BBD delay questions...

[cpm]

ok, once again a topic concerning BBD delay, at least i hope there are new questions...

i am working with the cheaper 3205 chips, the aim is a build friendly project: 9v, cheap 3205, 571, 4047 clock, etc...
also trying to obtain the best sound quality out of those 3205's and long delays.

First, i see all schematics using 3205 its powered at 9v, or close, with a 8.2v zener, but datasheet suggest 5v.
Certainly, noise figures and THD is minimised at 5v, but other rating at frequency response and insertion voltages seem to improve with some more Vdd.
- Is there a way to get a regulated voltage, lets say 6.5v from either 7805 or 7809 regulators?
- Is the 100R+8.2v zenner configuration safe for powering 3 or 4 delay chips?

Now, for best frequency performance, there is this graph on the datasheet (MN3205):


Its actually an inherent low pass filter on the way through the delay stages. -3db at 3kHz. So 2 chip in series makes up for a -6db at 3kHz
I guess this is the "analog vibe" of the repeats

On the old SAD1024 datasheets there is this arrangement:



according to it, using two delay lines in parallel, with differential inputs and combined at the output can get ride of most of clock glitches and noise by differential cancellation. Also nyquist freq go up to clock freq, so my guess is that its possible lo lower clock freqency to get equivlaent delay times as would be having two lines in series, but audio is only passing through 4096 stages, having less effect of that insertion loss implicit low pass filter.

would it still be applicable with two parallel 3205's?
would it really benefit from the standars series delay lines?

[WLS]

...
Is there a way to get a regulated voltage, lets say 6.5v from either 7805 or 7809 regulators?
...

Use a 7809 with a voltage divider.


Bill

[Mark Hammer]

The entire MN32xx series was deliberately intended to be usable at lower voltages.  Why?  The bias voltage presented at the BBD input pin should remain "valid" for the life of the circuit.  When a BBD is battery powered and the bias voltage is set by dividing the V+ from the battery supply, the bias changes as the battery ages.

HOWEVER...if the battery is regulated down to a stable 5V, and you both power the BBD and take the bias from a stable 5v supply, then whatever you set the bias trimpot to when the circuit is built is what the bias should remain at for the rest of its functioning life.  The 5V regulator will function properly until the battery drops down to around 7V, at which point the battery is likely to be too weak to power the circuit anyway.

So, my first question to you is: Do you intend to power this by battery at any point, or is your intent to ALWAYS have it powered by a dependable stable and known V+?

[cpm]

The entire MN32xx series was deliberately intended to be usable at lower voltages.  Why?  The bias voltage presented at the BBD input pin should remain "valid" for the life of the circuit.  When a BBD is battery powered and the bias voltage is set by dividing the V+ from the battery supply, the bias changes as the battery ages.

HOWEVER...if the battery is regulated down to a stable 5V, and you both power the BBD and take the bias from a stable 5v supply, then whatever you set the bias trimpot to when the circuit is built is what the bias should remain at for the rest of its functioning life.  The 5V regulator will function properly until the battery drops down to around 7V, at which point the battery is likely to be too weak to power the circuit anyway.

So, my first question to you is: Do you intend to power this by battery at any point, or is your intent to ALWAYS have it powered by a dependable stable and known V+?

i know you have stated that before... i think i know more from bbds from your posts than any other resource (thanks!)

the intention is obviously to use that stable 5v because bbd bias is the most sensitive reference in the circuit. main power would be a 9v external source, not battery, but for commodity DC adaptors you cannot asume they are regulated so it may actaully be 9v, 10v, 11..

the point of raising bbd voltage is just in case it improves its performance regarding frequency loss.
i see there is a 7806, so i may try that if i can source any

somewhere i read that 1 or 2 diodes in series with the ground pin on a 7805 would result on a 6v output, is this correct, safe?


by the way, Mark, do you have a readable schematic for the old MXR/Ross analog delay, those which uses transmission gates on the filters?

thanks indeed!

[Mark Hammer]

the point of raising bbd voltage is just in case it improves its performance regarding frequency loss.
i see there is a 7806, so i may try that if i can source any
somewhere i read that 1 or 2 diodes in series with the ground pin on a 7805 would result on a 6v output, is this correct, safe?

Check the datasheet of the 78Lxx series of regulators and you'll usually see a suggested circuit for achieving a variable output voltage

by the way, Mark, do you have a readable schematic for the old MXR/Ross analog delay, those which uses transmission gates on the filters?

thanks indeed!

I do, and it is posted, but I can't see the site from work to be able to give you the link.  I'll try later tonight.

...and you're welcome!

[g.]

you can do the diode trick as cpm said,
you add a diode(1N4004)  betwen the ground and the pin 2 of a 7805
and you have about a reg of 5,7V, add two diodes you have 6,4v
etc.

[Mark Hammer]

http://i414.photobucket.com/albums/pp228/Mark_Hammer/rossdelay.gif

[gena_p1]

http://shiftcustom.netova.info/tmp/ross_delayschem.jpg
higher resolution ross delay scheme

and you may also look at arion SAD-1 and aqua puss/boss dm-2 shemes, they are already 3205 low voltage

[Mark Hammer]

Thanks.  MUCH clearer than what I had.