EHX - Octave Multiplexer  [schematic]

All about modern commercial stompbox circuits from Electro Harmonix over MXR, Boss and Ibanez into the nineties.
User avatar
Bernardduur
Transistor Tuner
Information
Posts: 1217
Joined: 24 Aug 2007, 07:36
my favorite amplifier: Welagen - ODR
Location: Losser.........
Has thanked: 94 times
Been thanked: 732 times

Post by Bernardduur »

daemons wrote:Nice one Bernardduur. I guess I should finish my trace of the one I started. It uses 1x 324 and 2x 4558

Some of the values are different, and some connections aren't exactly the same. It'll be interesting to compare.

Daemons
Nice! Please do!
'No more....... loud music.......'
Follow my love for pedals and amps on https://bernardduur.blogspot.com and https://www.instagram.com/bernardduur1

User avatar
daemons
Information
Posts: 11
Joined: 23 Apr 2015, 18:39
Completed builds: Black Russian Big Muff Pi, Human Gear Animato, ...
Location: Canada
Has thanked: 11 times
Been thanked: 3 times

Post by daemons »

johnk wrote:I built one about a year ago (I etched a PCB and the enclosure for it) and I love it. I found that it could use a bit more output so I added a Jfet boost to the output with a volume control:
Looks great John, as usual. I'm jelly of your work.. always clean and professional. I think this is the nicest looking clone of this pedal I've seen.
I think I'll make one just like yours, etched and black paint. Hopefully mine comes out as nice as this.

Daemons

User avatar
Andy2No
Breadboard Brother
Information
Posts: 82
Joined: 23 Sep 2015, 16:33
Has thanked: 154 times
Been thanked: 7 times

Post by Andy2No »

Here are the guts shots of my XO(?) sized EHX Octave Multiplexer, ending with one where I've added a TL071 unity gain buffer / voltage follower, on a small piece of stripboard. There's a wire link you can't see, under the IC socket, connecting the output back to the inverting input.

The schematic for the buffer, sketched on the bit of paper, is shown in the same orientation as the actual buffer.
P1010281.JPG
P1010282.JPG
P1010284.JPG
P1010288.JPG
TL071 output buffer fitted (unity gain input follower)
TL071 output buffer fitted (unity gain input follower)
The pictures are a bit blurry, sorry. I don't really have the camera for it, or the hands, or the patience. It gives you some idea though - mostly SMD mini chips, and one man sized transistor, so a bit of a sod to mod, but not impossible.

I decided it had been in bits long enough, and I still don't feel like doing any more ambitious mods, so I settled for just fitting an output buffer, which I always felt it kind of needed.

Since the output is normally taken from the Blend pot wiper (which is 100k), the output impedance depends on the position of that pot and can be quite high - at a rough guess, say 50kOhms or more, in the middle position. That's a lot higher than a guitar coil, for example, so it could sound very different depending on what you fed the output into.

Now, it has a constant and very low output impedance, as it always should have had.

Too many of EHXs older designs have outputs like that (e.g. tapped off a 100k Volume pot or in this case, Blend pot) and I don't approve :horsey:

Before the output decoupling cap, the output signal comes off the wiper of the blend pot, so I cut the track there. There was a bit of empty space next to it, which helped with soldering on the output wire for the buffer. The buffer input wire goes to the wiper pin of the Blend pot.

I used a TL071 as a unity gain buffer with no extra components - I figured the output level was already biased, so there was probably no need.

I mounted the chip on veroboard / stripboard, with some UHU White Tack (poster putty) and a layer of thick paper to make the pins less likely to poke through. There was quite a bit of height available, so I used a socket. I wired the +Vcc pin to the +9V battery lead pad on the PCB (which is always connected), and scraped a bit of the blue coating on the ground plane to reveal a patch of copper to solder the -Vcc wire to.

Word of warning - when I first touched up the Blend pot wiper pin with the 60/40 solder I was using, it produced a whitish bloom as it cooled. I removed most of the original solder with a bit of braid, so I could put mine on cleanly.

I had plans to take the square wave signal from somewhere - possibly the base of Q1, which is the transistor you can see in the middle of the largely blank side of the PCB, and either mix it to the output or send it to the Dry Out socket so I could mix it externally. I don't feel like drilling a hole for another pot though, at the moment, and I find the Dry Output is useful for connecting to my iPad (via an iRig), as an always on tuner. So, for now, the output buffer is all it's getting.

