Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Dummies guide to building a 120watt 140 volt power supply to drive IN-9 Nixie tubes.

Status
Not open for further replies.
Humm... I guess you are correct in the sense that we are drawing from it not inputting into it .. but I would have thought the current draw on it would effect fields.

No I am correct. The field from any current drawn from a secondary gets reflected back to the primary so that the primary takes more current. With a transformer power in equals power out (there are some losses in practice but or the sake of explanation they can be ignored). That is how a transformer works. The direction of the additional winding makes absolutely no difference whatsoever.

No .. that was with 110Vac on it.

Was that 110V RMS on a 115V winding or on a 230V winding (2 of 115V primary windings in series) ?

As for operating voltage.
I have no idea what the nominal 110 voltage is 115 or 110 ?

We use 240Vac here .. not 230

I am just assuming 110V is a standard ... google here I come.

edit: quick google later .. it seems that 120v is taking over.
It also says that 230 is taking over instead of 240
Not here it hasn't !

The specified mains input voltage range is most important to establish because it radically effects the design, especially the dissipation of the transistors and resistors.

If you would be happy with a mains input voltage 230V-5% to 240V+5% and 110V-5% to 115V+5% that would help a lot, and would allow me to work out the thermal budget and optimize the supply rails.

I have been working on a design for a stabilized 170V supply line using SCR control on the secondary winding. If you can go with the above mains voltage ranges a stabilized supply line may not be necessary.

spec
 
Last edited:
Thanks for putting in so much effort on this spec.
It's really appreciated.

Answers for above.
Two windings in are in parallel. (primary side).
Thought it would be more efficient.

Please feel free to operate on the above voltage assumptions. They sound pretty good to me.

My mission is to get the voltage level at about 100 and get a working striker circuit.
But am going to have to wait for parts.

Will try with what I have and see how many caps I can make go bang.


Off to bed.
 
Thanks fire- have a good sleep.

(Just for the record, the normalized voltage per turn is 0.235 volts per turn)

spec
 
Last edited:
Hy fire,

Good news! With the new mains input voltage specification and the NOTES and ERRATA, the circuit of post 147 is fully tolerenced and is now a goer. No extra windings are required and the Les buck striking PSU is not required either.

Hope it works as anticipated.

spec
 
Thanks Spec,
I am sure it will, Though it still has the heat issue :(

I have spent the day learning a little eagle cad and have fleshed out the Les Jones Bucker Circuit (made it pretty).

This circuit (if it works) could be the answer to all the problems ?

The way I see it .. limited amps for striking through the fat resistor he has put in. Eliminating heat problem.
Diode protection so we don't go back to the 78v side.
And .. it's so sweet that even the main sustaining voltage is above spec and will have plenty of current.

Waiting on some caps before I test build. Have no idea what I need so am just going to try a few smaller HV types.

This is my first eagle file .. so ignore all the values for now.

Will put the popper ones in if I can get it to spark up.

Wotchathink ?
**broken link removed**
 
Last edited:
Thanks Spec,
I am sure it will, Though it still has the heat issue :(

That is true about the heat, but I would think the single supply bridge circuit of post #147 is a practical proposition as long as a good physical design to allow sufficient heat dissipation is realized.

The very absolute minimum heat dissipation using a switch mode PSU to generate a stabilized supply rail with no ripple and without any control would be 70V (tube operating V) * 12mA * 100= 84W.

With control, allowing 10V across the transistors, would give an additional 10V * 12mA * 100= 12W, making a total of 84 + 12W = 96W.

Assume that the switch mode PSU is 75% efficient would add 96W * 0.25 = 24W.

This gives a grand total heat dissipation of 120W.

So the message is that the tube display unit (TDU) is quite power hungry by definition.

I have spent the day learning a little eagle cad and have fleshed out the Les Buck Bucker Circuit (made it pretty).

Very impressive. It is not easy to get started wit EAGLE, although once you get past the initial stage it is quite easy (you should be running beta version 7.5.2)

This circuit (if it works) could be the answer to all the problems ?

The way I see it .. limited amps for striking through the fat resistor he has put in. Eliminating heat problem.
Diode protection so we don't go back to the 78v side.
And .. it's so sweet that even the main sustaining voltage is above spec and will have plenty of current.

Waiting on some caps before I test build. Have no idea what I need so am just going to try a few smaller HV types.

This is my first eagle file .. so ignore all the values for now.

Will put the popper ones in if I can get it to spark up.

Wotchathink ?

I don't like being a killjoy, but afraid the circuit will not meet your specification.

