Act 6 dimmer pack PAT test failure.

[dave-w]

Hi. One of our old but invaluable Act 6 dimmer packs (local analogue fader type) has failed a PAT test on insulation grounds. Anybody know of any component commonly causing this?

[kgallen]

I wouldn't normally do insulation test (at 500VDC) on any electronics because anything not completely ancient will likely have capacitors from line to ground and line to line for EMC purposes which (a) won't be rated at 500V so could be damaged and (b) will leak current to ground thus causing an issue with the insulation test. Some testers have an "IT" or 250VDC insulation test which helps with (a) but not with (b).

Thorough inspection, earth bond (to all outlet sockets) and then an earth leakage test (<1mA) is my usual approach with items like dimmers and these days with LED fixtures.

But I'm interested in what other experienced folk say too!

Edited June 27 by kgallen

[Edward- Z88]

Hi @dave-w

@kgallen is correct:

3 hours ago, kgallen said:

I wouldn't normally do insulation test (at 500VDC) on any electronics because anything not completely ancient will likely have capacitors from line to ground and line to line for EMC purposes which (a) won't be rated at 500V so could be damaged and (b) will leak current to ground thus causing an issue with the insulation test. Some testers have an "IT" or 250VDC insulation test which helps with (a) but not with (b).

From the ACT6 manual...

Quote

If, as part of routine maintenance, the packs are to be subjected to electrical safety tests applied using a ‘standard’ Portable Appliance Tester, the following points should be observed:

The nature of the electronic dimmer circuits means that an ‘insulation test’ using a voltage in the order of 500V will give a result of approximately 1.7MΩ. Although this may appear as a test failure, the apparently low resistance is normal for an ACT6 DIGITAL dimmer pack and is not due to failing insulation. User safety is not at risk.

(Any reading below 1.5MΩ indicates a problem which must be investigated).

Caution: A high voltage ‘Flash’ test must not be applied to an ACT6 DIGITAL dimmer pack. A high current Earth continuity test must not be applied to the signal ground pins of the control input connector(s).

[dave-w]

That's a great help folks.  Thanks so much. I shall get back to the tester suitably informed!

[dave-w]

Meanwhile the Tester's Report quotes an 'Insulation Test" FAIL at 0.38MOhms which is well below that mentioned in the Act 6 manual even if a standard Tester was used. So it would seem I do have a problem.  Any advice on where to look first?

[kgallen]

Do an earth leakage test (operation at 250VAC measuring earth leakage) and see what that says.

If you get a leakage over 1mA then you'll have to open her up and do a good inspection of cables for insulation issues, open up the supply plug and look for tracing. Get a schematic and look for components that have paths between mains and earth. Check the integrity of any insulation sheets between PCBs and metal chassis.

I've attached a schematic picture of an Act6 probably of the vintage of yours (I used to have one with the local faders). It looks like only C3, the Class Y cap, from N to E. So I think I'd be looking for insulation issues or main PCB to chassis around the input blocks. I think the only thing off-board of the main PCB are the triacs.

I assume the test was done with no lights plugged in to it so we can exclude any external load issues.

Regardless I can't see there should be a big problem that is not easily repairable. Like Betapacks, these Act6 will go on forever.

Edited June 28 by kgallen

[richardash1981]

4 hours ago, kgallen said:

Regardless I can't see there should be a big problem that is not easily repairable. Like Betapacks, these Act6 will go on forever.

Provided nothing falls off the PCB!
Many years ago I did quite a lot of repairs on a hire fleet of Act6 dimmers, and the usual failures (apart from rough handling destroying the faders) was the soldering of components on the main board failing and causing flickering operation (the big electrolytic filter capacitor for +15V was the worst culprit, but also the 2-pin headers where the flying gate drive leads for the triacs connected). A lot of screws to get that main PCB out!

[dave-w]

Many thanks kgallen. That schematic will be most helpful.

The suspect Act 6 measures 340KOhms by DMM between neutral or line and earth (regardless of polarity) and I have checked a spare Act 6 which comes in at a very close at 310K.   A simple ohms law calc would indicate under 1mA leakage to earth at 240V and a bit over over 1mA at 500V which is what the PAT electrician used when arriving at his reading of 0.38MOhm.

