Hacker News Reader

Comment by kragen | Hacker News Reader

ΔΣ converters ("1-bit ADCs") have been commonplace since the 1970s; typically they use some feedback tricks to get better signal-to-noise ratios than you would calculate straightforwardly. 1-bit DACs were common in CD players throughout the 80s, and the SACD format about 20 years back was a raw delta-sigma bitstream. Unfortunately the Wikipedia article is not very good: https://en.wikipedia.org/wiki/Delta-sigma_modulation

Fundamentally, the reason they're especially accurate is that most of the usual sources of error in digital–analog conversion just don't exist in a 1-bit ADC. INL? Zero. DNL? Zero. Nonmonotonicity? Please.

Clover, by contrast, is interesting to me for a different reason: a lot of ANN inference stuff seems to work fine at 8 bits of precision. It'll be interesting to see if those results can extend to 4 bits.

View on HN

Thanks for the info on ΔΣ converters. It connects several things I knew about.

Now I need to study up on the area of computation that clover lives in.

Just to note how common delta-sigma converters are: Arduino's got them for voltage sensing.

I dare say that the most common AD converters are simply delta-sigma converters, especially if 1MHz conversions is fast enough for your tasks.

To get into the 100MHz or above speed (ie: Oscilloscopes, Software defined Radios, etc. etc.) you need to use other techniques.

When you say "Arduino's got them for voltage sensing," are you talking about the ADC in the ATMega328 that powers most of the Arduino models from the last decade? Because that ADC isn't a ΔΣ, it's a 10-bit successive-approximation ADC.

Ah yes, it seems like I've been mistaken :-(

I distinctly remember that Delta-Sigma was a slow-type of ADC. And I guess I confused it with the one in the ATMega328p.

You're right, it's an ATMega328P, not an ATMega328 like I said.

The ATMega's ADC is pretty slow, like 15ksps at full precision, so I can see why you'd think that!

> To get into the 100MHz or above speed

That gets back to my original point. The spectrum analyzer went into the gigahertz range. I don't think it was delta sigma. It was some multiple sample thing that only worked in the frequency domain.

It might have gone into the GHz range, but I'm guessing with a narrow bandwidth. It's the bandwidth that determines the sampling rate needed.

You can make spectrum analyzers that don't digitize at all, of course.

Well, I think most popular DACs in consumer electronics today are 1bit SD, though I remember older CD players having a 16bit DAC

They require higher clock frequencies and aren't that accurate (though, as you mentioned, some other sources of inaccuracy are simply nonexistent)

Reply on news.ycombinator.com