Test Equipment Anonymous

It arrived this morning...

Remember the 80s? Pop groups with more hair gel than talent? Something else taste forgot was the styling of test equipment. The 485 is in a plasticky dark brown case with a significant hood over its 4 1/2 digit LCD so you need to prop it up on its wibbly-wobbly bail arm to read the display. It looks dreadful. Oh, and it has a captive mains lead (brown). However...

I warmed it up and warmed up the Keithley 6221 current source. Having zeroed the 485, I tested at 1.9 of each scale (the meter reads to 1.9999), so that was 1.9nA, multiplying by ten up to 1.9mA. On all but the lowest range, it was within 0.1% of stated current from the (out-of-calibration) 6221. There was a 0.06% systematic discrepancy, and if that had been tweaked out, all ranges would have been within 0.1%. It's a bench meter, so it has its input connector on the front. And it's a BNC, so you can actually afford to connect to it. More to the point, you can afford to make dedicated test jigs that hang off the front (minimising capacitance and noise gain). Round the back, it has a pair of 4mm sockets on standard spacing that allow you to connect electrometer output to an oscilloscope to check for hum. But you then need to put a thick book under the back to stop your BNC and adapter from lifting one side of the instrument. Monitoring electrometer output, it suffers less hum from its own mains transformer than a Keithley 6517B. All in all, it's a very good meter. Nice accuracy, shame about the face.

I am told that it isn't really a Keithley but a design inherited from when Keithley bought Racal-Dana. This seems likely because if you compare a K600A (valve electrometer) with a K6517B (silicon, and digits everywhere), some fundamental ideas about range switching are carried through. But they're not in the 485.

I've just had the covers off. Hmmm. There's the main board, complete with "wavechange" switches and mechanical interlock, then the (optional) GPIB board above it. The rear panel is not a separate part, but formed by the top and bottom halves of the case; so you can forget about neatly adding extra connectors (such as a BNC for electrometer out, or a BNC for GPIB trigger out). Inside the folded aluminium screening can is the sensitive analogue stuff. The front-end is wired between PTFE-insulated posts and there are three relays, presumably for the most sensitive ranges (haven't studied the diagrams fully yet).

The ribbon cable to the display passes over the input connector and there's a gap between its screen and the main screen, allowing the input wire to be seen. I expect some of the high frequency interference I can see on the electrometer output is due to that gap. A bit of earthed copper tape and lead dress might help.

Did Keithley buy any of RACAL-DANA? I know that case style, it's from Keithley's 179, 177 etc, except they were not brown (except the 175, which was, maybe others too). RACAL-DANA sold rebadged 179s as the 4002, so there were clearly some business relationship in the past. I had assumed Thales got all the many chunks of RACAL.

EC8010 wrote: ↑Wed Apr 22, 2026 2:22 pm It arrived this morning...

Remember the 80s? Pop groups with more hair gel than talent? Something else taste forgot was the styling of test equipment. The 485 is in a plasticky dark brown case with a significant hood over its 4 1/2 digit LCD so you need to prop it up on its wibbly-wobbly bail arm to read the display. It looks dreadful. Oh, and it has a captive mains lead (brown). However...

Glam Rock. Sweet, Slade, Mud and many more. Plus Gary Glitter and Jimmy Savile, but those are no longer mentioned in polite society.

TBH it looks much the same as many other pieces of TE from that era, from Fluke, Solartron etc. Some of the Solartron stuff was particularly ropey, e.g. the 7045, with its awkward to stack case and those internal plastic pillars, which break easily. Some of the best and most durable items came from Racal-Dana. They had extruded aluminium sides and substantial front and rear panels. This made heat management easier. The top and bottom were sheet metal, even the bail arms aren't too bad.

Despite the distressing styling, I'm pleased to hear it's a good instrument.

synx508 wrote: ↑Wed Apr 22, 2026 3:39 pm Did Keithley buy any of RACAL-DANA?

That's what I was told, and it seems to be supported by the non-Keithley electrometer range switching. But my main point is that it's a nice meter, and if you only need to measure down to pA, it's arguably better than a 6517. It's certainly way cheaper. Shame it looks so horrible, though.

Another not-to-be-mentioned would be Rolf Harris.

I decided to deal with the screen properly; not copper tape. I folded up a new screen 4mm longer than the original. The new screen snugs round the input connector and up to the electrometer screen. It didn't make any difference to the digital clatter at electrometer output but... It changed the electrometer offset. And in doing that and tweaking to zero when warmed up, the 2nA range now meets its 0.1% accuracy specification at 1.9nA. I've seen before with electrometers that you have to eliminate hum and interference at source rather than assuming that post-processing can remove it. I suspect what it comes down to is that electrometers aren't terribly linear, so high amplitude interference can be converted into a DC component. But the main thing is that twenty minutes of cutting, filing, and folding have not only made it look right internally, but it now meets its original specifications despite being forty years old.

