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. ;)
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.
--
Tim Wescott
Wescott Design Services
http://www.wescottdesign.com
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
--
"it's the network..." "The Journey is the reward"
speff@interlog.com Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com
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.
--
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
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.
--
Tim Wescott
Control system and signal processing consulting
www.wescottdesign.com
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.
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.
Unfortunately, the ones that are 50x better cost about 50 times more.
Best regards, Spehro Pefhany
--
"it's the network..." "The Journey is the reward"
speff@interlog.com Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com
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
--
"it's the network..." "The Journey is the reward"
speff@interlog.com Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com
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
--
"it's the network..." "The Journey is the reward"
speff@interlog.com Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com
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
--
"it's the network..." "The Journey is the reward"
speff@interlog.com Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com
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"
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
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)
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.