Laser Trim Into Ceramic (Chip Resistor)

Resistors are made from a ceramic carbonaceous or metallic powder paste, which is dried and then kiln fired.

Way back in the 1930's, it was worked out cheaper to make carbon resistors approximate in value, then measure and trim to the correct value.

Of course in those early days, the adjustment was a wire slid along the resistor body, then cemented into place when the correct value was located, then colour coded.

Later carbon resistors were circumferentially from one end spiral grooved on a lathe, to the correct value, then covered with a hard enamel, and the value colour coded.

Precision resistors were printed with the value.

So, in the interests of efficient manufacture, today's chip components are laser trimmed, until correct value is established.

Thus the answer to your question is: Yes.

Kind Regards....

If the answer is yes, will it weaken it mechanical properties and easy to chip into 2 pieces during surface mount pick and place? Those L shape trim mark started from edge of its substrate, and certainly will act as origin of crack line. How to ensure its reliability in lon run?

Hi TNT,

The manufacturer of the resistors, should have a great deal of data on how well the resistors behave in the board manufacturing process. They will probably even have suggestions for how to mount, reflow profile, etc. Soooooo... contact the manufacturer!

Sparkstation, thanks for the great info!

Best Regards,

Mike

Hi TNT,

when you look at the SMD`s dimensions you see that a 0402 body (one of the smallest ones) has a thickness of 350µm, the largest ones (>= 2010 size) are 550µm thick.

Thickfilm hybrids, where the resistors are printed with decadically graded pastes fom Ohm to MegOhm, mostly have a thickness of about 500 to 1000µm.

In all these cases the pastes thickness which forms the resistor is about 10 to 15 µm which corresponds to the cut`s depth.

The resistors surface is or (in case of hybrids) has to be glass passivated. We cover our hybrids with a lo-temp glass passivation which melts together with the paste during the trimming process so that you can see the L-shape, indeed, but the surface is sealed and free of cracks provided the process parameters are correct and the depth is not more than 15µm.

Sometimes, for special secure applications, there is an additional requirement of sealing the surface with an epoxid raisin but the main reason for this is to seal the solder connections whose "electrically open" surface is much greater as the thin L-cut.
Regards Uwe

I am understanding your statement as follows:

The chip resistor has been printed onto an alumina substrate and trimmed. The substrate is adhered to the main circuit board? If this is correct, then the penetration of the laser beam through the resistor into the ceramic 9 to 11 microns is too deep. Although the resistor is within tolerance now, the high power used could cause microcracks in the resistor which will cause the resistance to drift with time.

You should ask your supplier to insure that the laser is in focus, the power and frequency settings are correct, and to look in the program to make sure the bite size is correct. Going further back into the process, the supplier should have tight controls on the fired thickness of the resistor.

No, It's not normal to have laser cut deep into ceramic substatrate. I assume that these are thick film resistors; therefore, more laser power was used to cut through the resistors. However, the laser should go into the ceramic substrate no more the 2 to 3 microns. Laser marks of 10 microns depth into the ceramic is the result of too much laser power was used. Too much power during trimming can cause problems during processing or later on, such as too much micro crackings, rough edges along the cuts, etc... and this can can change the values of the resistors dues to drifting..

Khan Bui

khanb555@hotmail.com