USB powered Neon Indicator lamps?

[Fungus]

I was thinking of making a desktop sculpture using those little neon indicator lamps.



The question is: How to power it?

Those things need 90V to light up but then they drop down to about 50V in use. You need to put in a current limiting resistor.

They use about 1mA of current each and I'm thinking 20-25 lamps.

The easiest way would be mains A/C (in a closed jar with fuse+isolation transformer, obviously) but I'm wondering if it could be powered via USB. With USB power I could take it down the pub to show it off, etc.

What would be a simple circuit to do that? Would a joule-thief-type circuit work with those lamps or do I need to go the full hog with a little transformer, 555 timer, rectifier, etc?

Any ideas welcomed...thanks!


PS: I'd like the circuitry to be small/hidden if possible.

[Fungus]

PS: Some entertaining reading material:

http://www.nutsvolts.com/uploads/magazine_downloads/NeonLamp-Information.pdf

Be sure to read right to the end, there's some fun projects and plenty of funny old adverts   

[SeanB]

Buy some USB ionisers on eBay, and use the small royer converter in there to drive the neon lamps. Small, cheap and has a case with USB connector on it. Big Clive on Youtube did some review on them recently, where he took some apart. You just need to remove the voltage multiplier on the output of the small transformer and use connect the neon there.

[Fungus]

Buy some USB ionisers on eBay, and use the small royer converter in there to drive the neon lamps. Small, cheap and has a case with USB connector on it. Big Clive on Youtube did some review on them recently, where he took some apart. You just need to remove the voltage multiplier on the output of the small transformer and use connect the neon there.

I found this:


I've left him a question about the voltages, let's see what he thinks of the idea.

[T3sl4co1l]

A multivibrator or "ZVS" or Joule Thief-like circuit could be used just fine, followed by a voltage multiplier (perhaps an LC resonant tank).  They don't need DC.  Any CCFL circuit will do (since, that's what they are, with less mercury and phosphor of course).

Tim

[Fungus]

I just saw this on eBay: http://www.ebay.com/itm/231693965272

If it outputs 1000V from 6V input and a six stage multiplier then the transformer itself should put out about 150V with a 5V input - pretty good! 

(Or maybe even more than 150 volts - Cockroft Walton multipliers have a lot of losses in the diodes)

I could leave in one stage of the multiplier and have a rectified voltage without too much loss (or maybe leave in two stages - 300V would let me run 3 lamps in series and reduce the current demand by a factor of 3).

I doesn't say it works at 5V though and the photos are too crappy to see the chip numbers (does it say 'NE555' on the 8-pin chip?)

[Fungus]

A multivibrator or "ZVS" or Joule Thief-like circuit could be used just fine, followed by a voltage multiplier (perhaps an LC resonant tank).

Might be difficult to tune/optimize (for me!)

They don't need DC.  Any CCFL circuit will do (since, that's what they are, with less mercury and phosphor of course).

Aha! I was searching for the wrong thing. There's a lot of people running CFL light bulbs off batteries on the web and the circuits will be similar.

[uski]

PS: Some entertaining reading material:

http://www.nutsvolts.com/uploads/magazine_downloads/NeonLamp-Information.pdf

Be sure to read right to the end, there's some fun projects and plenty of funny old adverts   

Crap. I thought I discovered something awesome when I noticed some neon lamps turn themselves off with no outside light, and turn back on when outside light is applied. But no, it seems like this so-called "dark effect" has been known for decades

Thanks for the interesting read !

[SeanB]

Also appears in fluorescent lamp starters, which often need external ionising radiation now to start them, now that they no longer have an americium initiator in them. Some interesting articles on how this works in a deep mine, where the string of lamps does not light after the power is applied till you light a torch, after which they all light in sequence down the tunnel.

[alsetalokin4017]

Here's one of mine, lighting six NE-2s in series from a single AAA battery:

[Fungus]

Here's one of mine, lighting six NE-2s in series from a single AAA battery:

That's a big transistor...

[Fungus]

Update: "Big Clive" is too scared to measure the voltage in the USB ionizer for me. 

I think I'm going to order one of those ebay boosters and see what happens at 5V. I reckon that's an NE555 and some sort of MOSFET on the PCB so it should do something.

[LaserSteve]

Here is my favorite neon circuit, blinks the lamps in sequence. Interrupt the voltage for a second, and the flash order reverses.

This is the "Goofy Light" Radio Shack P-Box project kit from my youth...

It works best with polyester or other thin film capacitors.

http://my.core.com/~sparktron/130P6.JPG

Steve

[alsetalokin4017]

Here's one of mine, lighting six NE-2s in series from a single AAA battery:

That's a big transistor...

But it's a _small_ battery....   

The transistor is a common-as-dirt 2n3055. The "Jason Mask" heatsink is just for show. The components on the near edge of the board are a simple wireless receiver, so that, with the addition of a loop antenna, the unit can operate without any onboard battery at all, as long as it's in range of its wireless power transmitter. They are not needed if the battery is used for power. So the BOM of the basic unit is: One 2n3055 or similar NPN, one 10mH choke with an extra 20-turn winding on the outside, one ten-turn 50k trimpot, one 1n4007 diode, one SPDT switch, six neons, one battery holder.

[alsetalokin4017]

Here is my favorite neon circuit, blinks the lamps in sequence. Interrupt the voltage for a second, and the flash order reverses.

This is the "Goofy Light" Radio Shack P-Box project kit from my youth...

It works best with polyester or other thin film capacitors.

http://my.core.com/~sparktron/130P6.JPG

Steve

Ah yes, the old reliable neon ring oscillator. The power supply section may be just what Fungus is looking for. I've built a few of those myself but I use an even simpler JT circuit to power them.