The 55V transformer secondary windings have a tolerance of 57.25 +- 9% when all the variables are taken into account.

(1) 57.25 RMS - 9% = 52.095 RMS

(2) 52.095 RMS * 1.414= 73.67V peak - 1V for both rectifier diode forward drops (500mV each) gives, 72.67V peak.

(3) 72.67 peak - 5V ripple voltage = 67.67V trough.

This does not meet the tube running forward voltage of 70V. Neither does it give any voltage across the control transistor. If you allow 10V minimum across the transistor you need 85V minimum, taking ripple voltage, to run and modulate the tubes (I would put a tolerance of +-5% on the tube running voltage too).

Of course, the extra voltage required can be provided by an extra winding on the transformer core.

Because of the variation of the supply lines and the ripple voltage the circuit will not be as efficient as you might at first think. That was why I went for the simpler circuit in post #147

Incidentally, the Les boost strike supply can be reduced to one capacitor and one diode as a DC supply is not required to strike the tubes.

Al in all though, nice work

spec
 
Last edited:
Hi Fire,
My circuit is not a buck converter. I think a variation on spec's switch mode power supply would be a better solution. I would only us two (Not 3) 48 volts power supplies to supply the 100 volt rail. I think the extra one for the 150 volt rail is overkill as there will only be 50 volts across the 2,2 meg resistors. so each resistor will take about 23 uA as you will have 100 of these resistors the load on the 150 volt rail will be 2.3 mA So a 5 amp supply is a bit too much. (Being a switch mode it might not be happy running at almost no load.) We may also need to consider what the minimum load will be on the 100 volt rail for the same reason. A very small transformer and rectifier could supply the top 50 volts. ( A 24 volts transformer and voltage doubler.) I think building the circuit in post #206 would be a good idea so you can actually test how low you can go on the voltage rails using your variac rather than just using theory.

Les.
 
Hi Fire,
My circuit is not a buck converter. I think a variation on spec's switch mode power supply would be a better solution. I would only us two (Not 3) 48 volts power supplies to supply the 100 volt rail. I think the extra one for the 150 volt rail is overkill as there will only be 50 volts across the 2,2 meg resistors. so each resistor will take about 23 uA as you will have 100 of these resistors the load on the 150 volt rail will be 2.3 mA So a 5 amp supply is a bit too much. (Being a switch mode it might not be happy running at almost no load.) We may also need to consider what the minimum load will be on the 100 volt rail for the same reason. A very small transformer and rectifier could supply the top 50 volts. ( A 24 volts transformer and voltage doubler.) I think building the circuit in post #206 would be a good idea so you can actually test how low you can go on the voltage rails using your variac rather than just using theory.

Les.
Hey Les,
No I know it's not a buck converter but I think the circuit you drew up is valid.
The concept makes sense to me.

The ebay off the shelf 3 PSU solution is just too dang big, bulky, and even though I can see it's merits it's just not the way I want to go.

I think you are on the right track with the idea.
If I can strike the damn things why keep up the voltage ?

Anyway as you so rightly say, first I need to build it then test it.
 
That is true about the heat, but I would think the single supply bridge circuit of post #147 is a practical proposition as long as a good physical design to allow sufficient heat dissipation is realized.

The very absolute minimum heat dissipation using a switch mode PSU to generate a stabilized supply rail with no ripple and without any control would be 70V (tube operating V) * 12mA * 100= 84W.

With control, allowing 10V across the transistor, would give an additional 10V * 12mA * 100= 12W, making a total of 84 + 12W = 96W.

Assume that the switch mode PSU is 75% efficient would add 96W * 0.25 = 24W.

This gives a grand total of 120W.

So the message is that the tube display unit (TDU) is quite power hungry by definition.



Very impressive. It is not easy to get started wit EAGLE, although once you get past the initial stage it is quite easy (you should be running beta version 7.5.2)



I don't like being a killjoy, but afraid the circuit will not meet your specification.

The 55V transformer secondary windings have a tolerance of 57.25 +- 9% when all the variables are taken into account.

(1) 57.25 - 9% = 52.095

(2) 52.095 * 1.414= 73.67V - 1V for both rectifier diode forward drops gives, 72.67V

(3) 72.67 - 5V ripple voltage = 67.67V.

This does not meet the tube running forward voltage of 70V. Neither does it give any voltage across the control transistor. If you allow 10V minimum across the transistor you need 80V minimum to run and modulate the tubes.

Of course the extra voltage required can be provided by an extra winding on the transformer core.

