A huge 7 Segment display made out of 23 IN-28 tubes? Wack.

[celzey11]

https://www.ebay.com/itm/PCB-with-23-pieces-IN-28-nixie-tubes-slightly-used-working/114333271792?_trkparms=aid%3D1110002%26algo%3DSPLICE.SOI%26ao%3D1%26asc%3D20190711095549%26meid%3D77ba7a42f3f24b709ebe632fb079b086%26pid%3D100047%26rk%3D5%26rkt%3D12%26sd%3D112419561527%26itm%3D114333271792%26pmt%3D1%26noa%3D1%26pg%3D2047675%26algv%3DSellersOtherItemsV2&_trksid=p2047675.c100047.m2108&autorefresh=true

[Alex]

The pixel mapping on that is really quite odd, 3 tubes in all arms except the lower two bars where it is 4... Wonder what the logic was behind that?

Seem to recall someone on here bought a clock with a few of these boards in not too long ago?

Nice find, thanks for sharing.

On Sunday, 2 August 2020 01:47:39 UTC+1, celzey11 wrote:

https://www.ebay.com/itm/PCB-with-23-pieces-IN-28-nixie-tubes-slightly-used-working/114333271792?_trkparms=aid%3D1110002%26algo%3DSPLICE.SOI%26ao%3D1%26asc%3D20190711095549%26meid%3D77ba7a42f3f24b709ebe632fb079b086%26pid%3D100047%26rk%3D5%26rkt%3D12%26sd%3D112419561527%26itm%3D114333271792%26pmt%3D1%26noa%3D1%26pg%3D2047675%26algv%3DSellersOtherItemsV2&_trksid=p2047675.c100047.m2108&autorefresh=true

[celzey11]

Tempted to buy a few...

[Bill Notfaded]

Has anyone tried driving this other version of these vs. the type that Paul has?  I assume these from Van are all the same.  It looks like they have some wires connected to a line of pins that plugged into something at one point?  It looks like there are resistors for each tube on the board already.  There are 23 tubes on each board.  Has anyone figured out the pin out on the connector yet?

On Sunday, August 2, 2020 at 5:58:49 AM UTC-7 celzey11 wrote:

Tempted to buy a few...

[Christian Elzey]

Unfortunately I haven't tried to power it up yet, moving and starting a new job took all my free time. Unfortunately one of the 4 I received had a few cracked tubes, but the seller is sending me some IN-28s to replace them with. Can't wait to see it light up though. 

--
You received this message because you are subscribed to a topic in the Google Groups "neonixie-l" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/neonixie-l/bVoCKhPv7Dg/unsubscribe.
To unsubscribe from this group and all its topics, send an email to neonixie-l+...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/9b691c02-facc-4498-838b-c25d7cb41d8cn%40googlegroups.com.

[Richard Scales]

I have a few of these boards and have just started to look at how they are wired. Mine had small boards mounted on the back which I have removed.

They are wired as follows:

• All Cathodes to ground.
• All Anodes to +Volts via a series resistor
• Each group of tubes forming segments have the Grid held on a potential divider made up of a 1M and 3M resistor.

My understanding is that the thing will need 240V DC (which I am currently researching) and then the Grids can be controlled to light each segment (or not).

That's as far as I have got,

It seems that generating 240VDC is the hardest part of all this - A transformer with 170V secondary and sufficient VA, Bridge Rectifier and Smoothing cap is the old-school method - which does provide some isolation.

Is there a better way?

- Richard

On Tuesday, 13 October 2020 14:35:39 UTC+1, celzey11 wrote:

Unfortunately I haven't tried to power it up yet, moving and starting a new job took all my free time. Unfortunately one of the 4 I received had a few cracked tubes, but the seller is sending me some IN-28s to replace them with. Can't wait to see it light up though. 

On Tue, Oct 13, 2020, 9:25 AM Bill Notfaded <notf...@gmail.com> wrote:

Has anyone tried driving this other version of these vs. the type that Paul has?  I assume these from Van are all the same.  It looks like they have some wires connected to a line of pins that plugged into something at one point?  It looks like there are resistors for each tube on the board already.  There are 23 tubes on each board.  Has anyone figured out the pin out on the connector yet?

On Sunday, August 2, 2020 at 5:58:49 AM UTC-7 celzey11 wrote:

Tempted to buy a few...

--
You received this message because you are subscribed to a topic in the Google Groups "neonixie-l" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/neonixie-l/bVoCKhPv7Dg/unsubscribe.

To unsubscribe from this group and all its topics, send an email to neoni...@googlegroups.com.

