LTC6994-2 Delay?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Bo Deadly

Well-known member
Joined
Dec 22, 2015
Messages
3,266
Location
New Jersey, USA
This "TimerBlox" chip is interesting: LTC6994-2

Does anyone see a reason why it could not be used to make a "digital" delay?

Specifically, use the standard comparator / modulation of a Class-D amp input to convert audio into modulated pulses, delay that using this LTC6994-2 and then just LC it back into audio.

Such a device could also make very short delays of as low as 1uS. If the noise / fidelity was good enough, you could make multi-pole FIR filters continuously adjustable using a CV.

You could also use the modulated form of high frequency pulses with conventional logic gates for signal routing / switching. Meaning you switch the high frequency modulated form of the signal and not the audio itself. The modulated form might also have some noise / distortion immunity so it could be transmitted over long distances.
 
squarewave said:
Such a device could also make very short delays of as low as 1uS. If the noise / fidelity was good enough, you could make multi-pole FIR filters continuously adjustable using a CV.
It could work for delays as short as the period of the highest frequency to transmit. For more, you'll need a FIFO register.
 
abbey road d enfer said:
It could work for delays as short as the period of the highest frequency to transmit. For more, you'll need a FIFO register.
Ah, I knew when I looked at this again in the morning I would find that it was too good to be true. There are other delay line ICs that have FIFOs builtin but the delays are only a few nS. This one doesn't have FIFOs at all. It just delays a single transition. It doesn't buffer them. Bummer.

But delaying / manipulaitng a pulse modulated form of audio is a simpler problem that might be worth exploring. Unfortunately I'm not aware of a shift register that is vaguely large enough to get to mS of buffered delay. You can't use DRAM because you'd need to generate addresses in which case you might as well just do regular digital. It can't be random access memory, it needs to be more like 1 bit sequential access memory (aka giant shift register).
 
Back
Top