Posted to RORT 12/02/04

Awhile back Bob Giddings entertained us with his Intellipower/Charge Wizard problems.  The Intellipower finally quit and Bob sent it to me to do a post-mortem.  It arrived yesterday and I opened it up, having to drill out pop rivets in the process.

The first thing that caught my attention was just how simple this thing is.  No intelligence despite the name.  Only an analog switchmode power supply controller and an op-amp.  Particularly given the price, I was quite surprised.  Here are some photos:

http://bellsouthpwp.net/j/o/johngd/files/rv/Intellipower/

This 13.8 volt @ 30 amp unit is about 400 watts.  I enclosed a photo of a 500 watt PC power supply that I recently bought from a local PC vendor new in the box for $19.95 for comparison.  This thing has at least 3 output voltages, multiple connectors and the IPX intelligence, yet it sells for a tiny fraction of what Progressive Dynamics charges for the Intellipower.

One might think that the volume production involved in PC power supplies would account for the difference.  Problem with that theory is that of all the PC power supplies I've take apart, I've never seen two that were the same internally.  I'm going to convert that PC power supply to 12 volts only and see if I can interface it to the Charge Wizard to make a 40 amp converter for my rig.

The problem with Bob's converter was immediately obvious.  The incoming power filter cap had burst.  See the arrow in the picture.  I was quite amazed at the UNDER-rating of this cap.  It is only a 200 volt cap.  It charges to 1.414 times the incoming line voltage.  For 120 volts, that would be 170 volts.  That is uncomfortably close to the cap's rating.  But consider 130 volts in, not uncommon in a lightly loaded campground or generator.  That works out to 184 volts.  To put the full rated voltage on the cap, the line would only have to be at 141 volts, not at all uncommon during switching transients and generator overspeeds such as when a big load is dumped.

The next thing that caught my eye was the small capacity, only 820uF.  This is quite small for the watt rating of this power supply.  The smaller the cap, the higher the ripple amplitude.  The higher the ripple voltage, the higher the ripple current.  The higher the current the more internal heating.  Which brings  up the next point.

The cap is only rated at 105 deg C.  Mount this converter in an unvented space, apply full load (depleted battery) in the summer and you can easily hit 100 deg C or more.  Even if the thermal rating isn't exceeded, a rule of thumb is that the life of an electrolytic roughly halves for each 10 deg C rise.  Again, this is an uncomfortably design margin.

I would have specified (and will for the replacement) at least a 2500uF, 300 volt cap. The higher capacity will result in less heating so the 105 deg C rating will be fine.

Rather than bursting the end which is normal, this cap relieved itself by forcing the rubber lead seal out of the case.  Instead of a catastrophic failure (and bad stink), this cap gradually failed over time as its electrolyte gradually dried out.  As the cap dried out its capacity decreased and resulted in more and more ripple reaching the downstream components.  This gradually increasing ripple is what accounted for Bob's report of slowly increasing misbehavior of his 12 volt system.  It is also probably what smoked his Charge Wizard.

I'm going to take this unit completely apart, reverse-engineer a schematic and then repair it.  In the process I'm going to figure out the details of the charge wizard.  I'll post this stuff as I go along for those who are interested.

The simplicity of the Intellipower reinforces my opinion that the Charge Wizard can be used with most any 12vdc power supply to make a smart charger.  I'm going to try to do it with a PC power supply.  If that works, one could have a 40 or 50 amp intelligent charger and converter for as little as $50.

John