Hello from Sydney
Cobra007
My name is Michel, dutch from origin and living in Australia for
nearly 9 years now.
I am an electronics and software engineer by trade, re-design
electronic control systems for mainly pharmaceutical tablet counting
machinery.
These machines have been around for a while (over 40 years) and yes,
the first electronic ones used nixie tubes to display the numbers.
I got a couple of those old counters, and I am actually very impressed
with one of them which is entirely made up of transistors (transistor
BCD counters and transistor BCD to Nixie decoders).
I am also fascinated by wristwatches since I was like 5 years old. Had
a small repair shop called "Kopriso" when I was 12 (see "Kopriso
Watchmaking" on facebook).
At some stage I came across Jeff Thomas and also David Forbes nixie
watches and I thought it was a nice concept, but not very practical
because of the size of these watches. I also think the watch enclosure
from DF doesn't really match the beauty of the tubes and for sure
there is potential in improvement regarding battery life.
I played around with the tubes and batteries and found a way to reduce
the size, improve the design and get better battery life at the same
time. This design is now in a concept stage, I should receive the
boards within a few weeks to try it all out for real (see "Nixie
Watch" on facebook).
A mayor struggling point is the high voltage converter which, like DF
mentioned for his design, only has a 50% efficiency. I got 2 different
designs to try out, one should run between 75 and 85%, the other
between 80 and 90%. This is only possible using very specific
components, so I am really looking forward to do the final
calculations on the results.
Another project I am working on consists of the conversion of 2 old
batch counters (one has 4 tubes, the other 3 tubes) and let them work
together to form a clock with a 0.1second accuracy. This is of course
much easier than the watch, but just as exciting!
Anyway, will also browse through your projects here in the group and
see how creative you all are!
Best regards,
Michel
jb-electronics
welcome to the group! :-)
Your wristwatch sounds very interesting, is there any chance to see some
pictures without joining evil? ;-)
Best regards,
Jens
kay486
if somebody made new tubes using modern technology so they could have
much smaller digits, and being able to have them lit all the time,
just like normal digital watches.
Cant wait to see your design.
Cobra007
http://xiac.com/Images/NixieComp3.jpg
I am still in doubt about the watch band size, I think 26mm is too
big, so I am thinking of 22 or 24mm.
Best regards,
Michel
On Jan 29, 12:01 am, jb-electronics <webmas...@jb-electronics.de>
wrote:
Cobra007
exists so it's going to cost quite a bit to setup a new production
line for micro sized nixie tubes. I can't really see the micro nixie
tubes being used in anything else but a retro style application, so
the market for a tube like that will be very small.
jb-electronics
the NL-4998 which is basically the tiniest Nixie tube ever made. The
problem is the glow discharge, it is always larger than the cathode, and
you cannot make the cathodes too small because otherwise you wouldn't be
able to discriminate between the digits - it would be much too blurry.
Panaplex displays are much more suited to these dimensions, sadly. ;-)
Jens
Cobra007
the height of the original ones (12.5mm versus 25mm), you would end up
with a tube that had great potential for wristwatch applications!
On Jan 29, 8:51 am, jb-electronics <webmas...@jb-electronics.de>
wrote:
jb-electronics
The product of pressure and electrode distance determines the striking
voltage (Paschen's curves). It is roughly U_strike ~ [p*d]^2 where p is
pressure and d is distance. If you decrease the distance, then you have
to use a higher pressure to obtain the same striking voltage. The higher
pressure condenses the glowing and makes it not as visible as before.
The alternative - a higher striking voltage with the old pressure - is
also not a good idea since you will have sparkovers in your tube that way.
I am not saying it is impossible, I just want to make the point that
most dimensions found in a commercially made Nixie tube are results of
long development.
Jens
Tidak Ada
such a phosphor exists for that purpose). But I am afraid the construction
of an VFD display is way more complicated...
eric
Jens
--
You received this message because you are subscribed to the Google Groups
"neonixie-l" group.
To post to this group, send an email to neoni...@googlegroups.com.
To unsubscribe from this group, send email to
neonixie-l+...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/neonixie-l?hl=en-GB.
Cobra007
much to these distances and pressures. For example, if you would let
the cathode pins come out of the side of the tube rather than the
bottom, you'd probably save about 7-8mm already. So you don't need to
modify the current grid setup too much in order to come down to
something close to 12.5mm.
Michel
On Jan 29, 9:26 am, jb-electronics <webmas...@jb-electronics.de>
wrote:
Charles MacDonald
> May be, besides PanaplexR, VFD with an orange phosphor is an alternative (if
> such a phosphor exists for that purpose). But I am afraid the construction
> of an VFD display is way more complicated...
Back in the 1970's I worked in a shop that sold calculators, there were
lots of calculators which had a small 8 digit VFD display Some of them
used a bunch of tubes in a row.
the US*4 is an example of the multi-tube design
http://www.vintagecalculators.com/CommodoreUS4_1_b.jpg
Many had a all in one tube
http://www.vintagecalculators.com/CommodoreGL996R_2.jpg
A "watch VFD display" could probably be made as easily as a watch nixie.