[Bill Notfaded]

I think I'm going to use power supply just like myself and Paul have which he showed testing one with but hopefully a slightly smaller model some have four positive and negative DC outputs on them.  Seems perfect for four digit clock to me!!!  Now the control for the segments and software for the digits to figure out.  The PS Paul and I have goes to 500V but I've found some that goto 300 which seems perfect for this with 4 sets of outputs!

Bill

[Michail Wilson]

Where to find this power supply?

 

Michail

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/CADToqn3B0yBsFY5bak3b0BnW92iB%3D79QqtP4TLJONV%3D%3DHbUwiA%40mail.gmail.com.

[Richard Scales]

.... and where to find it in the UK!?!

I am looking at a custom transformer for UK mains to 170V Secondary, bridge rectifier and cap etc but it looks like i will need one per display !

... unless anyone has a better idea?

 - Richard

On Wednesday, 14 October 2020 13:38:11 UTC+1, M1 wrote:

Where to find this power supply?

 

Michail

 

From: neoni...@googlegroups.com <neoni...@googlegroups.com> On Behalf Of Bill Notfaded
Sent: Wednesday, October 14, 2020 4:28 AM
To: neonixie-l <neoni...@googlegroups.com>
Subject: Re: [neonixie-l] Re: A huge 7 Segment display made out of 23 IN-28 tubes? Wack.

 

I think I'm going to use power supply just like myself and Paul have which he showed testing one with but hopefully a slightly smaller model some have four positive and negative DC outputs on them.  Seems perfect for four digit clock to me!!!  Now the control for the segments and software for the digits to figure out.  The PS Paul and I have goes to 500V but I've found some that goto 300 which seems perfect for this with 4 sets of outputs!

 

Bill

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.

To unsubscribe from this group and stop receiving emails from it, send an email to neoni...@googlegroups.com.

[Paolo Cravero]

Hi.

  The PS Paul and I have goes to 500V but I've found some that goto 300 which seems perfect for this with 4 sets of outputs!

 

Maybe it is not what Bill meant, but electrophoresis power supplies come in ranges up to 300Vdc (minimum) and few dangerously hundred mA. Add a current limiting circuit for human safety as suggested by gergebert and you're set. Easy to find in US and UK over eBay. With some patience can be had with less than 100 USD/EUR/GBP.

Paolo

[Dekatron42]

You can usually find a good selection of tube amplifier transformers in this voltage range, 150-180 VAC plus usually one or more 5-6.3-12.6 VAC windings which can be used for the logic portion,  and possibly a bias winding too ,  on eBay or for instance at Ask Jan First.

/Martin

[Paul Andrews]

I’m pretty sure Bill is referring to an electrophoresis power supply. That is what I used to experiment with one panel. I don’t know that they would be practical for more than one panel. You would need one for each panel. Maybe multiplexed they would take less current in total?

What is this current limiter you mention? Although frankly, one panel needs about 200-250mA anyway.

The power supply I ended up using was five stacked Mean Well LPV-60-48, but for those countries with 240V mains, a simple isolation transformer with a bridge rectifier should do the trick. Just try not to kill yourself.

[Paul Andrews]

Ye olde current limiter

[Richard Scales]

On the whole isolation transformer thing - I've been trying to find something with 170V secondary with enough VA (for 6 displays) and have failed thus far.

An alternative approach might be to take off the shelf toroidal transformers - like mains to 35V then 45 V to mains - two transformers back to back. Standard parts, lower cost, plenty VA.

- Richard

[Grahame]

A common step down 230V primary - 115V secondary, rectified will give 163V - worth a try?

For custom transformers I have used, a more expensive option but you are only buying one(!?!)

https://airlinktransformers.com/

in Harlow.

Grahame

--

You received this message because you are subscribed to the Google Groups "neonixie-l" group.

To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/55d67cb6-0c62-48b0-88c2-9661d77581e0o%40googlegroups.com.

[Richard Scales]

.... But I think we need 240v Dc.

Richard

You received this message because you are subscribed to a topic in the Google Groups "neonixie-l" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/neonixie-l/bVoCKhPv7Dg/unsubscribe.
To unsubscribe from this group and all its topics, send an email to neonixie-l+...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/d71b6672-18b9-1ee6-b104-c7a75a381d0f%40googlemail.com.

[Grahame]

Right ... I had misunderstood the requirement. A query to a custom transformer manufacturers might be work a try, especially if you are starting to look at back to back arrangements.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/CA%2B9r3s_1tYoWORSKNx2O%3D6cAPRE2tBenDYACw2V08m85ZLO-xw%40mail.gmail.com.