I find it works quite well with my semi home made mini bass guitar - technically a bass ukulele, because of the very short scale length, but with real round wound bass strings, slack tuned in standard bass tuning. It tracks better than when I originally tried it, with a guitar.

Originally, I think I was trying to get a clean octave down signal from it, but that just isn't practical. It works best when it's just adding a bit of low end thickness - with the Sub switch on, to clean up the octave down signal a bit, and the blend knob favouring the original signal. That has nothing to do with me adding an output buffer, just my take on how to get some use out of one.

User avatar
ege_ergul
Information
Posts: 2
Joined: 11 Oct 2016, 11:27

Post by ege_ergul »

johnk wrote:I built one about a year ago (I etched a PCB and the enclosure for it) and I love it. I found that it could use a bit more output so I added a Jfet boost to the output with a volume control:

[ Image ]

[ Image ]
Wow, great work. I just find its ready to print pcb file , but i didn't measure it despite my numerous printing attempts. Could you explain how did you measure it or do you have exact dimensions of this pcb to build as same as your pedal. thank you.

User avatar
Ichabod_Crane
Resistor Ronker
Information
Posts: 354
Joined: 26 Apr 2016, 15:11
Has thanked: 82 times
Been thanked: 36 times

Post by Ichabod_Crane »

Hello. I need a serious help.
I built the modern version with JFET boost adn the two 4558s of the Multiplexer Octave.
Here's the layout I used:
Image

Schematic is in this post: https://www.freestompboxes.org/viewtopic ... 39#p255931

There's a big bump turning on and off the bypass switch, the effect seems ok enough, and the volume, thought the boost the volume is about the unit or slightly lower even with the volume is set at max.

Voltages

JFET booster 2N5457:
G - 0.00
S - 0.06
D - 5.19 (with the trimmer counterclockwise) - 6.64 (with the trimmer at half)

2N5457 (replace the 2N5459)
D - 3.98
S - 4.20
G - 0.00-2.20

LM324
1. 4.15
2. 4.15
3. 3.81
4. 8.31
5. 3.81
6. 4.25
7. 4.16
8. 4.15
9. 4.16
10. 3.52
11. 0
12. 3.70
13. 4.15
14. 4.14

4558 (upper one)
1. 6.98
2. 3.54
3. 4.11
4. 0
5. 4.12
6. 4.15
7. 1.60-1.25**
8. 8.31

4558 (down one)
1. 1.60-2.35
2. 4.15
3. 4.11
4. 0
5. 3.54
6. 4.11
7. 2.14-7.65
8. 8.31

4013
1. 8.31-4.10
2. 0.00-4.45
3. 0
4. 1.61-2.47
5. 0
6. 7.65-2.07
7. 0
8. 0
9. 8.28-4.07
10. 0
11. 0.00-4.29
12. 8.28-4.10
13. 8.28-4.10
14. 8.31

I checked the value of the componets, orientation, solder side and wires's contact. Everything seems ok.

Thanks for helping!

User avatar
Ichabod_Crane
Resistor Ronker
Information
Posts: 354
Joined: 26 Apr 2016, 15:11
Has thanked: 82 times
Been thanked: 36 times

Post by Ichabod_Crane »

*When there are two voltages = no hitting the strings - hitting the strings.

User avatar
PinzinKinzin
Information
Posts: 1
Joined: 22 Jun 2016, 15:10
Been thanked: 1 time

Post by PinzinKinzin »

The schem has two switches on it. "range" and "bass". Also next to the blend pot it says "Toe Down" I believe there is a version of the multiplexer housed in a wah enclosure. Could this be the schem for that version? Is there any difference?

User avatar
Ichabod_Crane
Resistor Ronker
Information
Posts: 354
Joined: 26 Apr 2016, 15:11
Has thanked: 82 times
Been thanked: 36 times

Post by Ichabod_Crane »

Can anybody help me with my Multiplexer Octave's trouble?

