Go Down

Topic: urgent, trying to neutralise flourescent tube prob (Read 2317 times) previous topic - next topic


can somebody please tell me if this thing has poles, or if i can connect it in either direction. or should it be parallel


Sep 26, 2009, 10:14 pm Last Edit: Sep 26, 2009, 10:15 pm by Grumpy_Mike Reason: 1
You can connect it any way round.

It should be placed across the load like it is shown.

Anyway your connecting it to AC so AC is alternate ways round 60 times a second. So you would have a job connecting anything that had a polarity. (not a pole  ;))



got 64 tubes hanging off 3x19 amps.

the tubes can potentially kick back up to 180 w

these rc modules are what i use for decoupling (from the other thread where you so kindly helped me).

i pray it holds...


It's not working

I am really running out of ideas here. These crappy tubes create a huge spark when turned off. I added a 7µF CAP which eliminated that spark, but now the tubes create a enormous surge when turned on (as the cap charges itself I suppose). This completely freezes up the relays.

The tubes have no protection being the cheapest kind, but I have no way of replacing them now. I gotta combat the spark.

Today I tried this little buddy: http://www.farnell.com/datasheets/338755.pdf
But it is far to slow to even react to my monster spark...

I have had a electrician look at it but he is dumbfounded. All he could tell me was that the spark is powerful enough for welding  :)

Any ideas?

Thank you



When you say sparks where are they going to / from, you shouldn't be getting sparks at all. It suggests you have something fundamentally wrong.
The other thing that can suppress transients are ferrite rings and clamps but I suspect they are no match for a spark. You clamp them round the wire and they absorb the high frequencies.

Can you get some to work properly and then a lot give you trouble or does one just give you trouble?


Oct 08, 2009, 12:56 am Last Edit: Oct 08, 2009, 12:57 am by fubbi Reason: 1

Excuse me, I didn't explain that I actually create the spark by touching the cables together and pulling them apart. I do this to test the size of the on-rush spark and the off-rush spark. Its very revealing.

The 4.7µF cap generates a massive on-spike and without it there is a large off-spike. The cap setup will cramp the relays up in open mode for sure. Setup works best with just the RC's described in my earlier post.

The smaller spark/spike with just the RC's will confuse the board and cause irregular switching. Not crazy bad, but it indicates wear and tare which might get worse and ruin stuff along the line. Also a programmer is writing a nice animation software for it and shouldn't be fragile. It must be able to rock.

Everything is x64 the trouble. Taking the caps back out took many many hours. Otherwise it is a pretty cool piece


some images, actually hangs free by now:


Oct 08, 2009, 03:18 am Last Edit: Oct 08, 2009, 03:18 am by ajb Reason: 1
You mean when you plug and unplug the main power cables you get big sparks?  If that's the case you shouldn't be making and breaking those connections under load (unless the connectors are specifically designed for it--very few are, and they tend to be big and expensive--not your typical mains plug by any means).  All you're doing is ruining the contacts and potentially creating a fire hazard.  Put a switch of appropriate voltage & current rating in series with your power plugs, and make or break those connections only with the switch off.


Ok well I had to laugh at the circuit board, it's very beautiful but not designed to withstand any sort of EMC incursion. That's what you get with letting artists design electronics. I am saying the your layout is contributing a lot to your problems.

Anyway lets see what can be done about this.
The sparks you are generating are a good source of interference, indeed in the old days before health and safety legislation and decent equipment I worked at a place where our standard suseptability test was to have a hole drilled over the live socket in a mains plug and insert a round file so it touched the live. Then we wiped a wire along the length that was connected to a large motor. The resultant sparks generated interference and we had to design the circuit to withstand this.

So while what you are doing is some sort of test it is not representative of the interference it will actually generate. The problem is switching a large inductive load, when the power is removed you get a large back emf. You are on the right lines with the capacitor what you need to do is to slow down the rate of change of current. As you say the capacitor will generate a large inrush of current but buffer or slow down the sudden stop. So stick with the capacitor but investigate putting some series resistance in line with it. It only has to be small because if it is too big then you don't get the beneficial effects of the capacitor as well. I would go for a value of around 10R to begin with and see what that gives you.

There are other techniques for spark quenching like back to back zener diodes. You can even get relays that have contacts sealed in a partial vacuum to reduce arcing.


Learning (the hard way) by doing! I learned a lot doing that PCB for sure. I welcome any critique.

A 100[ch937] R (2W) burned straight away. I have some cement 5W resistors, could it be that simple? Will a 10[ch937] resistor not burn?

Will this little resistor really save me?




Oct 08, 2009, 01:58 pm Last Edit: Oct 08, 2009, 01:59 pm by Grumpy_Mike Reason: 1
A smaller resistor will not get as hot because there will be less voltage drop across it, but you are dealing with a load of 57 Amps and that is a serious amount of current. I would have thought that you would be looking at resistors in the order of 100W like this one:-

Mind you when they blow that shoot the plug out of the end so put it in a box before you try it.

If you were to put 57Amps through a 10R resistor it  would burn 32KW but you are not doing it continuously so you don't need that much power rating.

Will this little resistor really save me?

It's not so little  ;)
But it's the best I can think of at the moment.


thanks a lot mike!

I am going to try a few things out


Go Up