Stacking ISD1000A's in Dean's Looper

[boyersdad]

Hey gang,

I recently came across a source of ISD1000A's for $6.25 CND each, and have built a few of Dean's Mobius Trip Loopers. They're great, and I even got them to fit into a 1590B.

I'm thinking about adding additional chips for extra memory. I'd like to stack the chips, as opposed to doing it dead bug style.

Has anyone here (Especially Dean) had any experience with this? Seems like I can tie the address pins together, the power, and just swap the sp+ and ana pins respectively.

Any thoughts?

[The Tone God]

They ICs don't work that way. Stacking them will not expand memory and you cannot added external memory. You could in theory chain them but then things get tricky with the control logic. You might as well use something like Payback with a larger time value IC.

Andrew

[boyersdad]

In the manual that comes with the chips it says you can slave other chips off it.

[The Tone God]

In the manual that comes with the chips it says you can slave other chips off it.

Thats what I mean by chaining the ICs which is really what they are talking about in the datasheets. You can do that as it uses some of the control signals of the previous IC to trigger the next IC. It is still kind of tricky. The control logic is one part of the adaption but you also have to update your audio section. Particularly the output interfacing. It gets really messy IMHO.

Andrew

[boyersdad]

Hmmm, maybe I'll see if I can find a few examples in other circuits first. Anyone else have anything to share?

[The Tone God]

Well my thinking is you can get a 25xx version that has more time and better sampling rates for less then your current cost for two 1100s and you don't have all the hassles of making it work.

There are a few other circuits out there including the Payback design.

Andrew

[Dean Hazelwanter]

Hi,

It gets really messy IMHO.

What Andrew (or Betty) said. 

Except if this is what you are talking about: Have 2 ISD1000s in parallel except for the CE pin. Connect a SPDT switch so that in 1 position the CE signal goes to chip 1, and in the other position the CE signal goes to chip 2. Then you could 'program' chip 1 with 1 loop, chip 2 with another loop, and then replay one or the other depending on the position of the switch. In this case, use a separate 47K pull-up resistor on each chip's CE pin that is connected between the CE pin and the switch. Dunno if it will load down anything in the audio section, but it's wacky enough that it just might work - give it a try!   

[Dean Hazelwanter]

Oh yeah - if it does mess up the audio section, use a 4PDT switch CE, SP+, SP-, and ANA IN.

[The Tone God]

Ah something like the multi-loop function of the Payback but with separate ICs instead of managing addresses. It could work. I don't think the inputs will have an issue as they seem to be high impedance but the output is a problem. When the IC goes into shutdown after playing back or recording it loads down the output so you would need to either switch the outputs or buffer then mix them down the line.

Andrew

[boyersdad]

Dag... I had the impression from the docs with the chip that you could "just" slave other chips with just coupling caps on the sp+ to the ana ins and have everything behave just like a single chip... I'mma look into the payback design.