EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: MattHollands on August 11, 2015, 01:17:39 pm
-
Hi! I've built a Nixie clock! (https://keepdevelopingprojects.files.wordpress.com/2015/08/finished.jpg) ::)
I have a slight concern though. At the moment, you could lift up the glass dome and touch the back of the tubes which are operating on a rectified 240Vac supply (ie 340Vpp) (not isolated). However, each tube is in series with a 100k resistor, so even if you were to short the tubes, the peak current is limited at 3.4mA. (You definitely cannot get to the "hot" side of the 100k resistor).
However, I am aware that if you were to touch this with your hand, you would actually be shorting this to earth and I assume that makes a difference because there could be common mode voltage between earth and neutral/live.
So what I want to know is, how much safety does the 100k resistor provide? Does it make any real difference, or will touching the back of the nixie tube be no better than touching the live wire?
[Disclaimer] I have no intention for anyone to ever touch any part of the circuit. I intend to permanently stick down the glass dome any day now because I am confident that the clock is working. Even if you say that it is perfectly safe to touch I will take precautions where ever possible to ensure there is no risk. [/Disclaimer] :phew:
Thanks :)
-
rectified 240Vac supply (not isolated).
Very very dangerous
So what I want to know is, how much safety does the 100k resistor provide?
None whatsoever. If it fails (and they do when there are high voltages across them as anyone who has repaired a vacuum-tube amplifier will tell you) then you have death waiting to happen.
Use an isolation transformer.
If an isolation transformer is not available at reasonable cost, use two standard transformers back to back. The first one steps down to, say, 9V AC, then the second one steps this back up to 240V for the Nixies. The 9V is useful to power everything too.
It's still dangerous, shorting a hand between both outputs of the second transformer will still give you a nasty nip. But adding a high-ish value series resistor between the back-to-back secondaries will significantly limit the current on the output 240V side.
edit - better still, use an AC adapter (a wall-wort that has an AC output) to power the thing, then use the smaller transformer to step the AC up to 240V. Don't know why I didn't suggest this immediately, this is how I built my nixie clock.
Or use a cockroft-walton multiplier. https://en.wikipedia.org/wiki/Cockcroft%E2%80%93Walton_generator
But DON'T rectify the mains!
:phew:
-
I thought so.
But DON'T rectify the mains!
The clock is already built doing that. I built it from Mike's design (http://www.electricstuff.co.uk/nixclock.html (http://www.electricstuff.co.uk/nixclock.html)). But I'm making absolutely certain there is no chance of any touchable wires. Everything is behind wood and glass.
-
I thought so.
But DON'T rectify the mains!
The clock is already built doing that. I built it from Mike's design (http://www.electricstuff.co.uk/nixclock.html (http://www.electricstuff.co.uk/nixclock.html)). But I'm making absolutely certain there is no chance of any touchable wires. Everything is behind wood and glass.
I think what Lewis is implying is that if you are asking these questions then don't build it the way Mike has. It's perfectly acceptable to rectify mains to get the high voltage needed, but not if you are inexperienced.
-
if you are asking these questions then don't build it the way Mike has
Fair enough. I'm still an engineering student and I'm fully aware of the danger of mains but at some point I need to ask these questions. I just wasn't sure what effect the 100k resistor had on safety.
Thank you for the answer :)
-
SteveyG :-+
I think it is a mistake of Mike to post a circuit like that, he knows what he's doing but it's a bit iffy to encourage hobbyists or anyone that hasn't has a nasty belt off the mains to build it. The AC adapter method I mentioned would be a lot safer. I know there's a disclaimer, but.......
He has the diagram for the redneck isolation transformer I was talking about on the page - use that as a matter of course, by default and all the time, not just if you're probing with a scope.
-
None whatsoever. If it fails (and they do when there are high voltages across them as anyone who has repaired a vacuum-tube amplifier will tell you) then you have death waiting to happen.
I would certainly go along with the isolation transformer recommendations. However if you are stuck with the resistor approach, for whatever reason, then you could mitigate the resistor failure risk by using two 51k resistors, or better still, 3 x 33k resistors in series. Here, failure of one resistor would not expose you to full fault current. A single resistor failure (less likely because of the lower voltage stress - 1/4W resistors can be rated as low as 250V) would expose you to worst case 7.4mA or 6.6mA respectively. This should be below the threshold of fibrillation in most (not all!) situations.
To be effective the resistors would need to be securely soldered end to end and enclosed in heatshrink to make an insulated composite part and insulated to prevent any possible exposure to the hot end wiring.
Not a perfect solution, but a lot safer than what you have at the moment.
-
I think it is a mistake of Mike to post a circuit like that, he knows what he's doing but it's a bit iffy to encourage hobbyists or anyone that hasn't has a nasty belt off the mains to build it.
