dr bob

Jeff--

FET devices are rated for several factors. Beyond basic current capacities and thermal limits, perhaps the most important rating is RDSon. This stretches out to "Resistance Drain to Source On", the impedance of the device when it is energised. In turn, this is a key factor in how much heat is generated and therefore how much must be dissipated through a heat sink and available airflow. The numbers are pretty amazing these days, especially for those of us who grew up on using bipolar silicon devices for power supply designs. Anyway, reducing that RDSon number by half doesn't cut the heat dissipation requirement by half, but the improvement is still substantial.

Tom in Austin

Put me on the list too ... too much Texas heat for man or machine these days ...

CJ8 2 928

John Put me down for one. I will pick it up at the next breakfast. :---)

worf928

I count 16 units in the "I'm interested" side of the column. A number of folks have posted one or more times but did not indicate interest - so I didn't count them.

Jim M.

I'm interested!

SeanR

I know I am going to want one on the shelf for testing proposes, so add me to the list please.

BTW John, it touched cloth.

jleidel

Well played sir... well played.

jleidel

All quick update on our progress. We just had the board under the scope in the lab verifying some of the resistor tolerances [its been awhile since our lab folks have had to read the color bands]. We're going to put the board on the scope and feed it signals per the 928 electrical specs in order to verify our circuit/resistor mappings. In speaking with the lab folks, they're highly intrigued at the use of the standard black cylinder caps. They're only rated to 105 degrees C [221 F]. Apparently, they also have a tendency [especially circa 1980's] to dry out over time. This may actually be part of the cause for the fets going bad over time. The fets are of automotive quality where the caps are not.

Either way, in the new design we're going to make a point to rid ourselves of these failure points.

To answer Soontobered's question, we have yet to find a replacement connector. The pinout is actually quite strange. Its an 8 pin connector that is split to 9 solder points on the board. The circuit tells us that they're using this to improve the incoming signal quality by splitting it across two pads. At this point, it appears that we're going to need to use the original connector [however, we're still searching].

...more info as we go along.

soontobered84

John,
If "we" use the existing housing for these units, then we will be limited by the number of "cores" that we can come up with. Likewise the connector. What if "we" have an 8 or 9 pin female connector that would fit onto the existing male harness connector with tail into our new housing? Would that make it more doable? Or would that even matter?

Either way is good for me since as you saw I have/had a nice supply of "cores" that I distributed at the last First Saturday Breakfast GTG. I'm just thinking about normal folk here.

GregBBRD

I rebuild a fair amount of these units...and I'd love to be able to just replace the "guts" with something new....especially since I'm seeing more and more corrosion issues with these controllers, as they get older. Regarding removing and re-using the existing inside connector....these connectors are also getting old, have corrosion, and have increased resistance where they are "riveted" together. Note that some of the units were soldered at these rivets and some were not.

You could certainly count on me for a couple units for testing purposes.....I'm pretty "****" about testing anything new, before using it on any customer vehicles.

Cost will certainly be a factor. To give you an idea of where the "costs" are, when rebuilding these controllers, I currently spend approximately equal amounts of time working with the electronics as I do preparing the "old" case and bottom cover for "water proof" re-assembly. The one problem that I run into is finding enough of those little white "filters" that allow the case to breathe, while not allowing moisture inside. It would be really great if someone could figure out what that little white filter is made from and where to get additional pieces.

Alan

Understand that the power dissipated in the controller is a very small percentage of the total power consumed by the fan when running on full. So even if the controller was twice as efficient you'd hardly notice the difference in total power consumed.

It will have a much bigger benefit to the controller's longevity and that would be the primary point.

Alan

Speedtoys

Got it..thats what I was missing..HOW much is lost in heat.

jleidel

Answering questions from both GregBBRD and Soontobered...

I agree completely on the source of the connector. I've searched through the Molex milspec catalog for an hour or so and I've found a few that might [I say "might"] be interesting candidates. The milspec connectors are the best bet as they're generally moisture tight and surface mountable [see below]. As long as its pin compatible with the harness, then we can split the signal and mimic the current configuration on the board. At which point, we can do more interesting things with a more modern chassis.

All these things add some amount of cost and time, but I believe if we spend the time now, then we'll have a better solution in the future. So far, we've sourced more modern surface mount parts for everything except the actual connector. What does this mean to us? More durability for the long term and less cost in the short term.

kingsmen11

I need two units, since I cannibalized from two of my cars. Maybe a third. Thank you for the work on this project.

TheoJ

supporting Dr Bob: if you use proper modern FET's the generated heat is very little. Have you guys ever checked how hot the power stage with modern FET's really gets? Modern FET's are capable of doing 45 amps each, and have an extremely low resistance when fully conducting.