7
\$\begingroup\$

Imagine I have a single-IC switching DC-DC converter which supports input voltage up to 16V, for example a LM2727.

Can I stack the inputs of such converter in series, for example 64 of them, and be able to apply up to 1024V across the combined input?

I need lots of stable 3.6V outputs anyway (for charging each cell in a multi-cell battery separately).

And on the output side, can I put a 3.6V cell between the outputs of each DC-DC converter and then connect all the batteries in series - will that be fine, or everything will blow-up?

\$\endgroup\$
4
  • 1
    \$\begingroup\$ Oh, you want a 1000V input... Where do you get this 1000V from? \$\endgroup\$
    – jpc
    Commented Apr 5, 2011 at 14:34
  • \$\begingroup\$ @jpc: I won't actually apply 1000V, but one or two hundred volts from a solar panel made of cells connected in series might be there. I am just trying to find out if this is a feasible way to do such conversions, and then I'll go on choosing the exact input voltage range. I also need something for rectified mains power input, but that'll probably need an isolation transformer, so things get more complicated. The idea is to combine DC-DC conversion and cell balancing in one go, otherwise I need to design a balancer anyway to each battery cell to go up to 3.6V while charging, no less, no more. \$\endgroup\$
    – ria
    Commented Apr 5, 2011 at 14:45
  • \$\begingroup\$ If cabling is not limiting you than maybe two (or three) separate solar panel circuits (with 500V or 340V output each) will be much easier to handle. Most popular cables are not even rated for 1000V operation. Unfortunatelly I am no expert on current balancing. \$\endgroup\$
    – jpc
    Commented Apr 5, 2011 at 15:00
  • 1
    \$\begingroup\$ Don't do it. There are various safety issues. \$\endgroup\$ Commented Apr 5, 2011 at 21:26

3 Answers 3

8
\$\begingroup\$

Connecting the chargers in series

Connecting several such switchers in series will be a problem since they would have to share the same supply current (all would have to draw the same current all the time). All the chargers have some input resistance which vary according to their immediate needs. If one of the chargers draws less current then:

  1. It's input resistance rises so it will get less current.
  2. This causes the voltage on this charger to rise.
  3. Other chargers get lower a little lower voltage and when they "notice" it they lower their input resistances to get the required current.
  4. This effect is highly unstable. Eventually only one charger (unfortunately the one that requires the least power) will get all the input voltage and (most probably) be destroyed.

To verify this logic assume that one charger finished charging and does not need any current at all. The same thing happens.

The only way out is to dissipate excess (unneeded at the moment) power but that is probably not something you would like to do. :)

High input voltage SMPS

A single SMPS supply working from 1000V input is also quite difficult to do since most MOSFETs and IGBTs are only rated up to 1200V. A 600V supply should be doable.

Output current regulation in multiple-output SMPS

You could do a multi-output transformer based switched supply and regulate the output voltages and currents with a magnetic amplifier (this Ferroxcube flyer is a proof that this is used in real circuits)

\$\endgroup\$
2
  • \$\begingroup\$ And what would happen if the battery behind one of the switchers draws less current then others? Will the others down-regulate to the lowest current draw among them all, or something else will happen? As to the input nothing over 600 V will ever possibly appear on the combined input. \$\endgroup\$
    – ria
    Commented Apr 5, 2011 at 15:07
  • \$\begingroup\$ @miernik I expanded the answer. Let me know if you want me to add anything. \$\endgroup\$
    – jpc
    Commented Apr 5, 2011 at 15:30
1
\$\begingroup\$

Edit: I completely misread your question. I think this is still applicable, though:

So, it sounds like you want to "stack" the inputs/outputs of these converter IC's, which I would describe as more of a parallel combination rather than a series combination, although I do not think that either term really describes this arrangement perfectly. From what I know about the various popular DC-DC converter topologies, my intuition tells me that this would fail, although to fully explain why would take a bit of digging and explanation and circuit theory (probably). Maybe someone else can provide a concise answer.

Edit: I am not actually sure all of the following would apply to the proposed arrangement, since your switches may share the combined 1kV and therefore no one switch would have the full magnitude of the input voltage across it.

A final note: 1kV is really high and generally super dangerous. In order to deal with this kind of voltage, all your switches (probably most of your other components as well) would need to be rated to this voltage, as would your PCB, etc. I'm guessing you were exaggerating on that number in order to get your point across.

\$\endgroup\$
2
  • \$\begingroup\$ nowhere I am trying to step-up the voltage here, maybe you didn't understand my question. What I am trying to do is to increase the input-voltage range of the DC-DC converter by stacking them in series. \$\endgroup\$
    – ria
    Commented Apr 5, 2011 at 19:00
  • \$\begingroup\$ @miernik: You are quite right, answer updated. \$\endgroup\$
    – Adam P
    Commented Apr 5, 2011 at 19:00
0
\$\begingroup\$

Tamura and CLC are companies making DC/DC converters with rated input ranges of 200-1500Vdc. Both are certified to UL1741 and/or IEC62109-1, which is required per article 690 of the electric code. Both have outputs that are rated as SELV (safe to touch), and are rated for overload and over-temperatures. If you are not an experienced power supply designer and attempt this and stuff goes bad, one or more of several things will happen:

  • your house and/or neighbor's house will burn down
  • member of your family or a first responder incurs shock or burn injury
  • your insurer legally abandons you
  • the local AHJ cites you for code violation, so resultant damage becomes civil and/or criminal issue.
\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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