Instantly Triggering Fluorescent Tubes

[Steve Adam]

I'm an amatuer musician and do solo gigs from time to time, and I've been looking at assembling a small scale light show using fluorescent tubes which could then be controlled automatically. I've seen it done (albeit on a larger scale) and although a little bit of research has - I think - pointed me in the right direction, I would welcome some advice as to exact means of putting it together.

The idea is to run twelve 4 foot tubes, each independantly controlled although they will be mounted in clusters of three. Power to each tube will come from a purpose designed lighting controller

which supplies separate channels of 5A at 230v to each tube. This in turn is triggered by notes played on a keyboard which can be sequenced in advance so that the show is precisely in time with the music. It is not essential that the tubes can be dimmed, simple on and off at normal brightness will suffice, although it would be possible to vary the input voltage if this would achieve a reliable dimming effect. The most important thing is that the tubes fire instantly.

I believe that the key to this is using a separate transformer (with a separate mains supply) to send a constant low level voltage to each of the tubes to keep the filaments warm for the duration of the show. Other than that I'm assuming I can simply use bog standard fixtures as you would commonly find in B&Q and the like, albeit re-wired accordingly. This is where I would welcome some pointers:

1) Am I correct? 2) What type of transformer would be appropriate? 3) Is is possible / advisable to preheat all three tubes in each cluster from a single transformer assembly? 4) How should such an arrangement be correctly wired? 5) Can / should the pre-heat transformer be mounted in a separate housing or could it be incorporated inside the existing fixtures?

Any help would be very much appreciated!

Steve

[Capitol]

1) Yes. 2) IIRC about 6.3V per heater was required. Each heater on a tube requires a separate winding. 3) One transformer can be used to power a number of tubes which will be fired together. 4) Very carefully! There was an application note published by STC/ITT around 1963 which specified in detail how to do this. 5) Can be local or remote, depends on the space available.

Sorry not to be of more help, but it's been a long while since I dealt in this area.

Regards Capitol

[Toby]

Yes, just like that(ish). Did this with UV fluorescents at a university hall. We kept about 1/3rd voltage on them all the time so they would strike immediately. The voltage must be low enough so they don't 'hook up'.

I can't decide whether this is all a bit 1980's, or you need an Arts Council grant? I suppose as they are going to be used in a commercial environment you plan to take all reasonable H&S steps.

[Ian Stirling]

You can cheat somewhat. You can wire all the heaters connected to neutral to one 6.3V transformer, and just use seperate windings for the ones connected to live through hte ballast.

[Steve Adam]

Thanks for all the replies, guys.

Forgive my ignorance, but is this something I would be able to lay my hands on, and if so who would I need to contact?

So that I understand you correctly, do you mean something like:

Switch / Ballast AC Live o------O/ O-------^^^^^^^^--+--------------+ | | | | | | | +6 | | +-+ +-------------------+ +-------+ | )|( | | | )|( | -|---|- Transformer )|( | | ^^^ | 2 * 6.0V 250mA )|( 0 | | | )| +----------------------+ | | Tube )|( | | | )|( | | ^^^ | )|( | -|---|- )|( -6 | | | +-+ +-------------------+ +-------- | | | | | _ | | | / \ | | AC Neutral o--+----------| |----------+--------------+ \_/ Starter

Also, although I've included the starter cap in the circuit, would I be correct to presume that it is no longer required since the filaments are now being preheated?

Ha! Right first time. That's exactly the sort of stuff I do, so yes, this is a deliberate attempt to achieve something a bit different/retro rather than settle for the everyday mobile DJ stuff you get at Maplins. But there's a practical side as well - 4 foot assemblies will fit nicely across my back seat and thus keep everything down to one car load and the whole thing should run much cooler than normal lights. But yes, it will probably look bonkers...

[Andrew Gabriel]

No. Try something more like:

Switch / Ballast AC Live o------O/ O-------^^^^^^^^--+--------------+ | | | | | | | | | +-+ +-------------------+ +-------+ | )|( | | | )|( | -|---|- Transformer )| +----------------------+ | ^^^ | | grounded 2 * 6.0V 250mA )| | | | metalwork )| Tube | | +---+ )| | | | | )| +----------------------+ | ^^^ | | | )|( | -|---|- | )|( | | | | +-+ +-------------------+ +-------- | | | | | | | | | | | | | | AC Neutral o--+-------------------------+--------------+ | | Ground o---------------------------------------------------+

Note, no starter, but you will need grounded metalwork in near proximity to the tube.

Note the two 6v windings will have 240V between them at some points, so they should be wound on separate bobbins. Most likely, you will need to use two separate transformers, as I doubt you'll find a suitable single transformer.

I did this some years ago, and I think probably 3-4V was plenty. I suggest you breadboard one lamp before buy the parts for lots.

Another variation you could try is to take the transformer voltage from the other side of the ballast, so when the tube is on, the extra filament heating is almost disabled. (This is the basis of Quick Start control gear.)