User avatar
Andy2No
Breadboard Brother
Information
Posts: 82
Joined: 23 Sep 2015, 16:33
Has thanked: 154 times
Been thanked: 7 times

Post by Andy2No »

Ichabod_Crane wrote:Hello. I need a serious help.
I built the modern version with JFET boost adn the two 4558s of the Multiplexer Octave.
Here's the layout I used:
[ Image ]

Schematic is in this post: https://www.freestompboxes.org/viewtopic ... 39#p255931

There's a big bump turning on and off the bypass switch, the effect seems ok enough, and the volume, thought the boost the volume is about the unit or slightly lower even with the volume is set at max.
I don't see a booster, in the schematic you linked to. I can't make much out from a layout.

I guess there is a jump in levels between the two switch positions. There might be a DC level offset too, due to charge on a couple of the capacitors. Does the position of the 10k P2 pot make a difference? It looks like it might be worse when it's set to zero ohms, connecting one leg of C13 to Vref.

Personally, I only found one of the switch positions useful, so I left it in that position (sub on). I guess if it works as expected, and the only problem is a pop when you move the sub switch, I'd just settle for only changing the switch while the effect is turned off.

User avatar
Ichabod_Crane
Resistor Ronker
Information
Posts: 354
Joined: 26 Apr 2016, 15:11
Has thanked: 82 times
Been thanked: 36 times

Post by Ichabod_Crane »

Thanks for the reply. :)

The booster is positioned in the end of the original schematic and it's been added to have a volume control and an eventually boost of the volume. The booster is a simple JFET style booster, just a few of parts in the left-bottom corner of the board, I'm sure you can find it.

I got the pop with the bypass switch, pretty loud. So I have absolutely fix it. No issues with the Sub switch that I found useful to change the effect. For the frustrating is a bit that I don't test the circuit, but I remember that everything works fine, except the pop and the overall volume, as if the booster doesn't work properly. I have to keep the volume always set at max, I'm pretty sure there's something wrong.

User avatar
Andy2No
Breadboard Brother
Information
Posts: 82
Joined: 23 Sep 2015, 16:33
Has thanked: 154 times
Been thanked: 7 times

Post by Andy2No »

I see you did say the bypass switch. I was looking at the sub switch.

I really work much better with schematics, but it looks to me like the drain of your JFET is connected directly to the output jack. If so, there's probably a big jump in DC level at that point, when you switch it on.

The first thing I'd try is decoupling the output. Pedals are not normally DC coupled, at either end. The size of capacitor you need depends on the impedance - your output impedance plus a guesstimate of the input impedance of the thing you will want to drive with it. As a worst case you could assume that's very low, and just use your output impedance.

Generally speaking, 10uF will do it, so maybe just try that. If there's loss of bass, you need a bigger one.

I made a similar JFET buffer, just to see what they do. I've tried is as an input buffer, direct from a guitar. I took the schematic from the input stage of the Ruby amp - with no level / gain pot, just fixed. The gain is a bit under one. Using the values on that schematic, it lost a lot of bass, so I made the capacitors quite a lot bigger.

A JFET buffer is really just for impedance matching, so a gain of less than one is normal, I think. If you manage to get a gain of more than one, you may have made something other than just a buffer - e.g. a filter, or a clipper, of some sort. My first attempt could be described as a high pass filter.

The simplest clean output buffer is probably the unity gain input follower op amp circuit, like I used. I might not have made it clear, but that's not DC coupled to the output jack either - I placed it between the Blend pot and the output decoupling capacitor.

If you want gain control, you could use a non-inverting op amp gain stage - which has a minimum gain of one. If you want gain to be able to be less than one, you'll need an inverting op amp stage, ideally followed by another one, to change the polarity back - though that only really matters if you want to mix it with the original signal. If not, either way up will do.

If this still doesn't help, or I've misunderstood what you did, I really would need to see a schematic. It doesn't need to be of the whole thing, just your additions. with the connections to jack sockets and the bypass switch. Hand drawn and photographed is fine, just so long as it's readable.

User avatar
Ichabod_Crane
Resistor Ronker
Information
Posts: 354
Joined: 26 Apr 2016, 15:11
Has thanked: 82 times
Been thanked: 36 times

Post by Ichabod_Crane »

This is the JFET booster schematic:

Image

In the veroboard layout R2 is a 50k trimpot.
Anyway you can read even this: https://www.freestompboxes.org/memberlis ... le&u=16501 probably you see it. The purpose of the JEFT thing is push the output volume, in case. The layout I built seems based on the PCB layout that Johnk used.

