Liste-Renn

Wow.

First off, props, kudos and a deep bow to you guys for the impressive R&D and ingenuity displayed on this and the excellent three-year-old Guide to Repairing a PDK Transmission sticky thread with almost 440K views.

You are literally writing Linux code for 997 through 992, Boxster, Cayman and Macan owners with PDK.

Most of the tech flies far over my head, despite a 38-year background operating every sort of Mil and Commercial aircraft in close earth aviation.

That said, the text and pictures you’ve posted are, to me, superior to the aircraft operating manuals and system description software of Boeing, Airbus, McDonnell Douglas, Lockheed, Northrop and Cessna.

No lie,
I’ve had to absorb from all of them over the last four decades. Awesome job.

Worth every penny…errr…oh, yeah:

I will gladly step up and purchase a sensor if a group buy becomes a reality. Unit cost is, as you stated, driven by production numbers. I am guessing you have a couple hundred thousand potential customers- who are driving a ticking time bomb PDK with an expired new, CPO or aftermarket warranty.

If any tech savvy readers out there are conversant with GoFundMe, I’ll gladly make the first donation.

In the interim, would be honored to buy you beers whenever/wherever you tank up. Deep thanks.

Your contributions to Rennlist rank amongst the most valuable of all time, and your financial interest in all this is freakin’ nil. That just doesn’t happen online, in the USA or anywhere on this planet outside of Firefox and Linus.

Well done.

Cheers and all the best in 2023 to you outstanding individuals.

PV997

Hey Sandwedge - I've been thinking about this for about a year once we had three aftermarket suppliers (one US, one Canada, one China). I held off as I was hoping to see some competition break out and a significant drop in prices. Unfortunately that didn't happen as they all continue to charge ~$2k each. I knew the basic design approach but we started digging into it in earnest and designing/fabbing demonstration parts about a month ago.

Regarding the group buy question there are really two ways we can go. If people are willing to accept a high-temperature nylon housing these can be built cheaply in small quantities without issues. A CNC machined housing is a different story though, as it would take at least 20 (preferably 50-100) to get costs low enough to really break the cost curve. Don't get me wrong, even in small quantities (e.g. 3-5) it is still way cheaper than the aftermarket parts. But in reality this is a $200 part compared to similar devices in non-Porsche vehicles and that's where the cost should be (at most). I'm sick of the Porsche Tax and want to shatter it here. We all laugh at $600 dealership oil changes and OEM parts marked up 4x because they are in a "Porsche" box. It's no different with these.

And of course everything will be open source in case people want to go for it themselves.

Hope to have more photos later today based on some test fits. Thanks for everyone's support.

PV997

Haha. I run Ubuntu Linux, use Libre Office, browse with Brave, email with Thunderbird, and designed the PWB with KiCAD (all open source). As you can imagine I'm a huge supporter of the open source movement, particularly when it takes on monopolistic enterprises like Microsoft. I don't begrudge a businesses ability to turn a profit (I have a job after all), but putting the screws to people when they are desperate (like a medical issue or a broken PDK for that matter) chaps my hide. Seriously, how is it different from price gouging after a natural disaster?

workhurts

Don't take it the wrong way, as I think what you're doing is fantastic and well executed but comparing fixing a pdk on a Porsche to water/generator sales after a natural disaster is a stretch

PV997

How is it different? Pricing something way beyond it's actual cost because people have no choice and are in a desperate situation is predatory. It may not be of the same magnitude but really, how is it different?

If you'll excuse me I have to get back to my reading of the Communist Manifesto (haha).

Edit: Just to be clear here I'm specifically talking about Porsche as they tell people the PDK can't be repaired, won't tell us what's inside the PDK, provide zero servicing info, and forbid their supplier from selling parts to the public. There's a concerted effort to prevent any alternative to full replacement. IMO that's predatory but I get @workhurts point that Porsche owners won't get a lot of sympathy particularly compared to someone who's without power during a blizzard.

workhurts

perhaps because Porsche and desperate typically aren’t perceived to go hand in hand.

