Think I need a new PEM Fan... Anything else to check?

[gregd]

Puzzled... In a moment of self-challenge, I thought I'd do a little experiment. Trying some small nails, I found some that were a tight fit into the damaged end of the connector on the PEM-end of the wiring harness. The PEM fan positive is toast, but the motor positive isn't in too bad a shape, and the nail seemed to make a tight fit.. Soldering some twin lead to the nails, with a power pole on the other end, I re-connected the "loose" pair of connectors at the blower motor end. 12v in at the PEM-end, and the blower spins normally, pulling about 11.8 amps. Nice.

The test: If my theory is correct, I should be able to tug and wiggle the supposedly loose wires and get the fan to cut out, or at least have the current drop markedly. Couldn't do it, at least it didn't seem so.

So, if the connectors, or specifically the positive pole of such, can't come loose, then the conclusion about it being the root cause of the failure at the PEM-end doesn't hold water. And in this test, I was supplying the entire motor load on one set of wires, vs the normal case where it's shared among the two sets. That should have made it easier to see the failure, yet, no failure was seen.

What am I missing? It all seemed so perfect.

 

[ML Auto]

What am I missing? It all seemed so perfect.

It will be so perfect if you would just fix the connector. It doesnt matter if you run your 11.8 Amps thru one wire or two. Originally each wire carried almost 14 Amps, so reducing current flow on one set wont change anything. The current is probably sensed at the APS output, so the car doesnt care how the current flows. Everything you have desrcribed and all your test results point to a bad connector, which you have visually verified. I dont see any more problems.

 

[gregd]

It will be so perfect if you would just fix the connector. It doesnt matter if you run your 11.8 Amps thru one wire or two. Originally each wire carried almost 14 Amps, so reducing current flow on one set wont change anything. The current is probably sensed at the APS output, so the car doesnt care how the current flows. Everything you have desrcribed and all your test results point to a bad connector, which you have visually verified. I dont see any more problems.

Oh, don't get me wrong. I understand the connector is fried and needs to be replaced. That's not the root of this thread. The issue is the next word in that sentence. "Again!"

This particular car has a history of having that connector fry, and I'm assuming that not every 2.x goes through a connector every couple of years. I've owned the car since the very end of 2014. So, call it 5 1/2 years. This will be it's 3rd connector repair. Something else is going on, and I'm trying to determine (and fix) whatever it is.

There are two approaches. One is to use a sledge hammer, replace the connector pair with something totally different, and call it done. Tempting. But before doing so, I'm trying to get to the bottom of why my car, and few others, seem to eat PEM connectors as if they're disposable. I thought I had it figured out, and maybe I do. But my engineering background reminds me that one should not dismiss contrary evidence. It may be convenient to do so, but ultimately it's not going to lead to the lasting solution I am after.

 

[ML Auto]

The connector is marginal in this application, and Tesla's Band-Aid fan upgrade is not much better. If you are at all familiar with Tesla's inexperience with car building, you will realize that the only lasting solution is a redesign of the flawed system. TPMS, PEM, ESS, 400V controller, brakes, headlights, radio, etc. Some issues were addressed, some are still ignored. The car was an experiment, not a long term vehicle.

Dig out the BFH and fix it right.

 

[X.l.r.8]

Anyone have a part number for the 90 degree socket. And if possible gold plated pins for the PEM side. I am going through something similar, I swapped out the plug side and put gold plated female pins on the plug, but would like to do the PEM side to.

 

[PV-EV]

Molex 0194270032 tin
Digikey WM17505-ND
Replaced mine in 2016. No trouble since. Molex has a good web site. You should be able to find the # for the gold plated ones.

 

[X.l.r.8]

019427-0107 for gold-plate, thankyou @PV-EV

 

[gregd]

Another quick update... I created a short pigtail cable, F1 spades on one end, digital amp meter, then F1 female on the other. Spades plugged into the (still loose) wiring harness, female ends attached firmly to the motor. Driving the original wiring harness at the PEM end from an external source, I can clearly see a variation in the current going to the motor as I wiggle the spades. It wasn't enough to shut down the blower motor, but I could hear a change in motor pitch and saw a measurable change in current depending on the position of the blades. With a higher load (this was only about 50%), and over a range of temperature, I conclude that the loose connectors likely played a role in this. There's no magic in them. They are simply loose on my car, with unreliable electrical behavior.

Next steps. I (finally) received in the mail some new Molex connector pins for the PEM-end. I'll be fashioning a test cable with the PEM-end pins on one end, and F1 female on the other, to wire around the original wiring harness. With reliable exposed F1s on the blower end I can insert the power meter and verify the balance between the two sides as driven by the PEM itself, and hopefully demonstrate the effect of the loose positive terminal's role in load shifting. This will also verify that I can get good contact with the PEM connector pins, and not need to pull the PEM to replace its connector as well. Fingers crossed...

 

[gregd]

Side note: Why are the Molex connector pins (female side) so bad?

I spent a little time with one of the brand new bare Molex pins, squirting a bit of DeOxit on it and fitting it over the PEM-side pin. The idea was to do a little metal-level cleaning of the PEM's connector. Wow. After a couple of wipes, I the found the connector pin simply slipping over the male end . Totally no friction at all. Took a closer look at the female pin.

To be fair, the pin was stuck on the end of a short stick of wood, in order that I could reach the PEM-side connector with the PEM still fitted to the car. So, there was poor alignment of the pins when pushing them together. But still...