Schematic of the original Multiplexer Octave in this thread posted early not by me. Some part could be slightly different.
Open it: https://www.freestompboxes.org/download/ ... &mode=view

The layout come from Guitar FX, generally those users work fine, I could be wrong, but though the layout should be verified I guess this time there's some confusion around this project. There's a vintage version with just quad op-amp and no boost; a vintage version with boost; a modern version with two dual op-amp which replace one quad op-amp, and it includes the booster. This is what I did, and the schematic seems to me a bit different.
The vintage version should be verified, too, but if I can fix that I did is better.

User avatar
Andy2No
Breadboard Brother
Information
Posts: 82
Joined: 23 Sep 2015, 16:33
Has thanked: 154 times
Been thanked: 7 times

Post by Andy2No »

Well, looking at the JFET buffer schematic, I can see it's not DC coupled, if that's how you built it, and I can't see why you'd get the thump / pop, just from that.

What I would suggest is that it's not going to give much gain, if any, and that it's not a linear response buffer - basically, a filter. The capacitor across the source resistor sets more gain for lower frequencies. The overly small input and output capacitors limit lower frequencies. It's not good, if what you want is a linear boost. Arguably, it's just not good at all. It might be okay as the input buffer of a pedal which has a reasonably high input impedance, but it still won't be linear.

I'd suggest increasing C1 and C2, by a factor of at least ten, and maybe even remove C3.

A good way to test for frequency response is to play a white noise sample through it, and look at the spectrum. There are good free Android apps you can use, and probably some not so free iOS ones, or there's lots of free software for Windows. It's often surprising, what you see, and also surprisingly easy to fix.

The main purpose of a JFET buffer is not to add gain. It's to take a signal with a high input impedance, and put out a signal with a much lower output impedance. If you're using it to drive an amp with a very high input impedance, like a valve amp, you may not notice much difference. If you're using it to drive a solid state amp, you might notice a lot of difference, which is mostly in frequency response (more linear, ideally), but you may also perceive a volume difference.

Here's a thought - what happens if you bypass your output buffer? Does it still pop/thump? It looks like you could do that fairly easily, by clipping a wire link or two. The XO design, which is the current production design, AFAIK, still doesn't have an output buffer. I added one to mine because I use solid state amps. It makes it work "better" (mostly for frequency response), but it still worked before I added it.

Another option is an external buffer, e.g. almost any pedal capable of giving a reasonably clean signal, other than just a "true bypass" one. Candidates include boosters, graphic equalisers, and so on, or a Boss style pedal where the bypass is buffered - just use the bypass. I'm not suggesting that has to be your final solution, just a diagnostic aid, to see if it's the buffer that's the problem.

User avatar
Ichabod_Crane
Resistor Ronker
Information
Posts: 354
Joined: 26 Apr 2016, 15:11
Has thanked: 82 times
Been thanked: 36 times

Post by Ichabod_Crane »

I tried months ago to bypass the JFET Booster (escuse me, I keep calling it BOOSTER, because so it's been called), and I tried to use just the booster. I have to repeat the test, but I remember it wasn't so good.

Now, I'm lloking the layout.
To exclude the booster I have to desolder the Vol 2 wire frome the switch, and the Blend 2 t will be the new output of the circuit, that will be go on the switch: SW 1-3.
To use just the booster I have to disconnect the SW 1-2 from that 100nf, the input of the Multiplexer Octave, and it will go on the Blend 2 line, the input of the Booster.

I had done it using some flying connection, with the clip alligator, but maybe something goes wrong.

It's really frustrating! THe circuit seems really perfect, maybe it's something with the wiring. I used this model, declared corrected, the board is another version, without the booster (it includes a mistake, now corrected, it doesn't care), but the wiring it's the same:

Image

PS the input sleeve is not connected at the ground, but I did it, of course.

I think I don't need to modify the schematic because it results verified, but maybe the modern version has some mistake, or there's some mistake I did but I don't know.
The effect works, the booster seems no good, and it could cause the pop. But I can't be sure of this.

