Act 6 dimmer pack PAT test failure.

[kgallen]

I've found some more Act6 schematics and have collected them all together into one pdf. Hope these help folk.

act6_sch_set.pdf

[David Duffy]

C3 would certainly be a suspect, but C5, C6 & C7 also need to be checked.

If any of them are the old silver RIFA block capacitors you should change them regardless. They crack over the years, then eventually fail in a smokey, stinky way.

[dave-w]

Thanks for all the input folks.  I'll be putting it all to good use as I try to get to the bottom of the problem.

[dave-w]

Good afternoon gurus!

Problem identified but not (yet) resolved!  Having eliminated anything to do with the triacs, or anything else outwith the PCB, and taking note of Richard's comments about ignoring anything LT side of the transformer, I concentrated on the mains input part of the PCB.  It soon became apparent that my single phase PCB (numbered 881/2) was significantly different from the later PCB (950/2) which you folks kindly found for me. Mine probably corresponds to the earlier pulse transformer type which Richard worked on.

I have traced the circuitry on my PCB and attach an image of the mains input and one channel output (components are numbered from the PCB).  Immediately obvious is the 330K resistor R1 in parallel with the 0.0047microF capacitor C1  between neutral and earth!  Coupled with the additional 330K resistor R8 directly across the  Line and Neutral inputs and the different arrangement of the 330K R3 / 0.22microF triac bypass, then, allowing for the transformer primary (540 Ohms), this can account for all the almost identical "leaks" to Earth (330K or 670K) from any Line (in or out) and Neutral points: albeit via different routes and/or 330K resistors, including the increases when F5 is removed.

These differences are not modifications, they are inherent to the 881/2 PCB.  So it would seem they were designed that way and would NEVER have passed PAT testing unless known to the tester as inherent to the design and safe. Do PAT testers have the discretion to make such  exceptions?  If not, lets consider a possible modification.

Of the three 330K resistors R1, R3 and R8, only R3 would appear to be essential. The function of R8 is obscure to me and R1 is exclusively the cause of the problem.  Equally, I do not see the benefits of R1 and if just removed (or perhaps increased to 2Meg) R3 and even R8 could remain in place and "Insulation" is restored!  C1 on my PCB clearly corresponds to C3 on 950/2 which has no bypass resistor.

Any views?

Woops. Just noticed I have missed a 0.47Mfd across mains in L and N in parallel with R8.

[DrV]

I believe those resistors are there to discharge the filter caps. If you were to pull out the plug when the mains was at a peak of its cycle the filter caps would be left with a charge on them. If you then touched the pins of the supply plug you'd get a small shock. Nothing dangerous but enough to give you a surprise and maybe drop something. The resistors just stop that happening. Changing them to 1 or 2 meg or so will still achieve the same effect but just take slightly longer to discharge the caps. Wait for someone else to confirm what I've said before doing anything. After all I'm just some bloke on the internet!

[dave-w]

Hmm. Thanks Dave but wouldn't the transformer primary perform the same function for Line and Neutral whilst Neutral/Earth voltages and the tiny capacitor would scarecely warrant such a precaution surely?

[sunray]

13 minutes ago, dave-w said:

Hmm. Thanks Dave but wouldn't the transformer primary perform the same function for Line and Neutral whilst Neutral/Earth voltages and the tiny capacitor would scarecely warrant such a precaution surely?

Not if the fuse has blown.

Not if the supply had a reversed L/N.

[DrV]

DOH! Yes, you're right about the transformer. It would discharge any X caps but the resistor is still needed for the Y cap (C1). I've certainly had a shock from an oscilloscope mains filter a couple of times. I don't know what value the cap was but this was definitely the reason. It's not the size of the jolt but the secondary consequences that could be a problem - dropping your hot tea down your leg etc🙂. Sunray's point about the fuse is also valid.

[sunray]

6 minutes ago, DrV said:

I've certainly had a shock from an oscilloscope mains filter a couple of times... Sunray's point about the fuse is also valid.

And whyelse do you think I thought to mention it?

Edited July 3 by sunray

[themadhippy]

If im remembering my AC theory correct from many moons ago  an inductor and capacitor in parallel  will never lose energy,the capacitor discharges into the inductor which discharges into the capacitor,guess the transformer secondary acts as a discharge path  thought

[David Duffy]

C1 should be a "Class Y" rated capacitor and would not usually have a resistor across it. It's very odd that they did that.

[dave-w]

Our PAT tester is saying he won't pass the dimmer unless the LINE can be shown to be "sufficiently" insulated from Neutral and Earth combined. With the 330K resistor and transformer primary both bridging L to N I cannot see him being satisfied. The later 950/2 board has neither R1 nor R8 so maybe the earlier arrangement was deemed overkill?  And I could remove both safely? (In order, metorphorically, to render it "officially safe"???)

And yes, David D, C1 is Class Y rated which rather suggests R! could be removed without issue?

[David Duffy]

Yeah, I would remove the offending resistor. It's somewhat of a design fail imo.

[KevinE]

out if interest, are the dimmers supplied via 13A plugs, or BS4343 (cee) plugs?

[dave-w]

13A plugs, Kevin.