Because of the variation of the supply lines and the ripple voltage the circuit will not be as efficient as you might at first think. That was why I went for the simpler circuit in post #147

Incidentally, the Les boost strike supply can be reduced to one capacitor and one diode as a DC supply is not required to strike the tubes.

Al in all though, nice work

spec


Dang.
Ok.... fair enough.
So... not enough volts out of my existing toroid.
And I have not done any maths. (not unusual)

Oh well. Looks pretty though.... right ?

Perhaps I should just use an old fashioned ballast and starter from a fluorescent light fitting.
Those buggers figured this problem out eon's ago !

Nah.. that won't work... look how hard it is to dim fluro's

The heat across those resistors is killing me.

Using a whole bunch of these with a FAT DC supply ?

https://threeneurons.wordpress.com/nixie-power-supply/hv-supply-kit/

Edit: still gonna try the Les circuit.
I have variac ... can turn !
 
Last edited:
Dang.
Ok.... fair enough.
So... not enough volts out of my existing toroid.
And I have not done any maths. (not unusual)

Dang is just what I thought when I worked through that approach. :banghead:

Oh well. Looks pretty though.... right ?

'pretty' is just what I thought. :)

The heat across those resistors is killing me.

Why? 1W is no big deal, with good ventilation.


Each of those boost boards would power 6 tubes so you would need 17 boost boards.

Depending on cost, it would be a practical proposition.

Edit: still gonna try the Les circuit.
I have variac ... can turn !

Of course- experimentation is the fun in engineering. :cool:

I have done an SCR control circuit which helps reduce the heat dissipated. The thing is that it uses a couple of techniques that I have no practical experience of, so it would be an off-the-boards (experimental) deign. This approach could be used to provide a stabilized 170V supply with no auxiliary strike supply rail or 90V stabilized supply rail using an auxiliary strike supply rail.

By far the best approach, as Les and I have been implying is, a 90V switch mode PSU. The circuit using the three PSUs was only conceptual and should not be analyzed in too much detail. I also did a conceptual design using 2A 24V PSUs which are more freely available on Ebay.

The cheap Ebay PSUs have a fairly flexible design. In the past one of the vendors has done specials for me. It takes about 4 weeks to do. They might be able to do you a 90V 2A version.
It may even be possible to modify a standard PSU yourself. The beauty of those power supplies is that they are dirt cheap. :)

In any event, there would be any number of ways to produce the low current tube strike voltage.

On a general point, are you building just one of the tube displays or are you looking for a production design?

spec
 
Last edited:
Spec, please remember the title ... dummies guide ... ok ?

So .. the basic idea then of getting them lit and using low voltage is sound.
We agree on that. 3 ebay power supplies or extra volts ... one way or the other.

As for heat.
100 Watts is just a lot to dissipate IMO.
Not without basically mounting the whole thing in a colander.
Or using a fan (a no no on this project).

I don't like the waste or lots of holes.
+ I live in a hot country already .. we don't need extra radiators round the place.

The tubes will be in air so that's ok.. but the circuit should be running efficiently and not need a lot of air or preferably none.

Umm... specials you say.

I have trawled through a few listings .. but yes I hear you.
You say I only need 90Vdc @2A ... and we just boost to strike .. diode pump like ah la Les or another transformer ? Or another supply.
Agreed there are plenty of ways to go up .. but just to strike is the tricky bit.

Gonna have another look at that ebay post of yours
 
Here is a saving the planet version of the tube PSU:

2016_04_13_Iss01_ETO_COLD_CATHODE_TUBE_PSU_VER_07_Sh04.png

NOTES
(1) Almost certainly a 48V 2A version of the power supply could be procured. This would be smaller and possibly cheaper.
(2) The 48V PSU are adjusted to 45V
(3) There are many ways of implementing the strike voltage generator. This should be simple in view of the very low current requirement.
(4) There are no heat dissipating collector resistors.
(5) The transistors would only be dissipating 226mW maximum each.

DATA SHEETS & SOURCES
(1) 48V PSU
**broken link removed**
**broken link removed**
(2) Transistor 2N6517
https://www.onsemi.com/pub_link/Collateral/2N6515-D.PDF

ERRATA
(1) For every 5 tubes fitted, place 1 high frequency decoupling capacitor between the 106V supply line and 0V.
(2) For every 10 tubes fitted, place 1 high frequency decoupling capacitor between the 158V supply line and 0V.
 
Last edited:
Errr... I can't come up with an efficient power supply can I ?
Gonna relax and look at digikeys transformer page again.
Really find that relaxes me....
You ?
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top