--
Charles MacDonald Stittsville Ontario
cm...@zeusprune.ca Just Beyond the Fringe
http://users.trytel.com/~cmacd/tubes.html
No Microsoft Products were used in sending this e-mail.
David Forbes
> Hello everybody here in the neonixie group.
>
> a small repair shop called "Kopriso" when I was 12 (see "Kopriso
> Watchmaking" on facebook).
> At some stage I came across Jeff Thomas and also David Forbes nixie
> watches and I thought it was a nice concept, but not very practical
> because of the size of these watches. I also think the watch enclosure
> from DF doesn't really match the beauty of the tubes and for sure
> there is potential in improvement regarding battery life.
>
> I played around with the tubes and batteries and found a way to reduce
> the size, improve the design and get better battery life at the same
> time. This design is now in a concept stage, I should receive the
> boards within a few weeks to try it all out for real (see "Nixie
> Watch" on facebook).
Michel,
Interesting work. The Nixie watch has amazed me in that no one has made
a similar design in the seven years since I made mine and published the
source code and schematics on the Web.
I'm also working on a smaller Nixie watch. There are batteries made for
Bluetooth that work well for this job. Recharging is something that
people are used to these days, what with phones and iPods etc.
I'm naturally driven by the Steve Jobs design ethic. Look up the iPod
aquarium story to see what I mean.
I'm still figuring out the buttons and power nibs to charge the watch.
We'll see what happens. A 3d printer figures in somehow.
--
David Forbes, Tucson AZ
dr pepper
I have a remotely similar background, only my experience for the past
15 years or so is cnc machinery installation and maintenance,
computerised lathes and mills, some of the very first digital readouts
used nixies and theres still the odd one in use.
I've made various time keeping projects including nixie clocks,
however watches arent my thing.
On Jan 29, 5:25 am, Cobra007 <mic...@xiac.com> wrote:
> Hello David,
>
> It is quite interesting but I came across your website while searching
> for a custom watch case for my domino watch, I think I sent you an
> email about that, or maybe I didn't actually send it out. At first I
> was quite interested in the nixie watch, I thought it was a nice
> concept and actually wanted to order one. That was until I read the
> story about battery efficiency and opening up the case to change the
> time. You wrote somewhere that it was difficult to achieve a better
> than 50% efficiency with off the shelve components, which is true, but
> it is not an off the shelve watch so why would you use off the shelve
> components? Another thing is that it was very big and I remember
> reading somewhere that someone called it a "hockey puck on a strap"
> which pretty much described how I looked at it. So I decided not to
> buy it but just see if I could make something better myself.
>
> While designing my domino watch, you stumble across issues that you
> would never worry about if the circuit wasn't battery powered, and it
> is challenging to make it all work and achieve excellent battery
> lifetime at the same time.
>
> I have seen you used some open source software in your watch, that is
> however not the most efficient way of software watch design. If
> someone would occasionally wear the watch, it would still drain the
> battery, no matter what. One battery will last for about 7 years if my
> watch is worn only 1 day per week and will last at least one year with
> daily usage. For this reason, it is not of much use to put a
> rechargeable battery in there. I'm not a fan myself to have a
> rechargeable battery in a watch, so for sure my nixie watch won't have
> one.
>
> Those Apple ipod watches ..... no, not something for me. I just think
> you can only call a device a wristwatch if the battery last for at
> least one year (without recharging), otherwise you just have a clock
> (or something else) attached to a strap :-).
>
> Anyway, looking forward to see your new design as well, I hope it is
> something more creative than the 3 1/2 digit one that Jeff Thomas was
> working on already.
>
> Best regards,
> Michel
>
> - Show quoted text -
David Forbes
Hi. I'd like to clear up a couple things that you don't seem to
understand about my Nixie watch design.
On 1/28/12 10:25 PM, Cobra007 wrote:
> Hello David,
> ...
> I have seen you used some open source software in your watch, that is
> however not the most efficient way of software watch design.
I do not know what you mean by this. I wrote the software myself, then
published it under the GPL. That is why it is open source - I opened it
myself.
Have you read my watch code? It's as efficient as I could make it. Feel
free to point out anything in the code that seems wasteful to you.
> If someone would occasionally wear the watch, it would still drain the
> battery, no matter what. One battery will last for about 7 years if my
> watch is worn only 1 day per week and will last at least one year with
> daily usage.
I'd be interested in seeing your cell life calculations, since you make
a big deal of how much more efficient your circuit is.
My 16LF722 processor draws about 10 uA while running at 32 kHz, and the
ADXL335 accelerometer draws 350 uA for about 1.5 msec, activated 5 times
a second, for an average current of 2.6 uA. That 13 uA average current
draw would use up the 750mAH CR2 cell in 7 years, except that the cell
will drain itself in about 7 years. So to be realistic, I expect the
cell to last 3-4 years without viewing.
