0
\$\begingroup\$

I'm relatively new to electronics and I'm trying to make a boost converter for a nixie tube watch that runs off of 3v.

I'm pretty sure that the most current that will be drawn is about 70mA. I'm not sure how to choose what components I need for it.

schematic

simulate this circuit – Schematic created using CircuitLab

The duty cycle will be provided by an atmega328p. If there's any other information that you need please let me know, any help would be appreciated.

\$\endgroup\$
  • \$\begingroup\$ Is it 70 mA at 170 V output? \$\endgroup\$ – Ale..chenski Nov 23 '18 at 5:21
  • \$\begingroup\$ I think that's total. The 170v goes to two 74141 transistors that draw 25mA each and they both output to a nixie tube which has a current draw of 1mA each \$\endgroup\$ – Some guy Nov 23 '18 at 5:42
0
\$\begingroup\$

As Wouter has already mentioned, the current drawn on the high-voltage rail will be far less than 70 mA.

I have built nixie clocks with a 160V HV rail and the power supply is only capable of 20 mA, which is more than sufficient for 6 tubes + 4 glim bulbs.

For completeness, here's a schematic of the 555-based boost convertor: Nixie clock step-up

Taken from this Belgian project.

If you want to use it, you only need to have a regular step up (3 to 9 volts), and you wouldn't need the ATmega's PWM (the power supply is an independent unit, only the ENABLE pin needs control).

\$\endgroup\$
1
\$\begingroup\$

This is not an answer to your question, but it has implicatipons on your design (and I can't put pictures in a comment).

A 74141 is not a transistor, it is a decoder + open-collector driver. The 25 mA drawn by the 74141 is from its VCC supply. (last row in the table)

enter image description here

The VCC is ~ 5V, so you don't need to design your 120V to supply it.

enter image description here

Your load on the 120V will be pretty much only what your Nixies draw.

PS designing with a nice capacitor charged to 120V is not something I would recommend to a beginner. Or are you in for a Darwin award?

\$\endgroup\$
  • \$\begingroup\$ So, the OP needs just about 8 mA total on high side. This certainly changes all considerations. It was 40+ years ago when I was playing with clocks and Nixie tubes, lost a bit of number sense here...a simple small 50-Hz AC-AC transformer did the job, I recall no problems then. \$\endgroup\$ – Ale..chenski Nov 23 '18 at 8:27
0
\$\begingroup\$

Load of 70 mA at 170 V output is a 20-watt design. The area of SMPS converters design is somewhat sophisticated and requires some skills. For this range, especially with so low input voltage, the design will have challenges, and would require more than a transistor and diode, but a more advanced controller IC.

I would strongly advise to look into industrial offering of boost circuits, say, starting at Texas Instruments SELECTOR. Plug in your initial numbers, 3 Vin, 170 Vout at 0.07A, and you will have a selection of controllers that will be capable to get these parameters. The list will show only two basic parts, LM3478 and LM3488. Which already indicates a serious challenge.

Then try to use their automated DESIGNER to generate the entire design, which will offer a suitable selection of all components.

The result is unfortunate for you:

We're sorry, we could not find a solution to match these requirements for the following reason(s): Can't find suitable FET for the design. Try reducing load current

Attempt of use more general TI WebBench Designer evaluates maybe a hundred of controllers, but with the same result - can't do this. If TI can't find suitable FET for the design, there are little chances that you can.

I would say that you need to reconsider your design parameters, likely to use higher Vin, or more realistically evaluate the load current, or use a transformer-based converter topology with plain output rectifier.

\$\endgroup\$
0
\$\begingroup\$

At 30 nanoHenry Inductance, at near 100% ontime, with 3 volts, what must the Inductor sustain?

I = 1/L * integral[ V(t) dT]

and with 0.03 uH and 1uSecond on-time (so we have a TIME number) and 3 volts, the current rises to

33/uH * integral (3 volts * 1uS)

or Current = 100 amps. With V = 3 volts, this is 300 watts peak, or 150 watts average (the current ramps up ideally in a smooth triangle).

How about 3uH, and run at 100KHz? That builds up at 1 amp/uSecond, or to 10 amps peak every 10uS. Does that support enough energy transfer?

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.