A small HV DC-DC converter for your average Nixie Tube applications

Similar projects worth following
µHVPSU v1.0

This is a small high voltage boost converter that uses an MC34063 buck-boost converter IC.
This µHVPSU can work from 3V up to 6V and generate a voltage from [input]V up to 300V MAX.

Perfect for Nixie tubes, Geiger counters and other sub-deciAmp applications

μHVPSU v2.0

[Yet to be written :) ]

Don't forget to like the project if you, well, like the project ^^
If you want to get notified of the future project logs, or if you are interested, or if you want to do me a favor, you can follow this project to see the logs & changes in your personal feed !!

EDIT : To be rewritten

Project Logs :

  1. First step : the schematic
  2. Second step : the PCB
  3. Ordering PCBs
  4. Quick thing
  5. UPDATE : This project is not dead yet !!!
  6. v2.0

old description [

I was playing around with Nixie tubes, and I was using a 10$ HV supply kit on eBay to generate the 200VDC needed to power them.

My friend Manoel (@M.daSilva) uses an assembled module, which also costs 10-12$. It's a reasonable price, but I don't really understand why it's "so" expensive when you see the few components on it.

A few months later, I found a quite interesting website (link) made by Radu Motisan ( profile), which has a lot of projects involving High Voltage radio tubes, geiger tubes, etc.

I came across a post with a really small HV boost converter, powered by a CR2025 battery (link).

This gave me the idea to make my own; the micro High Voltage Power Supply, or µHVPSU, for short.

My goal here is to make a small HV PSU that costs less than 5$, "labor" included.

The schematic is nearly entirely based on the one on the website, but I made some modifications (see project log 1)

Project progression : about 75%

If this project works, I will put a link somewhere where you will be able to buy the PCB directly on DirtyPCBs.

P.S.: I misread the MMBTA42 datasheet, so the maximum input voltage drops from 16V to 6V. For the next revision (V1.1) I'll probably use the STR1550, as V(eb) is equal to 9V, and the maximum V(eb) is 500V, which means that the maximum output voltage of the µHVPSU will be 500V !!



CadSoft EAGLE Schematic

sch - 338.08 kB - 01/03/2017 at 17:37


µHVPSU v1.brd


brd - 126.59 kB - 01/03/2017 at 17:37


  • 1 × MC34063 Power Management ICs / Switching Regulators and Controllers
  • 4 × 100nF SMD Capacitor (0603)
  • 2 × 2.2nF SMD Capacitor (0603)
  • 1 × 10nF 1kV SMD capacitor (1206)
  • 1 × 1R SMD resistor (0603)

View all 12 components

  • v2.0

    Frédéric Druppel09/27/2018 at 21:01 0 comments

    Okay, I owe some explanations about this project.

    For v1.0 of the µHVPSU, I blindly followed an existing schematic of a similar design, that worked (or at least seemed to). So I ordered my PCBs and parts, assembled one, and it didn't work. So I assembled a second one, being extra careful, and it didn't work either. I got so annoyed about it that I decided to shelve this project. Until now !

    After hunting on eBay for Nixie tubes, I stumbled on many nixie clock kits and PCBs, with documentation, and most importantly, a schematic. What was really redundant was the use an analog input pin and a PWM output pin of the main µC as the mais DC-DC step-up converter. So I decided to base my new design on one of these schematics.

    Here's what I'm talking about :

    This particular clock kit uses a PIC microcontroller, which is fine, but it's a bit too much of an overkill for this project.

    Instead, I will use an ATTiny44 as they are widely available on eBay at a cheap price (about 6$ for a pack of 10), and they allow me to program them with the Arduino IDE, which I'm already really familiar with ^^ So that gives me a lot more flexibility.

    So yeah, that's it for this log ^^ I'll post again when I have something interesting to say, or when all the new components have arrived :)

    P.S.: M.daSilva and I have decided to make a spreadsheet, in which all of our Nixie tube types and quantities are listed; we have gathered a total of more that 600 tubes (400+ if you don't count the IN-3's)

    I guess we'll have enough to make some interesting projects ^^

  • UPDATE : This project is not dead yet !!!

    Frédéric Druppel09/27/2018 at 09:36 0 comments

    It’s been a while ^^

    As you saw in the title, I decided to work on this project again, as a friend and myself have ordered over 200 Nixie tubes together, so we need some way to power them.

    For the second version, I thought of a *wonderful* process in order to make sure the design works; it’s called [insert WordArd] « Prototyping »

    Anyway, I thought of using a microcontroller instead of a dedicated dc-dc converter IC, as it would give me more flexibility (and perhaps custom features) I couldn’t have with the IC I was using.

    So stay tuned for the next log that I will publish as soon as the components arrive from China ^^

  • Quick thing

    Frédéric Druppel01/05/2017 at 18:04 0 comments

    Just a quick thing I would like to say ^^

    I made a document for the BOM which I will update a lot while I'm ordering, assembling and testing the µHVPSU. So in order for you to always have the latest file, I made a shared Keynote document, which you can find here.