The HV power supply uses about 200 mA to light the tubes, which are lit
for two seconds per viewing. This gives power for 6750 viewings, which
is four months at 50 viewings per day. That's more than one viewing per
hour. After all, how can a person resist reading their Nixie watch over
and over again?
I have lost track of how long the battery in my watch lasts. I just grab
a cell from the drawer of cells that I keep on hand to test watches
before shipment, so I never know how fresh the cell that I just put in
my watch really is. Perhaps some of my customers could chime in.
Frank Bemelman
Well, I am not so amazed that we have not seen
a similar design yet. Just too many problems for the
average guy.
Did you ever see the ipod nano act as a nixie watch, here
is a picture...
http://www.apple.com/ipodnano/features.html#clock-gallery-13
All the best,
Frank
-----Oorspronkelijk bericht-----
From: David Forbes
Sent: Sunday, January 29, 2012 4:08 AM
To: neoni...@googlegroups.com
Subject: Re: [neonixie-l] Hello from Sydney
Interesting work. The Nixie watch has amazed me in that no one has made
a similar design in the seven years since I made mine and published the
source code and schematics on the Web.
I'm also working on a smaller Nixie watch. There are batteries made for
Bluetooth that work well for this job. Recharging is something that
people are used to these days, what with phones and iPods etc.
I'm naturally driven by the Steve Jobs design ethic. Look up the iPod
aquarium story to see what I mean.
I'm still figuring out the buttons and power nibs to charge the watch.
We'll see what happens. A 3d printer figures in somehow.
--
David Forbes, Tucson AZ
--
kay486
better if anybody could program it. Ive been playing with the idea of
making realistic CGI nixie for a while.
On Jan 29, 9:15 pm, "Frank Bemelman" <bemel...@franktechniek.nl>
wrote:
>
> Well, I am not so amazed that we have not seen
> a similar design yet. Just too many problems for the
> average guy.
>
> Did you ever see the ipod nano act as a nixie watch, here
Cobra007
case doesn't match with the "tubes". If you wear that one, nobody will
ever suspect that this might be a nixie watch :-)
On Jan 30, 8:15 am, "Frank Bemelman" <bemel...@franktechniek.nl>
wrote:
>
> Well, I am not so amazed that we have not seen
> a similar design yet. Just too many problems for the
> average guy.
>
> Did you ever see the ipod nano act as a nixie watch, here
>
> All the best,
> Frank
J Forbes
> I have lost track of how long the battery in my watch lasts. I just grab
> a cell from the drawer of cells that I keep on hand to test watches
> before shipment, so I never know how fresh the cell that I just put in
> my watch really is. Perhaps some of my customers could chime in.
have two, one is the copper case prototype, the other is a sort of
early production black watch. I don't wear a watch all the time, I
expect someone who wears it every day would need to change the battery
more often. It's never been a problem for me to replace the battery
when needed.
I wonder how long the battery lasted in David's first digital watch,
an LED Timex he got in the mid 70s?
Jim Forbes, near Tucson, AZ
JohnK
wonder if you have picked up some of the local colloquialisms yet - 'having
tickets on yourself' and 'having someone load the gun for you but you
pulling the trigger' come to mind.
John Kaesehagen
Australia
[PS... Do a bit more research into the dynamics of this group :-)) ]
----- Original Message -----
From: "Cobra007" <mic...@xiac.com>
To: "neonixie-l" <neoni...@googlegroups.com>
Sent: Monday, January 30, 2012 8:48 AM
Subject: [neonixie-l] Re: Hello from Sydney
> Hello David,
>
> Thanks for the explanation. I guess you made a typo somewhere because
> I remember reading 200uA somewhere in your files rather than 13uA and
> 1 year rather than 7 years.
>
> Your claim is then not correct, because at 50 readings per day and 2
> seconds per viewing, the battery will only last less than 3 months.
>
> I didn't know you wrote the software yourself, I though you played
> around with someone else's software.
>
> It's truly going to be 16.5mm, including the components, yes. Your're
> also correct about the multiplexing, I do that so that it is easier to
> dim the tubes by means of PWM, but not in the way you think I will do
> it :-).
>
> I actually thought you had invented this nixie watch, so I didn't want
> to re-design it as it was your "thing". In one of the videos that I
> guess you had made up by a pommy it is stated that woz said it was the
> one thing he wished he had invented. So it really seemed it was your
> invention. However, when I came across Jeff Thomas and exchanged a
> couple of emails, it was clear that he had invented this nixie watch a
> couple of years earlier, so I didn't bother about designing another
> one and he didn't mind himself either. What I do think is very lame is
> to just take his later 3 1/2 digit tube design as well and try to make
> a buck out of it. http://www.cathodecorner.com/nw35/index.html I
> verified with him if he was aware of this latest Cathode Corner copy
> of his design, and he personally confirmed that he wasn't.
>
> I know this world is a dog-eat-dog business, but I guess these people
> are supposed to be your friends, right?
>
> Best regards,
> Michel
>
>
David Forbes
> Hmmm, I am in Australia too btw and have followed since your first post.
