Ultimate lamp limiter & Variac setup.

[Lucien Nunes]

The title makes a bold claim, but in this thread I describe the making of a very comprehensive device for test-powering old gear, and along the way explain the whys and wherefores and some electrical safety and constructional points for the newcomer to vintage electronics. The thread will be a slow-burner as construction will be done in spare moments in the workshop.

The unit needs a descriptive name, preferably one with a catchy acronym. Universal Power Limiting And Control Apparatus For Energising Long-Dormant And Suspect Equipment. UPLACAFELDASE. Eh, not catchy, you say? I'll work on it. Suggestions on a postcard please.

In short, there are four functions that are useful / important for test-powering equipment on the bench and under evaluation:

* Electrical Separation: This will be provided by a purpose-made isolating transformer.
* Overload and short-circuit protection: Selectable circuit breakers and fuses.
* Voltage control: Variable-voltage autotransformer a.k.a Variac (tm).
* Current limiting: Selectable filament lamps

In the UPLACAFELDASE these will be selectable and adjustable, because I work on a wide variety of equipment with different power requirements. The significance and functionality of each aspect will be explained as work proceeds.

The first component that I have requisitioned for the build is the variable-voltage transformer. This is a BERCo Rotary Regavolt of 2A rating up to 270V. It appears to be in good condition but in the next post I will inspect and test it electrically and discuss the correct overcurrent protection to prevent it being damaged in the future.

[Boulevardier]

Hi Lucien. Looking forward to your thread!

I'm currently thinking about whether to put such a piece of equipment in a metal case or a plastic one. Don't really fancy another large chunk of earthed metal on the bench - e.g casually resting one hand on the metal case when accidentally putting the other on a mains-live point on the DUT - 230V AC across your arms and thorax (heart)!

Mike

[GrimJosef]

If the gear you're test-powering has a valve rectifier in it then you might want to include a plug-in solid-state substitute for that.

We can inch a Variac up, degree by degree, into a mains transformer primary and all can seem well. But of course that's because the rectifier cathode is still too cool to emit any electrons. So there isn't any HT. By the time the cathode's a bit warm the primary voltage might be into three figures. Then quite quickly the HT appears, but it'll also be at a three-figure voltage aaaaaand ... puff ! ... just as quickly there goes the magic smoke from the not nearly reformed electrolytics .

A solid-state rectifier gets round this problem, but now we have to be sure that nothing is inadvertently over-volted due to the remaining valves drawing less current than the circuit designer expected and the solid-state rectifier having almost no Vf.

Cheers,

GJ

[Lucien Nunes]

This is a point I will return to in the thread and a good idea. I tend not to use a variac in that manner on general electronic stuff, for the reason you describe. I prefer to prepare the equipment so that it should withstand HT without too many surprises, then apply full mains voltage via a lamp limiter. The variable voltage aspect is as more to simulate non-standard supplies and conditions, and for applications other than valve electronics.

[GrimJosef]

I'm also a big believer in prepping as much as possible (measuring resistor values, reforming electrolytics, checking valves) before applying power.

I think of it as being at the 'patient' end of the 'how much can I be bothered ?' continuum. At the impatient (I really can't be bothered) end is 'let's just plug it in'. The various devices which let us plug it in while trying to limit the resulting damage lie at various points along this continuum. We decide how much patience we can muster and we make our choice .

As you say, a Variac is very handy if we want to run at something different from 240V. If we're repairing a piece of kit for use on the continent (quite often 220V) there's a lot to be said for checking it works decently at that voltage. Equally if it's been manufactured for use 'anywhere in Europe' it might well actually have been designed for 230V at best and 220V at worst. I have an amp on the bench at the moment which caught fire. Disassembling it I find the mains transformer has 220V written on it in big letters.

Cheers,

GJ

[trobbins]

Lucien, were you considering metering, and earth leakage protection capabilities?

[Lucien Nunes]

Basic metering, using a pair of small moving-iron instruments, is allowed for. 4mm sockets for shrouded test leads allow more sensitive meters to be connected.

Earth-leakage protection is a bit of an odd one because although it looks desirable, it really shouldn't be necessary. When the isolating transformer is in use, the RCD is redundant. By the time a device under test is let loose on real mains, it should have been tested properly with an insulation tester and leakage meter so should be scarcely more likely than anything else to give rise to a shock risk.

Realistically, in 2022 any wall socket it will ever be plugged into will be RCD-protected and there is no way of achieving discrimination between an RCD in the UPLACAFELDASE* and the RCD or RCBO in the premises wiring protecting the socket. It's simply a race to disconnect and either or both might trip in the event of a significant fault. One might argue that anything that reduces even a small probability of a trip in the premises circuit is useful, plus an RCD provides reliable DP switching and might serve as the main switch. So it might win its case for inclusion on grounds of practicality, rather than as a necessary safety device.

* UPLACAFELDASE: 1. n. An electrical device awaiting a proper name. 2. n. Acronym: Universal Power Limiting And Control Apparatus For Energising Long-Dormant And Suspect Equipment.

[trobbins]

My comment was related to the UPLACAFELDBASE secondary side circuitry, which is isolated from any premises wiring protection by the isolation transformer. The isolation transformer secondary winding is where an extra layer of protection from misadventure related to the secondary side (ie. DUT) circuitry could plausibly be of benefit.

I'd guess that the majority of forum viewers would not have an insulation/leakage tester, whereas the majority of viewers who have undertaken a range of testing, or come from an electrical background, or have advanced bench capabilities would have an insulation tester.

