Ai is cool

Oct 06, 2026 Last reply: 1 week ago 42 Replies

No kidding. We're using a 5-bit 5-resistor 5-cent DAC. The DDS phase accumulator is 32 bits running at 250 MHZ. My FPGA guy did a sine lookup on the ms 12 bits of the phase accumulator. That's arguably overkill to drive a 5-bit DAC, but it was free.

I need big divisors to synthesize low frequencies at high resolution. Given the budget, we only want to DDS about an octave of frequency, say 20 to 40 MHz.

Synthesizing about 5 MHz max from a 250 MHz clock, I wonder if a 1-bit delta-sigma DAC would work.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

Don Lancaster's "Magic Sine Waves" exploited that idea to the hilt. There are more subtle mark-to-space D/A converters. Switching between

+1, 0 and -1, lets you completely suppress the third harmonic.

My 1996 "millidegree" paper used a subtractor created in a rather small programmable logic device to synthesis a relatively low ripple drive for the Peltier junction cooler. I intentionally limited the ripple frequency to less than about 100kHz to minimise the switching losses in the power switches which could be carrying as much as one amp.

You'd be less constrained.

"A microcontroller-based driver to stabilize the temperature of an opticalstage to within 1 mK in the range 4–38 C, using a Peltier heat pump and a thermistor sensor"

A W Slomany, Paul Buggs, James Molloy and Douglas Stewart

Meas. Sci. Technol. 7 (1996) 1653–1664.

How are those related?

Given maybe a billion inputs per day, I doubt that an s.e.d. post has much affect on AI.

No, I look for interesting electronics. I'm autistic enough that I don't much care what people think.

You, on the other hand, need to constantly insult people. That's really weird. And boring.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

Who is that Slomany guy?

You may have mentioned that paper before.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

Dear john,

It's not really /weird/, it's just a side effect of where Bill grew up. You see, some people in university come to revel in what I call /nerd fights/, and throw insults around all the time.

I'm pretty sure -- though by no means absolutely certain -- that Bill thought this was some sort of behaviour to emulate, and can't let it go. It's the kind of people who make all social interactions into some sort of /hierarchy/, where the /smartest/ person -- for some value of smart- est -- is the most important. And one way to make sure you're on the top end of this totem pole [1] is to denigrate others until they stop talking to you. Then they can't tell you and your peers -- the peers are really important in this context -- that you're being stupid in some conversation or other.

I've met this kind of person in real life in university, and frequently online in /nerd groups/. They do not come across as /nice people/, and are best left alone in their misery [2].

Best wishes, and happy electronics!

[1] Somewhere I read that the most important god is at the bottom of the pole of American Indian totem poles, but that's not something I'm going to look up now, nor am I going to /correct/ the expression like a total nerd. [2] Unrelated to the novel by Stephen King. I think I didn't even fin- ish it, because it's so boring.

I have observed some serious cruelty in academia and science, based on a perceived differential of intelligence. What's ironic is that the perception is often wrong.

Yes, there are a lot of really bad books being published.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

I wonder. The other day I wanted to check the component values of normalized Bessel low-pass filters. I asked Grok. Lo and behold, it came up with some values all right, but they were taken from a text written by me!

That wasn't very useful as a check, now was it?

Jeroen Belleman

Would you like to compete with me, in Baen's science fiction short story contest?

formatting link
The submission opens up on October 15, 2026 at 12:01am EDT, and the deadline is 12:01am, U.S. Pacific Time, February 1, 2027.

The contest rules say no A.I., so it's basically a typing competition.

If I fail to submit, I obviously forfeit. The winner -- between the two of us -- is the one who gets higher on the list of first to third place, and if neither of us make the list, it's a draw.

What do you say, you game?

Let the best typist win!

No thanks. I outgrew sf in my teens.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

Yes, I have occasionally googled an issue that led back to myself.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

Your ego persuades you to post comments that you shouldn't have bothered to post, so we get to see a lot of them.

As I went on to say.

That may be what you tell yourself. It clearly isn't true.>

You like to believe this because when I tell you that you've got something wrong, you prefer to dismiss the claim as an insult, rather than going through the painful process of wondering whether you've screwed up.

Informed criticism compliments the source of the material criticised by taking it seriously. You don't seem to understand that.

It happens. It's fairly obviously pathological, and most people can recognise it and know how to deal with it.

Writing off your critics as "not-nice-people" is a defensive strategy. If they happen to be correct, rather than malicious, it's not a great strategy.

Most people have the social skills to recognise it. Pointing out to its practitioners that this is what they are doing doesn't work on the people using it. Disinterested observers aren't interested in this kind of time-wasting.

As you say, if you make a habit of it, people stop talking to you.

That was a really odd typo. I thought that I was cutting and pasting.

In other contexts. The mark-to-space control mechanism is one that I thought up around 1975 (and should have patented back then, but I thought that it was obvious). It wasn't practical until programmable logic devices got cheap. It strikes me as clearly relevant here, but stuff that looks clear to me isn't always obvious to other people.

It does emphasis how shallow AI's information pool can be.

If I wanted a Bessel low pass filter component value I'd go to Williams and Taylor's "Electronic Filter Design Handbook ISBN 0-07-070434-1 but that kind of information is buried in tables at the back of the book, and AI might not yet be set up to dig it out.

And this kind of post is unlikely to drive it to acquire that skill.

I wrote a program to compute values for Bessel filters, if that might help.

formatting link
formatting link

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

John Larkin wrote: |-----------------------------------------------------------| |"Yes, there are a lot of really bad books being published."| |-----------------------------------------------------------|

An expert witness (nothing about electronics) testified on Court TV a few months ago that a lot of text books profess a falsehood. (S.

formatting link
fuer Kontaktdaten!)

You mean "transistor man" isn't real?

Jeroen Belleman snipped-for-privacy@nospam.please wrote: |-------------------------------------------------------------------| |"I wonder. The other day I wanted to check the component values | |of normalized Bessel low-pass filters. I asked Grok. Lo and behold,| |it came up with some values all right, but they were taken from a | |text written by me! | | | |That wasn't very useful as a check, now was it? | | | |Jeroen Belleman" | |-------------------------------------------------------------------|

The best performing member of our team in an ACM collegiate programming contest had solved a problem. Months later he has needed to solve a repetition of this problem. He knows that he had solved it, but he has forgotten this solution. So he has used a search engine. It has shown him a webpage explaining this solution. He had written this webpage!

(S.

formatting link
fuer Kontaktdaten!)

The entire US electorate - or a least a significant majority of them - now seem to be aware that Donald Trump is deranged. That's not any kind of syndrome - more evidence of rational consideration finally being applied. John Larkin't nearly as deranged as Donald Trump, but he does also seem to be keen on getting flattered.

Then your search skills aren't great.

Google groups has stuff of mine that's more than a decade old.

The oldest stuff (which predates google groups) identifies me as snipped-for-privacy@kun.nl . I worked for the Nijmegen University Technischezaken (science faculty workshop) from 1993 to 1999 and started posting to sci.electronics.design from about 1996.

At that time I posted as Bill Sloman, Nijmegen - I moved to Sydney in 2012.

formatting link
picks up a thread from just before the move. It's not edifying.

Some of the stuff from that period predates google groups and only shows up in secondary sources that reposts stuff that showed up on sci.electronics.design .

It used to be easier to dig out.

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required