> I wonder if you have picked up some of the local colloquialisms yet -
> 'having tickets on yourself' and 'having someone load the gun for you
> but you pulling the trigger' come to mind.
>
> John Kaesehagen
> Australia
> [PS... Do a bit more research into the dynamics of this group :-)) ]
>
We have resolved this matter privately.
Move along, nothing to see here.
David Forbes
> Hi David,
>
> Well, I am not so amazed that we have not seen
> a similar design yet. Just too many problems for the
> average guy.
>
> Did you ever see the ipod nano act as a nixie watch, here
> is a picture...
> http://www.apple.com/ipodnano/features.html#clock-gallery-13
>
> All the best,
> Frank
I did see that. It has rather small tubes, eh?
There was an amusing video interview a few months back, of the Woz
receiving an honorary degree in Montreal. He was explaining to the
school bigwigs (everyone in their Ph.D suits, of course) how he wears an
iPod nano and a Nixie watch, to have the best of both worlds.
Cobra007
can't take back the things you post so I am sorry for that. I agree
that I should have done more research and this would not have
happened.
Like David said, we have solved this matter privately already, which
is good and we're both looking forward to each other's designs.
Anyway, I am not a bad guy, I just seem to upset people sometimes
without a grounded reason, that is basically the nature of a cobra :-)
Best regards,
Michel
threeneurons
> Well, I probably started on the wrong foot in this group and yeah, you
> can't take back the things you post so I am sorry for that. I agree
> that I should have done more research and this would not have
> happened.
>
enthusiastic sometimes.
As for your watch concepts, the big hurdles are packaging and power.
Coding a uC is really a non-issue. Its really arranging the tubes,
batteries, and the few components, in as small of a 3D space as
possible, with the tube still showing the digits. I'm really curious
about your HV power supply. I'm assuming you're going to use 3V. There
are uCs that can work down to 1.8V, as well as 5V. That same 3V can
power the uC directly, and pump the HV switcher. To get most of the
battery life, I imaging the switcher will still lite the tubes down at
1.8V in, and do it with reasonable efficiency. So far only Jeff &
David have done it. If it was easy, everybody would do it.
Please share your progress. This is after all a hobby group, and we
are on this web (still think its only a passing fad). I give high
brownie points to those who share their ideas. All the info on my
nixie projects, are free for anyone to make, given they actually want
to. Just Google threeneurons nixie.
Cobra007
voltage, the lower the currents will be. This will reduce some of the
losses in the 'HV switcher'. I know that most 1/2AA batteries cannot
supply the required current, but a few can and it's a good choice as
they are smaller than a CR2. Of course that is not the whole story,
the circuit itself is quite complex as I couldn't find a single IC
that would do the same job.
Apart from that, I added a daylight sensor and will PWM the tubes
according to the amount of daylight. I used the same circuit in
another design with LEDs. Believe it or not, but I went from 5mA in
bright daylight down to 50uA (40dB difference) in pure darkness,
always keeping the LEDs on a constant contrast compared to the amount
of ambient light.
I don't expect to get 40dB out of the nixie tubes, but I think 20dB
can be achieved. This would bring the current down from 200mA in
bright daylight to 20mA at night, keeping the display always readable.
Even 40mA would be good enough as this will already significantly
improve the efficiency of the HV circuit.
The processor will be put into deep sleep mode between readings
(drawing about 1uA) and wake-up either by a soft touch button or an IR
motion sensor. Not sure if I can fit the IR motion sensor on the final
design, but I try my best.
So all together, it will be very low power.
Best regards,
Michel
David Forbes
> I actually start off with a 3.6V 1/2AA battery, the higher the input
> voltage, the lower the currents will be. This will reduce some of the
> losses in the 'HV switcher'. I know that most 1/2AA batteries cannot
> supply the required current, but a few can and it's a good choice as
> they are smaller than a CR2. Of course that is not the whole story,
> the circuit itself is quite complex as I couldn't find a single IC
> that would do the same job.
It is nice to use a battery that can be purchased locally. The CR2 is
available worldwide. The 1/2AA is a special item. Customers don't like
to have to special-order a battery on short notice.
>
> I don't expect to get 40dB out of the nixie tubes, but I think 20dB
> can be achieved. This would bring the current down from 200mA in
> bright daylight to 20mA at night, keeping the display always readable.
> Even 40mA would be good enough as this will already significantly
> improve the efficiency of the HV circuit.
One thing about Nixie tubes that you may not be aware of is that they
need some time to start firing, because the plasma is triggered by an
external source of energy such as a photon from the room lighting or a
stray cosmic ray. The less ambient light there is, the more time they
need to start. If you reduce the duty cycle in the dark, you may find
that the display doesn't start reliably.
This is why some Nixie tubes were made with radioactive krypton in the
tube. But not the 5870 series, as they were designed to be used in an
office environment.
I would suggest experimenting with this dimming idea before committing
to it. You would hate to discover later that your watch display is
unreliable.
It may be possible to solve this problem by detecting when the current
starts to flow in the tube, and leave the tube powered up until the
current starts, then begin the PWM operation.
