4
\$\begingroup\$

We have specified some 0805 resistors to be 0.2ppm/C. On some boards/batches these are green. On others more "normal" looking. We specified Vishay. Is there some color code that indicates ultra-precision packages?

We have noticed different temperature sensitivity between the boards

\$\endgroup\$
  • 2
    \$\begingroup\$ What do the data sheets say? \$\endgroup\$ – Andy aka May 12 '17 at 12:25
  • 2
    \$\begingroup\$ Colours are whatever the manufacturer wants them to be like. Look at the datasheet of the parts you bought. \$\endgroup\$ – PlasmaHH May 12 '17 at 12:25
  • \$\begingroup\$ I don't know about markings for thermal coefficients, but some 1% SMT resistors are marked with a 3-digit alphanumeric letter code (EIA-96 marking) to save a digit (2 letters -> 3 sig digs) + exponent. \$\endgroup\$ – Nick T May 12 '17 at 17:01
3
\$\begingroup\$

There is no standard color code for the temperature coefficient of a resistor. A particular manufacturer may have a color code across their product line, or within a part of their product line. Look at the datasheets and/or consult the manufacturer directly.

Most likely the color is just a by-product of the manufacturing process, and not something deliberately selected by the manufacturer.

\$\endgroup\$
  • \$\begingroup\$ The problem is that it is very difficult to check temperature coefficient of the resistors when the manufacture of the board is farmed out. We just have to trust the honesty/competence of the assemblers. It seems in some cases this may have been misplaced. \$\endgroup\$ – Dirk Bruere May 12 '17 at 12:29
  • \$\begingroup\$ @Dirk: That's life. Manufacturers can't possibly print every spec on a tiny part. Note that 0805 and smaller capacitors usually come without any markings at all. You're already trusting your assembly house to get those right, and various other parts too. This is really nothing different or unusual here. \$\endgroup\$ – Olin Lathrop May 12 '17 at 12:31
  • \$\begingroup\$ It is given the price differential we are paying - about a factor of 1000. \$\endgroup\$ – Dirk Bruere May 12 '17 at 12:34
  • \$\begingroup\$ @DirkBruere: There is so much more to parts that is not printed, sepcially in smd. Tempco, tolerance and then some stuff for useful in RF applications... You are lucky if there is enough space to print anything at all, ever had 0201 packages? But really it doesn't matter much, if your manufacturer/supplied screws you over, they can as well screw you over with fake printings. \$\endgroup\$ – PlasmaHH May 12 '17 at 12:39
  • \$\begingroup\$ I get the feeling it was more like "Why do these guys need that really expensive resistor? - they won't notice if we just put this one in". And they are right. Most of the time. \$\endgroup\$ – Dirk Bruere May 12 '17 at 12:45
2
\$\begingroup\$

While all of the Vishay $$ ultra-precision 'Z' metal foil resistors I can recall buying are either robin's egg blue or molded, I can't state for sure that there are none that are not black, beige or whatever. Typically details as to appearance may not be precisely specified on the datasheet.

Your best best is probably to contact the maker (Vishay) to ask if the parts are genuine compared to the purported part number and get your PCBA supplier to confirm the exact part numbers used if you didn't specify the EXACT part number (a practice I would highly suggest you follow in future-- the next best thing to what I like doing- which is supply the critical parts for assembly plus ~5% for them to lose- costs more for sure, but there is much less question).

Measuring tempco has to be done very carefully to avoid errors from thermal EMFs. I would suggest carefully removing the part (the heated tweezer type of iron works well) and putting it in a 4-wire setup in a proper commercial environmental chamber and vary the temperature between room temperature and the two extremes specified on the datasheet such as -25°C and 75°C. The 4-wire ohmmeter (eg. 34401A must be outside of the chamber, of course, and allow sufficient warm-up and maintain temperature +/-1°C or better). If you just stick a heat gun on the resistor in open air your chances of getting a useful measurement are about zero.

Note that low value resistors have a tempco that is affected by the solder and terminations. All may be affected by PCB forces.

\$\endgroup\$
  • 2
    \$\begingroup\$ We intend to get some known genuine parts and put them on the boards that currently have too much temperature drift. If the drift ceases... \$\endgroup\$ – Dirk Bruere May 12 '17 at 13:51
  • \$\begingroup\$ Sounds like a plan. \$\endgroup\$ – Spehro Pefhany May 12 '17 at 13:51
1
\$\begingroup\$

In addition to what Spehro says, you need to consider your source. A question which presents itself is, where did you get your resistors? If you got them from either Vishay or a reputable distributor, the color difference probably doesn't matter. If, on the other hand, you bought from a parts broker in order to save time, there is an unnervingly high chance that you got some counterfeit parts and the color difference is important. I've been involved in projects using aerospace rated (and priced) parts which got burned despite the documentation, and the sort of price differential you mention is a big invitation to the unscrupulous.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.