Of BBDs and clock drivers (MN3102 vs CD4047)

[decc]

I'm thinking of building a chorus and/or an analog delay and so I've been trying to get up to speed on BBDs and their associated clock chips. Specifically why would anyone use a CD4047 such as in the Small Clone? The MN3102 is a smaller chip specifically designed for driving the BBDs, of which there is only one 1024-stage IC, while the CD4047 is larger and does not provide the V_GG voltage reference.

Could I make a Small Clone, or any other chorus, with a MN3102?

Why does the Small Clone not need V_GG? Is it just that the 14/15 ratio is so close to 1 that it doesn't matter?

Does the CD4047 have any advantages at all (for example can it drive multiple 4096-stage BBDs for making longer delays?)

Unfortunately the datasheets are all old and scanned so I may have overlooked something in there I couldn't read. Sorry if I missed something obvious.

Thanks

[Mark Hammer]

The 3102 is a decent chip that provides all the convenience you note.  It does not appear to provide the current drive, however, that permits it to overcome the capacitance on the clock input pins of the BBD at higher frequencies.  If one is clocking a chip for longest possible delay times, this does not pose too much of a problem, unless one is attempting to clock multiple BBDs with one 3102.  However, if the goal is for fast sampling rates and shorter delays, the 4047 does a better job of driving the clock pins on the BBD.

In commercial quantities, it's probably cheaper too.

[decc]

OK, thanks, that explains the use of the CD4047 for a chorus. Any idea on the V_GG thing?

I think I have figured out the question if can it be used to drive multiple MN3x05 chips by itself: There is a barely legible graph relating rise/fall time to load capacitance. While it does not extend out to the 2800pF range of the BBD it does look linear. Extrapolating (15V as the best case) I would guess that with one chip that would be a rise/fall time of 875ns. At the max clock frequency of a delay of about 100kHz (10000ns period) this is less than a tenth which seems, and I have no reason to justify this, acceptable. However the BBD datasheet has a max rise/fall of 500ns. Conclusion: the CD4047 fares better with high frequencies, but can only drive the smaller BBDs. A single 4096-stage is alone too much.

[StephenGiles]

No, EH used an unbuffered 4047 with an MN3005 in the 1980 Stereo Memory Man with chorus, but a buffered 4047 (4049) in the Memory Man with 2 x MN 3005.

[decc]

Interesting. I guess when you're dealing with an already strongly filtered & decaying delay signal you can get away with going past the limits a bit.