JohnK
tube. But not the 5870 series, as they were designed to be used in an
office environment."
Probably in bad taste, but how about getting the cases made in Japan? [Or
Russia].
JK
Cobra007
Humm, what time difference do we have? seems like they're up in Tucson
around the same time as here in NSW!
I did in fact take all of that into account, that's why I mentioned
I'd be happy to get 14dB dimming out of it (20 - 6)
The battery can be purchased through element14 or newark as you call
it in USA and only costs a few bucks. If you only need to replace it
every 2 years or so, I don't think that will be an issue for anyone.
Best regards,
Michel
Frank Bemelman
Seems that you know everything. It also seems
that your only interest is to show off how good you are
and how stupid everybody else is.
The proof of the pudding is in the eating. Sofar you
have only shown a meaningless 3d rendering of a
board and a lot of crap talk.
You would do _me_ a favour if you shut up until you have
something solid to report.
Frank
-----Oorspronkelijk bericht-----
From: Cobra007
Sent: Monday, January 30, 2012 9:04 AM
To: neonixie-l
Subject: [neonixie-l] Re: Hello from Sydney
Cobra007
:-) Well, Frank Bemelman from Lisse (which is about 5km from where I
was born). I think there is a difference between 'showing off' or
being passionate about something.
I do not just start designing a circuit without a proper background
check of the used technology. Most things I have tested already, just
not everything at the same time, hence the concept board.
Anyway, the boards should be ready soon, so I will have something
solid to report afterwards or do you the requested favor. Deal? :-)
Best regards,
Michel
On Jan 30, 7:42 pm, "Frank Bemelman" <bemel...@franktechniek.nl>
wrote:
Mefistofelix
Nick
Lets stick to the technical stuff and positive contributions. Only.
Moderators
GastonP
On Jan 30, 4:49 am, David Forbes <dfor...@dakotacom.net> wrote:
> On 1/30/12 12:06 AM, Cobra007 wrote:
>
> need some time to start firing, because the plasma is triggered by an
> external source of energy such as a photon from the room lighting or a
> stray cosmic ray. The less ambient light there is, the more time they
> need to start. If you reduce the duty cycle in the dark, you may find
> that the display doesn't start reliably.
<snip>
> tube. But not the 5870 series, as they were designed to be used in an
> office environment.
Somewhere buried in the archives (or a reference to an article, that
was specific to trigger tubes) is a nice calculation that shows that
by this time, all of the Kr85 has decayed so no kick-start help either
for those radioactively doped ones.
<snip>
> starts to flow in the tube, and leave the tube powered up until the
> current starts, then begin the PWM operation.
> --
> David Forbes, Tucson AZ
IMO, as the start of the digit ionization is what is unreliable we
would have a train of pulses of known duration but with random
separation (period, frequency, etc.). You know well that a PWM light
intensity modulation is based on turning on and off a digit, so once
the digit is turned off for the next cycle we will have to wait again
for a random time.
This is unless I got all wrong and only the first start of the
discharge is the unreliable one.
One alternative, discussed before IIRC is to use the decimal point as
a bias cathode... if fully lit of course it will upset all digit
voltages, but perhaps it could be used only as bias at low current as
some of the trigger tubes did...
Gaston
Lucky
don't think (I hope our fellow forum members) there is any need to be
rude! How about doing our best to keep it polite and friendly?
Michel I wish you the best in your venture and look forward to seeing
your results, you are obviously very passionate about your work, I
would suggest you try to be more empathetical/sensitive to others
work. I get where you're coming from mate (the ex aussie is coming out
in me now lol) others maybe a bit more more sensitive.
John Rehwinkel
I was thinking about this and display efficiency. Obviously, efficiency would improve if anode resistors are dispensed with, but then it's left to the power supply to monitor and control current through a negative resistance. David's watch does this directly, by incorporating current feedback into the power supply. However, David's watch is direct drive. I was thinking about how to do this with a multiplexed watch.
What I came up with is a little nuts, and probably would do horrible things to tube life, but I haven't tried it yet. I was thinking of using a circuit like the General Radio 1538 Strobotac, which has a cute way of building up charge for its flashtube by running a flyback type converter with a pulse train, charging up a storage capacitor incrementally until it has the desired amount of energy (they also claim nearly 100% efficiency, as their device can operate from batteries as well). This would be an interesting concept for a multiplexed nixie display: configure the cathode drivers, then dump a packet of charge into an anode capacitor. If the voltage is high enough, ionization should be gratifyingly fast. Since the capacitor is small, the total amount of energy would be limited. However, this would be running a nixie rather like a flashtube, with brief, high-current pulses. The duty cycle would be tiny, but the overload would be great. What this would do to tube lifetime, I don't know. It might be just fine, it might blast the cathode to pieces in hours. It would also be hard on the cathode drivers, but you can get SCRs in SOT-23 packages that can withstand 7A repetitive pulses (STM P0102BL, for example).
- John
JohnK
Maybe monitor the current and switch it off as some of the H-bridge motor
drivers do? Except that uses some wasteful circuitry to achieve it.