[Fungus]

Here is my favorite neon circuit, blinks the lamps in sequence. Interrupt the voltage for a second, and the flash order reverses.

This is the "Goofy Light" Radio Shack P-Box project kit from my youth...

Get one here: http://www.ebay.com/itm/291569456696

Ah yes, the old reliable neon ring oscillator. The power supply section may be just what Fungus is looking for.

Maybe. Is that a center tap transformer? I'm trying to figure out the ratio...looks like 200:1 but that seems a bit high.

I've built a few of those myself but I use an even simpler JT circuit to power them.

Schematic? 

[alsetalokin4017]

Yes, I think the schematic shows a transformer with resistance values given as 1k and 200k, and the 1k side is center tapped. This might be a miniature audio output transformer.

Sorry, I can't seem to find my schematic of that particular JT circuit I used for my ring oscillators. But I see that if you google "Neon Joule Thief" and then look at the Images, there are many usable variations shown there. Typically it will use one NPN transistor (high voltage, like a CRT horizontal deflection transistor or similar), a diode, and a home-made inductor wound on a generic ferrite toroid, and a base resistor, and perhaps a capacitor or two. A variable resistor (trimpot) for the transistor Base is a good idea.

[westfw]

I've successfully driven Neon bulbs with essentially this circuit: Single Alkaline Battery Drives White LED
(use high-voltage transistors like MPSA42 instead of 2n3904s.)  I have a PCB layout somewhere, if you want.

The guts from a disposable camera Xenon flash will also happily drive neon bulbs.  (though I guess they're not very USB-power friendly.)

Most of the EL inverters should generate approximately useful power for Neon, and you can get them pre-packaged with USB connectors for "case lighting."   (AC, though.  Rectify if needed.)  This MIGHT apply to the vanishingly tiny inverters aimed at backlights in watches (but it's slightly iffy: some of those only produce about 80V)

Most "ionizer" or CFL circuits are going to generate voltages much higher than you need.

[calexanian]

A simple inductive booster circuit should work. Does not need to be fancy.

[@rt]

If you just pulse DC to them (i.e. 555 and small audio former),
only one electrode in the neon lights, which is easy to see close up.
If you push/pull the transformer with Hbridge, then turns ratio applies and you don’t have enough voltage to light it.
You no longer have an off cycle for natural collapse of the field in the transformer.
There doesn’t seem to be a streamline way to get it right.

[Fungus]

If you just pulse DC to them (i.e. 555 and small audio former),
only one electrode in the neon lights, which is easy to see close up.

a) I that just aesthetic or does that mean they're less bright than they could be? I could arrange the sculpture so you mostly see them from the side...

b) Surely all transformers output AC.

There doesn’t seem to be a streamline way to get it right.

Yep, there's a lot of choices.

Update: I've ordered a couple of those eBay transformers to play with. I might end up building a few of these as gifts so anything pre-built that saves me a couple of mornings work is worth it. I figure I can tap them at different places to get 150V AC, 150V DC, 300V DC...etc. I'll experiment to see what works best.

On a side note, what determines the lifespan of those bulbs? Is it mostly down to build quality?

I think I might add some sort of auto-power-off function so people can't leave them on 24/7 (maybe a capacitive sensor to turn them on for 4 hours when you touch them).

Edit: According to this lamp life is due to damage at the cathode and they last a lot longer on AC than on DC.

[@rt]

It’s not that they aren’t as bright, just looks weird up close.
With the frosted glass colours you could hardly tell:

You can use a 1:1 audio transformer that way.
The circuit is relying on alternate build up & collapse of magnetic field.

Now to light it properly, the same transformer won’t easily do.
I have used a 1:10 transformer here and a H-bridge to use turns ratio to produce 120 Volts.

[Psi]

For $14 you can get one of these

http://www.digikey.com/product-search/en?vendor=0&keywords=811-1583-5-ND

They take 5VDC and produce +24, +48 and +72 output isolated from the input. (there's also a 12V input version)
If you only load the 72V output you can draw up to 42mA.

There is little to no voltage regulation so without any load the 72V output can be as high as 160V or so
But it quickly drops to around 72V once you load it.

With 42mA of available current at 72V and the neons only needing 1mA you could power quite a lot of them. More than enough for your 20-25 lamp project and all contained in one small module.

[Fungus]

They take 5VDC and produce +24, +48 and +72 output isolated from the input. (there's also a 12V input version)
If you only load the 72V output you can draw up to 42mA.

Not a bad option. Certainly easy to use...

Doing the math though: 72V @ 42 mA is 3W - more than standard USB can provide.   

(But still, I don't really need 42mA for what I'm thinking...maybe for version II)

There is little to no voltage regulation so without any load the 72V output can be as high as 160V or so

That's a GOOD thing for neons.

But it quickly drops to around 72V once you load it.

But ... that's a bit of a worry. Those lamps need about 90V to start up so if I connect a lot in parallel some of them might not light up if the voltage gets dragged down.

I don't have any experience yet though, I'm just reading stuff online and trying to figure out what I need. The first bulbs should arrive next week.

[Psi]

Use the 12V input version. It maintains a higher output voltage (which is > 72V guaranteed over the full output power range)

Then take advantage of the fact that they are isolated modules and connect the input to output so the output is ontop of the input. Isolation is cool

If you look at the graph you can see they actually output more than 72V. At 2W they look to be 75V.

So that gives you  75V+12V = 87V   I'm 99.9"% certain a 90V neon will strike reliably with > 87V.