Re: delay chips and design

[Matthias Grob <matthias@grob.org>]

Could it be that your computers watch is off, Lance? I somehow often 
find your post only later...

>ok. so the bucket brigade chip is actually a true analog chip that was in
>use in the 70s and 80s and now is in use in the moog analog delay.   would
>these work?:
>"SAD1024
>1024 stage, capacitor "bucket brigade" audio frequency delay chip.
>Discontinued, but Panasonic makes a full line of (apparently) equivalent
>chips in various sizes and configurations. Three that may be suitable as
>replacements are MN3007, MN3207, and MN3214. They are listed in the **
>current ** DIGI-KEY catalog (199901..03) on page 187 at $9 to $18. The 
>data
>book is listed as part number 9102B at $3.50"

yes, those are the ones! And they are cheaper that they were!
If I remember that MN3007 right, its also ancient. Not a direct 
replacement for SAD parts, but functionally the same or better.
so, go ahead and build your analog delay, its not so complicated...

If you are smart, you can adapt the clock frequency to create a tap tempo!

Say you have 2048 buckets.
So you want 2048 clocks to happen between the two taps.
You have a master clock some 100 times faster than the sample clock.
You count the master clocks between two taps.
You divide the result by the number of buckets, in other words shift by 
11bits.
The result is the number you have to divide the master clock with.
No?
Thats easy to do with a few 74HCxx chips...
Then you can still fiddle arround with the master clock to get the so 
desired analog speed effect.

I would put another delay line in parallel which is 4 times longer 
and runs on the same clock. So you get two tracks, a one bar rhythm 
and a 4 bar harmony... :-)

>next question: does the EDP "read samples" like the repeater and if so why
>does it not have a "pop" at the loop point of a looped sine.  does the 
>line
>6 "read samples" and is it's behavior more akin to the repeater or the 
>EDP?
>how prevalent is the repeater design architecture becoming in the
>delay -loop market?

The EDP is not sample oriented, its running through the whole memory 
just like a tape machine. It does not create pops because it does not 
jump in the memory (exept for some functions) and it fades in and out 
with a analog chip at the input. But the sine at the loop point is 
not perfectly spliced, just cross faded.

-- 


          ---> http://Matthias.Grob.org