    That's all ^^

    N.B.: Why a Keynote document instead of excel or something else ? Well I use the Apple "iWork" and "iCloud" environnement a lot and I'm really comfortable with that (yes you can say I'm a fanboy, yes..). Oh yes also, sorry for the adfly link ¯\_(ツ)_/¯

  • Ordering PCBs

    Frédéric Druppel01/03/2017 at 17:35 0 comments

    The PCBs are ordered !! As you can see, the size of them is about 14mm by 23mm,. that's small enough for me ^^

    As you can also see, the price was 13.95$ for ~10 PCBs (so 0.1395$ each). I will calculate the exact cost when I have received the PCBs, as there's a chance I get more than 10 :p

    If this project works, I will put a link somewhere where you will be able to buy the PCB directly on DirtyPCBs.

    If you want to see the DirtyPCBs Gerber renderings, you can follow this link ^^

  • Second step : the PCB

    Frédéric Druppel01/03/2017 at 16:40 0 comments

    As the name implies, it has to be small.

    I want all the components to be on the same side, so it's more convenient to solder by hand (as everything is SMD). It's also better if I want to use that module in future designs.

    For all the "Low voltage" components, I used 0603 components. They can be tricky to solder by hand, but it's doable. For the high voltage 10nF onboard cap, I used 1206 size (because I had no choice). I may have to use an externah HV electrolythic capacitor, as 10nF may not be enough.

    For the transistors I used SOT-23 types. The PNP transistor is a S8550 general purpose transistor, and the High voltage switching transistor is an MMBTA42 300V transistor (the best I found with a great voltage/current ratio), and it's also why the µHVPSU v1 is limited to 300V.

    The rectifier diode of the converter is a HER108, because, reasons.. (Okay, it's the one used in the original design).

    Here are some pictures :

    (3D renderings made with this tool (don't pay attention to the green color, the real ones will be white))

  • First step : the schematic

    Frédéric Druppel01/03/2017 at 16:20 2 comments

    As said in the project details, I based my schematic on the one on this website (link). Here's the said schematic, so you don't have to leave this page ;)

    However, I changed a few things in my design.

    First of all, I used classic 4pin-connector_type-breadboard-friendly-type-thingy to interface the module, so it is easier to implement in other designs.

    Secondly, I changed the feedback loop system. In the schematic above, there are fixed resistors, which give the module a fixed (theoretical) output voltage of approx. 370VDC. In my design there is an interface pin where you put a resistor between that pin and the HV output, so you can always control the output voltage.

    Oh BTW, here's the formula to calculate the output voltage (found on the MC34063 datasheet)

    In my design (I hate saying that, but anyway) R1 is has a fixed value of 1k. You can choose the value of R2 by putting a resistor between the "HV+" pin and the "Sense" pin.

    Here's my version :

    All the files can be found in the "Files" tab of the project.

View all 6 project logs

Enjoy this project?



brodericksonpriddy2 wrote 09/18/2023 at 19:01 point

Your endeavor is incredible. This endeavor has given me a lot to learn. I'd want to distribute it to the  sosomod  team so they can read it and apply something fresh to our ongoing efforts.

  Are you sure? yes | no

olexpeter123 wrote 04/06/2023 at 21:27 point

There has been a great deal of value to me in my involvement with the project. Would like to share it with the Billing Zip Code team so they can also read it and implement something new.

  Are you sure? yes | no

Per Jensen wrote 11/01/2021 at 18:04 point

I think some of your math is wrong. It is not practically feasible to boost voltage that much with a single step boost converter. You need around 9V input to reach 170V on the output. 3V -> 200V, not going to happen. You need a tapped inductor to do that.

  Are you sure? yes | no

Frédéric Druppel wrote 11/02/2021 at 23:08 point

Indeed. I learned a lot in my electronics engineering courses; maybe I’ll have the time to work on this project again after I graduate.. 

also, I now think it would ne better to use 12V (or minimum 5), as I don’t really use 3V levels in the projects this inverter would ne useful in; though the original idea I based this on works from 3V :

  Are you sure? yes | no

matseng wrote 06/21/2017 at 09:14 point

Did you ever get this working?  100mA @ 300V is 30watt. So if powering this at 3 volts then you'd need 10Amps (12-13 if considering losses) input to it.  At those kind of current and power levels most parts on the board will not be happy and will not do a particulary good job....

  Are you sure? yes | no

Frédéric Druppel wrote 06/27/2017 at 20:03 point

Nope, I didn't...  And I didn't really consider the power usage of the converter. I should maybe change the current sense resistor value, and maybe optimise some other component values 

  Are you sure? yes | no

Daniel wrote 01/03/2017 at 20:45 point

Have been working on a similar project.  I'm curious to see if this works very well.

  Are you sure? yes | no

Frédéric Druppel wrote 01/03/2017 at 21:39 point

Me too ! It's my first time working with "self-made" circuits with DC-DC converters. I hope it'll work !!

  Are you sure? yes | no

Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates