Threeneurons charge pump question!

[Chris Stalin]

Hey fellas,

Doing lots of lurking and reading, Ton of awesome info and intellect in here!

With my IN8 nixie clock nearly done as well as some IN16 clocks complete I think its time to venture into dekatrons.
I bought a matched pair of metal base OG4 dekatrons the other day, Wile they are on their way I wanted to ask a question regarding M-Mooreess charge pump.
http://threeneurons.files.wordpress.com/2012/07/nixie6c_sch.gif

I am using a horizontal taylor edge HVPS - BTW John and all his work is amazing, Not to mention unlimited support.
With my resister set for 200V output .... Why does Michael use a charge pump with a 4093 and not just a regular voltage doubler or tripler? What are the advantages of pumping the voltage up using the charge pump as apposed to the regular doubler or tripler?

I think I should read up a bit on them lol.

Thanks every one.

[threeneurons]

I released the nixie clock kit, long before Taylor made his own tripler. If you look at the charge pump schematic, the tail end of it, is a tripler. The 4093 stage, and the associated transistors, are there to create a high voltage pulse train. If you can tap on the pulses generated, directly from the switching supply, you don't need this additional circuitry. But with most HV supplies, available at that time, only DC is brought out.

A doubler. or tripler, will NOT work on DC. It needs a pulse train, or some time varying signal, like household AC.

One final note. The pulse train, on the clock kit's charge pump, is a relatively high duty cycle (>30%), and a relatively low frequency (600 to 1KHz). This means simple crappy low speed rectifiers, like the 1N4007 can be used. Ultra fast rectifiers (ie UF4007, MUR160, ...) are not required.

[Chris Stalin]

Thank you very much for your prompt reply.

I now understand the pulses needed for the charge pump. Makes sense as you didn't have one to tap into.
You said Taylor has a tripler for his HVPS, I tried searching but have to yet to come across it. Do you have any links or literature so I can skip the switching portion of the schematic and just go right to the tripler for ease of use with my dekatron.

Nice to know the selection of rectifiers that can be used. With component prices so cheap these days, Shouldn't be much of a headache locating what you need.

Thanks again for your time!

[John Rehwinkel]

> With my resister set for 200V output .... Why does Michael use a charge pump with a 4093 and not just a regular voltage doubler or tripler?

The charge pump is a voltage multiplier too. That lashup just uses the pins supplied with the 1363 power supply, which delivers DC. To make a doubler/tripler/multiplier/charge pump work, you need AC drive. Michael derived this from the DC with the 4093 oscillator and some switching transistors.

However, there is an AC signal available on the 1393 module (though not on its connector), as jt detailed here:

http://ppl.ug/IEd55TCltCo/

Michael's circuit predates that post, so it doesn't take advantage of it. Either approach will work.

> What are the advantages of pumping the voltage up using the charge pump as apposed to the regular doubler or tripler?
>
> I think I should read up a bit on them lol.

You'll find out that they're just different terms for the same sort of thing.

- John

[John Rehwinkel]

> You said Taylor has a tripler for his HVPS, I tried searching but have to yet to come acorss it. DO you have any links or literature so I ca skip the switching portion of the schematic and just go right to the tripler for ease of use with my dekatron.

Yeah, the link I posted showing the AC output also shows how to connect up a tripler:

http://ppl.ug/IEd55TCltCo/

> Nice to know the selection of rectifiers that cna be used.

Yeah, jt's power supply operates at a high frequency (so it can be compact), so a tripler for it should use fast diodes. Mike's chopper circuit (as he pointed out) is lower frequency and can work with ordinary diodes. The tradeoff is that, with the lower frequency, you'd need larger capacitors to get the same ripple. Happily, dekatrons don't draw much current at all, and are fairly forgiving of ripple, so this isn't a big deal. For reference, Mike's circuit uses 1µf capacitors, and jt's circuit
uses 0.1µf units.

> With component prices so cheap these days, Shouldn't be much of a headache locating what you need.

True enough, I appreciate being able to source parts quickly, easily, and cheaply.

