Anyone here tried running 2 BBDs with interleaved clocks ?

Started by DrAlx, February 19, 2014, 01:26:38 PM

Previous topic - Next topic

DrAlx

I've built a couple of flangers using the MN3207 BBD  (based on Thomeeque's EM3207 9V Electric Mistress Clone).

The BBD in that circuit is clocked by the (buffered) outputs from a D flip-flop, and the flip-flop is triggered by a comparator chip.

It occurred to me that there is an unused flip-flop on the board, and that it's possible to use that to clock a second BBD
in parallel to the first, but have the second BBD's clock interleaved with that of the first BBD.

You can produce interleaved clocks like this...

Flip-flop 1 has input D1.  Its outputs Q1 and *Q1  are used to clock BBD1.
Flip-flop 2 has input D2.  Its outputs Q2 and *Q2  are used to clock BBD2.

The existing EM3207 circuit has D1 connected to *Q1, so Q1 varies like this over time...

Q1: 1 0 1 0 1 0 ... (drives BBD)

To get interleaved clocks for 2 BBDs, you just need to rewire the flip-flop connections like this:   

                Connect D1 to   Q2
                Connect D2 to *Q1.

The outputs from the flip-flops are interleaved in time, but they run at half the previous rate...

Q1:  0  0  1  1  0  0  1  1  0 ...  (drives BBD1)
Q2:  0  1  1  0  0  1  1  0  0 ...  (drives BBD2)

Therefore the comparator chip that triggers the flip-flops needs to have its rate doubled in order
to keep delay time unchanged.  If I'm not mistaken, interleaving effectively doubles the overall
sampling rate, and should be able to lower overall noise, and provide less stringent requirements
for any anti-aliasing filters in the circuit.

I am wondering if anyone here has tried something like this before, and whether it would be worth pursuing ?

Scruffie


pinkjimiphoton

sounds like a cool idea... gonna sit back and watch <grabs popcorn>
  • SUPPORTER
"When the power of love overcomes the love of power the world will know peace."
Slava Ukraini!
"try whacking the bejesus outta it and see if it works again"....
~Jack Darr

DrAlx

Quote from: Scruffie on February 19, 2014, 01:31:29 PM
Look up parallel multiplexing.

Thanks Scruffie. Very helpful.
It sounds that if you want to use two BBDs to double your delay time at a particular clock rate, then parallel multiplexing works better than placing the 2 BBDs in series, but only if the clock rate is high (since you can clock the BBDs at half rate).  So not as useful as I first thought.




Fender3D

Quote from: DrAlx on February 19, 2014, 02:46:03 PM
Thanks Scruffie. Very helpful.
It sounds that if you want to use two BBDs to double your delay time at a particular clock rate, then parallel multiplexing works better than placing the 2 BBDs in series, but only if the clock rate is high (since you can clock the BBDs at half rate).  So not as useful as I first thought.

yes and no...

When par. multiplexing you have:
2 BBDs with the same delay time of 1 BBD AND double sampling rate with the same clock applied
"NOT FLAMMABLE" is not a challenge