'99 Explorer keeps eating coil. HELP!

[poppitb]

It's a '99 with the pushrod 4.0L and she's turned beasty since early summer. Started throwing misfire codes for cylinders 1 and sometimes 5. (And a minor evap leak code, but I found a torn vac line that I think took care of that one.) Plugs are one year old and replaced with Motorcraft platinum tipped (also weird - right bank showed normal wear (0.065), but left bank all wide open - like >0.090.) Wires replaced this summer but no help. Then did the coil in July and things were better, for awhile at least. Starting running rough again in October and as I was in Asia my wife took it to the garage. They did all kinds of tests and replaced the coil again. Well, guess what. Running like crap again and I'll check it tomorrow when my daughter comes by from school, but I'll bet it'll be the same two codes.
Anyone have a guess why or what is eating my coil? I really don't want to pay the garage another $300 to do what I did for $75 and have it not work again.
Thanks,
Tim

[shorod]

Well the most beneficial thing would be to get the current codes, if any. Is the CEL currently illuminated?

If you again have a misfire on those two cylinders, and you're certain that the wires are good and seated well, I'd suggest you check the compression on those two cylinders relative to the others. Maybe they have somewhat low compression and a new coil works well enough to overcome the low compression and ignite the compressed air fuel mixture, but after the coil wears a bit it doesn't produce quite as much secondary voltage and the lower compression makes the air/fuel mixture less ready to ignite. Even to me that seems pretty unlikely.

It seems more likely that you have a couple of plug wires that are not making good connection and get reseated for a short time when the coil is replaced. It would be quite a coincidence that the same two wires were bad with the old coil and the same two new wires are also bad. However, it those two are in difficult locations to make sure they snap on completely, it may be a bit more plausible.

-Rod

[rhandwor]

Use a DVOM and check the voltage to each coil. It should be close to battery voltage. The computer pluses the ground. Use a noid light and check if the light is blinking properly. Wiggle the wires and check for a bad terminal.
Otherwise buy a compression gage and check compression as suggested.
http://www.autopart.com/tools/toolsmain/tool/T_2004.htm
http://www.autopart.com/tools/toolsmain/tool/T_3401.htm