[gregebert]

You could use a voltage-doubler. That will solve the voltage problem but it will double the current. Is the VA rating of the transformer sufficient for your project ?

[Paul Andrews]

You don’t need DC to drive these things. Full wave rectified will do. Also, you want the average rectified value to be 240V. Rectifying 170V RMS won’t give you that. Smoothing it will get you closer. So I think you should just use 240V RMS full wave rectified.

On Oct 15, 2020, at 11:42 AM, gregebert <greg...@hotmail.com> wrote:

You could use a voltage-doubler. That will solve the voltage problem but it will double the current. Is the VA rating of the transformer sufficient for your project ?

--

You received this message because you are subscribed to a topic in the Google Groups "neonixie-l" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/neonixie-l/bVoCKhPv7Dg/unsubscribe.
To unsubscribe from this group and all its topics, send an email to neonixie-l+...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/77e9c049-01d9-4386-b736-3dd337823550n%40googlegroups.com.

[Richard Scales]

OK, I think I'm getting there - I now understand that in the UK, mains voltage is 230V RMS - I had assumed that the quoted value of 230V was a peak value - shame on me (brain is not what it was 40 years ago when I learnt all that stuff).

Given that UK mains is around that value - are we saying that full wave rectifying that and smoothing it is pretty much all we need?

... and if so, a suitable isolation transformer should be employed for safety/sanity.

... and a suitably sized electrolytic for smoothing.

 - Richard

To unsubscribe from this group and all its topics, send an email to neoni...@googlegroups.com.

[Paul Andrews]

I’m saying that I think just plain full wave rectified will do, but you might need to add some smoothing to bring the average voltage up. If you want to design a power supply, I suggest you download psud2 from here http://www.duncanamps.com/psud2/

On Oct 16, 2020, at 12:28 AM, Richard Scales <ric...@scalesweb.co.uk> wrote:



To unsubscribe from this group and all its topics, send an email to neonixie-l+...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/2af17b15-73a1-49a0-b178-b6318e8abd75o%40googlegroups.com.

[Bill Notfaded]

This is exactly correct Paul and I have both shown the 500V models we have here.  I bought a second 500V one.  I like the fact that some have four outputs too.  These are used for genetic research.

Bill

--
You received this message because you are subscribed to a topic in the Google Groups "neonixie-l" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/neonixie-l/bVoCKhPv7Dg/unsubscribe.
To unsubscribe from this group and all its topics, send an email to neonixie-l+...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/CABj2Vaaso61iFnB9fD80BUCD1TepGQzNeuXnD8PUG3oiuX9Kkg%40mail.gmail.com.

[Bill Notfaded]

I found some of these power supplies now that put out 0-500VDC 0-2A with 4 sets of outputs.  This seems perfect for a four or probably even 6 digit clock if the tubes are driven PWM.  The power supplies even have 1V resolution.  These power supplies have a fan in them so I'm interested to see how loud they are.  I don't really want some loud power supply fan but if you think about it 500V @ 2A o.O that's some 1000 watts we're talking about.  So say the power is good now.  I'm interested now in the switching of the HV.  I'm curious if PCA9685 could be used to PWM using some HV NPN transistors?  What we need is a PCB design that can route all these control lines for all these tubes.  This is a LOT more than a usual 4 digit clock with 0-9 cathodes to control.  One thing that seems nice about the PCA9685 is that the phase of the PWM can be controlled so we could make it so all the tubes don't turn on and off at the exact same time.  Has anyone played around with this kind of a configuration before?  It would also be nice if the transistors could come in smaller package because geesh we need a F' load of these things on the PCB.  The PCA9685 is 16 channel.  So 6 would be needed because 23 x 4 = 92 tubes and 92 / 16 = 5.75.  I'm kinda thinking out loud here so if I'm totally off on this idea someone please let me know?  It seems like I2C control of the tubes duty cycle and phase would be pretty sweet for dimming right?

Bill

[Bill Notfaded]

Anyone have any PNP transistor recommendations?  Seems like  SOT-323 size would be possibly ok if I could find some that can take enough voltage.  How high of voltage should I be thinking about for these?  You have to bear with me because I'm a CIS not an EE.  For me the I2C parts make much more sense than the analog EE parts.  Would  ON MSD42WT1G work or is 300V not enough would you think?  Wait they can't carry enough current can they... ugh!

Bill

[Paul Andrews]

On my boards the cathodes for each segment are wired together and presented at the edge of the board. The grid is permanently connected to HV via a large value resistor. So each segment is like a cathode in a regular nixie tube - pull it to ground and it turns on, let it float and it turns off.

