Marshall - ED-1 Compressor [schematic]

[Tonetweaker]

Thanx !

Would be great if you could start a retracing thread for the MXR Sustain and post good photos of both sides of the pcb.

Looks like I had beaten you to the punch when you posted this...

I will definitely crack it open again this weekend though to get the values.

Incidentally, the pedal definitely works as advertised... It added some obvious sustain with both my Strat and my Epi LP, but really didn't color the sound at all or add much in the way of noise, even when I turned it up a bit. I was pleasantly surprised.

No need to apoligize I'll download it twice to make up for the good work you did.

Thanks! I'm actually an admin of a small music forum... so I'm used to being able to edit anything I damn well please.

Bin playin around with the *EH Black finger* but it's a split supply,, Grrr.
Not to worry all good practice for my ever expanding electronic skills.

I've at least made headway with the Bfinger as it's been tweaked down to a very low noise level and I'm happy with it,,
But I will certainly take these schematics and see how they all stack up.

For me it's not about making the worlds greatest Comp circuit I simply want more sustain without the dreaded hiss that comes with most of them pedal types.

Having purchased a stock ED (Marshall Comp) and being able to improve it's usrless noise issue by just changing 2 things then I feel there is hope for a better pedal sustain unit.
So if the worlds longest sustain comes with 10 times noise then I'll give up but I have a hunch that there is a circuit out there waiting for me.

Grin,, Cheers Phil.

Occasionally, I start looking at a circuit and think, "No way do I want to screw with this." Those are usually the circuits that I end up learning the most from. Of course, in my case, it's usually because I tinker with it... break it badly... and then have to figure out how the hell to bring it back from the dead.

Comp/sustain circuits are the type of thing I don't really understand as well as I'd like... I guess that's because when you've got a good circuit that does what it should, you just don't think about it much. It's not as "glamorous" an effect, but I certainly notice it when it's gone in some situations.

It's very cool that you were able to soup up your Marshall to rein in the noise. I'll bet with a little further tinkering, before long you'll really have what you're after. That MXR might actually go a long way towards helping as well... which serves as motivation for me to try to get you guys some better info on the inner workings.


Cheers...
Steve

[phatt]

After much cleaning of shed,,, I found this,, see pic.
God knows what it's part of but it is quite obviously the compressor part of an Amp.

Anyone happen to know a Name for this thing?
Again it's a split supply

Meanwhile I've actually built a discrete CA3094 and although the compressor I built around it did not work very well at least I got a signal to pass.

Now to move onto a discrete LM3080.
Jeez this breadboard testing can wear you down and the time really starts to add up.

It would seem that these circuits are very touchy as the control voltage is very critcal with most of them.
i.e. just pressing a resistor that feeds pin5 of an LM3080 can induce more gain.
I could actually control the gain of one circuit this way just by the resistance pressure of my finger.

Without a scope it's hard to know the inner secrets but obviously the control voltage needs to be working DEAD Right.
If not the release can be jumpy or intermitant or just plain lumpy volume jumps as it raises the gain back up. Yuk.

Anyway I've canned it all for a few days as this W/end is Muso Club metting so better go practise some songs.
Cheers,, Phil.

[audiohub]

This is from the Gibson (Norlin) Lab series amplifiers, built in the early 80's. This circuit was used throughout several models, which varied mostly in the speaker configuration (and maybe output wattage). Nice amps (I have a preamp board from the L5 in storage somewhere that I used to use to feed into a channel of my keyboard mixer when I wanted to play guitar...sounded pretty good!) and I believe that the Pearce amp which came a bit later was a refined version of this series. I always wondered if Dan Pearce was one of the original Lab series designers and went off on his own at some point.
The Lab amp was the secret weapon of Ty Tabor from King's X who apparently used to house them in Marshall heads to keep them a secret.

[phatt]

This is from the Gibson (Norlin) Lab series amplifiers, built in the early 80's. This circuit was used throughout several models, which varied mostly in the speaker configuration (and maybe output wattage). Nice amps (I have a preamp board from the L5 in storage somewhere that I used to use to feed into a channel of my keyboard mixer when I wanted to play guitar...sounded pretty good!) and I believe that the Pearce amp which came a bit later was a refined version of this series. I always wondered if Dan Pearce was one of the original Lab series designers and went off on his own at some point.
The Lab amp was the secret weapon of Ty Tabor from King's X who apparently used to house them in Marshall heads to keep them a secret.

