New Looper Design

Started by Jaicen_solo, October 26, 2005, 02:19:59 PM

Previous topic - Next topic


Having recently acquired some new enthusiasm for this project, I've started work on a rough layout for a new looper based on the ISD2560 chip. This has 60s of record time, and a frequency response of up to 8k, with a 3db low pass at 3kHz.

Layout is HERE:

I've been working from the data sheets here:

Which is where the schematics can be found. So far, i've only laid out the ISD portion of the circuit, no external mixers etc.
At the minute, the layout is very rough and far too spacious but that can all be refined later. I Just wanted to see if anybody was interested in this new design, and could perhaps cooperate to check for errors.
Let me know what you think, all comments welcome.


If I can get hold of the chip I will be glad to help you "refine" the PCB. After you have the full version schem drawn up.

It is about time we had a looper pedal that didnt rely on a NOS chip that is too expensive.

Count me in!

My project site
Winner of Mar 2009 FX-X

The Tone God

I had a looper design awhile back that used the 25xx ICs as I thought the old design was silly to use an out of date IC. Maybe I'll look into porting my mods to this design. We'll see.



Sounds like a plan. I intend to just make it a straight looper for now, no FX loop etc.. Say a looper Lite edition.
Once it's finished maybe you can help us refine a more complicated luxury edition?  ::)
I figure this way, those less confident will be able to build a simple looper with easy to get parts, while the adventurous can work on something a little more challenging. I chose the 2560 because it was for me the easiest to get hold of, both maplin and RS stock it in the uk.


Ok guys, thought I'd share this morning's work with you all.

I've laid out a rough looper using just 3 transistors, a 2560 and a TL074. So far, the mixer design only includes Input gain and blend controls, but that can probably be improved upon with FX loops and tone controls for example.
I think we can probably make this layout fit into a Hammond BB box with a little refinement, but at the minute I'm shooting for a B size box (still not done the power supply).
I'm going to draw up a schem for the mixer etc in a minute. Basically, the first input stage (IC1D) is a gain stage using the 500K pot. This feeds straight into IC1B which buffers the signal and passes it into both the input of the 2560 chip (IC2) and the DRY/WET blend control.
IC1C just buffers the output of the 2560 (because it was spare basically) and passes the output to the other side of the 100K blend pot.
The middle lug of the blend pot is attached to IC1A which buffers the blended output and outputs via a 10uF cap and 1M to ground to the output jack.
All of this has been done off the top of my head, so if anybody spots any corrections let me know. I'd also be interested in eliminating a few components, I went with the kitchen sink approach, so any thoughts??
Another thing to note, the layout is done as if the IC's Are surface mount. It'll need to be mirrored to use as an actual PCB. I just couldn't get my head around all those IC2 pins backwards!  :icon_eek:
Anyway, the artwork is shown below.
NOTE: I've tried to keep the Digital and Analogue grounds seperate. If anybody spots any errors there let me know. (I spotted the Pin28 error, that's now connected to +6v)

Layout Here:

PCB Traces Here:


can the Winbond chips be made to play forward/backward?  Variable speed?  Overdub?  These are features I would be requiring in a looper if I were to make one.  It seems there's something to be desired in their sampling frequency rates as well.

Peter Snowberg

I think it was Z.Vex who posted about modulating the speed of these chips by modulating supply voltage.

Sorry, they cannot play backwards or overdub.
Eschew paradigm obfuscation


Ermm, no, no and.. No. Sorry!

These chips just aren't designed for something like that. The record and playback functions are mutually exclusive, unless you get into daisy chaining multiple devices. That's not really what this project is all about.
The sample rates are not too bad really, once you start putting a little crunch on the amp it's going to sound pretty much ok. The only thing that really suffers is the high frequency response or lack of. That said, because the chip dulls the tone of the recorded sound, it helps it to sit back in the mix much better, which enhances the illusion of two different players.
You're welcome to design a looper that satisfies all your criteria, just don't think of using a winbond chip. That's going to need some memory chips and some complex microcontrollers. Check out Manecloopers, then compare his products and you will see it's not really in the same league. It's like comparing a Ruby amp to a Flextone.
Yeah Pete, I spoke to Z about that a while ago after he posted his stuff. I understand the theory of doing it, but i'm not going to for two reasons.
Firstly, i'm not stepping on any more toes, and secondly I don't want to deal with all the noise problems it's going to create.
I think it would be easier to just design in an FX loop.

The Tone God

From my reading I belive there is another cleaner way to vary the speed of the playback on the fly but it is not for the common DIYer. The power supply trick is kind of a hack but it works and is easier to impliment.

I attacked my looper design from the control angle instead of the audio. My looper splits the memory up into multiple banks for recording. It uses some really basic logic to do it. So you can turn that 60 seconds into four 15 second samples. It will work with any of the 25xx series so you can pick your time range if you wish.

I didn't finish my audio design but I went about that in a different way also.

Jaicen, do you have a schematic of this looper version handy ?