User avatar
Andy2No
Breadboard Brother
Information
Posts: 82
Joined: 23 Sep 2015, 16:33
Has thanked: 154 times
Been thanked: 7 times

Post by Andy2No »

I'd start by just clipping the wire from the original circuit to the input of the booster, and routing it to the output jack instead. Bear in mind that there is a capacitor between the output of the Blend pot and the jack socket. That's important.

Try it like that and see if it still pops - if you have already, I've misunderstood.

I was calling it a buffer because what you have there is a fairly standard buffer circuit, that someone's messed with. It may have been tweaked to try to give it more gain, but a good buffer isn't designed to give gain - a gain of one or less is normal. As you described, that's what you were getting, so what you seem to have is a buffer followed by a pot, rather than a booster. The pot undoes a lot of the good work done by the buffer, assuming the buffer works well, which I doubt.

I don't mean to be critical for the sake of it, but those capacitors seem too small, especially on the output, and are going to rob it of bass. The capacitor across the source resistor (in parallel with it) seems to be an attempt to replace the bass robbed by using overly small input and output capacitors. It just looks like a half-assed bodge to me. I realise you didn't design it, and I don't mean this as criticism levelled against you.

If you want a booster, take a look at the MXR MicroAmp schematic. That's how a good, flat frequency response, booster is made. You could also just look at any article on op amp circuits and take the basic non-inverting op amp circuit (which gives a gain of one or more, but never less). It would work well enough, and better than what you have there. It only takes a few parts.

That sort of buffer design was originally meant to take a signal from a single coil guitar pickup, and produce a nice lowish output impedance signal, which could drive almost anything reasonably well - e.g. a solid state practice amp, with an op amp based preamp, using an inverting amp circuit as the input stage. My Peavey Backstage 50 fits that description. It's a good preamp, but it has quite low input impedance. A well designed buffer helps, with something like that.

But, the point of a buffer is low output impedance. If you then put a 100k pot across the output, and tap your output off that, you're back to high output impedance. High output impedance into higher output impedance is fine. Presumably, that's what EHX assumed when they made a lot of these designs, originally. High output impedance into low input impedance is not fine, because it messes up the frequency response, robbing it of bass.

As you'll be aware, almost all guitar pedals use capacitors to decouple DC at the inputs and outputs. That effectively makes the input and the output a high pass filter. The parts of the signal below a certain frequency are "rolled off", meaning the amplitude decreases logarithmically as the frequency decreases, typically at 6dB per octave. The frequency where that starts to happen noticeably (down by 3dB) is determined by the input or output impedance and capacitance. That includes the output impedance of whatever's supplying the signal to the pedal, and the input impedance of whatever it's driving - it's part of the circuit.

The formula is fairly simple - F = 1 / (2 * pi * R * C)

- where C is your actual physical capacitor, plus a tiny bit extra for stray capacitance, which isn't normally important. R is the effective impedance.

Using it to calculate a frequency isn't always needed - it just helps you see what's likely to happen. If you increase R or C, the roll-off frequency goes lower - that's what we want. We want it as low as we can reasonably get it. There are usually reasons why R can't be increased too much (especially on the output), but C can be as big as you can afford, or physically fit in the case. Ideally, you'd only make it as big as you need, but if you find your frequency response isn't flat, and is lacking in bass, you make it bigger.

As I see it, the best use for a JFET buffer is on the input of a pedal, not the output. It can give you an input impedance of more than 1M Ohm, which is great, and it can give you some nice soft clipping. Putting it on the output is less useful, because you're not going to get the really low output impedance you'd like, and you're probably not going to get the benefit of the nice clipping sound, so easily. Using it with a fairly high value output pot is just pointless - you're back to high output impedance, when you turn the volume down, so you start to lose bass again.

If you like the idea of JFETs, you still have quite a few options for op amps with JFET inputs. The input of the op amp is the place to have one too - that's where it does the most good in terms of giving a nice high input impedance for that part of the circuit, and a nice soft clipping sound.

Op amps shouldn't be thought of as characterless or boring. They're really well designed things, make great pedal and preamp stages, and some of them have a really nice character, mostly to do with how they clip. Some don't respond well to clipping, so sticking to one's other people have used in pedals is not a bad idea. If you put them in sockets, you can always experiment with different ones, later - they're largely pin compatible.

