good cheese

message

In the case of my little LDO, the fets are operating deep in their ohmic regions, with just a few tenths of a volt D-S. If they were operating as proper followers, the dynamics would be more obvious, and I wouldn't have Spiced it at all.

I can't see that it's any crime to tweak a circuit in Spice, as long as you understand what's happening.

I was staring at an ADC driver circuit, one of those diff-in/diff-out things with a Vcm input and a zillion resistors inside. I was driving one input from an LC lowpass filter, the other input from a DC offset voltage, the outputs driving a diff-in 250 MHz ADC. Currents sloshing everywhere, impedances vague. So, instead of a page full of algebra, I Spiced it and fiddled. Sometimes it even makes sense to design voltage dividers with Spice.

--

John Larkin         Highland Technology, Inc

jlarkin at highlandtechnology dot com
http://www.highlandtechnology.com

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators
Custom laser drivers and controllers
Photonics and fiberoptic TTL data links
VME thermocouple, LVDT, synchro   acquisition and simulation

Late at night, by candle light, Jamie penned this immortal opus:

message

No, once I have the idea of the circuit there's no problem calculating values. It's fun, too. Most ideas I simply file away but some look interesting enough to copy to spice and see if they actually do what it says on the tin.

Occasionally the topology's good but I need it for something else, so I'll recalculate and do some tweaks to see how far I can take it. The Baxandall circuit is a good example, lots of fun playing around with it and I did learn a thing or two.

- YD.

--
Remove HAT if replying by mail.

[snip]

Very likely. I often annoy customers by dragging my heels when I have any unease. In my 50+ years at this game I've had only one or two gotchas in literally multiple hundreds of chip designs. I prefer my standard "walk-on-water" image O:-)

It's just not analysis, it's meticulous design details to cover any eventuality... I can't tweak my final products.

I beg to differ. I have 18 patents and two more in the application phase. How many do you (or Larkin) have? How many products did you design 50+ years ago that are still being sold?

Indeed. My stuff must work right out of the womb... no touch-ups, no plastic surgery... and no wasted power to make circuits settle quickly... my customers worry over micro-Amperes... try designing cell-phone touch-screen drivers/receivers some time.

So I'm a picky asshole... and proud of it. ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
I love to cook with wine.     Sometimes I even put it in the food.

Late at night, by candle light, Bill Sloman penned this immortal opus:

message

Yes, if you know what you're doing. My own first estimates, without the calc, tend to be pretty accurate. I meant those who'll just sprinkle values nilly-willy until the whatever does approximately what it's supposed to do.

- YD.

--
Remove HAT if replying by mail.

message

Maybe you should put that shotgun to better use!

Jamie

for

=A0 =A0 =A0 =A0...Jim Thompson

SO8

ks.

he

o 1M t h

or

e e

But his reputation - and thus his income - does depend on how well his designs work.

You do like to think so.

It's not about the quality of the Spice simulators you and Jim use - though Jim does seem to have access to rather better simulation programs and device models than you and I do - but about the quality of the simulations you run, and the depth of your investigations of what the simulations tell you.

And those three aren't going to last long in the market. Nothing stops idiots from trying to sell rubbish into mass markets, but non-idiots have long since worked out that it's worth the trouble of designing mass-market products so that they are as good - and as cheap - as prolonged development can make them.

Details?

-- Bill Sloman, Nijmegen

e

I've got three patents - but I worked on low volume products where the cost of a patent is rarely justifiable. I've got no idea if any of my old designs are still in production - I no longer work for the companies that used them.

As far as your own designs that have lasted 50 years, I used some of them when they were new, and dumped them as soon as something better - if more expensive - became available. The 555 is still selling in huge volume, but not because it was an inspired design.

You are okay at what you do, but no Bob Widlar nor Barry Gilbert.

Integrated circuit design does call for that kind of personality. Your variant of picky asshole has a higher asshole content than most.

-- Bill Sloman, Nijmegen

in message

ith-I...

gy Inc

ted

=A0 =A0 =A0...Jim Thompson

's

need

a

I'd love to, but Dutch employers reading my CV gets as far as my date of birth, then throw it in the rubbish bin.

But I ordered the parts for my low distortion oscillator yesterday - I may not impress Dutch employers but I may yet impress the literature.

-- Bill Sloman, Nijmegen

Obviously. So does mine.

Spending years on an unfinished and likely useless 2-transistor oscillator doesn't sound very productive. What is your goal, one more paper in one more peer-reviewed journal before you die?

