Boonton RF Probes; Diode Replacement

[Bazz4CQJ]

In the past, there has been some discussion of the "magic diodes" apparently used in the Boonton and some other makes of RF probes.

This discussion on their replacement is interesting;

https://antiqueradios.com/forums/vie...p?f=8&t=393043.

The Boonton voltmeter used some kind of mechanical chopper on the output from the probe; is there some nice semi-conductor device that could replace that now?

B

[orbanp1]

Hi Bazz,

In the "Communication Quarterly" there is an article about replacing the mechanical chopper with a chopper stabilized low offset opamp:
https://chrisgrossman.com/manuals/bo...Voltmeters.pdf

It is basically a total rebuild of the equipment.

Regards, Peter

[Radio Wrangler]

I think the mechanical chopper chops light onto LDR cells.

A minimalist workaround would be an astable multivibrator flashing LEDs over those same LDRs. I think a second chopper is used as a synchronous rectifier so its LEDs go in series with the input chopper LEDs. Nothing gets touched in the actual measurement circuitry and high efficiency LEDs can be harvested from dead LED light bulbs.

David

[karesz*]

Dear David,
sorry, but as I see it, thes chopper-application functions an otherway...
sincerelly, K.

[Radio Wrangler]

You are indeed right!

It's switch contacts doing the switching. I've seen Boonton and HP use photoelectric choppers with motor-driven spinning blades chopping light onto photocells. But no, that one bangs contacts together. I didn't have the handbook so was working on memory so I was going on general memories.

The nice bit about optical choppers is that you don't get gate capacitance imbalance injected AC which gets rectified to cause trouble. This is the plague of FET choppers.

Sorry!
David

[Jon_G4MDC]

Thinking about these mech choppers. Aren't they very ripe for replacement by something using reed relays? I heard these could manage to switch at low audio frequencies if they were small.

[karesz*]

Yes, its possible to replace a mech-chopper with Reeds, but what abot reliability?
You have to calculate with i.e. 10 millions of correct switching-and how many hours are there if we assume a switching frequency of some 100 Hertzs?
We can find an impulse generator from -I think- Ortec they has Reed switched circuit for Nuclear Physics and is an old development from later 50s or early 60s but its a relative unknowed equipment and nowadays is no more reliable...

[Jon_G4MDC]

It's a very hazy memory but I seem to remember a circuit that used small reed relays (maybe encapsulated in about 3/4" or so of glass) as the basis of a Dalek voice machine.

The longer 2" ones were used in early computers too. I'm thinking of a Xerox place in Elstree about 197x. They worked in Octal.

As I recall the reliability of reed relays was quite good.

[Bazz4CQJ]

Thanks for the comments. Perhaps I should have said in the OP that I don't own a Boonton, but my interest in them started 5 or 6 years ago at the time that 'Skywave' (SK) inspired a couple of threads about homebrew probes and RF voltmeters, none of which seemed to get remotely close to the sort of performance of the Boonton or the comparable Marconi units.

At the time, I think even less was known about the types of diodes used in the probes than is perhaps known now.

I think there were 5 or 6 published designs (in ham magazines) for RFVM that got discussed in the thread, and those that were investigated and all worked much less well than the published articles claimed.

Re choppers, that's a subject I have no previous experience with, so cannot offer any input.

Were it the case that any hobbyist wanted to build an RF probe /VM today, I'm not sure that any new ideas have appeared in the last few years that would suggest a promising way forward?

Thanks to orbanp 1 for the link to the Boonton re-build article which uses a 'photo MOS relay' to do the chopping; must go off an find out what they consist of! That re-build article is 25 years old, so even that may be ready for a rebuild

B

[Jon_G4MDC]

Just for fun I looked up a common or garden reed relay that RS sell.
Switching time including bounce 0.7mS which is probably good enough.

The snag seems to be that it's rated for 100 million switching cycles. Chopping at 90Hz I made that about 300 hours or a bit under 2 weeks of continuous operation.

[Bazz4CQJ]

Originally Posted by Jon_G4MDC

Just for fun I looked up a common or garden reed relay that RS sell.Switching time including bounce 0.7mS which is probably good enough.

The snag seems to be that it's rated for 100 million switching cycles. Chopping at 90Hz I made that about 300 hours or a bit under 2 weeks of continuous operation.

But if it did the job very well, surely that could be got round? My budget DMM switches itself off after x minutes to conserve the battery.

B

[WME_bill]

Choppers.
There is a very useful discussion about the various types of chopper in the article on DC microvoltmeter by D. Bollen in Wireless World May1967. He settles on a reed relay.
See also the manuals for the Marconi TF2603 RF millivoltmeter. The 1965 germanium transistor version uses a relay CB1711 as chopper. The 1977 Silicon version uses a relay chopper C1417 and the 1984 version uses FET switches (2N4416).
The 1965 version has a section upon testing the chopper relay.
I also looked up the Marconi TF1313 bridge. This uses photocell choppers, Clarex CL703CL, driven by the light from a neon relaxation oscillator. The method RadioWrangler suggests.
The Boonton 91C uses a chopper relay, Stevens Arnold CA12 or Bristol C1417-8.
Diodes.
The Boonton 91C uses two gold bonded germanium diodes, no details given. However the Marconi TF2603, with an almost identical circuit, uses in all versions two germanium diodes in the probe,type AEI CG91H. These became Mullard AAZ13 or GEC GEX66. Similar are Hughes HD1871 or STC DK21.
The Electronic Engineers Reference Book, by Hughes and Holland 1967, with abstract into www.technicalscientific.com/techinfo.htm and quoted by Electrojumble.org in 2012, is an interesting discussion upon design practice used in Marconi TF2603 of heating the probe diodes resolve temperature dependancy and also of the noise level of the various types of chopper.
At one time. I worked upon the chopper relays used in Kent or Honeywell chart recorders for chemical plant control. Used Clare 3 pole change-over mercury wetted contacts, relay HG3A 1036, plugging into an 11 pin octal valve holder. Another one used a RadioSpares mercury wetted reed relay.
I suspect that mercury wetted reed relays are much more reliable than a standard reed relay, as the mercury will prevent contact bounce and possible arcing.
I have no details of these, but have the manuals and articles I mention for all the others if you cannot find them. Also some spare diodes AAZ13 or similar.
wme_bill

[Bazz4CQJ]

Bill, you mention the GEX66. do you think the probes had "run of the mill" components, or do you think they could have been carefully selected out of the production batch? Just as a a guess, they had the best forward/reverse resistance ratios, or some other parameter(s) which might have been found in just a small percentage of the production?

B

[WME_bill]

Matching. I think they were matched for similar forward resistance (or current at some small voltage). The forward to back ratio is less important at those low voltages.
The GEX66 was often specificed as a high frequency mixer diode. It was made by GEC, which then became AEI (hence I suspect CG91H). Then GEC/AEI merged it's semiconductor business with Mullard, hence AAZ13.
All gold bonded germanium with very low capacitance, low stored charge so very fast and about 8v rating.
wme_bill

[Bazz4CQJ]

OK re matching. And judging by the comments made in the link in Post#1, the diodes aren't just matched with each other, but also with the main instrument.

What about the use of the heater? Is that simple there to minimise the negate the warming effect of the hand on the probe, or is there any additional benefit in having the diodes operate at a slightly elevated temperature?

B

[WME_bill]

Boonton 91C or Marconi TF2603 (1977/84)
Bazz. The references I quoted seem no longer available, so herewith the original posting from Electrojumble- Roy Johnstone, with thanks. The TF2603 (Si version 1977-84) still seems available on the web.
wme_bill