With the boards that Richard has, all the cathodes on the board are connected to ground via a resistor. All the anodes are connected together. It is the grids that are grouped together and presented at the edge of the board. With IN-28s you pull the grid controls whether a tube will light up or not when the cathode is pulled to ground. If you want a tube to light up, you have to pull the grid high. Once it is on, you can let the grid float - not sure if it also stays lit if the grid is pulled to 0V. So in practice what you do is this:

1. Turn all the tubes off
2. Let the grids float for those segments that should stay off, pull the others high (e.g. up to the anode voltage) for those segments that should be on. Note that in Richard's case, his boards have a pullup resistor on the grids, so you have to actively pull them down for those segments that you want to stay off and let them 'float' for those that should turn on.
   
3. Turn all the tubes on - the segments with a high grid should light up

With pulsed DC (un-smoothed rectified AC), all the tubes turn off at twice the mains frequency, because the voltage falls below the maintenance voltage. So you can just permanently connect the cathodes to 0V and the anodes to HV. When the voltage gets high enough, only those tubes with the grid high will turn on. So in this case, all you need to do is pull those grids low for the segments that what you want to be off. The boards effectively read this state around 120 times per second. Pretty simple.

If you have steady DC, you have to turn the tubes off yourself, by letting the cathodes float before you change the state of the segments using the grids. If you do this, you will have to be able to handle between 200mA - 250mA per board with whatever is pulling all those cathodes to 0V.

So you can see that using un-smoothed, rectified 240V makes this a lot easier.

As for what you can use to pull those grids low, I have no idea. You probably don't have to drop them all the way to 0V and I am pretty sure you don't have to let them go all the way up to the anode potential.

I guess, in some ways, my boards are easier to drive because we all know how to drive nixie tubes, and my boards just look like very big 7 segment nixies.

[Dekatron42]

Be careful if you decide to switch the grid electrode to ground as you might destroy the tube if you pass too much current through the grid electrode. Normally you should not leave the cathode floating (unconnected) while the anode is seeing full anode voltage and then connect the grid electrode directly to ground as all current will then flow through the grid electrode. Always use a high-ohm resistor, usually in the megohms range, connected to the grid electrode to limit the current (the grid resistor is then working as a voltage dropper for the current drawn through the trigger electrode).

There are circuits which you can design which will let the tube self extinguish a short time after it has been triggered, this due to the use of a capacitor across the anode/cathode in conjuntion with either an anode or cathode resistor - some of these boards look like they have large non-electrolytic capacitors mounted on them which could then possibly be used for this effect. If you set the time constant short enough of this R/C combination you can select your own frequency to drive the grid electrode with, in effect dimming the tube depending on the frequency used. This is commonly used in the drive stage of dekatrons with a trigger tube like with the GTE175M and also the Z70U.

You can read more on this in "electronic counting circuits" by J.B. Dance and other books on trigger tubes, page 82 mentions the self extinguishing mechanism in chapter 4.3.5 with the "40 c/s driving circuit".

/Martin

[Paul Andrews]

Good information Martin. The board as described to me by Richard has the grid resistor you mention, as does mine. Certainly, if someone is wiring up one of these tubes directly - in this case as colons for example - they need to be aware if this.

On Oct 16, 2020, at 3:32 PM, Dekatron42 <martin....@gmail.com> wrote:



--

You received this message because you are subscribed to a topic in the Google Groups "neonixie-l" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/neonixie-l/bVoCKhPv7Dg/unsubscribe.
To unsubscribe from this group and all its topics, send an email to neonixie-l+...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/42c46a0d-95d2-4179-aa54-71250a24b638n%40googlegroups.com.

[Bill Notfaded]

Thanks Paul and Martin-

Sounds like I need to rethink this.  I ordered a sealed box of new IN-28 from Ukraine today.  I'm sure I'll have some extras if anyone needs them.  I got them to replace any tubes in mine that don't look right or are bad.  I'll take some pictures of my boards and post.  Hopefully you can help me identify.  If they really have the digits pre wired so they work more like a Nixie tube digit that makes for much less control lines but would also mean less effects we could do with tubes.  There won't be any counter flashing digits every other tube or snaking light though the digit effects... Stuff like that.  This would also mean one line would need to support the digit 8 all tubes turned on which would be much more current.  I bought two power supplies today.  They're VWR AccuPower Model 500:

Features of the VWR AccuPower Model 500 Electrophoresis Power Supply include:

- Voltage range 10-500V with 1V steps

- Current range: 0.02-2A 0.02A steps

- Microprocessor control

- Fully regulated output of constant current and voltage

- Automatic crossover from voltage to current

- No-load detection

Bill

[Bill Notfaded]

Pics...