5Kv flex

[Harry Bloomfield]

I've been repairing a very compact 35w HPS wall mount fitting, which had a failed e-ballast. The original ballast, fit precisely, inside the case of the fitting, but replacements of the original ballast are very expensive/near unobtainium. This, is to replace a LPS fitting, high on my house wall, which provides light on my very dark drive.

A member, Rocky, has very kindly donated a electrically suitable ballast, which I've tested, and works just fine, but despite my best efforts, I've been unable to make it fit in the case of the fitting. My plan B, if wouldn't fit in the case, was to mount the ballast indoors, just the other side of the cavity wall, upon which the fitting is to be installed. That would involve running the ballast to lamp wiring through the wall.

My concern, is that the ballast to lamp wiring, initially strikes at around 4Kv, until the arc is established, when it settles to around 100v. I'm not too bothered about getting it strictly right, but what could I get away with, which I might have to hand, for that through the wall connection? The ballast also suggests, it must be <50pf?

 

[JohnW2]

My concern, is that the ballast to lamp wiring, initially strikes at around 4Kv, until the arc is established, when it settles to around 100v. I'm not too bothered about getting it strictly right, but what could I get away with, which I might have to hand, for that through the wall connection? The ballast also suggests, it must be <50pf?

You might have difficulties finding flex 'rated' at 5 kV, but, to be a bit more pragmatic, it's perhaps worth bearing in mind that PVC seemingly has a dielectric strength in the range 15-30 kV/mm (sometimes said to be appreciably higher) so, in electrostatics electrical terms, PVC insulation under 0.5mm thick ought to be OK for 4-5 kV.
Edit: 'predictive text' nonsense (I presume that's how it happened!) corrected!

 

[Harry Bloomfield]

You might have difficulties finding flex 'rated' at 5 kV, but, to be a bit more pragmatic, it's perhaps worth bearing in mind that PVC seemingly has a dielectric strength in the range 15-30 kV/mm (sometimes said to be appreciably higher) so, in electrostatics, PVC insulation under 0.5mm thick ought to be OK for 4-5 kV.

I was thinking maybe two lengths of 2c, with each end doubled up, so between each pole there would be insulation, sheaf, sheaf, insulation.

Assuming I use just one two core, have you any idea what the typical capacitance value would be, if 1m long?

 

[endecotp]

the ballast to lamp wiring, initially strikes at around 4Kv, until the arc is established, when it settles to around 100v. I'm not too bothered about getting it strictly right, but what could I get away with, which I might have to hand, for that through the wall connection?

Any undamaged dry mains wire that you have will withstand 4 kV without breaking down. The dielectric strength of PVC is maybe somewhere around 20 kV/mm, so 4 kV would break down at a separation of 0.2 mm.

You might not want to touch it though. I’d put it in a good solid bit of pipe.

The ballast also suggests, it must be <50pf?

Back-of-the-envelope suggests you’d be OK:

Consider parallel “wires”, 1mm wide, 2mm apart, 1m long, in vacuum. Area = 10^-3 m^2.

C = e A / d = 9x10^-12 x 10^-3 / 2x10^3 = 4.5 x 10^-12 = 4.5 pF

That gives you an order of magnitude to play with.

 

[JohnW2]

I was thinking maybe two lengths of 2c, with each end doubled up, so between each pole there would be insulation, sheaf, sheaf, insulation.

You could, if you wanted some 'belt and braces'.

Assuming I use just one two core, have you any idea what the typical capacitance value would be, if 1m long?

Very low - endecotp has beaten me to doing some quick sums.

 

[SUNRAY]

Halfords?

Centre core of UR67?

 

[JohnW2]

Halfords?

do ignition lead cables have a significant conductor CSA?

 

[blup]

Edit: 'predictive text' nonsense (I presume that's how it happened!) corrected!

If predictive text is a tool, are you the good workman?

 

[Harry Bloomfield]

do ignition lead cables have a significant conductor CSA?

They used to have a decent, copper core, but some modern ones use a sort of string, impregnated with carbon, to reduce interference.

 

[JohnW2]

They used to have a decent, copper core, but some modern ones use a sort of string, impregnated with carbon, to reduce interference.

That's what I suspected/feared!

 

[JohnW2]

If predictive text is a tool, are you the good workman?

It's actually worse than that. I've just realised that I typed that message on my laptop, which does not generally have any 'predictive text' functionality - so it must have been my brain and/or typing fingers that somehow made an attempt to emulate bad 'predictive text'

 

[Harry Bloomfield]

I've just got around to bench testing this....

I used a 1m long piece of ordinary three core, pvc flex, linking the e-ballast, to the lamp. The EHT, only occurs when the lamp is switched on, to start the discharge, so tested it with several cold starts, and absolutely no problems at all. Next job, is to await a decent day, and a roundtuit, to replace the old LPS, with this much smaller, and better looking, HPS light fitting.

Thanks All, for the sanity check..