Standard non-inverting opamp with a cap across + and - terminals...... Why?

matching

an

Right. A few ten puffs usually suffices.

Before enlightenment: chop wood, carry water. After enlightenment: chop wood, carry water. ;-)

Cheers! Rich

We used to call that getting an education.

On Apr 16, 6:48=A0am, John Larkin wrote: [...]

I think it for the same reason that machinists generally don't design very good circuit. I saw a circuit that was about 6 inches by 8 inches of PCB. It was a dead time corrected rate meter designed by a machinist. It had the cook book circuit for this connected to the cook book circuit for that connected to the cook book circuit for the other thing including some analog multiplier chips etc. It could be replaced with basically this:

------ ----- CD4053 IN ---! ONE !-----!. ! ! SHOT ! !. O-!------ -VREF ------ --!--> ! ! ! O-!----------- \\ ----- ! / ! \\ ! ! ! +-----!!-------------+ ! ! --!-\\ ! ! >--------------+---Out GND------!+/

I'm sort of seeing the same patterns. A cookbook differential pair with large DC common-mode outputs, driving an insane level shifter, driving a cookbook 3-opamp diffamp. All of which could be collapsed.

And a lot of the "physics" circuits have insanely expensive, useless parts like $180 wideband transformers, or rows and columns of relays switching trivial stuff.

This is one of my favorites, seen again and again in RSI-type papers:

Then they publish the stuff and others think it's the way to go.

Like wanting to replace a single 0.05ppm resistor (my design) with EIGHT op-amps and lots of other stuff, including more resistors, of course, and claiming it will be just as stable. 8-(

Best regards, Spehro Pefhany

In a sense, I was using it an incantation, but the only "magic" I was invoking was the Ebers-Moll relationship - if you understood how that works in practice, we wouldn't be having this discussion.

-- Bill Sloman, Nijmegen

the

Because the bias current changes produce smaller offsets than the nonlinearity of the large signal voltage transfer. You are making a fuss about a mouse and ignoring the elephant.

Probably true, but irrelevant, since I've designed a lot of real electronics, some of it precision stuff where "small" offsets presented big problems.

It's not their primary job. Back when I read Review of Scientific Instruments regularly I'd send them a comment every year of so, pointing out that one of their articles included electronic nonsense. A few of them ended up getting published. They show up on scholar.google if you search on "A. W. Sloman".

Back when I was getting my Ph.D. in Physical Chemistry, I ended up having to design some circuits to monitor the chemical reaction that I was studying. They weren't very good, but they worked. Since then I've had more practice.

-- Bill Sloman, Nijmegen

On Apr 16, 8:40=A0pm, John Larkin [..bad circuits..]

Yes I have seen it. :( It is extra good with not very regulated supplies and carbon-comp resistors.

The test procedure written by the same guys were also very creative. They had the test technician doing math on a bunch of numbers they measured. Fortunately the technician involved knew enough math to be able to translate things like:

Measure A __________ Measure B __________

Calculate Y =3D A/B _________ (Minimum 1.9, Maximum 2.1)

The technician noticed things like the values of A and B being within a 2% span and translated the requirement into a limit on A and B.

A much better test for the same thing was just to apply the signal from a calibrated generator and look at the A value. The B was just the input that got to the board.

I don't mean to imply that the folks involved weren't darn good when they worked in their own field. Imagine trying to thread the inside of a 1.5 inch tube with 1/16th walls without distorting the tube. These same guys who didnt do a good job of working out circuits figured out how to do that machining operation.

Just imagining the resource-waste that implies is stunning. The design, the results not gotten, progress delayed, the infrastructure being ill-used...

The SNR of science is appalling. The signal-to-money ratio is worse.

James Arthur

Overblown egos. They can't admit that they aren't God, or even a jack of all trades.

=A0!

=A0!

=A0!

=A0!

Only if you assume that scientists know what they are going to find out before they do the research.

Innovative engineering is expensive for much the same reason.

In fact, the signal-to-money ratio n science is remarkably good. A lot of scientific instruments look badly engineered because they were designed to work adequately for a short period while being run by a skilled scientist who was watching the gear like a hawk. The scientists involved aren't really aware of this bias, and pass their stuff over to engineers thinking that putting it into production just involves wrapping the circuits in a pretty case.

-- Bill Sloman, Nijmegen

Undersized budgets. Thye can't afford to hire John Larkin - or even Michael Terrell - as a consultant to design their one-off equipment.

-- Bill Sloman, Nijmegen

"Spehro Pefhany" a écrit dans le message de news: snipped-for-privacy@4ax.com...

Or is it percent?

-- Thanks, Fred.

Nope. They only _think_ they can't. I've had numerous requests from universities but many just fizzled. Budget limitations and yada, yada, yada. Then they bought some freaking expensive modules at the usual suspects, the lab suppliers where all starting prices are four-digit. After blowing $100k or more on this it still doesn't work reliably. For less than half they could have had a custom design done by a consultant that actually works and let them concentrate on what they are really good at, the experiments. Most of all, we could have included some of their students so they can learn hands-on instead of just from textbooks and lectures. But no, the budget only allows materials, lots of materials.

Depends. a few years ago I was called in to help a medical electronics start-up at Georgia Tech... at the behest of the investors ;-)

...Jim Thompson

Off the shelf metal foil (but that's drift per Kelvin, not tolerance).

Best regards, Spehro Pefhany

I never do one-offs. We make our money on the 2nd floor (production) and not on the third (engineering.) I often do the engineering for free (as I'm doing now, on part of a big analytical instrument) so we can boost the sales of the instrument maker and share in the long term.

John

Same here but that's different. A start-up means they've got outside funding from business-minded folks. A.k.a. investors. They will, of course, want to make sure that the technical bases are covered.

Once a start-up has received outside funding it's not really the university anymore that calls the shots.