Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Getting Really Frustrated

This site may earn commission on affiliate links.
"$10k to replace the PEM with a refurb when it could be something like a blown fuse or a dry joint is not right"

I couldn't agree more. I had my entire PEM replaced for a "loose connection" two years ago (luckily, under the CPO warranty). Makes me dubious about getting the $29K 3.0 battery... what's the point if it becomes impossible to maintain the PEM or some other critical part? Not to mention the big problem of fewer and fewer trained techs and the lack of parts? Getting information is like pulling teeth, too. For instance, whatever happened to the other parts of the 3.0 upgrade? Silence there. Mr. Musk could step in and address all of this, but he hasn't yet....

O.K. done ranting. I still love my Roadster, and there's nothing out there that comes even vaguely close to it.
 
  • Informative
Reactions: neroden
I'm wondering how hard it would be for a 3rd party to make a brand new open-source PEM that looks functionally identical to the rest of the car (replicating CAN bus messages etc) but using newer, smaller and not obsolete components.

That would be great. A compatible PEM, with maybe a possibility to liquid-cool it using the existing battery cooling circuit would be a great improvement. From my point of view the PEM is THE weak part of the car that I'm afraid of :(
 
  • Like
Reactions: BartJ and dpeilow
For my daily use the air-cooled PEM works very well, its a 1.5 so I don't have issues with dirty PEM fans and clogged ducts. They went air-cooled for I believe its cheaper, easier and cheaper to implement. Remember they were not a full-bore auto manufacturer. They did learn and have more time to cool all the PEM electronics that get heated, by moving them down into the liquid cooled motor area, leveraging off the liquid cooling.

Racing is what pushes the car to its cooling limit, however the first thing to heatup and cause power limiting is the air cooled motor, PEM is second.
 
  • Like
Reactions: AtAge19
Does anyone know WHY the PEM is not liquid cooled? This seems like an absolute no brainer...

Primarily to save weight which contributes to better range and performance. If you don't track your car the air PEM cooling is more than adequate. The real question is WHY they don't have a better filtering system for the cooling air. The only time the air cooling system is inadequate for aggressive street driving is when it's dirty.
 
I'm wondering how hard it would be for a 3rd party to make a brand new open-source PEM that looks functionally identical to the rest of the car (replicating CAN bus messages etc) but using newer, smaller and not obsolete components.
I am sure that a third party could make a new PEM and offer it for sale. But there would be significant costs associated with its development and the potential market is very small, maybe a few hundred units over the next decade. The economics don't work.
Does anyone know WHY the PEM is not liquid cooled? This seems like an absolute no brainer.
Definitely the Roadster's Achilles Heel
People sometimes forget that a decade or more ago, when Tesla was designing the Roadster, they were a tiny company with very finite resources and on the verge of bankruptcy. They could not optimize every part of the car. They knew that liquid cooling of the PEM was not needed for street use and they knew that making the car track ready would involve more time and expense. The Roadster is not a race car.

As Henry and others have pointed out, the air cooled PEM is fine for street driving.
 
  • Like
Reactions: nsxpert
I'm not convinced.

The second half of your post backs up my thinking that a small company could do it.

Even if half the Roadsters still likely on the road need a replacement and they charge $5000, that's $5000000 for someone. They wouldn't necessarily have to even produce another housing or even replace all the PCBs in the existing PEM to start with.
 
I'm not convinced.

The second half of your post backs up my thinking that a small company could do it.

Even if half the Roadsters still likely on the road need a replacement and they charge $5000, that's $5000000 for someone. They wouldn't necessarily have to even produce another housing or even replace all the PCBs in the existing PEM to start with.
Think of the demographics f Roadster owners. Why would I pay some random company $5,000 when I could pay Tesla $7,000 or $10,000? Roadster owners have a certain amount of loyality and dedication to the company. I doubt an outside company would have demand for 100 units. I bet they'd produce closer to 50. So, revenue of $250,000 instead of your $5MM projection. And that's over time. And requires a significant investment in skills and tools.

I can tell you it's not a venture I'd undertake. Would you?
 
Back in 2013, I lost my PEM when I was on a long distance trip, about 1,600 kms from home.

it blew with just under 60, 000 kms on the odometer, so Tesla processed the warranty claim for me.

When I returned, I immediately purchased both the battery and the "remainder of vehicle" extended warranties on my Roadster.

Did any of the other posters on this thread purchase the extended warranties?
 
A third party has already created their own Roadster PEM. Drive eO created one for their Pikes Peak Roadster. Their car was 360kW (483 hp) and 800 Nm (590 ft lbs) peak. In addition to being drastically higher power than Tesla's, their PEM looks smaller.
Tesla Roadster 360 by Drive eO

I disagree with the sentiment that air cooling is adequte for the PEM. That might be true in moderate climates, but in places with a hot summer (like Texas), the PEM cooling is insufficient. My PEM hits the red first, then my battery and finally my motor.

I think the ideal solution is to replace the Roadster's whole drivetrain with a Model S drivetrain. Even better, wait a while longer and stick in the Model 3's drivetrain. Then everything is liquid cooled and more reliable.
 
  • Informative
Reactions: AtAge19 and Rolf68
Think of the demographics f Roadster owners. Why would I pay some random company $5,000 when I could pay Tesla $7,000 or $10,000?

Well, I'm hearing in person and in various threads on this forum that people are starting to get a bit tired of being told they have to cough up $10k for any minor failure. If someone created a more modular, more serviceable PEM without waiting lists or delays for replacement, I am sure there would be some interest. And as had been said, everyone will face this failure in due course.

