Boiler power control

[Hogarth]

Oh, wait, the one I linked to above does need a remote filter. This one doesn't.

[Dr. Dextrin]

Interesting; this integrated controller doesn't seem to require an external filter - I've been looking at the PSR-25, and whilst good value in itself, its data sheet suggests its used with an RFI filter (one with an in-line choke, not just a simple parallel network), and that costs about £55!
Kev

Yes I was a bit shocked at the cost of that RFI filter but the controller Hogarth links to above includes an RFI filter for a pretty reasonable price, so they're obviously not actually that costly to make.

[Kev888]

Indeed, looks like a good solution, though I'm concerned about the current ratings; their graph 'looks' like the current rating is 7A or extended to 10A (maximum) with an additional heatsink. That may 'just' do a 2.4kw element (if it can cope with the extra initial/cold switch-on current of an element), but even then its a serious concern for my 3kw/13A element.

Whereas hogarth's previous link to a similar device is extended to 15Amps with the right heatsink.. As far as I can see (which isn't that far at this time of night) it doesn't say there's a need for the external filter, like the PSR-25 data sheet does, but neither does it say that there isn't..?

Of course the RFI may or may not matter according to what sensitive equipment you have and how your mains circuits transfer it; legislation for retail products aside, I really don't know how important that is likely to be within one's own home.. or maybe a slightly less costly parallel RFI supressor would do; still not cheap but can be had for quite a bit less.

Cheers
kev

[Hogarth]

As far as I can see (which isn't that far at this time of night) it doesn't say there's a need for the external filter, like the PSR-25 data sheet does, but neither does it say that there isn't...

It's buried deep in the 'Technical Data Sheet' under 'CE Marking': 'These phase angle controllers require a suitable remote filter.' (1)

That may 'just' do a 2.4kw element

I hope so ... I just bought one! (2)


Footnotes:
(1) the QVR/S-RFI 17
(2) the QVR-TB-RFI

[Kev888]

Ah, I see, thanks!
I'd imagine it will be fine for 2.4kw if you give it a healthy sized heatsink and keep things well ventelated; it looks like its got very useful tolerance for inrush current and so on which is ideal with big elements.
The very best of luck with it!

Cheers
kev

[Dr. Dextrin]

The device I bought on eBay is still being advertised here:

http://www.homerweb.com/vintage/38KW-40 ... 89113.html

I notice the data sheets for these type of things also recommend a semiconductor fuse (not exactly sure what they are, but another expense presumably). I imagine that's because if your load goes short-circuit a traditional fuse isn't fast enough to stop the thyristor frying. Something else to keep in mind.

As regards RFI - I tested mine (with no filter or shielding) using a portable AM radio. It picked up a lot of interference within about 3m, but beyond that the interference rapidly became negligible. I don't think it would have worried the neighbours, and presumably when it's inside a metal box the problem will be less. However, I don't know how much was going up the mains cable and that might be more of a worry. I'll need to find a way of assessing that.

Reading around on the internet, it looks like a home-made RFI filter should be possible quite cheaply. Radio amateurs seem to have it sussed.

[Kev888]

Ah, very interesting, thanks. Possibly the expensive filters are only 'really' needed if you have to comply with the CE type stuff then..

I may either do away with the filter, unless I can detect any problems, or buy/make a cheaper version perhaps. Its unlikely to be worse than the f'ing mopeds that go past my TV aerial every ten mins..

Cheers
Kev

[Hogarth]

I notice the data sheets for these type of things also recommend a semiconductor fuse (not exactly sure what they are, but another expense presumably). I imagine that's because if your load goes short-circuit a traditional fuse isn't fast enough to stop the thyristor frying. Something else to keep in mind.

Yeah, I noticed that too. I wish they'd gone into a bit more detail, though. Can anyone help us out here?

[Kev888]

I think it means a fuse for protecting semiconductors, rather than one that is itself a semiconductor. I can't say that I've ever used one but I believe they're just 'very' fast acting fuses, particularly the short circuit rather than over-current ones.

With triacs in the past I've just used normal fast-blow fuses, but I don't know how effective they'd actually be if there were a real problem..

