keithley 414S

Рет қаралды 271

Күн бұрын

This Keithley 414S Picoammeter was built in the late 60's/early 70's and was in production until the early 80's. It can measure 10mA FS, to .1nA FS in 17 ranges. The .1nA range is also 100 picoamp range and the meter is divided into 100 little steps, each a picoamp (a millionth of a millionth of an Amp!).
When it arrived, I saw the cord was cut off of the unit, but could find nothing else that was obviously defective. I replaced the cord, plugged it in and turned it on, only to blow my GFI.
I reset the GFI and tried again, with the same response. There was only the primary side of the transformer and a fuse/power switch on the electrical inlet. I checked for shorts, even high resistance ones, but found nothing. I looked at the schematic and saw that the transformer had a shield wire on the primary and figured it was possible that enough current collected there to set off the GFI. I know, with some experience, that some of the older equipment has been known to trigger GFIs. I connected the unit to my isolation transformer and turned it on, and everything worked fine. No blown fuse, no problems noticeable anywhere. So, I started looking for a small isolation transformer, found one and ordered it.
When the isolation transformer arrived, I decided how to best hook it up into the circuit. I kept the fuse and power switch on the primary input side of the isolation transformer and hooked the secondary to the existing transformer input.
That all worked fine! Now to test it’s operation and ranges.
Pulling out my Keithley 225 current source and connecting it to the input, I began to check the zero and span on each of the ranges. I found that the 414S switches were a bit erratic, so I shut everything down and gave them a good cleaning. I went back to checking the ranges. I found about half were out of spec (+/-) 2 to 4%, placing the span on the range either low, or high, by 2 to 4 least significant digits on the meter. Altogether, there were 7 range resistors that were out of tolerance. I determined which of the resistors I had on hand, and then replaced the range resistors in the unit with them, each time finding the results to be satisfactory. I then ordered the remainder of the resistors, except for the lowest two ranges, which required a 1% 10Gig ohm resistor (these cost about $100 each). It was at this point I realized that I had no satisfactory means to measure the 10G ohm resistor on my bench. Something I needed to remedy! (See “Extreme Resistance Instrument”.) When the resistors showed up to the house, I installed them in place of the defective resistors (on the appropriate ranges) and again saw good, satisfactory results. Finally I got down the .1na range ( 100 picoamps). I did not want to spend over $100 on a new 10G, 1% resistor. I noted that the reading was low about 3% of the meter span. I calculated that 3% of 10G would be about 300M ohm. But, am I 300M ohm low, or high in resistance on the 10G ohm? I looked at the schematic and saw that there was a “op amp” circuit that fed off the voltage drop across the input resistors R123 and R124. In the 10G (.1na) it was only R124. R123, R124 and R125 all measured good. In order to “push” a tiny bit more current through R124 and get the 3% greater reading, I would need to increase the resistance of R122 (10G) by 300M ohm. I found a suitable 300M ohm resistor on Ebay for about $8 shipped and I ordered it.
Getting to R122 is difficult, unless you remove the suppression range switch assembly. That assembly is held in by a wire, an aluminum wire rod, and a hex nut under the knob.
With those removed, the suppression assembly slid right out and I had (mostly) unobstructed access to R122. I cut one end of R122, and pulled it outward somewhat. I decided to first put a shrink tubing around R122, to act as an additional insulator to the 300M resistor I was installing in series with it. Then I made little wire extensions to the 300M ohm resistor leads, which I coated with shrink tubing as well, leaving both ends protruding from out of one end of the resistor. Then I slid a large tube shrink tubing around R122 and the 300M resistor, leaving the cut end of R122 and both ends of the 300M resistor protruding. I connected one lead of the 300M resistor to the open end of R122, and pulled the connection to the side, getting it away from the other 300M lead. Then I connected the remaining 300M lead to the switch wafer-connector that the R122 used to connect to. I heated the shrink tubing and it was done. Finally, I reinstalled the suppression switch unit and buttoned up the case.
I plugged the unit in, fed a .1nA signal into it, and dialed the range switch down to the last two range settings of .3nA and .1nA. The meter dial responded satisfactorily now! At this most sensitive level, I did notice that the unit is very sensitive to any movement of my body nearby, as I took the readings.