All the stuff I design - lots of it - works. I don't simulate a lot. The last sim I did, last month, was an e/m simulation of an edge-launch SMA connector on a 4-layer PCB. It worked:

I used ATLC2, a little weird but free.

The Samsung is about $4 with the cable, and the others are closer to $1. They will probably sell because people who shop online have no way of knowing what's good or bad, so buy cheap. If it doesn't work, most people will throw it away and not bother to get a refund.

Nothing stops

Nice rectangular package that doesn't overlap the neighboring outlet on a power strip; really nice flexible cable and connectors; good accuracy and regulation; low noise; looks good; nice flat surfaces where we can stick our labels.

--

John Larkin                  Highland Technology Inc
www.highlandtechnology.com   jlarkin at highlandtechnology dot com   

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators
Custom timing and laser controllers
Photonics and fiberoptic TTL data links
VME  analog, thermocouple, LVDT, synchro, tachometer
Multichannel arbitrary waveform generators

You are OK at what you don't do, but no John Fields.

--

John Larkin                  Highland Technology Inc
www.highlandtechnology.com   jlarkin at highlandtechnology dot com   

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators
Custom timing and laser controllers
Photonics and fiberoptic TTL data links
VME  analog, thermocouple, LVDT, synchro, tachometer
Multichannel arbitrary waveform generators

[snip]

You keep saying that. Cite which you used.

Inspired has no meaning. What made RS-232 inexpensive... my original

1488/89... which still sell, along with my first OpAmp, MC1530/31... because it's stable, come hell, high water or capacitive loads ;-)

You have no way of knowing that. Since 1973 my work has been mostly private custom stuff for which I can't exuberate. However, there is my USB work for Intel... the first inexpensive and compact work in that area... I specialize in making things work... cheaply and forever!

I'm so proud O:-) ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
I love to cook with wine.     Sometimes I even put it in the food.

--
Geez, talk about obsession...

Larkin: Obsessive/compulsive manic/depressive... spares no effort to keep his name in print... no matter how repulsive he appears. ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
I love to cook with wine.     Sometimes I even put it in the food.

My LDO does *not* oscillate at low loads. It does *not* have "too much gain." You were just wrong.

--

John Larkin         Highland Technology, Inc

jlarkin at highlandtechnology dot com
http://www.highlandtechnology.com

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators
Custom laser drivers and controllers
Photonics and fiberoptic TTL data links
VME thermocouple, LVDT, synchro   acquisition and simulation

I didn't say it would "oscillate". I said it was unstable, which fits with your description of overshoot. It _does_ have too much gain _and_ you clearly blundered the compensation... which it why it overshoots... but, if you're happy, I'm happy... it isn't my _wasted_

240mA.

You're wrong, as usual, but you'll never 'fess up. So I'll not waste my time further, other than to include your pdf in my examples of amateur design. ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
I love to cook with wine.     Sometimes I even put it in the food.

--
How easy to say, but how difficult to prove since, if it overshoots,
then it's not critically damped and, consequently, has too much gain
until it settles down.

ou

MC4024/MC4044. I did look at the MC1495 fairly carefully, but only long enough to decide that it wouldn't work in our application

Actually, because they were designed into enough products on early that they became legacy parts. I once wrote an "approved op amp" document for Cambridge Instruments - mainly to demonstrate that having a list of "approved op amps" was a silly idea - and ended up listing some 150 parts with comments about applications where they were pretty much the only part that you could use. The MC1530/31 didn't feature -

+/-6V isn't a popular choice of power rail, and an input stage that you can blow up by putting more than 5V across isn't a popular feature either. The uA709 could at least take +/-15V rails.

Dream on.

If you were in the Bob Widlar/Barry Gilbert class, someone would have made it worth your while to stick around. Admittedly you are pretty obnoxious, but Bob Widlar was at least as bad and a drunk to boot.

o

We've noticed.

-- Bill Sloman, Nijmegen

ed

m

pose

ut

n

ng

.

ve

ou

Bizarre formulation. The idea of assigning competence to skill which isn't practised is lunatic - if I don't do what ever it is, how can you know if I would do it well were I to take it up?

And why is John Fields relevant - he knows all that there is to know about the 555 (except that it's obsolete) but in every other respect he is depressingly ordinary.

-- Bill Sloman, Nijmegen

Oh. Excellent point.

--

John Larkin         Highland Technology, Inc

jlarkin at highlandtechnology dot com
http://www.highlandtechnology.com

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators
Custom laser drivers and controllers
Photonics and fiberoptic TTL data links
VME thermocouple, LVDT, synchro   acquisition and simulation