John K.
----- Original Message -----
From: "John Rehwinkel" <jre...@mac.com>
...clip....
I was thinking about this and display efficiency. Obviously, efficiency
would improve if anode resistors are dispensed with, but then it's left to
the power supply to monitor and control current through a negative
resistance. David's watch does this directly, by incorporating current
feedback into the power supply. However, David's watch is direct drive. I
was thinking about how to do this with a multiplexed watch.
What I came up with is a little nuts, and probably would do horrible things
to tube life, but I haven't tried it yet. I was thinking of using a circuit
like the General Radio 1538 Strobotac, which has a cute way of building up
charge for its flashtube by running a flyback type converter with a pulse
train, charging up a storage capacitor incrementally until it has the
desired amount of energy (they also claim nearly 100% efficiency, as their
device can operate from batteries as well). This would be an interesting
concept for a multiplexed nixie display: configure the cathode drivers, then
dump a packet of charge into an anode capacitor. If the voltage is high
enough, ionization should be gratifyingly fast. Since the capacitor is
small, the total amount of energy would be limited. However, this would be
running a nixie rather like a flashtube, with brief, high-current pulses.
The duty cycle would be tiny, but the overload would be great. What this
would do to tube lifetime, I don't know. It might be just fine, it might
blast the cathode to pieces in hours.
...clip...
David Forbes
>
>
> Hi David...
> IMO, as the start of the digit ionization is what is unreliable we
> would have a train of pulses of known duration but with random
> separation (period, frequency, etc.).
> discharge is the unreliable one.
My experience is that it is only the first pulse that has trouble
starting. After you get the tube to light, there are enough ions
rattling around in side it that subsequent PWM pulses work just fine.
But this may not be the case for very short PWM pulses, as the
oscilloscope reveals a millisecond or two of variation in the voltage ->
ionization delay in a multiplexed tube.
John Rehwinkel
I'd be interested in seeing how much that variation varied with voltage. I'd also be interested in seeing if "priming" the tube helped. Eventually, I'll probably hook everything up and try it myself, but I'm deep in another project at the moment.
- John
David Forbes
> Id really like to see the finished watch! It would be really awesome
> if somebody made new tubes using modern technology so they could have
> much smaller digits, and being able to have them lit all the time,
> just like normal digital watches.
>
> Cant wait to see your design.
>
I'm not sure what modern technology you are thinking of. LCD, perhaps?
Nixie tubes use a lot of power because they make their own light.
Lightning in a bottle takes juice. There's no way around that.
David Forbes
> Hello everybody here in the neonixie group.
>
> My name is Michel, dutch from origin and living in Australia for
> nearly 9 years now.
> I played around with the tubes and batteries and found a way to reduce
> the size, improve the design and get better battery life at the same
> time. This design is now in a concept stage, I should receive the
> boards within a few weeks to try it all out for real (see "Nixie
> Watch" on facebook).
>
I am interested in your thoughts on your proposed tube mounting method,
which appears from your renderings to be soldering the tubes to the PC
board through tiny plated through holes, with the tubes pressed against
the PC board.
It is safe to say that anyone who has worked with vacuum tubes will tell
you, "Don't do it!"
Here are three reasons:
1. The tube leads are not consistently plated at the tube exit point, so
expect to not be able to use more than half of the tubes that you buy.
I've personally examined the lead plating on a thousand of these tubes
under a microscope. Some of them are shiny, and others look like they
have spent their life in the salt air.
2. Soldering the tube that close to the glass will produce thermal
stress that is likely to vent the neon on some noticeable fraction of
the tubes you install. This, combined with #1, will make for a very
frustrating assembly experience.
3. You will not be able to replace a tube. The tubes do go bad now and
then, so replacement is necessary. There is no way to extract the tube
from the PC board if you have to remove the solder from 14 tiny holes
and break the lead-to-hole solder connection on all of them at once. You
can try sucking out the solder, but you won't get it all. Plated-through
holes will pull out, and the board will be ruined. Again, I speak from
experience removing parts from PC boards.
A better way would be to cut a slot in the PC board where the mass of
leads passes through, and fold the leads over to both sides, soldering
them to narrow rectangular pads like a surface mount device. That would
solve all three of the above problems.
threeneurons
too, was born in the Netherlands. About 10mi (15Km) north of Eindhoven
(Phillips HQ). Though I've spent 50, of my 53 years, here in the Los
Angeles area. Don't send me any correspondence in Dutch. Most 3-year
olds can't read, and I was still not too keen on that whole potty
training thing. Still think its a women's conspiracy.
GastonP
On Jan 30, 11:21 am, John Rehwinkel <jreh...@mac.com> wrote:
<snip>
> - John
and keep a low ionization current from anode to "glass" like the
outgassing testers or plasma balls do ?