Great work,, Thanks heaps.
Could this also be the *MOOG* Lab Amp? BB King used one I'm told.
When time permits I just might be crazy enough to test this circuit and see if it's an improvement on some of these pedal compressors.
My Marshall ED is still coming out wrong in live use, it's insanely hard to find a balance between the bypass sound and when it's engaged it's ever so touchy with the dreaded hi gain hiss.
Which is not much better than the old Boss CS3 I once owned.
Phil.

[mictester]

My Marshall ED is still coming out wrong in live use, it's insanely hard to find a balance between the bypass sound and when it's engaged it's ever so touchy with the dreaded hi gain hiss.
Which is not much better than the old Boss CS3 I once owned.
Phil.

The problems are the same as with all these types of compressor.

The first two constraints are that everyone seems to want their pedal to work from a 9 Volt battery, and that battery has got to last for weeks. So - we've got little headroom and little current to play with.

The next problem is the transconductance amplifier IC. The 3080 wasn't ever intended to pass audio, so it's a really poor choice. It's distorted and noisy. It also doesn't really work very well at lowered supply voltages. Also, the whole of the signal goes through it the whole time the effect is engaged.

Next we have the rectifier that derives a varying DC control voltage from the instantaneous level coming out of the 3080. The recifier will inevitably suffer from "overshoot" so either the gain of the 3080 will fall to almost nothing as the note arrives cutting a hole in the sound, or the "designer" will have slowed the attack response down to mitigate this and ended with little or no real compression. Also the range of gain control is frequently much too large, leading to ever increasing noise as notes decay.

The phase splitter in the "traditional" Dynacomp circuit also provided a low impedance output signal from the emitter of the transistor. Large signals would be clipped (in one direction) by the lower rectifier diode, and give rise to rather nasty distortion on peaks. My "Improved Dynacomp" elsewhere on this forum took the audio output straight from the 3080 through an additional emitter-follower transistor. This eliminated that nasty source of distortion. The values around the retifier were also tweaked, and there was a measure of pre- and de- emphasis added to mitigate the worst of the noise. It still isn't a very good design!

If you look at the Morley compressors, you'll see a different "gain cell" topology. They've used a good quality op-amp to pass the audio, and then varied its gain by varying the negative feedback around it with a transconductance amplifier. The range of control is reduced (it's about 18 dB max in this type of circuit), but it's enough for very effective guitar compression (which is, after all, what we're after!). You have to remember that most of the audio is passed by the good quality op-amp. The distortion (and noise) is only in the feedback loop, so is much less critical. Since they've got a quad package of good quality op-amps, they've implemented a good recifier circuit which should overshoot rather less.

Morley still get away with a 9V supply, but the performance can be radically improved if you go for a 15V supply.

My current compressor uses two gain cells cascaded, controlled by two rectifier circuits. The whole board has a 4066 for signal routing, a pair of TL 074s and an LM13700. It sounds great, and has huge amounts of sustain.

If the silliness is over on this forum, I'll put it up here later this week.

[roseblood11]

Thanx for that explanation mictester!

There´s another thread about the MXR Sustain pedal now, could you please have a look at that? Help is needed to identify the transistors...
https://www.freestompboxes.org/viewtopic.php?f=1&t=12177

[phatt]

Hi mictester,
I really appreciate the time You've put into this.
All up I'm getting sick of of comp,noisermaker circuits so I might go back to my original thought which was to shoot for a more touch responsive preamp where I have a better chance of success at keeping noise **DOWN**

I'll soon post my testing of that under your *RSO* schematic.
Phil.

[phatt]

Just an update,
I had the chance to try out a brand new Boss CS3 compressor and it is very quite.
My Marshall Ed of course has been modified and *Now* has about the same noise floor as the CS3 but a side by side shootout reveals the CS3 certainly has the longer sustain.

I did hear a very slight glitch in level as the gain came back up but probably good enough for most players.

Only real gotcha on the CS3 is the tone control.
Turn past middle for more treble and Ouch the Dreaded Hiss comes back hard but it's unlikely you would need to use treble boosting.
(that of course is on my setup which is well balanced for treble but other setups might be different)

At least it shows me it is possible to get a half decent Comp to run from 9Volts or one that gives decent sustain without massive hiss.
I have no doubt that rack studio gear would be better but I might invest in the Boss and put my bread board to better use,,we will see. Phil.

[phatt]

Guess no one had an answer to this old thread???

I recently picked up a CP10 dirt cheap on the bay, and in some ways like the sound. BUT, I realized quickly why I have always quit using any compressor. They all seem to rob the low frequencies out of the signal, leaving a thin sound (at least to me). I am guessing this is because the lower strings/frequencies are a hotter signal than higher frequencies, so the output to the control input (pin 4, ba6110) is a relatively hotter signal causing the compressor to compress lows more than higher frequency signals. I hope this thinking is correct!!!

