Page 178 - Troubleshooting Analog Circuits

Letters to Bob I65

stick often proves handy. I’ve got a whole box of capacitors and resistors mounted
this way, which makes for a very efficient and portable trimming kit.

John Ardizzoni
M/A-COM
Lowell, MA

Dear Mr. Ardizzoni:

And if you look in the right magazines, the ones catering to RF designers, you may
find advertisements for similar “tuning sticks” that are available for purchase in
values from 0.1 pF to lo00 pF. You can buy these ready-made for about $50 for a set
of 20 from: American Technical Ceramics Corp., 1 Norden Lane, Huntington Station
NY 11746-2102 (516) 271-9600. Obviously they think your idea is an excellent
one-they believe their customers will pay for ready-made versions, with a neat
plastic shaft instead of a Popsicle stick. But when you make your own, you can make
up any value you want, in just a few minutes’ time.

RAP

Dear Bob:

I use Tempilaq temperature-indicating paint from Tempi1 (South Plainfield, NJ) to
monitor the temperature of power devices in metal cans and plastic packages, such as
TO-220 devices. I use the paint both in development and in units shipped to the field.
The substance comes in various temperature grades. I anoint the top of each device
with a range of paints. When and if a device heats up, each temperature-grade paint
melts and fuses at a specific temperature, permanently altering the substance’s ap-
pearance.
I also use the paint to ascertain the relative effectiveness of various thermally
“conducting,” insulating washers. These washers isolate a power device from its heat
sink. I measure the time for the paste to change state after applying power to a device.
I discovered that silicon-based, elastomeric washers didn’t work at all with TO-
220 cases. The mounting screw of these cases is off center. One end of the case is
tight, and the other is loose, which causes poor heat flow. I don’t understand why
these washers are even manufactured. Didn’t the makers try one out before deciding
to offer them?
Washers aren’t the only problem. One sample TO-220 device got hot so quickly
that I performed a post mortem on it. I opened the package and pried the chip loose
and observed that only about 15% of the die had been bonded to the substrate. Since
that experience, I have steered clear of that particular manufacturer.
Bill Sturgeon
Sturgeon Engineering Co.
Petrollia. CA