Craig Anderton Compressor

jrc4558 · October 21, 2004, 01:55:34 AM

http://www.fortunecity.com/tinpan/humperdinck/223/comp.gif

Does anybody knows what is the "301" IC that drives the LED?
I searched internet, but without prefixes it's no use... :?
Can it be replace with, say, LM741?

Thank you!

niftydog · October 21, 2004, 02:06:58 AM

LM301

a 741 would probably do.

Mark Hammer · October 21, 2004, 09:37:38 AM

I do believe Craig uses an LM748 in the book, and those should be available to you over at either of our mutual favourite hangouts on Queen Street.

The key thing here is the external compensation, which I gather plays some role in the response of the envelope follower.

If you haven't built the unit before, it's pretty darn clean.

jrc4558 · October 21, 2004, 12:32:31 PM

Thank you gentlemen!

I'm somewhat tired of Flatline, need something new.

So I unearthed this compressor from my archives. Wish me luck.

Mark Hammer · October 21, 2004, 01:34:06 PM

By virtue of having a recovery stage, the Anderton design provides some nice opportunities for overcoming some standard issues with compressors.

One of those issues is that compressors often seem to lose treble. because the Anderton design uses an inverting op-amp as the recovery stage, that makes our task even easier.

Whenever Roland/Boss or someone else wants to introduce pre-emphasis/de-emphasis for purposes of noise reduction or whatever, they almost invariably do it with inverting op-amps. Here, we'll use just the pre-emphasis feature. Stick a smaller resistor and small-ish cap in parallel with the 8k2 resistor. Since the gain of the inverting op-amp is set by the ratio of the feedback resistance to the input resistance, this parallel path with have a higher gain at all of the pot settings, than the other existing path. If you restrict what can pass through that lower input resistor to high end by means of a small-value cap, then you end up with more gain for the treble than for the mids and bass.

For example, if the parallel path is a .01uf cap and a 4k7 resistor, then the lowend rolloff for the added path is just under 3.4khz (1/[2*pi*4k7*.01uf]). With the pot and feedback resistance dimed, you impose a gain of 12.2 on the *entire* signal, but a gain of 21.3 on content above 3.4khz. That will likely restore the bite that gets eaten away by compression. If it isn't precisely the sort of compensation that suits your taste or instrument, you can try other cap values, or even stick a low-value fixed resistor and variable resistor in series with the cap to adjust it., like 1k5 fixed resistor and 10k variable resistor. At max resistance, the entire signal will "prefer" the other path so no high-end boost will be imposed.

Note as well that the circuit easily lends itself to a ducking function if the 2.2uf cap ahead of the Sustain control is fed with an external signal.

PB Wilson · October 21, 2004, 07:12:24 PM

I was also considering building this compressor to compare to the Dyna-Comp/Ross clones and the Orange Squeezer. Anyone care to share their thoughts?

ragtime8922 · October 22, 2004, 08:52:19 AM

Quote from: PB WilsonI was also considering building this compressor to compare to the Dyna-Comp/Ross clones and the Orange Squeezer. Anyone care to share their thoughts?

PB, how do the squeezer compare to the Ross? I built the ross and I love it but I want to build the Squeeze. Also, I want to try the Colorsound SupaSustain and the flatline.

Where is the schematic for the Craig A comp?

jrc4558 · October 22, 2004, 09:32:18 AM

Quote from: [email]ragtime8922@aol.com[/email]
Where is the schematic for the Craig A comp?

Like, maaaaan, like look at the first post, maaaan...

(c) Homer, those were our 'personal' vegetables!!!

Mark Hammer · October 22, 2004, 11:20:43 AM

Quote from: [email]ragtime8922@aol.com[/email]PB, how do the squeezer compare to the Ross? I built the ross and I love it but I want to build the Squeeze. Also, I want to try the Colorsound SupaSustain and the flatline.

The Anderton one is pretty darn transparent. One of the few pedals I've made where I felt that a status LED was absolutely essential. In part that's because of the sizeable headroom afforded by a +/-9v supply.

At maximum squish you can tell it's "on", but at medium amounts of compression you get no obvious coloration. I might point out that it doubles nicely as a clean booster. Just turn down the compression and crank up the gain.

The Ross/Dynacomp sounds great but has a certain coloration that may or may not suit your style. I don't think that makes it a LESSER compressor than Anderton's, just different.

The bipolar supply on the Anderton one is a bit of a downer for those who like to use wallwarts on a pedal-board, and it also poses certain issues for powering up. Mine uses a stereo jack for both in and out, so I have to plug in both sides to completely power up and unplug both to completely power down. It probably could have benefitted from a low-profile slide switch that would have let me turn both batteries on and off simultaneously without being susceptible to accidental switch movement. But the machining is a real PITA, given the chassis its in. probably a good time to try RG Keen's solution to powering up two batteries with a single stereo jack.