So, I am wondering if someone can tell me, would it be possible to add a simple highpass filter to ONLY the signal to pin 4, causing the compressor to react more to highs and less to lows??? If this would work, can you suggest the values for a filter??

Or, have I TOTALLY LOST MY aching old brain??

Any help would be much appreciated!

Dave

The Marshall ED 1 Comp has a trick you might like. Look for the *Emphasis* control.
It sets the amount of freq response that drives the control voltage. (it's only subtile but works well)

I own a ED 1 and it's OK once I modded out the dreaded hiss but a side by side test with a Boss CS3,, there was no contest the CS3 is far better. (much longer sustain)
They seem to have a whole lot of issues nailed down with that one.
Rather than messing about with bread board circuits I might just buy a Boss CS3.
(but watch the tone knob on CS3 as it's a direct copy of that used in the TS9,, high treble settings will bring back the much hated hiss)
Phil.

[adammartinmusic]

Here it is:

"Tonesucker" wrote that he made some changes after he took the photo: a) Changed the wiring to shunt the input to ground in bypass mode, b) added a 2,2M pulldown before the input cap.

Hi there, I'm trying to mod my ed-1 to true bypass, it looks like I could copy the wiring in the picture shown by roseblood11 but how do i shunt the input to ground in bypass mode? and where do i place to the 2.2m pulldown resistor?

I apologise as I'm a beginner when it comes to mods!

Thanks!

Adam

[roseblood11]

Try to understand this wiring scheme:
http://www.aronnelson.com/gallery/main. ... ewsIndex=1

[vastman44]

Hello all,

I'm definitely new to all of this and can't read schematics to save my life.

I have an ED-1, but the compression knob broke off. I was wondering if anyone could tell me what kind of potentiometers are used for this pedal. I've searched ebay and google and have so far come up with nothing.

Thanks for any help.

[roseblood11]

The schematic on page 1 says that it's a 1M reverse audio taper ("C1M"). If you can't find that, you could use a A1M (audio taper) and wire it backwards (swap lugs 1 and 3).
If you'd post a picture of the pot, it would be easier to recommend a pot that fits physically.

[pedalgrinder]

hey roseblood do you have any more guts shots of yours when you did you mod to true bypass? I'd like to see more guts shots if i could cheers

[roseblood11]

Sorry, I sold the pedal

[deltafred]

Lot of filtering caps at the power input... are all really necessary It makes a big difference
Even a cheap chinese wall wart usually got a 1000uF cap

The smaller caps improve the filtering at higher frequencies. But I don´t understand why they use 4 identical caps. I read somewhere that it would give the best result, if each cap was 10 times smaller than the preceding one, say 100µF, 10µF, 1µF,...

Good design practice that is not usually seen in pedals.

A large electrolytic for power supply (low frequency) decoupling.

A smaller ceramic mounted as close as possible to each chip's supply pins for noise suppression. Standard procedure in digital design and worth doing in analogue if noise is to be minimised.

Usually not seen in stompboxes for one or more of the following - ignorance, trying to cram too much into a small box, or just downright cost cutting. (IMO. IME, YMMV etc)

[pedalgrinder]

if it was causing a problem people would do it but filtering doesn't hurt it's a good thing just depends how anal you get.

[deltafred]

if it was causing a problem people would do it but filtering doesn't hurt it's a good thing just depends how anal you get.

Have you ever heard a mobile phone "bleeping" through a guitar amp or PA? To me that is a problem, one caused by at least one part of the signal chain where the "designer" was not being anal about filtering.

It is totally avoidable and just shows a lack of good design practice. You can call it "how anal you get" if you prefer but it has the same effect, unwanted interference.

[pedalgrinder]

in most cases good shielding and filtering to a point is fine if you really want to get down to design theory more isn't necessarily better it what cleans the ripple entirely with what time constant does the exact job. So yeah it how ever anal you want to get.

[deltafred]

in most cases good shielding and filtering to a point is fine if you really want to get down to design theory more isn't necessarily better it what cleans the ripple entirely with what time constant does the exact job. So yeah it how ever anal you want to get.

We're not talking about just ripple here, but lowering the impedance of the supply lines so one chip does not influence or interfere with others. They also help minimize any interference brought into the pedal down the power supply leads.

You are quite correct, more is not necessarily better, if they were all placed together as the drawing implies they are next door to useless. Strategically placed they make a huge difference.

As that is a Marshall pedal it no doubt had to comply with UK interference standards (both interference it produces and susceptibility to external interference). Hence the chokes on the inputs and outputs and the extra capacitors.

I'm not a betting man but I would put money on that one being immune to mobile phone interference.