Using a variac when testing always wise?
 

After a bit of reading this morning, I decided to pose this question.

Is it always a good idea to use a variac on "things containing electronics", rather than components alone, like just testing say a power transformer to find windings or voltages.
Some advice here may be of assistance in the future for variac users, especially new users.
I often wonder what could happen if say some complex electronics has maybe 70% mains applied? Or as voltage is gradually increased from zero to full? Do regulators regulate correctly, not at all or allow over voltage output up to such a point of input voltage? With say several regulated supply rails, I sometimes wonder what happens if say the -ve rail isn't there but the +ve rail is partially, or to any varying degrees caused by low voltage input.
Stages with dc feedback applied might be damaged? Audio output multiple device circuits might be harmed?
Processors and ics may not be protected?
SMPS with auto changeover from 115 to 230 could be a big problem as well?

If anyone has actual experience of such pitfalls, it may be very helpful to others, especially newcomers, to read about.

I must admit, about the only time I use my metered variac is on valve only gear. I am quite reluctant to risk possible damage to solid state circuitry.
Testing transformers is another use.

One good idea I did read this morning is replacing a valve rectifier with a silicon one, so that there are no heaters to delay the ht, allowing a true variac control of ht rails. (Posted by Silicon, post #5 in the Quad II thread) A high value resistor in series to limit current for reforming as well. This is on my to do list now. I had never thought of doing this before.

Rob

Re: Using a variac when testing always wise?
 

Switch mode power supplies actually draw more current when they have a lower input voltage, they approximate constant power loads. Some such supplies can be damaged with too low an input voltage, running them up on a Variac is a bad idea. Some are OK, but unless the service manual suggests it, I wouldn't

Re: Using a variac when testing always wise?
 

A lot of equipment contains switchmode supplies which behave badly when run undervoltage, so you need to check first - not just new stuff, but some stuff now considered old also have them.

Re: Using a variac when testing always wise?
 

I tried to run up my HMV 2711 (Thorn 3500 chassis) tv on a variac when I first got it, before I knew any better! It made some pretty awful noises when it got to a certain voltage.

Regards
Lloyd

Re: Using a variac when testing always wise?
 

The classical approach to using a Variac to start equipment not used for some time is to bring it up slowly; I'd maybe go 0-240 over 30-40 seconds.

In some cases, with modern electronics, I use the Variac but go 0-240 in maybe 2 seconds.

I've never had anything fail while being "woken by Variac". Maybe my Variac (a monstrous 8A beast which I struggle to lift now) is lucky; I saved it from the scrapheap.

B

Re: Using a variac when testing always wise?
 

+1 for not using this method to run up anything with a switch-mode power supply in it: As has been stated they will try to maintain their output into the output load by drawing as much extra input current as they need to, to compensate for the lack of input volts.

If the input voltage is really low then the current drawn is almost certain to exceed the maximum input current that the PSU was designed for.

It is why so many devices expire during a mains power brown-out where the lights fade right down for a second or so. Some SMPSUs do have low input voltage detection and will shut off if the input is excessively low, many older ones do not.

Re: Using a variac when testing always wise?
 

Many items with switchmode supplies these days have active power factor correction and these get very annoyed at variacs and/or lamp limiters.

Re: Using a variac when testing always wise?
 

Good question from Rob #1, and interesting replies.
John

Re: Using a variac when testing always wise?
 

I very rarely bother with a variac, and certainly not with anything with a switcher inside. I find a plain old lamp limiter to be much more useful.

Re: Using a variac when testing always wise?
 

I never bother with a Variac or lamp-limiter; it seems to me to be perverse to introduce a new fault [low power-supply voltage] if you're expecting there to be other faults.

It's not just SMPS that dislike being fed fron current- or voltage-limited supplies; it can be very bad for some directly-heated valves to operate them for any time in a condition where the emission is constrained by insufficient heater-temperature, and may result in a thing sometimes called 'cathode stripping'.