David · October 22, 2004, 12:14:46 PM

Constantin:

Please post how you do with this one. I was eyeing it last winter when my Flatline kept going flat. I still don't know if I want to tangle with that diode bridge on perf!

jrc4558 · October 22, 2004, 04:41:42 PM

I shure will!

gorohon · October 23, 2004, 01:36:41 AM

I've built this one (EPFM compressor). I liked it when it worked. I don't know how it happened, but one day I turned up the gain/output real high and it stopped compressing. The damn CLM6000 burned out. This really pissed me off and I banished it to my closet. I probably will dig it out of my closet and get it working again--too good to let it collect dust.

Anybody have an opinion on how I could've burned out the CLM6000 on this?

Nasse · October 23, 2004, 03:00:03 AM

There was a post on this forum about somebody fixed a clm6000 which had a burned led (too much current?) by cutting it half and fixing a new led maybe it can be found by search

Paul Perry (Frostwave) · October 23, 2004, 05:18:37 AM

FWIW I have cut a CLM6000 in half & used the LDR part as replacement in a Morley optical wah, for some reason it worked 'just right', unlike all the other mixed LDRs I had accumulated.

David · October 23, 2004, 06:04:26 AM

Quote from: Paul Perry (Frostwave)FWIW I have cut a CLM6000 in half & used the LDR part as replacement in a Morley optical wah, for some reason it worked 'just right', unlike all the other mixed LDRs I had accumulated.

Maybe I'm stupid or lucky (probably both), but the one time I made a DIY opto, I didn't seem to have a problem. I took a standard 5mm green LED and combined it with the small black LDR from the Rat Shack assortment (hope they continue to stock those for a little while yet!). The tough part was remembering to mark the LED's anode before I patched it into the circuit.

jrc4558 · October 23, 2004, 02:03:41 PM

Take a 5 mm LED. Take some 200 grit sandpaper. Gently sand the rounded top of LED untill you sand all the rounded portion off. That will give you a cylider with a flat top, dia ~4.3mm.
Now you can either glue it on the top of the LDR, but do it right the first time. Superglue doesn't forgive mistakes.

Or you can simply drag some 5mm shrink tube over the LED and put the LDR in from the other side and after heating they will stay together till death do them part.
HOpe it helps.

gorohon · October 24, 2004, 02:51:30 PM

I've cut open a CLM6000 before, and a while back, started a thread about it. All of your suggestions also help. I just don't know why this compressor would work fine for a month, then after turning up the output, fry the opto-isolater? I guess I'll have to do a check with the old DMM and see what's what. Is there something in the design that does not make certain levels of output safe for some components, or, more likely, I've got a bug somewhere that I've got to chase down? I was digging the whole boost thing with the output on this thing, and when it was cranked, it was almost like overdrive. Then when I returned to using it as a compressor, no compression. Is such gain, as I described above, normal for this project?

jrc4558 · November 09, 2004, 11:38:16 PM

Ok, here's the report.

It's amazing. There's barely any coloration tothe sound. Input stands hot humbuckers without any distortion (bipolar 9 volt supply, man. double the headroom) I used a backwards wired 100k audio taper pot, giver better control over degree of sustain effect.

highly recommended. Use 741 opamp instead of LM301, that way you can do it without the 10 pf cap between pin1 and pin8.

Good luck to all.

David · November 10, 2004, 05:57:32 AM

ALL RIGHT!

OK, now tell us how you implemented the bipolar power supply...

This is cool, Constantin! Thanks for the analysis! Hey, maybe you should file it with the actual build reports.

jrc4558 · November 10, 2004, 09:20:47 AM

Maybe I would, son. Maybe I would (c) Homer Simpson

Ok. Use two stereo jacks for in and out. Get two battery clips. Wire one clip to jack one w black side. Wire other clip to jack two with it's red side. Connect the free ends of the battery clips to respective (+) and (-) pads on the PCB.
When jacks are niserted, there's a connection to groun of both the positive and the negative part of the DC supply.

Hope this helps.

PS.

It really is an EXCELENT clean booster when compression is at minimum.
Just think of it: how many commercially made BOOSTERS are using a bipolar 9 volt power supply? That's DOUBLE the headroom :idea: of a single 9 votl battery operated unit. Which means approximately DOUBLE the potential voltage gain from the cirquit! Which ultimately means more and more overdrive from the front of an amplifier. :twisted:

News:

Craig Anderton Compressor