The 485 picoammeter now has the measurement jig it needs. The nice thing about the 485 is that its input is on a BNC, and that BNC is on the front panel. Yesterday, I found my sole remaining chassis mount male BNC and fitted it to a 150 x 150 x 50 shortbread tin that originally contained Waitrose Duchy organic shortbread. Avoiding an intevening lead minimises shunt capacitance and electrometer noise gain, reducing uncertainties. The tin's a bit big and looks slightly odd hanging off the 485, but it gives me plenty of room to test large components. The tin screens the measurement from hum and air currents, and keeps my fingers away from the nasty voltages. Also fitted to the tin is a BNC for incoming high voltage. I'm ultimately intending to test at high voltages, perhaps even to 5kV. (Yes, I know, I should really have fitted an MHV connector, but I don't have any.) With the combination of a 5kV supply and a picoammeter, what could possibly go wrong? The supply is current limited by a 10 giga ohm 6kV resistor and the input of the electrometer is protected by a shunt HLMP3401 amber LED (forward biased). I have previously tested these LEDs and their I0 is of the order of aA (I forget the exact number, but it's small). Final holes were bashed this morning and components wired in place. Piocammeter was zeroed.

A subminiature Soviet phenolic tag strip was tested for leakage current between adjacent tags with 100V applied. The 485 was triggered and logged every second over GPIB and the results taken to a spreadsheet where leakage resistance was calculated and plotted. I find it amazing that it is possible to clearly define the deficiencies of a bit of insulation between two adjacent tags. The result clearly shows serious dielectric absorption, which is to be expected from phenolic. All in all, I'm very pleased with this new (to me) instrument. Even if it does look horrible.

It's worth looking out for BNC sockets, tag strips, offsets and other things like that, in the boxes of assorted junk at swapmeets. Sometimes they turn up real bargains.

Indeed they do and I always check the junk boxes. That (used) BNC was from a junk box and I think I paid 50p for it at the recent Audiojumble. It's the large square flange variety intended for those horribly expensive (Suhner?) blue anodised in-line boxes.

And here's a picture of the 485 and test jig in all its gory glory.

Zenith wrote: ↑Tue May 12, 2026 12:04 pm It's worth looking out for BNC sockets, tag strips, offsets and other things like that, in the boxes of assorted junk at swapmeets.

I've never seen offsets in junk boxes. How do you spot them? Are they labelled "1.6mV offset" for example? I think you must mean something different to what I'm thinking. Please explain.

My mistake. Standoffs or spacers. Little plastic or metal things, usually tapped or with a threaded section. Used to mount circuit boards, tag strips etc. Sold by Farnell etc, but often surprisingly expensive.

Zenith wrote: ↑Wed May 13, 2026 8:27 am My mistake. Standoffs or spacers. Little plastic or metal things, usually tapped or with a threaded section. Used to mount circuit boards, tag strips etc. Sold by Farnell etc, but often surprisingly expensive.

Gotcha. If you think spacers are expensive, try looking up the price of stand-offs made by Jacksons. I was alerted to the history of Jacksons recently and found that their products can still be bought if you have a really thick wallet. There's a single Jacksons Type H-S 4 BR 6.3kV stand-off in the shortbread tin https://www.mainline-group.com/collecti ... insulators.

Just flipping through CPC online, it seems a lot of standoffs are very expensive for what they are. The Jackson Bros tag standoffs look very high quality. The kind of thing you'll find in military kit from the 50s/60s. Tag strips were a mainstay of electronics in the valve era. In the last days of Maplins they sold a limited range. Their prices were rather steep. I never bought any as I had a stock picked up at rallies. I don't use that many, but when I want one, I don't want to have to go to suppliers of bits for valve hi-fi.

You've probably seen this.

https://en.wikipedia.org/wiki/Jackson_Brothers

At the bottom of the page there are a few links.

It looks as if Jackson Bros was bought out, asset stripped and went bust. Mainline bought finished stock, parts, tooling and drawings and has restarted production.

They are indeed good quality. The insulator is ceramic (aluminium oxide), but if you want to big-up your product, you refer to "sapphire" insulators. I had not seen the Wikipedia page, but I was aware of the final demise. I don't think production has restarted, just that a load of parts were rescued from the skip. A skip filled to any depth with ceramic stand-offs would be very valuable. If you look at the Mainline page, you will see that they have a section Rally's (greengrocer's apostrophe) where they say they will be at a 2016 rally, indicating that the webpage hasn't been properly updated for a decade.

There are a few links at the bottom of that Wikipedia page and a couple talk about the Mainline connection, but there are no recent dates. It says they rescued a load of stock and tooling. It was in the Mainline HQ and warehouse in Leicester. In 2005 there was a fire and much of the stock was destroyed, but the tooling was at subcontractors. Products were assembled in Croydon.

https://g3zpf.raota.org/articles/JacksonBros.pdf

It's hard to know whether they still produce some items or whether they are simply selling stock.

I've a memory of most of the Jackson Bros stock being destroyed in the fire at Electrovalue after the move from a shop in Englefield Green to larger premises in Egham. After that Mainline bought what was left of Jackson Bros, tooling and some stock that had escaped destruction or something. I've forgotten what Mainline's owner, John Higgins looks like, we did vaguely know each other years ago, but Mainline's still on ebay with at least one account (IIRC anonalouise and some number is also Mainline).