5
\$\begingroup\$

I was researching different ways to make a high voltage DC-DC power supply in the 20 kV ~ 30 kV range, with output of a few tens of micro amps. The purpose is for testing isolation and resistance to ESD on panel mount components.

One common factor I seem to run across in the DIY electronics world is people using transformers with very high turns ratios. They can't be made by hand and are often salvaged from old CRT monitors/TVs. While that works for a one-off design, it's not really suitable for designing a reproducible product.

Why not just use several transformers in series with lower turns ratios instead of one transformer with a very high turns ratio? For example, instead of using a 1000:1 transformer, use three 10:1 transformers.

10:1 transformers can pretty easily be wound by hand. It's likely cheaper to make (at least from a coil winding standpoint), and as a bonus you get multiple stages of isolation between your input and output.

I am sure there is some practical reason for it, but it's not yet obvious to me.

enter image description here

\$\endgroup\$
8
  • \$\begingroup\$ did you check for "cascading transformer"? What is the missing information in that search? \$\endgroup\$ Commented 15 hours ago
  • 1
    \$\begingroup\$ Have you considered flux saturation density and what that means regarding the windings (i.e., number of turns) for T3 & T4? \$\endgroup\$
    – qrk
    Commented 15 hours ago
  • \$\begingroup\$ it seems simple only if you're using ideal transfomers. \$\endgroup\$ Commented 14 hours ago
  • \$\begingroup\$ @qrk All good comments, it would be nice to see an answer with concrete math. Obviously the second and third stages deal with higher voltages but must operate at the same frequency as the first stage. To keep the current under control and avoid saturation one has to increase the inductance by adding more turns of wire. I am hoping it doesn't work out with something dumb like needing the last stage to have 1000X the number of turns (compared to the first one). \$\endgroup\$
    – user4574
    Commented 13 hours ago
  • 1
    \$\begingroup\$ "not really suitable for designing a reproducible product" but if you're designing a product, you wouldn't be hand-winding anyway, you'd be buying from a suitable supplier. Even for small runs on a commercial scale, custom windings are feasible. There are also DIY coil-winder designs; while I've only done that for magnets, some would handle HV transformers \$\endgroup\$
    – Chris H
    Commented 3 hours ago

1 Answer 1

12
\$\begingroup\$

Why not just use several transformers in series with lower turns ratios instead of one transformer with a very high turns ratio?

Every magnetic core has a maximum volts per turn that can be applied at a given frequency without saturation. If your core has, say, a maximum of 1 volt per turn, at your desired frequency, then you must have 10000 turns on any winding that has 10000 volts on it. If your several transformers have the same sized core, there is no advantage turn-wise for not putting your 10000 turns on first transformer, and be done with the step-up process.

\$\endgroup\$

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.