I don't have a complete schematic as of yet, I basically laid out the design from the data sheets on the fly. I will produce one soon, but until then here's what I was working from:

I considered using the addressing logic in the chip myself, but apparently when you use any address other than 0 there's quite a large click at the beginning of playback. Something to do with it taking 25mS to address. Also, it seemed like a headache, so I won't bother!
Re: Playback speed, are you talking about modulating the external clock?? That supposedly causes some nasty audio aliasing if done during playback. Sounds nice to me, but I'm not going to make a hfo for it!

The Tone God

My analog signal section used a small transformer on the speaker output to interface to the rest of the output circuitry. Most designs I have seen at this point spit the output into a opamp or something simliar which I think would account for everyone complaining about sound quality or various other issues.

Quote from: Jaicen_solo on October 27, 2005, 02:14:05 PM
I considered using the addressing logic in the chip myself, but apparently when you use any address other than 0 there's quite a large click at the beginning of playback. Something to do with it taking 25mS to address. Also, it seemed like a headache, so I won't bother!

The click is similar to the click at the end reset of the loop. I guess its a matter of taste. I came up with a quick and dirty way of addressing the logic needed. In the end I used a bunch of resistors, DIP switches / jumpers, two 4013s, a 4066, and two SPST momentary switches for bank select and playback/record. I also had a switch for selecting record / playback modes.

Quote from: Jaicen_solo on October 27, 2005, 02:14:05 PMRe: Playback speed, are you talking about modulating the external clock?? That supposedly causes some nasty audio aliasing if done during playback. Sounds nice to me, but I'm not going to make a hfo for it!

Something in that space. As I said I don't think it is in the common DIYers space. I suspect that varing the power supply causes the clock to vary it's speed due to it's internal design.



Ok, here's the schematic for the mixer and input buffer of the looper. I'd really appreciate it if someone could check over the layout to make sure there are no errors etc.

Edit: Pedal input connects to the 1uF/1M network on the non inverting input of IC1D. I forgot to label it, can't be bothered to fix. Off to bed now.  :icon_eek:


I've uploaded the complete layout and schematic to the link below. It's alos available as a pdf if you PM me. I can't host pdf's on this crappy MSN space.
Anyway, i've now modified it slightly so that it loops at address 0 which simplifies things slightly.
The layout also includes LED indicators for play-record, which I couldn't be arsed to add to the schematic.
Again, I forgot to add the input to the +'ve input on IC1D.
Is anybody excited by this design or is it just me?? Any feedback welcome!!!


kinda small and hard to read.....


edit: Previous layouts contained stupid errors, which hopefully have now been corrected.
Contact me via a PM for the new layouts, i'm not hosting them online anymore.


you konw not to step on any toes here but if you do a search you should come up with a long thread about the loop junky in which a few people mentioned the same chips that zack said he uses.  also there is a mention that i am reiterating now.  for expiraments theres a kit for under 25usd from circuit specialists that already has all the basic functions in it. and it gives 40seconds sample and loop.  as well as its already on a pcb board.  i haven't looked in about three months but when i played with one i was quite impressed took about an hour of wiring headaches to get the thing to loop andrecord with two TPdT switches.

In a corner of the churchyard, Where the myrtle boughs entwine, Grow the roses in their poses, Fertilized by Clementine.


Last week I found another of those kit's with PCB and layouts online. But that's not the point.
What you'll find is that those kits are operating in push button mode, which is not really needed for this application.
Also, I don't want to be hacking a commercial kit, which is likely to be using the mic inputs etc.. Not needed.
The design I want is just barebones looper with a high impedance input stage and mixer for the audio. If you buy one of those kits, you need to add an external mixer.
As for cost, in terms of parts I can build this for £15, of which £13 is the chip itself! But that's not the point. I'd rather build something myself which I can be proud of.

edit: I just found an ISD2532 chip, with half the record time, and half the price. Makes this a sub £10 project (in terms of PCB components, not switches and jacks etc). I also found a quad opamp designed to run on 6v, so that should work a little better in this application. Can't remember what it's called right now, but I'll try it at some point.


Ok, below is a revised layout for this project drawn up in express PCB (much easier!).
I decided to use Bill Bergman's Analogue circuit, as it's known to work, if it isn't broke..
Digital circuit used is probably the same, though I haven't checked. I kept the digital half that I designed, so that the indicator LED's can be driven from transistors. My design also puts the ISD chip into standy when not playing, using the bypass switch, (Using a 3PDT Switch, the loop will start when the pedal is turned on, whilst still having true bypass). Connecting pads D & E will reset the pedal. 

Thoughts and ideas???


The circuit bending approach: rip the guts out of a children's toy with record/play function and make it go.


I'm really interested in this. All I think I really want out of a looper is the ability to record and play simple as that. Then what I would do is have an output jack for both the loop and then my guitar. This way if I wanted to add any effects to either one I could do that before recombining the signal and eveyrthing would be solved.