0.01% resistors

[Raveninghorde]

What are people paying for 0.01% resistors?

I wanted to reduce production tweaking so thought a few 0.01% 0603 or
0805 resistors would do the trick. Most suppliers have parts in the
many dollars range. Digikey do have a few Stackpole resistor values
for £0.57, about 85 cents.

 

[Bill Sloman]

What are people paying for 0.01% resistors?

I wanted to reduce production tweaking so thought a few 0.01% 0603 or
0805 resistors would do the trick. Most suppliers have parts in the
many dollars range. Digikey do have a few Stackpole resistor values
for £0.57, about 85 cents.

Farnell has got a few parts at a slightly higher prices - 10k, 1k, 100R.

 

[RobertMacy]

What are people paying for 0.01% resistors?

I wanted to reduce production tweaking so thought a few 0.01% 0603 or
0805 resistors would do the trick. Most suppliers have parts in the
many dollars range. Digikey do have a few Stackpole resistor values
for £0.57, about 85 cents.

Once, long ago, someone decided to use 0.01% resistors. We found the
rpcing in US dollars at $1 to $1.50
Delivery data, now this was the 'interesting' part, 6 to 9 months, but
NEVER arrived !!!!

It was cheaper to simply put in excessive parts and let a laser
automatically cut unwanted ones out. That was assembly was constant, and
the tester program did the 'adjustment'

In retrospect, I'd push for the 'design-it-out' principle. Go back to
Engineering and beat them about the head and shoulders until they had
designed out any problem like that in Manufacturing.

 

[Spehro Pefhany]

If there's a microprocessor in the loop, make sure that you keep things
within range of your ADC (or whatever), and calibrate by changing
constants in flash.

It does require tweaking, but it can be automatic tweaking.

The big dollars are not for close tolerance resistors but for
resistors that earn that close tolerance number by having commensurate
low tempco. Without that you're going to get noise and nasty
transients as temperatures change, at best, even if you calibrate out
the temperature drift.


Best regards,
Spehro Pefhany

 

[Bill Sloman]

True. I always thought that there's a market for matched resistors on
the same chunk of ceramic, like matched transistors. Either the market
isn't as big as I think, or it's just not possible to get good enough
matching to make it worthwhile.

Farnell stocks a few precision resistor arrays, so your first thought is correct. I bought a couple for my low distortion oscillator last year.

 

[John Devereux]

Farnell stocks a few precision resistor arrays, so your first thought
is correct. I bought a couple for my low distortion oscillator last
year.

These are pretty impressive, unique parts as far as I know.

<http://www.linear.com/product/LT5400>

0.2ppm/K ratio matching.

 

[John Devereux]

What are people paying for 0.01% resistors?

I wanted to reduce production tweaking so thought a few 0.01% 0603 or
0805 resistors would do the trick. Most suppliers have parts in the
many dollars range. Digikey do have a few Stackpole resistor values
for £0.57, about 85 cents.

I have the same requirement (let me know if you find anything!)

There seem to be Vishay "bulk metal foil" types north of $10.

Farnell have a few cheap "0.01%" resistors of unknown origin and
specification, very peculiar.

0603 size seem thin on the ground, they seem to be 0805 or above.

Rhopoint seem to specialise in precision resisistors. There are some
manufacturers of lower cost (than Vishay) bulk metal foil precision
resistors, as well as other types.

<http://www.rhopointcomponents.com/components/resistors.html>

It is very tempting here to use Larkins susumu types. Problem is if you
take the worst case datasheet limits on anything bar the metal foils,
they always seem incredibly pessimistic. But it seems ridiculous to put
say two $10 chip resistors on a board otherwise full of sub-$0.001 parts
and a $4 120MHz ARM.

 

[Spehro Pefhany]

The 30-cent Susumu thinfilms seem to stay below 10 PPM/K, and you can
get a few-PPM voltage reference for a couple of dollars.

Unfortunately, the ones that are 50x better cost about 50 times more.



Best regards,
Spehro Pefhany

 

[Spehro Pefhany]

True. I always thought that there's a market for matched resistors on
the same chunk of ceramic, like matched transistors. Either the market
isn't as big as I think, or it's just not possible to get good enough
matching to make it worthwhile.

They are out there.. matching "typical" +/-0.1ppm/°C, but not cheap
(~$20 pair).

It's strange when the resistor part of a BOM cost dwarfs that of the
semis.


Best regards,
Spehro Pefhany

 

[Spehro Pefhany]

I have the same requirement (let me know if you find anything!)

There seem to be Vishay "bulk metal foil" types north of $10.

Farnell have a few cheap "0.01%" resistors of unknown origin and
specification, very peculiar.

0603 size seem thin on the ground, they seem to be 0805 or above.

Rhopoint seem to specialise in precision resisistors. There are some
manufacturers of lower cost (than Vishay) bulk metal foil precision
resistors, as well as other types.

<http://www.rhopointcomponents.com/components/resistors.html>

It is very tempting here to use Larkins susumu types. Problem is if you
take the worst case datasheet limits on anything bar the metal foils,
they always seem incredibly pessimistic. But it seems ridiculous to put
say two $10 chip resistors on a board otherwise full of sub-$0.001 parts
and a $4 120MHz ARM.

It's not hard to have $1,000 worth of resistors on a board with semis
worth 5-10% of that.


