AoE x-Chapters - 1x.2 Resistors

Ooops, perhaps a bad link, try this one:

First one worked find here.

On a sunny day (4 Aug 2019 08:49:18 -0700) it happened Winfield Hill wrote in :

Nice, many digipots listed.

Excellent stuff on non-ideal effects and digipots.

If you have time (which I'd guess is unlikely) you might mention trimpots. They are frowned on in some circles, but sometimes they are what works.

Yes, use them all the time, adhering to Bob Pease's suggestion that they not be relied upon for better than 1% trimming resolution and stability, even the 15-turn ones, IIRC.

What would you advise our readers?

What mostly interests me is using pots to set gain in wideband DC-coupled circuits, which is difficult to do otherwise. Some pots have bizarre, high-inductance geometries inside, and some are good to a couple GHz. But that's kind of niche-y.

We don't use multi-turn pots; they are if anything less settable than singles. My ROT is that a decent 1-turn pot is settable to about 0.1% and stable to a few times that. 1% of the pot's range is probably prudent as a production limit. Dipswitch + pot is extreme but sometimes the best way to go.

You could also mention r-packs, given all the time you have on your hands.

The Susumu surface-mount resistors are cheap and really good. The guaranteed 25 PPM parts are 10 or 20 cents, and measure in the single digits of tempco.

=1

I liked page "24" "The ever-creative John Larkin needed to find out the best choice for a 0.33 ? resistor for a pulse-stress application.... " and then went on to describe how your RUT (Resistor Under Test) treated poor John's once-of-a-kind test fixture. For shame! (ducking). (I merely

misplace/lose lent equipment for a year (or three) in my shop lab...)

Thanks for this fascinating read on the humble resistor!

John :-#)#

Hi, Win -

Immediately below figure 1x39 is the following: "...and a tempco of ratio of 5 ppm/?C max."

It reads better to me with only one "of" in the sentence. Personally, I would leave out the first "of" and make it "...and a tempco ratio of 5 ppm/?C max."

Just my 2 cents.

John

P.S. Excellent paper!

Here's the H+H improvement to the zapper,

which I should have done from the start. The hogged-out bit in the corner should have been an adequate warning.

Win, It may be out of scope, but just in case: Something I came across a long time ago was a degradation mechanism of spiral cut film power resistors at high common mode voltage. It appeared that corona discharge was happening at the edges of the spiral cuts (even though they were protected with some sort of paint or glaze) causing erosion of the film and early failure. I don't know whether the film was carbon or metal.

John

Interesting stuff, thanks. D-pots are fairly hard to use well in signal applications, as you usefully point out.

BTW a 'nicety' is a fine or subtle distinction, not something nice. (Cf. 'enormity'.)

Cheers

Phil Hobbs

Excellent stuff on non-ideal effects and digipots.

Especially if you need low noise. Transconductance-based VCAs are super noisy at low gains.

Wirewound pots are a bit like a mechanical DAC--they have a staircase characteristic with about 500 steps.

Especially the use of common-centroid design, e.g. making a 2:1 divider with a quad pack thin-film resistor wired as (R1 + R4) and (R2 + R3). That helps with both temperature and resistance gradients.

Cheers

Phil Hobbs

Any generated ozone attacks carbon film, or even composite, resistors. HV focus resistors are infamous for failing OPEN, even while dozens of non-HV similar components survive.

Don't forget this chestnut:

On Sun, 19 Jan 2003 12:46:30 -0800, John Larkin , In Newsgroup: sci.electronics.design, Article: , Entitled: "Don't do this! Just don't!", Wrote the following:

OK, every once in a while we need to have a trimpot crank in a little bit of DC offset into an opamp or something, and there are those nice

Fig 1x.29 -- what about capacitance of noninductives?

Depends actually which type, Ayrton-Perry should be fine but bifilar (shorted at one end) will be awful.

Speaking of Ayrton, she was not only one of the first EEs as such, but one of if not the first female EE. I don't know if that's footnote-worthy, but it's not well known I think!

Thin film resistors -- I've heard paranoia about these corroding. Any insight? (Not AoE-specific, anyone's experience will do.) Maybe was a legitimate concern, now outdated?

Fig 1x.30 is very useful! It looks visually noisy though; could you get that formatted as a heatmap, or like, drawing dotted or hashed curves around the general area of each family?

May be worth a note, Fig 1x.31 -- you can flatten out a messy resistor, sure, but you're just moving the power dissipation from one resistor to the other. Fatal for signals with a large RF component!

Do you have any data on TO-220 and other power film resistors, in terms of capacitance to heatsink, or equivalent circuits? And JL's measured significant "drool" apparently from induction into the heatsink I think, in such parts?

So many directions to explore with only so much time to go around, I know. :-S Looking forward to the finished book!

Tim

a very minor point...

"You often use a high-resistance voltage divider to monitor a high-voltage dc supply,"

do I?

Fig 1x.30: I'd suggest adding construction type to the list printed on the graph. Fig 1x.32, lower graph: scales marked -0.5% and -50%! Fig 1x.37: if you move the B to above the C it would be clearer

"Its failure mode, like the carbon-film CMA, includes a low-resistance current surge and a fail-short endpoint."

Some readers may misinterpret that as meaning that it shorts permanently. In fact they can go very low R but recover & resume normal service.

Figure 1x.39 Including social stuff is only going to irk those who don't know you from Adam and have no interest.

Motor driven panel pots were still in mass production in a minority of consumer stereos in the 80s.

Fig 1x.43 lacks any scale markings.

p32 'niceties' - I'm not sure those are niceties according to a dictionary.

NT

Thanks Win... really good stuff!

George H. (is there a capacitor X-chapter?)

Cermet material? Do you have a favorite brand?

We use a bunch of panel mount pots.. For DC and low freq the Bourns

We would also mention Phil H's loaded pot technique. (Which I've only used once.. but nice to have in your bag-o-tricks.) (You tie a two smaller resistors from the pot ends to the wiper.. makes a lower value pot with very fine adjustment.)

Yeah.

George H.