MOSFET testing

[Ian Stirling]

Where?

[Mike Tomlinson]

In article , Ian Stirling writes

Don't recall. (I knew I'd regret not bookmarking the link.) It was from a post in alt.consumers.uk-discounts.and.bargains. An Aldi or Lidl special, I think, or one of the internet shops

perhaps.)

Do note I said the ad *claimed* 1kW - this in real life means 600W peak, and 300W continuous running, esp. if it's a cheapy. :)

has a 1kW peak, 300W continuous inverter for ukp69.

Another source of power inverters is UPSes - at the RSGB radio rally in Blackpool in March, one trader was selling APC Smart-UPS 600s with dead batteries for ukp20. One of those and a car battery...

[Ian Stirling]

Maybe two batteries. A thing to watch for if doing this is to ensure that the ventilation is adequate. Some UPSs have very poor cooling, certainly inadequate for long-term running.

[Dave Plowman (News)]

Indeed. Rapid Electronics are reasonably priced, and their current price for the 300 watt model I'm fixing is 49.80 plus VAT.

It's a 50Hz modified sine wave type.

[The Natural Philosopher]

They should NOT show any perceptible resistance between source and drain.

And on a typical meter, not much between the gate and any other terminal.

There is a parasitic diode btween some terminals but I forget which.

Most failures are meltdown in the drain source cahannel or punch through the MO layer to give excessive gate current.

[G&M]

Ah - I'd seriously consider replacing every MOSFET. Those that haven't gone yet probably will do, taking the new ones with them again.

[Scott Mills]

I would construct a very simple circuit.

Tie the Source to ground. Put a 1K* resistor to the drain side and connect to the +iv**e supply Slowly increase the voltage*** applied to the gate, while monitoring the voltage aross the resistor with the 'scope

• Think V=IR and chose a sensible resistor - ie current must be in spec for the FET. Assume no voltage drop across the FET for this calc.

** Use a supply that is in spec for the MOSFET. 10-15V at a guess. ** If you don't have a variable source use a potential divider - you can assume the gate draws no current. Monitor the pd with the other channel on the 'scope.

Have a play, you should be able to see it switching quite nicely and cleanly.

Scott

Pin layout for a FET:

[Scott Mills]

If you are really bored make one. I made one for my degree project. They are really easy and very cheap to make - especially if you build your own timing circuit and wind your own transformer. The output had loads of harmonics, but ok for bulbs etc... A project if you're bored anyway.

Our one delivered about 500W+ @ 240V, 50Hz spot on from a 12V car battery. Thick cables on the input side - not quite cig lighter fitttings though.

[Dave Plowman (News)]

Have a look at the URL below for a simple go/no go testing method.

again to Bob for supplying it.

[Dave Plowman (News)]

You're restricted to about 10 amps from a cigar lighter, so say just about

100 watts at 230v.

The battery cable supplied with this one has a 35 amp fuse - not that it stopped the damage through, I'd guess, wrong connection.

There's room for a relay inside to prevent it happening again.

It's working ok now - I had all the bits needed. Total cost of parts needed would have been about 6 quid.

As regards building one, I think you'd be hard pressed to make one in a suitable box for what they cost to buy.

[Ian Stirling]

To make one that's 'competitive' (weight, space, ...) with the existing ones, and can cope with wierd loads (say a power-drill switched to reverse while on, shorting the output, ...) It can easily result in lots of broken FETs before you get it working.

[Mike Tomlinson]

In article , Scott Mills writes

I built one from a circuit in one of the popular electronics magazines some years ago. It used a couple of transistors, resistors and caps in a circuit self-oscillating at about 50Hz to drive the secondary of a standard toroidal transformer. Worked fine - I still have it.

The author said he used it to run his mains shaver so he could shave at the lights on his way to work. Don't think that would be too popular with Plod if he saw him nowadays.

[Dave Plowman (News)]

Vellerman still do a kit that uses a standard mains transformer in reverse

- I've got one as standby for the CH in event of a power cut. But it costs more to make than to buy a cheap one - although it does produce a rather better sine wave than most.

Crikey. Must be old. My brother got a present of a re-chargable razor when he started shaving. He's retired now.

[G&M]

Bit old-fashioned nowadays. Newest ones use a small high frequency transformer running at 5 to 10kHz, the primary side drive oscillator supply being fed from a pulse width modulator running at 50Hz. A simple LC on the output gives good 50Hz at up to a kW or more.

[Dave Plowman (News)]

Yup - the one I've fixed is of that design. Not sure how tolerant they are of overload, though.

[G&M]

Far better. Driving a short the high frequency transformer saturates and the transistors turn off the next cycle, i.e. 100uS. Shorting a 50Hz transformer is much more difficult to detect and the transistors can end up driving a short for 100s of mS.

Same problem in audio amplifiers - always the base amp that blows.