7
\$\begingroup\$

I need a DC-DC convertor to bring a 5V input up to 160V DC output. Does anybody know off any of the shelf parts that would manage this? The high output needs to supply minimal current (uA) we really just need it to supply a potential.

We have a part for this currently but it's going out of production. We'd need the replacememt to have a similiar foot print (60mmx30mm).

I know linear tech have a good range of DC-DC controllers such as the LT3757, one of these may do the job.

Has anybody used these for something similiar?

Also I read an application note which which suggested it's possible to stack voltage references which may be another way of doing it as we don't need current. Has anyone tried this?

I'd be grateful for any suggestions.

\$\endgroup\$
  • 2
    \$\begingroup\$ 53 CR2016 cells in series. \$\endgroup\$ – Jasen Oct 10 '15 at 7:10
5
\$\begingroup\$

Not really a whole off-the-shelf part, but... A small step-up converter with an autotransformer sounds just right (flyback converter without isolation).

Also, you could try something like figure D1 on page 93 of Linear Tech's app'note 47 (without Q1 and the three resistors and the capacitor around it, just to the point that says "avalanche bias"). Adjust the divider to the FB pin for 160 V instead of 90 V. This little circuit is a combination of a step-up converter and a charge pump.

\$\endgroup\$
4
\$\begingroup\$

You're in luck - There are many ways of achieving this.

Stacking voltage references will not work - they are a means of reducing a voltage to an well defined level.

If this is a one off then buying one of the many many many products available may be easiest.

If DIYing - various boost converter iCs exist. Most of these do not have the voltage rating to handle 160 VDC directly, but use of a transformer will give you N:1 stepup.

I usually start with recommended and olde but goody smps IC MC34063. There are many better but this about anything at low ICcost.
MC34063 datasheet

Fig 19, page 14 does what you want. If you want REALLY low idle power there are better IC.s As you want only uA you want an IC that goes into burst mode at low load. Run a long long time on batteries if needs be.

I just noticed that they have pin 1 connected to output collector. That limits Vout to 40V. Instead return pin 1 to Vin via a resistor and Vout can be "high".

enter image description here


Here is a circuit that uses a more modern IC an that achieves 150-250 VDC with essentially the same core circuit as above, but with improvements. Uses Maxim MAX1771 IC. Datasheet here . Be wary of single source and availability - but seems to do your job exactly. Their design is aimed at 12V in but the IC will do 2V+. IC uses PFM to get low idle current at low load - high in your context and better can be done. If necessary a system that shuts down to essentially zero drain most of the time under uA loads can be implemented.

From here - Nixie HV Switching PSU

Note in text their comment on layout affecting efficiency. Stray capacitance and other AC losses matters at this boost ratio. Note eg lack of ground plane near inductor. Using their supplied PCB design would be wise.

enter image description here

enter image description here

\$\endgroup\$
  • \$\begingroup\$ Thank you for your reply, these have given me a better idea where to look, i'll have a closer look at the circuits mentioned. As for complete spec hhmm, 5V DC in from a swicther, output 160V DC, current output max 0.001 uA. ripple +/-30mV. Load will be steady . We really use the potential to direct ions \$\endgroup\$ – Conor O'Callaghan Mar 20 '12 at 14:29
  • \$\begingroup\$ With a single inductor, the duty cycle D=ton/T would have to be D=1-1/(Vout/Vin)=1-1/(160/5)=96.875%, leaving 3.125% as the on-time. Does the recommended IC really tolerate this, considering dead times and such? \$\endgroup\$ – zebonaut Mar 20 '12 at 15:27
  • \$\begingroup\$ @zebonaut - I noted "...use of a transformer will give you N:1 stepup." BUT, yes, (AFAIK :-) ) if you set the system running and it does not have enough energy to provide Vout in one cycle it will run for the cycles required until the final cycle delivers Vout >= limit and shut down till Vout drops. So it can have a very long off cycle if there is no load. Ton set by IC, Toffmin set by energy transfer. Tidle = wait for IC to catch up, repeat. Limit on Vout is the stray capacitance present. With no load Vout "rings" until 1/2Li^2 = 1/2C_stray.V^2. 5-> 160 is high but I believe is doable. \$\endgroup\$ – Russell McMahon Mar 21 '12 at 1:20
  • \$\begingroup\$ desmith.net/NMdS/Electronics/NixiePSU.html \$\endgroup\$ – Russell McMahon Mar 21 '12 at 1:24
  • \$\begingroup\$ @RussellMcMahon: Yeah, the first thing I thought when I had just read the headline of the question was: Yay, Nixies! \$\endgroup\$ – zebonaut Mar 21 '12 at 8:21
1
\$\begingroup\$
  • Use an oscillator with power transistor
  • to drive the 6V side of an off-the-shelf 230V/6V transformer
  • and rectify (diode) and smooth (capacitor) the HV output

If needed use a feed-back loop to regulate the output voltage.

EDIT
I removed the example circuit because of flaws mentioned by Olin.
Still outline above shows an easy way to generate 130V from 5V.

\$\endgroup\$
  • \$\begingroup\$ The left diode prevents this from working in flyback mode. However, for this large voltage ratio you probably want flyback mode. A single ended drive like this is not good for forward mode anyway. \$\endgroup\$ – Olin Lathrop Mar 20 '12 at 15:53

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.