Resistor Measuring Jig

[sodderboy]

I needed to inspect 3000 resistors for 3% tolerance, so I made v2.0 of my resistor jig. Longer, inner-stiffened springs for more leverage, a longer base for stability, and a Belden jacket handle. In operation, it only needs to move between 10 and 40 degrees, so measuring goes very quickly.

Mike

 

[Script]

Very useful

 

[trobbins]

Well done, as it is a pain to use clip leads.

How do you know your meter is within x.x% for the nominal resistor value?

Are you batching the resistors for tolerance bands as well ?

 

[Evertide]

Well done, as it is a pain to use clip leads.

How do you know your meter is within x.x% for the nominal resistor value?

Are you batching the resistors for tolerance bands as well ?

With a 10 % meter you will just select to 0,1 % of the reading , you will be within 1 % of the desired value between all selected resistors ... ( even with 20 % resistors ... )

 

[trobbins]

Ta, so it is 3% of tolerance of matching, not absolute value.

 

[RuudNL]

I think in that case I would have bought 1% resistors...
(Would have saved me a lot of work!)

 

[trobbins]

Fluke 8060 could be close to cal confidence, which is down to 0.1% for 8060A and 1kohm.

 

[pinebox]

Clever. Could you use two parallel metal plates and rest the leads across them or is the pressure of the spring necessary for the accuracy?

 

[sodderboy]

You can test sufficient pressure yourself. Drop a lo value rizzy in the holes of your meter and measure it sitting there, or sitting on top of stabilized probes. Repeat 9 times and see how different the measurements can be, if the reading is even stable. The spring tension removes any doubt regarding surface oxides on the components. I could use smaller diameter springs and get similar results, but I had 1/4" plastic round stock so designed around that. The other idea was gluing big clip leads on wood and a popsicle stick across, but I figured there would be too much chewing on the leads.
Regarding the component under test, I need carbon composite, non-inductive resistors for RF (DC to 20 GHz) termination, 51Ω ±5%. I find that composite can drift over time, in product or sitting in parts, so I screen for ±3% and save time re-working out-of-tolerance parts. The techs are measuring on a calibrated Fluke 8060A in REL, and the numbers are easy, 49.5 to 52.5.
Mike

 

[jkc78]

Photo shows my resistor testing jig. It uses 2 pieces of brass strip, with a narrow notch cut in each one. Drop the resistor into the notches, and if necessary rotate slightly to cut through any corrosion on the resistor leads. Faster than springs or spring clips.

 

[JohnRoberts]

Something like a "zero insertion force socket" provides good electrical contact and is designed for lots of insertions.

JR

 

[sodderboy]

I was thinking ZIF, two,side-by-side with the plastic removed and the levers joined.
The brass jig is good, but I would have to put pressure, either with gloved fingers or a piece of neoprene or something. Laying a resistor on pieces of metal with no pressure is not a complete connection in my experience. You would be surprised if you lay tcw on that jig hooked to a milliohmmeter. Not necessary for rock n roll, but I need tighter readings for what I'm doing.
Mike

 

[jkc78]

When I built this I wasn't worried about a few milliohms. I have a large collection of NOS carbon composition resistors that I've had for several decades, and I decided that I would sort out the ones that were still within 5%, and trash the rest. For anything over 10 ohms, I figured this jig was good enough for rough sorting. But your comment got me wondering how good or bad this really was. Photos show a 4 wire resistance measurement of 18 gauge wire resting in the jig: 2 milliohms. Better than I expected, considering that I was *not* pressing on it. Normally when I'm using this I'm doing really fast sorting, so I never let go of the resistor -- just pick it up, press it into the jig, get a quick measurement, and toss it into the appropriate bin. Yeah, if I was trying to match resistors below 10 ohms to 0.1%, I'd want something better, like a jig that's 4 wire all the way to the resistor lead.

 

[JohnRoberts]

The optimal solution depends on the quantity and precision required. Back in the 70s I tested IC's 100% because industry quality wasn't that good, and I wanted to be able to tell my kit customers that I knew the ICs were working when I shipped them. I found zero insertion force sockets very useful. Since then IC processes have gotten better.

JR