BTW when I was a kid, my father gor a Strobotac and it was mesmerizing
to look at while "stopping dead" a fan, while still kept blowing the
air :)
Gaston
Cobra007
I think that if the tubes wouldn't strike for whatever reason, those
problems can always be solved in the software. Another thing that
might help striking them quicker in darkness is the 3 backlight LEDs,
I can let a small current flow through them while the tubes are off.
Since they are very close to the tubes, there will be lots of photons
entering the tubes. The nature of the design is that should the tubes
not immediately strike, the voltage will slightly increase (up to
about 10%) until they do.
Anyway, I have not tested this in every detail, but I trust that there
is always a solution around the corner in case something doesn't work.
Michel
Cobra007
> you, "Don't do it!"
>
> Here are three reasons:
>
> 1. The tube leads are not consistently plated at the tube exit point, so
> expect to not be able to use more than half of the tubes that you buy.
> I've personally examined the lead plating on a thousand of these tubes
> under a microscope. Some of them are shiny, and others look like they
> have spent their life in the salt air.
>
> 2. Soldering the tube that close to the glass will produce thermal
> stress that is likely to vent the neon on some noticeable fraction of
> the tubes you install. This, combined with #1, will make for a very
> frustrating assembly experience.
>
> 3. You will not be able to replace a tube. The tubes do go bad now and
> then, so replacement is necessary. There is no way to extract the tube
> from the PC board if you have to remove the solder from 14 tiny holes
> and break the lead-to-hole solder connection on all of them at once. You
> can try sucking out the solder, but you won't get it all. Plated-through
> holes will pull out, and the board will be ruined. Again, I speak from
> experience removing parts from PC boards.
think it can be done without doing damage. For sure I would prefer the
cut out slot as you mentioned below. This board is only for trial, and
I won't know unless I try. I have used quite a few vacuum tubes before
and always in sockets. What I remember is that you can't solder a wire
to the pins of those tubes, or at least I couldn't do it with the
tools I used at that time.
jb-electronics
you need something to ionize the gas atoms inside your glass bulb. I am
not sure if a simple LED will do the trick, usually you need energized
particles in the order of 60eV, that is electron volts, and a blue LED
will only give you 3 eV.
Jens
Cobra007
On Jan 31, 8:18 am, jb-electronics <webmas...@jb-electronics.de>
wrote:
>
> you need something to ionize the gas atoms inside your glass bulb. I am
> not sure if a simple LED will do the trick, usually you need energized
> particles in the order of 60eV, that is electron volts, and a blue LED
> will only give you 3 eV.
>
> Jens
>
so if the tube was designed for an office environment, from which
source would it receive this 60eV?
jb-electronics
Jens
Cobra007
wrote:
> Cosmic rays :-)
>
> Jens
>
So basically you're saying it is more difficult to strike the tube at
night time because the sun has gone under?
jb-electronics
sky and your tube thus diminishing the rate of ionizing radiation.
But yes, some parts of this ionizing radiation are caused by protons and
alpha particles that arrive here from the sun at decently high energies.
They cause particle showers in our atmosphere and those particles then
pre-ionize our beloved Nixie tubes. Naturally, when earth is blocking
the way, those primary solar particles cannot reach us, and all we are
left with is the rest of cosmic rays that will most probably not do the
job at all times.
Jens
Cobra007
I put the tubes in a permanent dark room, they should then strike
quicker during the day and slower during the night.
At the same time, I can see if lighting the backlight LEDs will bring
any change in that. It could be that many 3eV particles will have a
similar result as a few 60eV particles, it's all in quantum mechanics
I suppose.
On Jan 31, 9:03 am, jb-electronics <webmas...@jb-electronics.de>
wrote:
jb-electronics
> It could be that many 3eV particles will have a
> similar result as a few 60eV particles, it's all in quantum mechanics
> I suppose.
quantum physics tells you that it actually will not work, sadly. You
need to have just the right energy in one particle (photon, that is).
Jens
Cobra007
fact. Nevertheless, the LED test is simple to try out and may still
have some positive effect for reasons previously unknown :-)
On Jan 31, 9:17 am, jb-electronics <webmas...@jb-electronics.de>
wrote:
Nick
Nick
GastonP
wrote:
> Cosmic rays :-)
>
Actually, background radiation is enough, and that comes from a
variety of materials that surround us. But the usual source is just
light photons. Lots of it.
What happens is that it's all statistical... the less the light, the
less probable a photon will help start the ionization. And the more
dependent we are on high energy natural particles.
There's a project of ring-counter based nixie clock whose author
helped the neon lamps he used as active devices to reliably trigger
through the use of a couple blue leds.
One could perhaps select nixies based on low light condition
starting...
OTOH, cosmic rays are not bothered much for a couple flimsy concrete
blocks :)
Cobra007
cheese at all, not on bread, not in pizza, not on a hamburger, it is
all not not not :-). I only make 1 exception, and that is on a flight
if I feel terribly hungry and the only thing they have is a ham-cheese
sandwich, I will eat it (just for the energy) but keeping my nose
closed :-).
