KMS]

Good news on the OS.

I got it to give me apparent sustain for most anywhere on the neck as long as I want it, endless sustain, and nearly immediate attack where it chops off the peaks.  The very first note in a run will have a slight peak but any other notes like 16ths following the first will not have a peak.

Part of this is due to some mods I made on my own which I will post later because the mods might change as I box this thing up.  I was able to put the bias pot to a maximum setting (not full on but full compression).  This setting causes unwanted noise when the notes are not being played but only when in front of distortion FX.  I put my MXR noise gate clone after the OS in the chain and the noise problem is solved.

Fantastic FX this OS and I almost gave up on it.......glad I didn't.  I can now play the last lead guitar part for Skynyrd's "I need You" where Rossington holds an A for about 10 bars (4th string 9th fret) and then slides in and out that A on a very simple riff but never uses a pick for any of it.  If you know what song I'm talking about (most people don't) the note sounds like it is done on the organ with a manual pitch shifter, but it is not .....it is Rossington on the guitar.

The "clean" sound is unbelievable! With the mods I made I can tell no difference from the OS and true bypass to the tone (no distortion).  Maybe just a slightly less treble with the OS but the OS ads a nice little ring. The difference in tone is insignificant, if it even is there.....could be just my imagination.

Also, my OS is sensitive to RF like mobile phones but I have not put it in a steel box yet.

It might take me a while to finish boxing this thing up but I'll come back and post all my mods when I'm done.   

Thanks all for everything..............I love the OS........I need to stop playing with it now and box it.


Later

[KMS]

OK...I'm going to post these mods because it looks like others are working on this too.

Keep in mind that I get unwanted noise with this set up in front of a distortion FX  with the gain on the distortion turned up, but I built a MXR Noise Gate clone that solves that problem with the noise gate between the OS and the distortion FX. The noise gate also has all the mods on it from tonepad.com for which the attack and release mod although are hardly on....help out a lot.

I get endless sustain and also the big peak on the first note is dampened almost immediately and then completely dampened on any 16th notes played in succession after the first note.

I get no noise using the OS straight to the amp which I use a Peavey Blazer 158 as a preamp before any of my big amps.  Note that the Peavey Blazer 158 has some very mild compression built into it.

On my OS from GGG..........
D1 is 1N34
Changed C1 to .01uf ceramic.
Q1 and Q2 are 2N5485
Changed C3 to 10uf electrolytic
R7 set at 3.68K
Changed R9 to 200K
Added 51pf ceramic in parallel with R9
Changed R10 to 6.8K
Changed C6 to 1uf NP
Changed R11 to 1K
Changed R12 to 1M
Added 100pf ceramic across pins 1 and 3 of R13 (not the wiper).

Don't use a head set to test or adjust the Orange Squeezer.......high frequency oscillation will result and endless sustain will not happen with a head set.

Some of this might change as I put it in the box, if so I will post it.

[Mark Hammer]

Thanks for the updated information.  These strike me as very sensible and coherent changes.  Coherent, in the sense that they all fit together in a strategic way.

It's a nice circuit for tinkering with.  Glad this all worked out for it.

[KMS]

Thanks Mark....for everything.

I hope someone can make use of the mods I used.

Your right about the tinkering....and I learned  about how this circuit works too as many changes I made did not work.....sort of an experiment session...for me that is.

I tried all of your mods....and I liked them but I didn't want all those knobs on the surface of my box....I'm needing as much room on my box as possible and did not feel like trying to rig trim pots on the GGG PCB so I settled for some improvisation to some of your mods.

One thing that threw me off big-time was using my head set.  My wife was having a fit over all the noise so I started using a head set..................wrong way to experiment with the OS fore sure...but there are some interesting noises on the head set that you don't get from the amp...ha!

I have also be eyeballing that optical compressor from Colorsound  "Supasustain".   It looks so simple and I have a whole bunch of 3904(s) laying around.  I also have a bunch of CLM6000(s) too.

Not sure if I need one but would be fun to play with. 

The only area on the neck I have trouble with using the OS is way up......past fret 15.  The OS works but does not hold endless sustain on each and everything I play up there....but still does a good job up there.  I currently don't need endless sustain up that high.  Also I have not yet cranked the amp.....and all that will likely change once I get that amp cranked.   

[Mark Hammer]

Okay, I guess this brings us to the next set of experiments! 

Thicker longer strings provide greater amplitude rectified (envelope) voltages than thinner shorter ones.  Adjusting  a single half-wave rectifier to respond optimally to both lower and higher notes is unlikely to be satisfactory.  Heck, even units that use full-wave rectification will often split the entire spectrum up into two or more bands and compress them separately.

The "shortcoming" of the OS (though let's be clear and recognize that the late Dan Armstrong designed it and all the other modules in his old series to be simple and inexpensive ways of achieving a variety of efects with minimal fuss) is that the gain stage used for volume recovery is also the same stage used for amplifying the signal enough for producing a suitable envelope signal to drive the FET.  Whatever you do to provide macximum bandwidth of the audio output interferes with appropriate envelope sensitivity (across the fretboard), and whatever you do to shape the relative fretboard sensitivity screws up the tone of the output.