I beg to differ. He already has two warning paragraphs right at the top of the page.
-
If the unit can be considered double insulated, direct mains operation (no safety ground) is fine.
I don't think grounding would matter in this instance, because the potentially exposed surfaces aren't, in turn, protected by metal anyway. Unless you want to put a grounded mesh cage over it, which would probably be ugly.
Whether a glass dome is sufficient, I'm not sure. Impact resistant polycarbonate (or even full-on laminated safety glass) would probably be quite reasonably safe.
I don't think development counts towards UL/etc. Obviously, be careful not to touch it while being grounded, or probe it with grounded equipment (e.g., scopes).
Tim
-
Your basic problem is that you can just lift up the dome. Find a way to fix the dome in place and the clock will be basically safe (not safe enough to sell to the public in the open market, but for personal use).
-
First - a nixie doesn't need 240V to fire; 180V should do fine.
Second, why not supply your circuit with e.g. 9V DC (using an external adapter), and use a simple boost circuit to boost it up to 180V. This is much saver, and quite simple to do. Numberous example boost circuits are only a google search away, and can be very cheaply built (e.g. using an almost-free MC34068, see: https://threeneurons.wordpress.com/nixie-power-supply/hv-supply-kit/). Not only is it saver, it's arguably also a lot more interesting.
Good luck!
-
Anyone can do low power switchers. That's not interesting. Resonant converters are interesting! ;D
(btw. he doesn't have 240 Vdc - he'll end up with >300 Vdc +- ripple after rectifying mains)
-
(btw. he doesn't have 240 Vdc - he'll end up with >300 Vdc +- ripple after rectifying mains)
While it is rectified, it isn't smoothed at all, so there is a 100Hz flicker on the tubes, but that's obviously not visible.
Find a way to fix the dome in place and the clock will be basically safe
This has now been done :) I don't want to sell it.
why not supply your circuit with e.g. 9V DC (using an external adapter), and use a simple boost circuit to boost it up to 180V.
Maybe for V2 ;)
-
A resonant converter in form of a royer converter is a real alternative to a boost / flyback converter. At least it looks more old school.
Powering from a wall wart is still not that simple, as one needs the line frequency signal - so a DC supply will not work.
-
Powering from a wall wart is still not that simple, as one needs the line frequency signal - so a DC supply will not work.
- There are wall warts without rectifier
- Or you could detect the ripple on the DC and use that for a frequency reference
- The circuit uses CMOS 4017 counters to divide the 50Hz (or 60Hz) mains frequency down to 1Hz. Easily replaced by a 32.768 kHz watch crystal and a 4060. Long term stability is probably not as good.
-
Hi! I've built a Nixie clock! (https://keepdevelopingprojects.files.wordpress.com/2015/08/finished.jpg) ::)
I have a slight concern though. At the moment, you could lift up the glass dome and touch the back of the tubes which are operating on a rectified 240Vac supply (ie 340Vpp) (not isolated). However, each tube is in series with a 100k resistor, so even if you were to short the tubes, the peak current is limited at 3.4mA. (You definitely cannot get to the "hot" side of the 100k resistor).
However, I am aware that if you were to touch this with your hand, you would actually be shorting this to earth and I assume that makes a difference because there could be common mode voltage between earth and neutral/live.
So what I want to know is, how much safety does the 100k resistor provide? Does it make any real difference, or will touching the back of the nixie tube be no better than touching the live wire?
[Disclaimer] I have no intention for anyone to ever touch any part of the circuit. I intend to permanently stick down the glass dome any day now because I am confident that the clock is working. Even if you say that it is perfectly safe to touch I will take precautions where ever possible to ensure there is no risk. [/Disclaimer] :phew:
Thanks :)
What you could do is look up the applicable product standard, e.g. for computers EN60950, there will be a different one that applies to household clocks, but all of the ones relating to household appliances are broadly similar. You can probably access these standards in a university library which is worthwhile as they are very expensive otherwise. The university near me has a subscription that allows the standards to be viewed online from their computers.
The standard will probably give a method for measuring touch current, and I suspect that 3.4mA would fail this test meaning that the insulating (glass) cover would need to be affixed such that you can't remove it without tools. There is probably also an impact test, meaning that if the glass could be smashed then you might have to change to a material that passes the impact test, like polycarbonate - but then it would need to pass a flame retardence test, and so on!
If you changed the resistor value so that the touch current would pass the test, then you might also have to ensure that it still passes when any one component has been shorted out - in which case you would have to use two resistors in series that each alone would pass the touch current test. You would likely also have to ensure that each resistor was rated for mains voltage, etc.
Anyway from a legal perspective, it doesn't matter what anyone on here says - just read the standard. If you don't care about the legal perspective, then from a practical perspective, take whatever precautions are necessary to make sure nobody dies - this will depend a lot on the people who are going to be around the device.