A third party has already created their own Roadster PEM. Drive eO created one for their Pikes Peak Roadster. Their car was 360kW (483 hp) and 800 Nm (590 ft lbs) peak. In addition to being drastically higher power than Tesla's, their PEM looks smaller.
Tesla Roadster 360 by Drive eO

Super interesting, thanks.

Looks like they built a whole new battery too Electric eO PP02 at Pikes Peak International Hill Climb 2014 – Drive eO
 
Last edited:
  • Informative
Reactions: AtAge19 and neroden
Well, I'm hearing in person and in various threads on this forum that people are starting to get a bit tired of being told they have to cough up $10k for any minor failure. If someone created a more modular, more serviceable PEM without waiting lists or delays for replacement, I am sure there would be some interest. And as had been said, everyone will face this failure in due course.

Ideally, a more modular car would be ideal, but given Tesla's founding need to create something to prove the base technology and to wake up the industry to what an EV could be, modularity certainly had to take a back seat. Presumably the newer cars have much better serviceability than our beloved Roadster.

The reality is that we (both Tesla and the Roadster owners) need to deal with this. In my opinion, there are two components to the solution:

1. Tesla should create a transparent repair pipeline for each of the Tesla-unique subsystems. Battery, PEM, Motor/Inverter, under-dash and under-hood stuff (computers, air conditioning, etc). The key word here is transparent. Price based on actual component-level repair, consistent with the not-a-profit-center charter for the Service Center part of the business. I suggest a pipeline, rather than a repair-in-place strategy, in order to shorten the cycle time for a repair, and to eliminate the unrealistic need for trained technicians at each Service Center. A faulty PEM, for example, would be diagnosed, swapped out with a refurbished unit, and billed based on an estimate of the what the actual component-level repair should be. The failed unit would be shipped back to the factory for the actual repair, and within limits, the bill might be revised up or down upon the final repair. Labor, unfortunately, is not something we can do much about, so we'd be on the hook for that. But I believe that's also to be expected. The key point here is that we actually have this in place now, except for the net billing of the actual repair. Further, I have at least one real-life example of this working (my PEM repair from last year), where we (Factory, Service Center, and I) diagnosed and repaired my 1111, 1144, and 1146 errors (the burned fan motor connector contacts on the bottom of the 2.x PEM).

2. Publish an official list of the non-Tesla-sourced components, and how to swap them. These are the things that were stock on the heritage Lotus Elise, as well as stuff "borrowed" from other makes, or available from a non-Tesla catalog. We (the owners) have figured out a lot of this out on our own, but having an official cross-reference would help with locating parts, and prevent accidentally putting under-spec'd parts into service (this is both a safety and reliability requirement). This should enable a lot of the otherwise-mundane repairs to take place properly, independent of and augmenting the increasingly diffuse team of Roadster-trained service technicians. Note: much of this may already be in the restricted-availability service manuals; I haven't seen them! Obviously, they'd need to be made more broadly available.

The only area I think will be a problem would be for unique items that aren't repairable. Body panels, for example. Not sure what the right answer is... Perhaps open or license the CAD drawings?

Ok, so that's three things. Thoughts?

Greg.
 
IPresumably the newer cars have much better serviceability than our beloved Roadster.
Nope. Replacing large assemblies and entire drivelines is fairly commonplace when there's even just a weird noise or something. Seems like there are still very few serviceable parts that the service centers can work on with respect to the major systems. I'm not an expert on this subject but I've been reading the Model S forums for a few years.
 
Ideally, a more modular car would be ideal, but given Tesla's founding need to create something to prove the base technology and to wake up the industry to what an EV could be, modularity certainly had to take a back seat. Presumably the newer cars have much better serviceability than our beloved Roadster.

The reality is that we (both Tesla and the Roadster owners) need to deal with this. In my opinion, there are two components to the solution:

1. Tesla should create a transparent repair pipeline for each of the Tesla-unique subsystems. Battery, PEM, Motor/Inverter, under-dash and under-hood stuff (computers, air conditioning, etc). The key word here is transparent. Price based on actual component-level repair, consistent with the not-a-profit-center charter for the Service Center part of the business. I suggest a pipeline, rather than a repair-in-place strategy, in order to shorten the cycle time for a repair, and to eliminate the unrealistic need for trained technicians at each Service Center. A faulty PEM, for example, would be diagnosed, swapped out with a refurbished unit, and billed based on an estimate of the what the actual component-level repair should be. The failed unit would be shipped back to the factory for the actual repair, and within limits, the bill might be revised up or down upon the final repair. Labor, unfortunately, is not something we can do much about, so we'd be on the hook for that. But I believe that's also to be expected. The key point here is that we actually have this in place now, except for the net billing of the actual repair. Further, I have at least one real-life example of this working (my PEM repair from last year), where we (Factory, Service Center, and I) diagnosed and repaired my 1111, 1144, and 1146 errors (the burned fan motor connector contacts on the bottom of the 2.x PEM).

2. Publish an official list of the non-Tesla-sourced components, and how to swap them. These are the things that were stock on the heritage Lotus Elise, as well as stuff "borrowed" from other makes, or available from a non-Tesla catalog. We (the owners) have figured out a lot of this out on our own, but having an official cross-reference would help with locating parts, and prevent accidentally putting under-spec'd parts into service (this is both a safety and reliability requirement). This should enable a lot of the otherwise-mundane repairs to take place properly, independent of and augmenting the increasingly diffuse team of Roadster-trained service technicians. Note: much of this may already be in the restricted-availability service manuals; I haven't seen them! Obviously, they'd need to be made more broadly available.

The only area I think will be a problem would be for unique items that aren't repairable. Body panels, for example. Not sure what the right answer is... Perhaps open or license the CAD drawings?

Ok, so that's three things. Thoughts?

Greg.