Ideally, then, what one wants is a separate gain/filter stage between the output of the one that is already there and the rectifier circuit, that could be used to tailor the high and low end to achieve roughly equal sensitivity across the fretboard when normal differences in note amplitude are taken into account.  Alternatively, one uses a simple frequency-dividing network to split mids/highs from lows, and adjusts both the gain and time constants of each path to provide more consistent ripple rejection and envelope handling, then combines the two envelopes in some manner.

That is certainly feasible, but it is obvioously going to require a different board layout than what is presently posted around.

[KMS]

I have been thinking along those same lines for some time now......divide and conquer.......conquer the frequency problems, that is.

It was/is my hope that by building my box with multiple parallel channels (ABY splitter then back into a mixer) that I would be able to over come this problem.

I was hoping to build multiple compressors and put some EQ in front of each (forming separate bands) to separate the signal low-mid/mid-high on two channels or use three channels for low/mid/high and then try to find compressors or mod some compressors to optimized each band, then mix back together.

I have already built a 6chanel splitter and it is connected up to a 10chanlnel stereo mixer all in one box with room for about 10-16 other FX depending on how I arrange things.

Actually the OS works so good...it almost solved this problem for me....but still lacking on the top of the neck.

And now I see that what your talking about is not just limited to frequency alone...the thickness  and length of the string play an important roll too.  Thus the top of the neck has issues.....primarily above the 15th fret.

Signal strength is really the issue not frequency...me thinks.

So.... I guess I will still get plenty of versatility from my parallel channels but what I need to solve this problem would be some type of signal strength detector.

Question.....does your Tangerine Peeler address this issue?   

Or will I have to try and cut this path on my own?...which I'm not sure I could do.  I guess my target is to incorporate a compressor-compressor system that will drive the amp just enough to keep the strings vibrating such that the strings keep driving the compressor.  This is really the key to endless sustain.  Then the issue of keeping the treble tone and the attack (I love the attack because it levels out the volume of my hard-picking and soft-picking hammer ons and slides..... those notes that are so fast that it is hard to bring them up to volume).

OT...talking music for a moment.......I play lead and a lot of Lynryd Skynyrd stuff.......and I have been having problems with Sweet Home Alabama.  I can play the song, just like the original....but some of the notes in those real fast runs like the very first lead riff for example are hard to keep the volume up so all notes are heard at equal volume......The second lead riff has a section at the end of the riff also (just before boo hoo hoo).   The OS solved that problem....so I am one happy camper right now.....but I still want the sustain up on the top of the neck....I will no doubt need it in the future.

I don't mind building another compressor...obviously I had already planned on that.....but it would be nice to wrap this up with just one more compressor...mods or no mods.

The build report I read at Colorsound for the Supasaustain is not to encouraging...but you can't really judge a circuit by a build report alone.

[Mark Hammer]

The Tangerine Peeler is just a response to a thread last year about feed-forward vs feed-back control of gain.  Limiters use feed-forward, and compressors generally use feed-back.  I decided to co-opt the level control circuitry for both purposes since it seemed easy enough to do.  It will NOT, however, address the problem you raise.

The approach to divide-and-conquer that I suggested was really meant at providing a sort of differential sensitivity adjustment across the fretboard, in an effort to compress all the notes the same amount, whether they were up high or down low.  Of course, in the absence of high volume, the natural sustain characteristics of higher notes are pretty weak so there are realistic limits to what even improved compression can do for them.  Divide and conquer can also create smoother compression.  Because the ripple from low notes is more evident than it is from high ones, if you can smooth out the ripple from each band in a tailored way you can achieve the smoothness without having to sacrifice attack time to do it.

The Marshall ED-1 is rather clever in that it includes a sort of Big Muff tone control feeding the rectifier circuit.  This way, if you want it to be more sensitive to higher notes, you rotate the control in the treble direction, and if you want it to squish low notes but leave dynamics of higher notes intact, you rotate it the other way.

[KMS]

My brother has a Marshall (can't remember the model) and it compresses real nice too, but not as good as this OS I built.

It is finished now, in the box and working great.  I had some problems with the sensitivity pot and a bad bypass switch. Also had to hook the noise gate up in the box which took some time. 

Now that I have all the bugs worked out (including a new 20 turn 10K trim pot for sensitivity) the noise is gone but it is nice to have the noise gate for stacatto notes and to stop the hiss from the BOSS OD2 when I'm not playing.  I took the bypass off the OS...... don't need it.

I get about 5 to 7 seconds of sustain at the top of the neck with the amp NOT Cranked.  The attack is plenty fast as I mentioned before. 

I was never able to get much out of the OD2 with the turbo off before but with the OS in front of it.....wow what a difference.

A lot of new sound discovery for me in the last two weeks....all good discovery.

I put a 47pf ceramic in place of the 100pf due to all of my 100pf have a portion of the metallic disc protruding out of the ceramic.

The 20 turn pot makes it so easy to adjust the OS to exactly the right impedance.  That sweet spot that everyone talks about occurs through the course of 1/2 turn on the 20 turn pot (which is actually a 26 turn pot 10K).  Also the main change too the sweet spot occurs in 1/8 turn (about 60 ohms).  I can adjust that pot 1 ohm at a time so it is optimized to the minimum setting that still gives max compression.

That's it......my OS is done.

One more thing...I played around with the supply voltage (I have an adjustable PS) and for my set up 9.1VDC seemed to work the best. 

Cheers!