With insufficient supply volts, regulators can't regulate properly, oscillators may not oscillate, PLLs may not lock, bias-supplies may not produce enough bias to protect high-power valves etc...

Getting the power-supply lines up to their intended voltage is, for me, one of the first steps in any fault-finding.

Only places I _have_ u sed a Variac are when checking the performance of a piece of newly designed/built gear against supply-voltage variations, and one of my field-techies was famous for finding elusive 'intermittent' faults in transmitters by feeding them from a Variac, winding it up to 300V, and running the TX 'key down' into a dummy-load while he went to lunch.

Re: Using a variac when testing always wise?
 

Considering the numerous and lengthy threads on the forum about reforming capacitors, I'd argue that the use of a Variac with old valve equipment is worthwhile. I have never possessed any type of capacitor reformer, and have very rarely "warmed one up" on a variable power supply. Never had a lamp limiter either.

B

Re: Using a variac when testing always wise?
 

Me too,

Re: Using a variac when testing always wise?
 

I agree. My methodology is to stealthily use my eyes and ears and after switching on be prepared to switch off quickly. And I only switch on after having done a considerable amount of testing and/or replacement of caps, testing of valves, cleaning switches etc.

Re: Using a variac when testing always wise?
 

>use my eyes and ears

Don't forget your nose!

Re: Using a variac when testing always wise?
 

Yes, what a friend calls the "Rhino test" - useful for detecting the acrid smell of cooked transformers/resistors.

Re: Using a variac when testing always wise?
 

This thread has a kind of "split personality" which is present even in the opening post. We are on a forum about Vintage Radio Repair and Restoration, but the conversation is drifting considerably towards servicing modern switched mode devices.

So, I think we have to distinguish that what's a good approach for dealing with a 1950's radio which has been sitting in a garden shed for some time could be appreciably different from the approach used to service a modern unit which failed in service yesterday or last week.

No one has mentioned the use of the thermal imaging camera; there's a neat piece of equipment for dealing with all kinds of equipment and far better than a nose, often showing up hot components when the Variac is only supplying 60volts..

B

Re: Using a variac when testing always wise?
 

A lot of what is now 'vintage' contains SMPSUs. I've got 40-year-old test gear and computers that use them.

I do use a lamp limiter with an SMPSU, _but_ I run the supply into a dummy load and make sure the lamp limiter bulb is way in excess of the load. For example a load drawing 1A from the 5V output of the supply (so 5W) and a 100W bulb in the limiter. I have never had any problems doing this (and I work on such things a lot) on the other hand it has saved transistors hitting the ceiling (they still fail, but at least I don't have to worry about bits in my eyes), PCB traces melted, etc if something does go wrong.

Re: Using a variac when testing always wise?
 

Yes, even my 1965 vintage Racal RA217 receiver uses a SMPS, and there were several 1960s-era TV receivers that used thyristor-based SMPS.

I tend to think of SMPS as distincly old-hat these days.

Re: Using a variac when testing always wise?
 

I think it’s fine for valve equipment, though not for solid state equipment. Things like CRT television or monitor time base oscillators might not start oscillating and draw a high current through the output stage. Also degaussing circuits could run much longer than designed cooking the positstor. There are certain “smart” power supplies that could see a low mains voltage and switch to working at 110V ac. Then the mains voltage increases above this and damages the power supply.

Christopher Capener

Re: Using a variac when testing always wise?
 

I used to work in a place that made linear and switchmode power supplies. Both types were routinely tested using a variac. It was usually possible to see very quickly if something was fundamentally wrong with the power supply before subjecting it to full mains.

Well-designed switchmode power supplies will shut down if too much current is drawn, or simply not start up until there's sufficient input voltage. However, some early colour TVs for example didn't like being run on a variac. It depends on the design of power supply. Linear power supplies and valve equipment shouldn't be a problem.