User avatar
Andy2No
Breadboard Brother
Information
Posts: 82
Joined: 23 Sep 2015, 16:33
Has thanked: 154 times
Been thanked: 7 times

Post by Andy2No »

If it doesn't sound good, without the booster/buffer, put something else between it and the amp that has an active buffer - at least, just while you test it.

People go to great lengths to make things "true bypass", but quite often the bypass buffers in pedals that have them, work quite well. Mostly, because it's not that hard to design one that does. Of course, if you have a long chain of pedals, all rolling off the bass just a little, or colouring the sound just a little, due to a not quite flat frequency response, it starts to add up

IMHO, the EHX Octave Multiplexer doesn't sound good into a solid state amp, or at least not any I've tried it with. As I see it, it's because of their annoying habit of tapping the output off a 100k pot, without anything after it to lower the output impedance.

That's why I added the simple input follower / voltage follower op amp buffer, just before the output capacitor. In terms of number of components added, I don't think you can get better than that - it was one :)

User avatar
Andy2No
Breadboard Brother
Information
Posts: 82
Joined: 23 Sep 2015, 16:33
Has thanked: 154 times
Been thanked: 7 times

Post by Andy2No »

Despite wittering on at some length, I'm still over simplifying why it's important to have a low output impedance. From what I said before, you might be thinking more total impedance will mean better bass... and it might, though it will mean lower volume, at the very least.

More to the point, the input impedance of whatever you're driving won't be constant - it will depend on the frequency. You can easily end with a very bumpy / humpy / notchy over all frequency response, as a result. The way to minimise that is with low output impedance, so the output voltage of your pedal is affected as little as possible by the load that it's driving changing with frequency.

It's probably best to just google output impedance and frequency response, if you want to learn more. Here's a good start:

https://www.soundonsound.com/techniques ... -impedance

User avatar
JimSalabim
Breadboard Brother
Information
Posts: 53
Joined: 02 Jun 2015, 23:00
Has thanked: 7 times

Post by JimSalabim »

Hi! I have just finished my Octave Multiplexer, but something is not right. The tracking doesn’t work well at all and the controls act strange: Crackling and sudden volume drop/boost when the Blend and Bass filter controls are set to maximum and when the High filter control is set to minimum in one switch position.
See video:
https://www.dropbox.com/s/p0jf5abrtqymn ... r.m4v?dl=0

The knobs in the video are from left to right: High Filter, Blend, Bass Filter

Could it be that I wired the DPDT switch in the wrong combination? Or does this sound right?
I have used this PCB layout:
1458411992_plata_octave_multiplexer_v3_lay.jpg
The DPDT switch is connected as follows:
1: Pad D
2: SW BASS FILTER Pad (at Blend pot)
3 Pad C
4: Pad I
5/6: Pad H

Pad B is not connected, I didn’t know what it could be there for.

Any ideas?

User avatar
JimSalabim
Breadboard Brother
Information
Posts: 53
Joined: 02 Jun 2015, 23:00
Has thanked: 7 times

Post by JimSalabim »

I also have a strange beep noise that will not stop in one small range of the Blend ad Bass Filter pots. I cannot find any build errors. Is there something wrong with the layout?
https://www.dropbox.com/s/zfd36tqxvyx2m ... e.mp3?dl=0

User avatar
Andy2No
Breadboard Brother
Information
Posts: 82
Joined: 23 Sep 2015, 16:33
Has thanked: 154 times
Been thanked: 7 times

Post by Andy2No »

JimSalabim wrote:I also have a strange beep noise that will not stop in one small range of the Blend ad Bass Filter pots. I cannot find any build errors. Is there something wrong with the layout?
https://www.dropbox.com/s/zfd36tqxvyx2m ... e.mp3?dl=0
A beep noise from a filter is self oscillation - the output of the filter is feeding back to the input, somehow, and the Q of the ilter is high enough to make it only happen around one spot frequency - whether the Q should be that high; I don't know.

I can make a Behringer ADI pedal self oscillate, just by connecting a cable from the input to the output and adjusting the knobs (using a splitter so I can hear the output)... It's quite a fun thing to do.

Could you post pictures? Also, if you want to show videos, it would be better if they were on Youtube or Vimeo. Personally, I'm not inclined to download a video. Youtube lets you have them unlisted, if that's an issue - you can still provide a link.

Post Reply