> Most 3-year
> olds can't read, and I was still not too keen on that whole potty
> training thing. Still think its a women's conspiracy.
jb-electronics
> OTOH, cosmic rays are not bothered much for a couple flimsy concrete
> blocks :)
you are right about that :-) But during night time the sun does not work
as a decent source anymore because 12000km of stone are a pretty good
shield, at least for solar particles. Then again, Nixie tubes have
problems in striking in dark rooms at daytime as well. So I guess it is
not the solar particles after all :-).
But it has to be something different from light that is scarce in dark
rooms since gases have an ionization energy of roughly 60...100eV and
ordinary light photons will not do the trick.
Jens
John Rehwinkel
Since it's easily demonstrable that visible light does encourage ionization, it is apparent that photons do not have to carry the full ionization energy to help. Given this, it's worth thinking about how the process could occur.
I suspect that passing photons just kick electrons to higher energy orbits*, thereby reducing the amount of energy required for full ionization (completely dissociating an electron from an atom). Further, this process can be repeated, with additional photons kicking the electrons to higher orbits. Via this process, you can gain ionization by a series of small steps. Quantum mechanics does not require it to happen all at once, just that no forbidden energy states are occupied.
This is all made less clear by an energy level diagram:
http://raptor.physics.wisc.edu/talk/neon_e.gif
* effectively reversing the light-emitting situation, and using the same energy, so the energy required would be the same as the photons emitted by neon - in other words, plain old orange photons will suffice.
- John
David Forbes
David Forbes
http://www.cathodecorner.com/
jb-electronics
yes and no. You are of course right about the wavelengths. What I meant
earlier was the fact that a photon has to have just the right energy to
excite an electron in a shell of an atom. If the gap to the next level
is, say, 0.25 eV, and your photon has 0.30eV, there is a good chance
nothing will happen at all in a statistically relevant number. That is
the nature of quantum physics. Photons can either interact or not
interact, at least in this setting at these energies.
What I completely forgot, though, is what John pointed out. 60eV is the
range of the ionization energy, sure. But we do not need to knock out an
electron from the very core, it is enough to knock out an electron from
a higher shell that has an energy that is closer to zero. So ordinary
light, including many many wavelengths, will provide the necessary
photon for some of these transitions. We won't need a 60eV photon for
that. I stand corrected :-)
Jens
Cobra007
On Jan 30, 5:08 pm, threeneurons <threeneur...@yahoo.com> wrote:
> To get most of the
> battery life, I imaging the switcher will still lite the tubes down at
> 1.8V in, and do it with reasonable efficiency. So far only Jeff &
> David have done it. If it was easy, everybody would do it.
>
> Please share your progress. This is after all a hobby group, and we
> are on this web (still think its only a passing fad). I give high
> brownie points to those who share their ideas. All the info on my
> nixie projects, are free for anyone to make, given they actually want
> to. Just Google threeneurons nixie.
Your clocks are great, I actually never heard of decatrons until I
started digging deeper in the nixies :-)
You could convert a battery voltage to a nixie voltage relatively easy
with a MAX771 circuit. It will only not give you the maximum
efficiency. What I have done is basically design a circuit (using
multiple ICs) that together function in a similar way as the MAX771,
but slightly different in order to increase the efficiency. So it's a
bit more complicated than just using 1 IC, but I hope it pays off the
effort.
Once finished, I am going to compare my circuit with the standard
MAX771 and see how the two compare to each other.
Michel
Nick
> You could convert a battery voltage to a nixie voltage relatively easy
> with a MAX771 circuit. It will only not give you the maximum
> efficiency. What I have done is basically design a circuit (using
> multiple ICs) that together function in a similar way as the MAX771,
> but slightly different in order to increase the efficiency. So it's a
> bit more complicated than just using 1 IC, but I hope it pays off the
> effort.
deprecated) and has been so for several years (maybe 10?). The MAX1771
is its replacement. The 771/1771 are not good at low input voltages -
its a topology that would need bootstrapping, which then begs the
question: "Why use it in such a specialised design". As size is such a
key issue here, I would probably go for something that switched in the
MHz range so the inductor is far smaller. The TPS61040/61041 are
really nice devices that I've used a few times - http://www.ti.com/product/tps61040
http://www.ti.com/lit/ds/slvs413f/slvs413f.pdf .
Have a look at this EDN Design Ideas - "Cascode MOSFET increase boost
regulator's input and output voltage ranges" for some ideas... One
basic problem with straight boost converters of this topology is that
they require a high voltage FET and the physics of such FETs means
that they are generally not logic-level driven and mechanically large
- other topologies work far better in this application - have a look
at "flyback converters" - they use low voltage FETs with low Vgs and
which are physically small.
Nick
Cobra007
just wonder what the total efficiency would be, according to the
datasheets it runs at about 80 - 85% in standard configuration, so the
cascoded circuit is probably 70-75% as you loose quite a bit of
efficiency in the gate charge @ 1MHz and the on resistance is not all
that low either (around 260mOhm @ 6V).
I can't change the current design of the concept boards as they are at
the PCB manufacturer already. I do expect to reach a higher efficiency
than 80% in my design, but we'll see.
Michel