Best regards,
Spehro Pefhany

 

[Spehro Pefhany]

These are pretty impressive, unique parts as far as I know.

<http://www.linear.com/product/LT5400>

0.2ppm/K ratio matching.

Thanks, that's interesting, and they're relatively inexpensive.

One of the dangers of using parametric search is that I'd missed those
due to their rather poor absolute tempco (25ppm/K), but that tempco
(typical 8ppm) is matched typically to within +/-2.5%.

Not quite as good as the Z-foil types (0.1ppm/K typical matching with
1:1 values), but typically better for ratios other than 1:1.


Best regards,
Spehro Pefhany

 

[John Devereux]

It's not hard to have $1,000 worth of resistors on a board with semis
worth 5-10% of that.

Simplifies the buyers job. Forget everything else, just shop around and
negotiate for a couple of resistor types.

I have recently (belatedly) realised that is already the case for ICs.

 

[John Devereux]

Thanks, that's interesting, and they're relatively inexpensive.

One of the dangers of using parametric search is that I'd missed those
due to their rather poor absolute tempco (25ppm/K), but that tempco
(typical 8ppm) is matched typically to within +/-2.5%.

Not quite as good as the Z-foil types (0.1ppm/K typical matching with
1:1 values), but typically better for ratios other than 1:1.

I wanted to make precision op-amp gain stages (single ended). The
LT5400s come in a handful of value combinations, so I got to modify my
"parallel-series resistor combination" program to work with the possible
networks of 4 resistors.

What combination of 4 resistors come closest to a specified ratio?

I sketched them and got 19 combinations (plus their reciprocals)


"a+b",
"a + (b+c)",
"(a+b) + c",
"a+ (b+c+d)",
"(a+b) + (c+d)",
"(a+b+c) + (d)",
"(a|b) + c",
"a + (b|c)",
"a + (b + (c|d))",
"(a+b) + (c|d)",
"a + ((b|c)+d)",
"(a+(b|c)) + d",
"(a|b) + (c+d)",
"((a|b)+c) + d",
"((a|b) + (c|d))",
"(a) + (b|(c+d))",
"(b|(a+c))+d",
"a + (b | (c+d)) FB from c-d",
"(b | (a+c)) + d FB from a-c"


"|" means "in parallel with"
"+" means "in series with"

Then a dumb C program to try them all.

Later extended to 2 and 3 gain stages.

 

[Phil Hobbs]

True. I always thought that there's a market for matched resistors on
the same chunk of ceramic, like matched transistors. Either the market
isn't as big as I think, or it's just not possible to get good enough
matching to make it worthwhile.

If you use a common-centroid layout, you can get pretty good ratio
stability with garden variety ones.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net

 

[Spehro Pefhany]

If you use a common-centroid layout, you can get pretty good ratio
stability with garden variety ones.

Cheers

Phil Hobbs

Pease on this stuff:-
http://www.datasheetarchive.com/files/national/htm/nsc03933.htm

Irritating that it's so difficult to even come close to the tempco of
a well-designed and well-constructed rack mount ratio transformer.
(< 1ppb/°C typically)

 

[Bill Sloman]

Pease on this stuff:-

http://www.datasheetarchive.com/files/national/htm/nsc03933.htm

Irritating that it's so difficult to even come close to the tempco of
a well-designed and well-constructed rack mount ratio transformer.
(< 1ppb/°C typically).

It would be more irritating if there were more applications where you couldexploit this. I've known about ratio transformers for about forty years now, but have yet to find an application where I could use one.

And they don't have to be rack-mounted to be that good. A ferrite core wound with Litz wire could be hooked up as a ratio transformer, though making the connections would be both fiddly and labour-intensive. I've long cherished the idea of winding one with round-to-flat cable. You couldn't put on all that many turns, but you wouldn't have to sort out the wires.

 

[John Devereux]

This is my version:

https://dl.dropboxusercontent.com/u/53724080/Parts/Resistors/R-pak_values.JPG

I think that misses at least

r+(r|(r+r))

to use my "notation".

I think I have one somewhere for voltage dividers, too.

Yes that is what I needed, otherwise you don't get to take advantage of
the ratio matching of an integrated network. (Or of resistors from the
same reel possibly?).

Also some of the LT5400 parts have two values in the package, so you
need to actually get the formulae so you can plug in the values.

 

[Phil Hobbs]

It would be more irritating if there were more applications where you could exploit this. I've known about ratio transformers for about forty years now, but have yet to find an application where I could use one.

And they don't have to be rack-mounted to be that good. A ferrite core wound with Litz wire could be hooked up as a ratio transformer, though making the connections would be both fiddly and labour-intensive. I've long cherished the idea of winding one with round-to-flat cable. You couldn't put on all that many turns, but you wouldn't have to sort out the wires.

Well, they're used all the time in capacitive gauges.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net

 

[John Devereux]

I think that's my 1.666 case.

No no, in my sketch the bottom resistor is moved over underneath the
right hand one!

Oh all right, sorry. I plead a cold coming on.

 

[Richard Henry]

What are people paying for 0.01% resistors?



I wanted to reduce production tweaking so thought a few 0.01% 0603 or

0805 resistors would do the trick. Most suppliers have parts in the

many dollars range. Digikey do have a few Stackpole resistor values

for £0.57, about 85 cents.

My question is why? Even brewing beer does not take that kind of precision..