[Radio Wrangler]

Originally Posted by Lucien Nunes

Earth-leakage protection is a bit of an odd one because although it looks desirable, it really shouldn't be necessary. When the isolating transformer is in use, the RCD is redundant. By the time a device under test is let loose on real mains, it should have been tested properly with an insulation tester and leakage meter so should be scarcely more likely than anything else to give rise to a shock risk.

Hmm This means that the only way you can be sure that an item is safe is after the whole set of tests, involving different rigs, is completed. So, precautions have to be taken to make sure that anything partially tested can't be mistaken for something completely tested.

David

[Bazz4CQJ]

No offence intended, but this thread is making me think of this;

The phrase; “keep it simple, stupid” (KISS) is thought to have been coined by the late Kelly Johnson, who was the lead engineer at the Lockheed Skunk Works (a place responsible for the S-71 Blackbird spy plane amongst many other notable achievements). It is worth noting that Kelly’s version of the phrase had no comma and was written “keep it simple stupid”.

https://www.interaction-design.org/l...sign-principle

B

[wireman]

Originally Posted by Lucien Nunes

Realistically, in 2022 any wall socket it will ever be plugged into will be RCD-protected...

I suspect there are many homes that have not been rewired or had work done since the regulations changed. I'm writing this from one at the moment.

[Lucien Nunes]

the only way you can be sure that an item is safe is after the whole set of tests, involving different rigs, is completed.

Indeed, and all appliances that pass my bench, new, old or ancient, are subject to a thorough visual inspection and sequence of electrical tests on completion. The safety test process is independent of the repair / overhaul itself.
The 'soft-start' control functionality used during the repair is that of the UPLACAFELDASE (terrible name.) The electrical test functionality used during the final safety test (high-current earth continuity, DC insulation, leakage) is that of a PAT tester (terrible name.) The two complement each other.

“keep it simple, stupid”

Yes and no. Many bench instruments, not least the above-mentioned PAT tester, incorporate multiple functions selectable and configurable with knobs, dials and menus. At the moment I can't do that with with power control and limiting devices. If I want to change the limit on my lamp limiter I have to physically change the lamp, which is clunky. If I want to apply full mains but protected with a 250mA fuse, I have to fish out a separate inline fuseholder. The proposed device simply integrates a handful of functions in a single unit to save fiddling with test leads and plugs.

My comment was related to the UPLACAFELDASE secondary side circuitry, which is isolated from any premises wiring protection by the isolation transformer.

The transformer is made to EN 61558 part 2 and can be trusted to provide electrical separation. The only way an RCD on the secondary side could ever trip would be in the case of a double fault, one upstream of the RCD and one downstream. The probability of an earth fault in the minimal amount of wiring between the transformer and the DUT is very low although if the variable transformer is in this position, which is (in its present form) Class I, there is some justification. I do like the idea of using the RCD as a DP isolator on the output too.

I suspect there are many homes that have not been rewired or had work done since the regulations changed. I'm writing this from one at the moment.

I agree, the statement was too much of a blanket one.
My device is intended to be used in one or other of my workshops which are not at home, but others might use such a device connected to unprotected circuits in homes with older wiring.

Thus for a variety of reasons the RCD seems to have earned its place on the DIN rail. I suppose my original point was about the fact that relying on an RCD in the unit implies incautious work bypassing the isolating transformer when it should really be in circuit.

[The Philpott]

Soz, the best i can come up with is CLAVICLE.
Current Limiting Apparatus (for) Variable Input Control (of) Legacy Eqpt.
Dave

[Alistair D]

Lucien, you mentioned that changing bulbs on your lamp limiter is not ideal.

I have attached the circuit for my own LL. It uses 2x60 watt bulbs and a switching arrangement that allows 2 bulbs in series, 1 bulb, bulbs in parallel and bulbs shorted out, all done by 2 switches.

In the 10 years I have been using this circuit I have never needed to change the bulb wattage. Perhaps it is just luck that the 4 setting have covered my needs.

In addition the the LL function sockets SO2 and SO3 allow the bulbs to be used as lamp loads when testing high voltage power supplies. There are a couple of caveats that need to be taken care of when using it in this manner.

1) Pretty obvious really. Never use the shorted out setting when testing a current unlimited supply. There is a fuse to protect the LL but still not a good idea.

2) The mains plug live pin will be at the same potential as the SO2 socket. The cure for this is to plug the plug into SO1 and switch it off or in some way make the plug pins inaccessible.

You probably have a better circuit than mine but I thought I would run it past you anyway

Al

[Lucien Nunes]

I work on / play with a wide variety of stuff including industrial electronics of high power rating. Quiescent load could be anywhere from 1VA to maybe hundreds. I was thinking of perhaps 40W, 100W and 275W lamps or similar, haven't experimented with series operation yet.

CLAVICLE is much better. I wonder whether we can go shorter still. I was toying with ACME: Adaptable Current Management Equipment, but it's not descriptive enough.

[trobbins]

Ahh, I see the difference in configuration of the isolation transformer secondary is the root of my concern, as I would typically earth one side of the secondary at the winding to maintain a TN-S format through to the DUT.

[Ed_Dinning]

How about a single , say 40W lamp limiter and some low R, higher wattage resistors switched in parallel

Ed

[Lucien Nunes]

Because we like the non-linear foldback current limiting behaviour of a hot filament.

[PJL]

How about including a current limited HT supply for reforming?

[Lucien Nunes]

I considered including a reforming function but decided against it. First, I have a couple of devices for the purpose already; a Reforming Unit No.1 and a Labpack with variable HT output up to 330V. Secondly, electrolytic reforming is something that can go on in the background which ought not tie up the UPLAC... sorry, CLAVICLE, when wanted for foreground repair action.

It might be worth adding a rectifier with optional smoothing, perhaps accessible only from separate output sockets.