Switch Ballast / AC Live o--------^^^^^^^^-----+------+----O/ O-----+ | | | +------------------+ | | | | | +-+ +-------------------+ +-------+ | )|( | | | )|( | -|---|- Transformer )| +----------------------+ | ^^^ | | grounded :: : : : : metalwork [etc]

Another path might be to investigate using a dimming electronic ballast. These use a 0-10V control voltage to adjust brightness, and mostly you can just use a pot as they supply enough control voltage for just themselves anyway. In your case, you could use a switch here instead. I don't know what the instantainous response to the control voltage is like though, and it might vary between makes.

Starter can't be used in this sort of circuit.

[N. Thornton]

The OP's circuit could never work.

So how is the starting HV pulse derived in this circuit? Or is it not? Is it an LC thing with the metalwork?

I know some tubes will ignite with no start pulse, but most wont IME. I had a 2' fitting once that was just a capacitor, no choke, no starter, no preheat, and it worked with some tubes.

Getting back to the requested info, I've used fl tubes in lightshow kit, and it canbe done much simpler than this if you wish. Taking a standard fitting, replace the starter with relay contacts, put power Rs across the filaments to dim them and allow them to run continuous, and replace the choke with 2 in series of much higher power rating. Both options work.

Regards, NT

[Andrew Gabriel]

No HV pulse is required to start a 4' or 5' tube on 240V providing the filaments are at thermionic emission temperature, and the tube itself is not down at freezing or below. (Actually, I think this is true all the way up to 8' tubes, although 8' tubes would have to be at room temperature. However, not experimented much with tubes longer than 5', nor with T8's, which might be different from T12's in this respect.) The grounded metalwork helps by capacitively coupling the area around the live electrode to ground to draw off the first few electrons and get the arc started down the tube (like the starting electrode in some HID lamps). Often you won't need the grounded metalwork, but in portable apparatus I would certainly include it, as you might find it worked well without in your garage, but not in some venue you take it to.

You used to be able to buy tubes for Quickstart and Rapidstart control gear which had a thin metal strip running the length of the tube connected to the metal part of the end caps which was grounded via the lampholders. I haven't seen these for perhaps 20 years now though, which probably reflects the transition from the more expensive and better of the older control gear designs over to electronic control gear today.

Purely capacitive ballasts are a disaster for tube life. When the tube starts conducting each half cycle, there's a high current pulse as the capacitor voltage charges through the tube which quickly blasts the emission coating off the filaments, causing the tube ends to quickly blacken and premature tube failure. Technically, this is known as driving the tube at a low crest factor (ratio of peak to average current flow).

Sounds fine (I'd put a resistor or another ballast in series with the relay, rather than resistors across the tube filaments, but you'd have to play with values to find one which reliably extinguished the tube arc).

Oh, and whilst I'm posting again, I'll correct a stupid mistake in my earlier second suggestion (switch in the wrong place)...

Switch Ballast / AC Live o--------^^^^^^^^----+-O/ O-+--------------+ | | | +-----------------+ | | | | | +-+ +-------------------+ +-------+ | )|( | | | )|( | -|---|- Transformer )| +----------------------+ | ^^^ | | grounded :: : : : : metalwork [etc]

However, I suspect my first suggestion is likely to work better.

[N. Thornton]

I'd heard it can be done with little ones, didnt think it went as far as 5' tubes though. I'l take your word for it, I'm not really sure myself.

having said that though I found with 2' tubes that some would light and some not - ah, now I remember, ok this is getting clear now, my fitting had no preheat.

Ahh!

I learnt that at some point, and put an LC ballast in.

I dont think that would work well, it would substantially reduce light output during on time if it reduced power enough to keep the filaments dark.

Basically this approach keeps the rig in start mode most of the time. During run you have a choke & tube load, during start you have choke load only, so the choke dissipates well above what its rated for. Hence the need to use 2 chokes in series to avoid breakdown or fire. Use much higher powered chokes for 2 reasons: 1 to get more light out,

2 because 2x 80w chokes would behave very approximately like 1x 40w choke.

Regards, NT

[Tim Mitchell]

In article , Andrew Gabriel writes

I've done some jobs with these ballasts (Philips and Tridonic) and they do not provide instant on. There is no flicker, it's a snap start to full brightness, but it takes about 1/2 second for the tube to light. The dimming response is also somewhat smoothed out and delayed.

[Steve Adam]

I was initally pointed in the direction of these things - there are companies who can supply them with in a DMX controlled version rather than the usual 10V which would make controlling them from a professional lighting console a breeze - but the comparative high cost put me off, plus I have become somewhat suspicious of the term 'instant on' in relation to fluorescent tubes as from what I can tell this seems to cover quite a broad range of definitions!

I have a small batch of transformers on their way and I'll try out Andrew's suggested circuit on my 'test-tube', hopefully tomorrow. Thank you all again for your help.

[Tim Mitchell]

In article , Steve Adam writes

The design I did was for a DMX controlled version but it had the same limitations as the 0-10V version. I work in the professional lighting industry and I do not believe there are any truly "instant on" products available.