After looking at the pin under a strong magnifier, I see the problem. The "spring" part of the pin is simply made from the same material as the rest of the part, in the form of a metal flap that is bent around and into the body of the connector. Specifically it's a metal that is easily deformed and designed to hold its shape after being deformed. That makes for a really nice crimp on one end. And a poor connector on the other. If the metal was somehow treated for more "springiness", it certainly failed to do so. Very fragile. I suspect that one can damage the connector when mating them, if any sort of wiggle or angle is applied, so be careful when cycling the connector during the annual maintenance!

I "fixed" the pin by shoving a toothpick into the wire-end (no wire was there yet), and under the metal spring, to lift it back into position. I'm thinking of inserting something similar when rebuilding the connector, to keep it from flattening over time. Is this a good idea? It would need to be the right thickness, of course, and provide just enough "springiness".

 

[X.l.r.8]

Maybe a piece of gasket rubber, or a snip of closed cell foam?

 

[KJA2001]

So there is a project for someone, design and make more durable female pins to replace the Molex ones. It would be for the greater good!

 

[PV-EV]

The last pins I bought were less than a dollar each and are easy to replace. Not sure why a person would monkey around trying to fix them. If my better half was stranded on the side of the road because I saved four dollars I would never get to drive her car again.

 

[gregd]

This isn't about cost, it's about making a fix that will last. The trick is doing it in a backwards-compatible manner, such that Tesla service won't balk at working on the car.

 

[KJA2001]

This isn't about cost, it's about making a fix that will last.

That

 

[gregd]

#$%$%&^%^*$##$^&*!!!

(Is that strong enough language?)

So I got the new test cable made, and installed, and the voltages at the other end are the SAME as when all this started. So, yes, the connector at the PEM of the wiring harness is fried, but fried pins or not, they were apparently making good enough contact to send current through, if current was being sourced. The problem is upstream, which means, I think, it's the CIC board. Back to page 2 of the thread.

#$%$%&^%^*$##$^&*!!!

Now what?

 

[jason.tv]

#$%$%&^%^*$##$^&*!!!

(Is that strong enough language?)

So I got the new test cable made, and installed, and the voltages at the other end are the SAME as when all this started. So, yes, the connector at the PEM of the wiring harness is fried, but fried pins or not, they were apparently making good enough contact to send current through, if current was being sourced. The problem is upstream, which means, I think, it's the CIC board. Back to page 2 of the thread.

#$%$%&^%^*$##$^&*!!!

Now what?

would it be an option to send the main board connector to Gruber to get fixed and upgraded just for this one piece? Looks like they can fix this.

Fast forward to 14:40 mark and again at 17:20 mark; seems like they have a fix for the fan connectors wearing out.

 

[gregd]

Hi Jason,

Email already sent to Peter with the update, and we had discussed this in the past. My preference is to not dig into the PEM, though that might be unavoidable at this point, and sending the entire PEM to Gruber is apparently best done using the car as the "shipping crate". So this could be a big deal...

I'm also going to give the local SC another try, though the last email to them requesting a pricing on swapping the CIC board (presuming I needed a new connector) went unanswered. I can throw a 12v battery in the trunk and run the fan from that for the 20 minute drive to the SC.

 

[gregd]

ACK!! Never mind! All is fine.

I woke up this morning remembering the charging sessions early on, where the 114x beeps didn't come right when charging started, but a minute or two later. Duh! The PEM was still cold, and didn't need cooling. I just ran a test withOUT the blower going from an external source, with just the 6W lamp on the PEM side, and a resistor on Motor. No beeps at wake-up, then about 2 minutes into the charge, the lamp turned on! and I measured 11v across the resistor. Of course, neither load would look like the SPAL motor in terms of draw (so I got an 1144 beep in the cabin - the car noticed that too), but it confirmed that the CIC board appears to be functioning properly. Sorry for the distraction!

So, back to Plan A. Next charge cycle (tomorrow) I'll hook the PEM back to the motor, with a meter on one of the legs. Measure the current. Next cycle after that swap to the other leg. If both are about the same, that will confirm that the new wiring harness is making good contact with the PEM, so I can move forward with just the final wiring harness (connecting the two sides together above the SPAL motor), and a new blower assembly (while I'm in there). Also create a mesh screen over the blower, and put it all back together.

YAY!

 

[gregd]

So, looks like I might be good. Each side was putting out just a bit over 7 volts and about 3.5 amps. The readings bounced around a bunch, and I could hear the motor change its sound slightly as it ran. Not sure if that's normal or not. Might be a sign that the motor is tired.

I'll do a few more measurement runs, just to get some statistical data. If the motor side is consistently higher, that might mean the oxidation on pin #1 (PEM Positive, where most of the burning was evident) might be a problem that would force a repair on the CIC board. Email request to the SC for a quote has so far gone unanswered.

Assuming all is ok, I expect I'll be swapping the blower assembly out, and using a custom cable with a single feed line spliced together near the PEM, terminating in a single higher power connector at the motor. Then, while I'm in there, some open-weave filter material over the whole thing to keep out the leaves and such, and button it all back up.

Oh, and to the discoveries of the "R-C circuit behavior" back on page 2, I am thinking now that it's a low-power circuit test. Waking the car up with both sides tied to almost anything (even a small lamp) is sufficient to keep the car happy, so long as it doesn't need to spin the motor. But disconnect either side, and with a few seconds there will be an alert.

 

[X.l.r.8]

I still have the 2 separate fans and my speeds jump around all over the place. I would be happy if they were a straight on and off, but they seem to be variable, I have not even looked at my new fans to see if they have multi or variable speeds, I’m just happy they are working again