- John

[Chris Stalin]

Hi John,

Wonderful reply. I really appreciate the link and the explanation.
I now understand why Michael mentioned his method uses simple parts, In comparison to JT HVPS - I haven't a clue where I can source 4700pf 3kv ceramic capacitors hah!
Is that last 3kv ceramic to smooth things out?

In some respects the 4093 derived pulse and cheap parts almost beats playing with slightly harder to find parts using the AC pulse, And trying to solder to the tiny foot print of the HVPS.... Rather interesting.

Is the inductor choice critical in terms of current or would a simple axial lead inductor work?
Does one adjust the voltage resistor to get 200V DC output from the HVPS and then add the 100Kohm preload or what is meant by preload?
Would 1N49378 diode suffice?
"For doubler DNI C3 and bypass CR3" - Is that electrical engineer terms for omit C3 and CR3(Diode)? lol
As for AC out routing comment, How far did he anticipate a person should or would place their connections in relation to the HVPS it self?

More questions then I came in with originally, Always some thing new!
Thanks again John.

[John Rehwinkel]

> Wonderful reply. I really appreciate the link and the explanation.
> I now understand why Michael mentioned his method uses simple parts, In comparison to JT HVPS - I haven't a clue where I can source 4700pf 3kv ceramic capacitors hah!

I get things like that from old television sets, computer monitors, and bug zappers.

Or you can buy 'em from Digikey/Mouser/Newark or various surplus outfits (Electronics Goldmine often has high voltage capacitors), for example:

http://www.digikey.com/product-detail/en/S472M69Z5UR63L0R/1290PH-ND/2356826 ($3.78 for ten)

> Is that last 3kv ceramic to smooth things out?

Looks that way.

> In some respects the 4093 derived pulse and cheap parts almost beats playing with slightly harder to find parts using the AC pulse, And trying to solder to the tiny foot print of the HVPS.... Rather interesting.

It is easier to just use the built-in connector on the HVPS instead of tacking wires to it. Like many of these things, it depends on what's important to you.

> Is the inductor choice critical in terms of current or would a simple axial lead inductor work?

The current is tiny, and the waveshaping function is non-critical, so an ordinary axial lead inductor should be fine.

> Does one adjust the voltage resistor to get 200V DC output from the HVPS and then add the 100Kohm preload or what is meant by preload?

The HVPS has trouble staying regulated without a load on its usual DC output - that's what the 100KΩ resistor is for - just to provide a minimal load. Attach it first, then adjust for 200V (for Vadjust to ground, this works out to around 470Ω).

> Would 1N49378 diode suffice?

Yeah, a 1N4937 should be fine. I just buy UF4007 diodes in bulk (100 for $17) and use 'em for everything.

> "For doubler DNI C3 and bypass CR3" - Is that electrical engineer terms for omit C3 and CR3(Diode)? lol

Yeah, DNI means "do not install", and "bypass" (in this context) means "replace with short circuit".

> As for AC out routing comment, How far did he anticipate a person should or would place their connections in relation to the HVPS it self?

Since it's a high frequency, high voltage signal, it's going to suffer capacitive and possibly corona loss fairly easily, so it would be best to keep such wires short. His routing comment refers to the fact that the high voltage AC output can couple back into the power supply's feedback circuitry and cause trouble - so the wire from the AC output should be routed away from the board.

> More questions then I came in with originally, Always some thing new!

Questions and answers are how we all learn (both the people asking them and the people answering them). Bring 'em on!

- John

[Chris Stalin]

John,

Funny you should say that about the od TV. I just retired my old tube TV and decided to keep the electronics.
I did find some 2KV ceramics caps but the others are either too small in capacitance or don't indicate voltage ( they are quite large though 3/4" in diameter )

I resort to places like taydaelectronics goldmine ebay dipmicro etc. So I have to keep my part selection narrow or according to what I can find.

Thanks for the detailed information and reply as always.

Take care John

[Instrument Resources of America]

A number of old scopes (vacuum tube scopes), especially Tektronix and H-P will have them as part of the high voltage supply.   Ira.

--
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 post to this group, send an email to neoni...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/72a9d4e7-5a56-4c86-93c3-69e0cb2135fc%40googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.