Ruggedizing an Iota Power Supply for EV Use
Like every other vehicle, conversion EVs need a source of 12 volt power to operate lights and accessories. There are three ways that hobbyist EV converters typically do this. There are advantages and disadvantages to each approach. For various reasons, the third option is probably the most popular. Some converters choose the Iota DLS range "battery chargers," mainly for their low cost. These aren't exactly what most people think of as battery chargers -- for one thing, they don't shut off when the battery is fully charged. They're actually just regulated switching power supplies. But they have a big advantage for conversion EV use: they can run on DC input just as well as they do on AC.

In the following narrative adapted from posts on the EVDL, longtime EVDL contributor Lee Hart lays out some thoughts on the Iotas' suitability for EV use, and how to improve them. The photo below will help you follow as you read his comments and suggestions.

Iota DLS-55

Photo: Rod Hower

I recently had an Iota apart to see how it was built, and wasn't impressed. These things are not EV quality DC:DC converters, they are low quality indoor AC power supplies. I wouldn't recommend one for anything over a 156vdc battery pack. Here are the problems as I see them, and what I did about them on the ones I've used.

  1. It only uses 200vdc capacitors. A lead-acid battery will exceed 2.5v/cell during charging (195v for a 156v pack). The two 1000uF 200vdc black filter capacitors at the top left literally exploded in an EV with a 156v pack during charging. The maximum voltage on a 192vdc pack would be even higher, exceeding 240v! I replaced them with 330uF 250vdc electrolytics. I also epoxied a piece of PC board material across the tops of them to tie them to each other for support (they were held only by their leads).

  2. There is no input fuse. You'll need to install your own input fuse, rated for DC use, at your pack voltage or higher. Without one, you risk a serious fire or explosion if the input diodes, capacitors, or other parts fail. I added a 15-amp Bussman 3AB or Littelfuse ABC type 250vdc fuse.

  3. The various large capacitors, inductors, and power resistors are not securely mounted. They depend on their wire leads and a dab of silicone rubber. This isn't enough in a vehicle, which shakes and vibrates. A friend's Iota failed when the large toroid inductor broke its solder joints.

    1. I secured the big blue core toroid in the foreground with a piece of blank PCB material through the center, with a screw at each end through the PC board. I then epoxied it to the board and to the side of the transformer.

    2. The yellow boxed-in toroid at the left is also held only by its leads. I cut off the yellow tape, and epoxied it to the PC board.

    3. Several other parts with large heavy leads were not properly soldered. I resoldered them.

  4. The bridge rectifier was held by the screw immediately above the yellow boxed-in toroid. In mine, it was poorly soldered. I removed the bridge and replaced it with two GE Sensors CL-30 inrush limiters, mounted in the 4 holes where the bridge used to be. With the bridge removed, there was no need for its heatsink. Removing the heatsink allowed more room for the inrush limiters, to keep them and their (normal) heat away from surrounding parts. This made the input DC ONLY, but added inrush protection.

  5. The TO-220 transistor right between the two toroids is Q6 (a P16NF06 MOSFET on the DLS-45). It regulates the voltage for the fan. It has no heatsink, and runs rather hot. I added a small bolt-on heatsink. I epoxied the heatsink to the PC board for support.

  6. The TO-220 part up against the two big electrolytics is U1 (an LM317T 3-terminal adjustable voltage regulator). Pop rivets are a terrible way to mount a transistor to a heatsink! They can be so tight that they warp and crack the case, or so loose that they hardly do any good. In this case, it was too loose. I replaced it with a clip-on heatsink, and epoxied it to the capacitors for support.

  7. The big white square thing standing up in front of the fan is a 47 ohm 10w resistor for the RC snubber on the power MOSFETs. It seriously overheats with low input voltages (as voltage goes down, current goes up, so snubbing losses get worse). I replaced it with a 47 ohm 20w resistor. I used a metal cased chassis mount resistor, mounted to the metal end plate left of the fan and behind the two big electrolytics.

  8. The insulator between the MOSFETs and the case is a simple piece of sleeving. It was placed over the MOSFET, and then the MOSFET was clamped to the case with a spring clip. The ends of the power supply's extruded case weren't deburred. When the assembler who put mine together slid the PC board into the case, the case ends cut through the sleeving on one MOSFET. The breakdown voltage between input and case was thus very low, just the air gap in the cut sleeving. I turned the sleeving around so the cut place wasn't between the MOSFET and case/heatsink. I'm sure this doesn't happen to every supply, but it's still something you should check for in yours.

  9. There is no moisture protection. Many high voltage traces have very small spacings. There was an attempt to conformally coat the board, but most of it was not coated. I coated the entire board with Dow 1-2577 conformal coating. This is a thick clear silicone rubber based coating that is easy to remove if servicing is needed.

  10. The output fuseholders are a cheap home-made affair. Not much I could do about it. :-(