Bucket Brigade Delay Cascading (Longer time for ad3208)

[Rectangular]

Hey guys

I've built an ad3208 project (general guitar gadgets), and was wondering how plausible it would be to double the delay time by adding another dual-BBD section in series.

my only real concerns: can the 3201 clock driver support four mn3205 chips, or would I need to add a new clock driver chip, and if so, how do I sync the two clock chips ?

if anyone with some experience with BBD's can give me a hand, it'd be really appreciated.  just a note that I'm already using 3205's, so its not possible to increase the delay time by swapping out for larger chips

here's the section I believe to be isolated and duplicated, circled in red:

[nelson]

Buffer the Clock signals going to the MN3205 with a CD4049, feed the 4049 with the regulated 5v supply. I would also include a switch to switch out the two extra stages so you can get short slapback delay with the original two stages and add an LFO for modulated repeats.

[Rectangular]

hey nelson, thanks for the input

I dont do much cmos... assuming I add a duplicate stage (the one I have circuled in red) ,  am I feeding all four of the BBD's a clock signal through the 4049 ? because I'll have four BBDs and only one clock, should I place the buffers before I split up the clock signals, or after ? (ie: giving each clock -> BBD signal its own buffer)

I was considering adding an lfo, I need to dig around for a simple opamp based lfo with maybe two or three waveshapes, then use an optocoupler in series with the delay time pot.  if you can recommend a simple lfo, let me know.

[frequencycentral]

The square may need attenuating when compared to the triangle

[brett]

Hi
the LFO in the EA tremolo sives quality sine waves and is used in several DIY projects.
cheers

[nelson]

hey nelson, thanks for the input

I dont do much cmos... assuming I add a duplicate stage (the one I have circuled in red) ,  am I feeding all four of the BBD's a clock signal through the 4049 ? because I'll have four BBDs and only one clock, should I place the buffers before I split up the clock signals, or after ? (ie: giving each clock -> BBD signal its own buffer)

I was considering adding an lfo, I need to dig around for a simple opamp based lfo with maybe two or three waveshapes, then use an optocoupler in series with the delay time pot.  if you can recommend a simple lfo, let me know.

Why bother with an optocoupler?


I would buffer the signal then just feed it to the 4 BBD's. No need for a separate buffer for each BBD.


There are lots of LFO schematics around.

[Rectangular]

hey nelson

other than modifying commercial choruses and "fake-surround" systems, this is my first BBD project, so I'm unfamiliar with injecting control voltages into them, or their clocks. where would I be feeding this LFO signal into, the clock or the bdd ?  can you name another project that does this ?  (perhaps a chorus)

[nelson]

You would feed the LFO signal into the clock - look at the deluxe memory man schematic for ideas. 

[Praying_V]

Will adding more BBDs in series double the amount of signal degradation at long delay times?

[Entrant_21]

Will adding more BBDs in series double the amount of signal degradation at long delay times?

It shouldnt as instead of extending the delay time of one chip it is fed into another chip of separate delay time. so if you set the first stage at 320ms delay and the second at 150ms then both chips are within their acceptable audible THD threshold and the total delay will be at 470ms. as opposed to driving one chip to produce 470 on its own creating a great deal more THD than separate stages.

[Praying_V]

...Then how come no commercial pedal exists (that I'm aware of) with more than maybe 400ms of analog delay?  It seems so simple...  I know that BBDs ain't cheap anymore, but 20 years ago, you'd think SOMEONE would have been manufacturing a longer analog delay...

[snap]

they are dirt cheap currently and each single stage adds to the degradation

there are cheap tricks to reduce the total disturbances of cascaded multiple BBD chips
and those tricks are obviously used in some of the older commercial pedals
but those tricks haven`t been discussed here yet.

[Rectangular]

praying_v : look at the Moogerfooger and maxon AD999 delays, they both weigh in at about 1 second maximum delay time. I'd be interested in seeing what those circuits are like inside. if all it means is selecting certain components and tuning a bunch of trimpots on a scope, its no hassle for me, especially if it saves me 800 bucks

reading into this stuff more, I'm intrigued to see that the Memory Man does away with those BBD clockchips entirely, and controls the BBDs with a 4047 multivibrator, which seems pretty smart . especially if these bbds want a pulse waveform. I was looking at the mn3202's output and its practically a sine/triangle wave. very smooth and sloppy.  using a 4047 in the first place would omit the need for that 4049 buffer I'm adding

it seems this 4049 buffer can accept 9v supply voltages, any reason I should run it at 5v  instead of 9 ?

[nelson]

praying_v : look at the Moogerfooger and maxon AD999 delays, they both weigh in at about 1 second maximum delay time. I'd be interested in seeing what those circuits are like inside. if all it means is selecting certain components and tuning a bunch of trimpots on a scope, its no hassle for me, especially if it saves me 800 bucks

reading into this stuff more, I'm intrigued to see that the Memory Man does away with those BBD clockchips entirely, and controls the BBDs with a 4047 multivibrator, which seems pretty smart . especially if these bbds want a pulse waveform. I was looking at the mn3202's output and its practically a sine/triangle wave. very smooth and sloppy.  using a 4047 in the first place would omit the need for that 4049 buffer I'm adding

it seems this 4049 buffer can accept 9v supply voltages, any reason I should run it at 5v  instead of 9 ?

Silly me - I hadn't looked at the schematic properly - just power it off the same supply as the BBD/clock chips - through the 100ohm resistor.

[flo]

You can use a MN3205 (4096 stages) instead of a MN3208 (2048 stages). That doubles the delay time also and you have to change nothing. The new Coolaudio ones are cheap and ok:

Small bear electronics:

IC V3205
Coolaudio clone of MN3205 4K BBD
PRICE: $4.95

IC MN3205 (the orginal)
Low Voltage 4096-stage BBD
PRICE: $15.95     Shipping Wt. 0.03 oz

[flo]

Ok, sorry for that previous post, you are already using 3205 chips should have read the first post more carefully... 

[StephenGiles]

The Yamaha E1010 used 4 MN 3005 chips with a 4049 bufferred cmos clock.