(I don't have a PAT tester, Would a DMM in series with the earth lead to the mains plug be a safe indication of leakage to earth?)

However, all these readings being well below the 1.5MOhm limit identified by Edward in the Act 6 manual, I have opened the dimmer up.  All credit to Strand, it looks pristine. I shall need to study your diagram for suspect areas to examine.  All I have found so far is that the resistance from the the channel line outputs measures 660 to 670KOhms to earth whether or not its respective triac is isolated from the circuit.  This rather suggests the fault (if indeed it is a fault) lies on the main PCB.

Unhelpfully the dimmer passed two previous tests carried out by a different tester. (Who maybe used the 250V IT setting???)

Any further suggestions would be most welcome.

Edited June 28 by dave-w
clarification of 340K resistance points.

[dave-w]

Thanks Richard. Nothing fell off or out! It all looks fine. The dimmer works faultlessly. Its just this PAT insulation problem we have to resolve somehow.

[richardash1981]

https://www.theatrecrafts.com/bhc/equipment/strand-act-6 has much clearer versions of the scans. However they may not match your completely, as I'm sure the ones I worked on were pulse-transformer fired where as this one is opto-coupled.
The only component shown here which could leak to mains earth is C3 (bottom left) as kgallen identified. This should be open circuit at DC, so the fact the multimeter reads 340k (I assume you aren't touching both sides!) indicates damage somewhere.

I'm wondering if an earlier / subsequent design may have added surge protection, in the form of varistors, somewhere? These could account for the low resistance and failure, even if working correctly, and are also like to degrade with age. They are usually disc-shaped, and very often blue for some reason. Would rationally be located near C3 in the mains input area of the PCB.

[dave-w]

This does get more confusing!  My Act 6 does not appear to have any surge protection and our PAT tester admits to having to isolate any domestic dimmers when testing fixed wiring installations...... while condemning our commercial dimmers!! 

We do have a spare Act 6 and it measures slightly less (310KOhms) between line or neutral and earth while otherwise working normally.

C3 is missing on our motherboards PCBs. The two large 1000microF electrolytics C19 and C20 seem (in circuit) to be OK which leaves the transformer TX3 a possible suspect?

Any other potential leaks spring to mind?

[richardash1981]

Anything on the secondary side of the transformer is not going to be an issue, so C19/20 (clearly not quite the same as on the schematic, where C19/20 are 100uF) are not going to be the issue. The transformer isolates all this circuitry from the mains - one of the points of it. TRX3 could be the problem, which could be tested by lifting R4 (PTC self-reseting fuse) and seeing if the line to earth insulation changes or not (there is probably no easy way to disconnect it from neutral).

I would be surprised at a transformer fault, because whilst I have seen that sort of loose-wire transformer (the windings are neither taped nor varnished) fail, it has always been catastrophic - in that the insulation failed, the resulting shorted turn vapourised and what was left was open circuit (sometimes with a nice coat of metal deposited onto neighboring components ...).

To ask an obvious question, have you opened up the plug (assuming it has one) and made sure there is nothing nasty lurking in there?

[dave-w]

Thanks Richard. Yes, plug is not the issue! 

I'll report back in due course.

[kgallen]

Sorry, the JPG schematic I attached earlier hasn't come out very well on the forum. I've made a pdf, hopefully this comes out better and is easier to zoom into and pan around.

act6_sch.pdf

[kgallen]

One other thing would try is unplug all of the triacs (I believe they are all on pin header connectors) and see if the insulation reading improves. If so, plug connections back in one by one, repeat the insulation test, and see if a triac has an insulation problem. If I remember correctly they are screwed to the chassis, but probably have mica insulation sheets between the trial body and the chassis.

Regarding "PAT Testers", I realise you don't have one, but it is something you might find useful when maintaining equipment as you seem to do with your group. I've got a Megger PAT101 which I think is a good one for AmDram as you get the meter for a direct reading (rather than just a pass/fail) and this is useful for looking for deterioration (over time) of insulation, or when doing TRS cables. Keep an eye out on eBay for Megger PAT101, PAT2 and similar "old style" testers. Look to pay no more than about £50. I've also attached here an "Applications Note" from Strand about PAT testing of (theatrical) equipment.

strand_pat_test.pdf

Edited June 30 by kgallen