Maker Pro
Maker Pro

Resistor distribution

G

George Herold

Jan 1, 1970
0
I was killing some time last night before going to my first maker
space meeting, watching some Dave videos.

http://www.eevblog.com/2011/11/08/eevblog-215-gaussian-resistors/
http://www.eevblog.com/2011/11/14/eevblog-216-gaussian-resistor-redux/

In the first video Dave finds that some Philips resistors have a ~
+/-0.5% distribution right around the ‘correct’ value.

In the second he finds that some cheaper (Xicon?) also have a 0.5%
‘spread’, but the average is a bit (~0.35%) lower than nominal.

This raises a bunch of interesting questions.
Do any resistor makers publish this sort of data?
Does buying resistors from a ‘better’ manufacturer lead to resistors
with a better mean.
I use mostly cheap Xicon 1%ers. A few times I’ve gone hunting through
the parts bin with an ohmmeter trying to find some particular value.
Though I don’t have any data I did strike me that the average seemed a
bit low. (When looking for a 10.0k ohm I found many more 9.9X k ohms
than 10.0X k ohms.)

So does anyone know how 1% resistors are made? I find it hard to
believe that they trim each one.
(Do they trim the 0.1% ers?)

George H.
 
G

George Herold

Jan 1, 1970
0
Most 1% or better surface-mount resistors are laser trimmed. You can check them
under magnification and see the trim cuts. If it's done very quickly, themean
could well be a bit off for any given reel.

OK Thanks, I guess I can believe the trimming, but do they measure
each one?

George H.
 
G

George Herold

Jan 1, 1970
0
   How can they trim them if they don't?

I could imagine that they measure a few to set the laser cut and then
only do spot checks. But I clearly have no clue how it's done!
It sorta blows my mind to think someone can make it, measure it, trim
it, and then sell it to me for under a penny.

George H.
 
G

George Herold

Jan 1, 1970
0
Sure; they laser cut until the value is right. It's often an L cut; the cross
axis is coarse trim, and the longitudinal cut is final fine trim.

Well I scrapped the silk screen off the surface of a few and an L cut
was exactly what I saw.

I think there is a Vishay resistor plant somewhere near me... I wonder
if I could get a tour?

George H.
 
I could imagine that they measure a few to set the laser cut and then
only do spot checks. But I clearly have no clue how it's done!
It sorta blows my mind to think someone can make it, measure it, trim
it, and then sell it to me for under a penny.

More like six to a penny.
 
C

Cydrome Leader

Jan 1, 1970
0
In said:
Have you ever seen an operation like DigiKey? ...or Amazon? ...or
NewEgg? Do you really think they have people chasing parts?

amazon sure has people doing the picking.
 
R

Robert Baer

Jan 1, 1970
0
Michael said:
By computer? Yes. I haven't bought electronics parts in person or
over the phone in over 20 years. :)
Back in the "good old daze" '70s to '90s) 5% carbon comps ran
typically 2% high with a rather tight spread - maybe +/- 1% (to +/- 2%
at most).
 
G

George Herold

Jan 1, 1970
0
A laser trimmer must be cool to see. The parts are flying through some
handler/measuring gadget and being laser blasted at some number per
second. Time matters, so there is probably some empirical
second-guessing of subsequent thermal and other creep-back.

--

John Larkin         Highland Technology, Inc

jlarkin at highlandtechnology dot comhttp://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- Hide quoted text -

- Show quoted text -

Hey, I was scraping the blue coating off some through hole MF's
Do they laser trim those too, or something else?
I couldn't see much, but when I was done, 10k had become 10.5k.
(I could do in circuit, tweaking with an exacto knife.)

George H.
 
G

George Herold

Jan 1, 1970
0
   Back in the "good old daze" '70s to '90s) 5% carbon comps ran
typically 2% high with a rather tight spread - maybe +/- 1% (to +/- 2%
at most).- Hide quoted text -

- Show quoted text -

Interesting, Don't CC's have a terrible tempco. (We used them as low
temp sensors, maybe better near RT.)
 
S

Spehro Pefhany

Jan 1, 1970
0
Michael A. Terrell a écrit :

Selling too?

I recall one of the local disties back years ago decided to stop
carrying resistors-- their bean counters figured it was something like
80% of their costs for 20% of their volume, and they wanted to
concentrate on high-markup semiconductors. No longer around, natch.


Best regards,
Spehro Pefhany
 
G

George Herold

Jan 1, 1970
0
Spiral grinding of cheap leaded resistors:

http://www.mfrelectronics.com/cut.jpg

Neat, Thanks. (I take it those are spitting out 1% metal film
resistors.)
If I was them, I wouldn't be talking about the mercury contacts.

Hmm, how else are you going to make contact to the spinning lead?

(there's some indium gallium... ? tin? alloy that melts below 100C.)

Got it, Field's metal. (I assume named after our own John
Fields :^)

George H.
 
G

George Herold

Jan 1, 1970
0
Neat, Thanks.  (I take it those are spitting out 1% metal film
resistors.)




Hmm, how else are you going to make contact to the spinning lead?

(there's some indium gallium... ? tin? alloy that melts below 100C.)

Got it, Field's metal.  (I assume named after our own John
Fields :^)
oops.. nope it is indium gallium and tin... Galinstan.

Geo
 
R

Robert Baer

Jan 1, 1970
0
George said:
Interesting, Don't CC's have a terrible tempco. (We used them as low
temp sensors, maybe better near RT.)
The more modern Ohmite carbon comps were about 5 times better that
the really old ones that had the leads spiral wrapped around the ends..
The 47 to 100 ohm Ohmites were very nice as fuses / overload
indicators when placed in the cathode of series power regulator tubes.
Discoloration "level" would indicate a certain amount of overload,
and next cracking to open, and then carbonizaton and finally "busted black".
 
P

Phil Allison

Jan 1, 1970
0
"George Herold"

In the first video Dave finds that some Philips resistors have a ~
+/-0.5% distribution right around the ‘correct’ value.

In the second he finds that some cheaper (Xicon?) also have a 0.5%
‘spread’, but the average is a bit (~0.35%) lower than nominal.

This raises a bunch of interesting questions.
Do any resistor makers publish this sort of data?


** The tolerance percentage includes a nominal service life - so a nominal
1% tolerance part is usually much tighter when new. That is all that is
being claimed.

Does buying resistors from a ‘better’ manufacturer lead to resistors
with a better mean.

** Pointless even reseching it as it would be a moving target.


So does anyone know how 1% resistors are made? I find it hard to
believe that they trim each one.


** That spiral cut does not get there by itself and the cutting machine
stops when the value is right.

Tighter tolerances ( than say 1%) requires slower operation and or the use
of a laser - so the parts cost more.


..... Phil
 
J

Jasen Betts

Jan 1, 1970
0
Neat, Thanks. (I take it those are spitting out 1% metal film
resistors.)

carbon film also has a spiral cut.
Hmm, how else are you going to make contact to the spinning lead?

how about carbon, or copper, brushes: two each end so you can make a
kelvin measurement.

Whatever you use cutting is going to cause localised heating so you'll
have thermocouple effects messing with your measurements, for small
resistors you'll probably have to use AC to counter that.
 
Top