EDIT: like these or for 25Amp this one

Cheers
kev

[Dr. Dextrin]

Possibly someone might be interested in this, which is available cheap as a clearance item at the moment. It's an IEC mains socket with a built-in RF interference filter. I've just bought one but have yet to try it out, so no guarantees on how effective it will be. But it looks just the job for supplying a phase angle controller and at £7.86 it's way cheaper than anything else I've seen.

http://www.rapidonline.com/Electronic-C ... 0a-26-4902

Nominally, it's rated at 20A (the 20EEJ1 model), but as it uses a C19/C20 connector I think it's probably only usable up to 16A, but still OK for most homebrew situations.

There are a few negatives, though. It's quite long (about 80mm) so needs a big enough box to put it in. It doesn't say on-line I think, but it's usable up to 40 degrees C according to the label that's stuck on it. That might be an issue if it's in the same box as a triac. Also, you'll need a C19 connector for it (available from the same site at £4.25, or on eBay at around £4 delivered). P&P from Rapid Online will also cost you almost as much as the IEC connector, although it's free if you spend £30 ex VAT. They do have quite an impressive range of stuff, including more metal boxes than you can shake a stick at.

I've never used them before and impressions are mixed. Prices seem pretty good and delivery is fast (next day). But they sent me an angle grinder disk labelled "20A IEC socket" on their first attempt, which wasn't 100% impressive. Still, they've sorted it out quickly with no fuss.

[Kev888]

Ah, interesting - many thanks!

I used rapid occasionally for work in the past (many years ago) - IMHO they were not quite up to the very high standards as RS and farnell, but they were still better than most and tended to be keener on price back then.

Cheers
Kev

[Hogarth]

Well, my phase angle controller finally arrived from Farnell after various f***-ups by UPS. Looks very nice. Now I need to get a fast-acting fuse for it -- probably one of the Cooper Bussman ones recommended by Kev. But I'm not sure what rating it should be. Can anyone help?

Here are some of the controller specs (ie, those concerning amps)

Triac limiting RMS. On state cycle: -- 25A
Peak one cycle surge @ (10ms): -- 320A
Max load at 20°C: -- 10A
I2t limiting value for fusing: -- 112A2s
Unit limiting RMS. On state current: -- 15A

Out of my depth here -- any help appreciated!

[Kev888]

Hi,
just to clarify, those were examples of semiconductor fuses, rather than actual recomendations.. as mentioned I've never actually used one as they always seem almost as expensive as whatever I'm trying to protect..

EDIT: not sure about the best rating. Conventional rule of thumb would say 'at least' 125% (up to 300% for inductive loads) of the expected current even for normal fast blow fuses. However, elements can draw higher current than you'd imagine when cold, and these fuses are fast to respond and bl@@dy expensive so 'nuisance' blowing would be £painful..

I'm no expert with this sort of stuff, so I think were it me (and in the absense of any more informed info) I'd probably chicken out and go for a very high value (like 25A or something) and accept that the fuse will only be protecting the controller against short circuit (i.e. very high current failure) rather than over-current conditions. I'd be fearful that if you chose one to protect from over current, it could be a low enough rating to blow by accident..


Cheers
Kev

[Dr. Dextrin]

From what I've read recently, I think the way to do this is to find the A^2s value for the triac you're protecting (that's amps squared times seconds and specifies the amount of energy the triac can handle safely on a brief overload). This is often specified on the data sheet. Then you try and find a fuse to match or slightly exceed that value.

Personally, I went for one of these:

http://www.rapidonline.com/Electrical-P ... se-26-5658

rated at 16A - which I plan to use with a load up to 3kW. I'm not sure what the A^2s value is for my triac, but from the fuse datasheet (see the graphs in the PDF referenced from the link above) it looks like it'll take a very long time to blow with any reasonable startup surge. It's really only "Ultra Rapid" once the current gets seriously big.

As Kev888 notes, though, the fuse is almost as expensive as a new triac. OTOH, if you're using an integrated package things would be different and a good fuse could be worth having.

EDIT: Just noticed Hogarth's post includes an I2t (A^2s) value.

EDIT2: Let me try again... You need a fuse whose current rating slightly exceeds the expected max current and whose A^2s rating is no greater than that of your triac. Not what I said above.

[Kev888]

Ah, that sounds 'much' better; a reasonable compromise between the likely vs fused currents, but now with enough of a fuse-to-triac cost differential to make it well £worthwhile protecting things.

After all , I would suggest the most likely faults in our applications (assuming you got the ratings right to begin with) would either be open circuit (no problem) or something major like the element (or something else) failing and connecting live to earth or neutral - subtle overloads seem relatively less likely.

Cheers
Kev