MXR carbon copy mystery part (clock driver? transistor?) - with schematic.

[ExpAnonColin]

I just got a carbon copy and being curious to see how they crammed that much delay into a small box I popped it open.  It uses 3208 of some kind (not MN, BL, or V) for those curious - 4 of them.  The interesting and smart part is that it doesn't use 4 3102's, it uses a 4047 (like the DMM) and two transistors which I can't find any info about.  I imagine they are being used to buffer/drive the clock signal because the 3208 clock inputs sink a lot of current.  They are very small, 6-pin surface mount chips that say something along the lines of "AB7" on them.  I'm guessing they are transistors because they are labeled Q1 and Q3 (designators for transistor elsewhere on the board) and there is very little support circuitry for them.  Any ideas?  It could be A137, or AB4, or AB1 - the lettering is tiny, I am using a magnifying glass here.  None of the above turns up anything useful on alldatasheet.  I heard Mr. Huge designed this, is he around to shed some light?  It would be useful to be able to use a 4047 and a couple of transistors rather than a bunch of 3102's in cascaded BBD designs.

-Colin

edit - schematic below.

[Processaurus]

a two pack of inverters?

[ExpAnonColin]

a two pack of inverters?

I didn't think so because they are labeled with "Q"s and not "U"s, most inverters can't drive four of the 3208's clock inputs (I think they drain 4 or 5 mA and most inverters are rated for about 2 or 3), and the V+ and GND pins are in the wrong places.  Here is a quick schematic -



It looks like a dual "something", with the inputs and outputs tied together, probably to allow it to source more current.  I mean I can't imagine that it's anything but some sort of transistor pair but I haven't encountered any transistors/transistor configurations that have allowed that kind of driving.  Just so everyone is on the same page, pins 10 and 11 on the CD4047 are inverted versions of each other (which the BBD's need) and the 3208 pin 2 and 6 are the clock inputs - again, which need to be 90 degrees out of phase.

-Colin

[ExpAnonColin]

One thing it could be if it is an array like this:

http://www.diodes.com/datasheets/ds30563.pdf

is a dual transistor configuration where one transistor has it's collector high, the other has it's collector low, then the emitters get tied together and the bases get the input.  Does this seem plausible?  I think I have seen something similar before.  Sort of like a push-pull emitter follower, but with two NPNs?

I guess it could also be a push-pull configuration if it was a complementary pair:

http://www.diodes.com/datasheets/ds30627.pdf

-Colin

[Mark Hammer]

Well you are correct that this is Jeorge's baby, but that's about all I know.  EHX appears to be coming out with a competing pedal, coyly named the Memory Toy (because the smaller version of the Memory Man is the Memory Boy, and Toy rhymes with Boy  ).  It also comes in a 1590B-sized package with a switchable modulation source: http://www.effectsdatabase.com/model/eh/nano/memorytoy

[ExpAnonColin]

Well you are correct that this is Jeorge's baby, but that's about all I know.  EHX appears to be coming out with a competing pedal, coyly named the Memory Toy (because the smaller version of the Memory Man is the Memory Boy, and Toy rhymes with Boy  ).  It also comes in a 1590B-sized package with a switchable modulation source: http://www.effectsdatabase.com/model/eh/nano/memorytoy

Indeed, I saw that!  Quite the slew of low-price analog delays coming out because of all of the BBD reissues.

Here's a question - if it is indeed a complementary pair, then wouldn't the transistors clip the 4047's signal to a certain extent, which I always thought was unacceptable for BBDs?

-Colin