I don’t disagree though that Porsche tax is extremely annoying but I can’t have too much sympathy for pdk drivers

PV997

Got some new photos here from our fit check. Note that this is a low-resolution inexpensive dummy housing for a fit check only. If we do go with a nylon/carbon fiber housing it will have much higher print quality like that seen in the ASA/ABS part in the main post.

Here's the bushings and alignment pins installed, bushings are a nice tight press fit.



Fit check within a Cayman PDK. Note that the end bolt locations are intentionally thickened from what the factory unit uses to improve strength. The center bolt location is unchanged from factory as there's no clearance available to enlarge it (though we did beef up the top side of it inside the electronics cavity).



Fit looks good with no interference problems, thanks to @jjrichar for testing this. Magnet clearance to the housing is about 1 mm which is the same as the factory ZF unit.

Next step is to install the latest PWB and pot it in the same printed housing seen in the main post. I'll then send it to @jjrichar in Australia for testing in the car. It's ABS so we'll be limited in temperature usage but it will let us know if we are in the right track. Hope to have some high temp nylon/carbon fiber housings in several weeks.

Prairiedawg

I would recommend going with the aluminum housing for the sensors. Yes, they cost more but it is tried and true. There is definitely a cost savings going with plastic but there are certain unknown risks with doing so. In my field, the hvac industry, I see first hand all of the systems and components that have switched from metal to plastics and they just aren't as resilient. Yes, they're cheaper, for the manufacturer, but not the end customer if it fails prematurely

There are costs involved with replacing the distance sensor beyond the sensor itself. Several thousand dollars to pull the transmission, dissect, reinstall, calibrate, etc. To have the new sensor(s) fail due to some unforseen issue due to the plastic would be unfortunate and double the cost to replace, again. Remember, this is a harsh environment with heat, oils, vibrations, bumps, etc. I'd hate to have it break or shed and get particles meshed in with the gears.

Bite the bullet, go with aluminum, it doesn't make it that much more expensive anyways, especially at scale. Plus, do you think ZF would make them in plastic if they thought they could get away with it and save a couple nickels per unit?

cwheeler

I'm on board with buying units for a transmission I don't even own. As I love the problem solving.

But also, let's not begrudge anyone or company that is out to make a profit. We all wish to profit. As we should. It is the moral way to be.

I sell medical equipment, and let me tell you, we get crushed on margin. I get crushed by the company and the facilities.

We should be making profit. What is fair? What ever the market will allow. Takes two parties to consent.

Cw

PV997

@cwheeler Appreciate your comments. I'm good with profit too, businesses without one don't last long. My issue is 1) with Porsche's "no repair" policy and 2) with $2k aftermarket price tags for what is reasonably a $500 part tops. Was just kidding with the Communist Manifesto crack.

Great insight, I can't recall but I'm guessing you must be an engineer too as virtually everyone of those scenarios went through my head. (In particular the piece of plastic breaking off and getting ground up in the gears.) It's our job to visualize what can go wrong and prevent it. What if this happened, what if that happened?

In all honesty though, I think those are mainly perceived risks rather than real risks. The chemical, mechanical, and thermal properties of these engineered plastics are well understood as they've been around quite a while now. I'm not a material properties expert though, and would welcome input from someone that is. Who knows why ZF didn't use it but it may just have been corporate conservativism in a part designed fifteen years ago. All that being said, perception is often reality though, and if people are afraid to use it what's the point of making it?

The problem though is that we'll need at least a $5k buy to get to the sweet spot on the aluminum housing curve. If we can get 50 people to pony up $100 each then we are golden, but I'm not going to put out $5k and potentially get stuck with a bunch of unused housings. If I was a business selling these at $2k a pop then it would be a different story. But I'm not and won't be.

So will people be prepared to try a unit made of an HTN-CF housing? Depends on their risk tolerance and cheapness. People went with an unproven $2k aftermarket distance sensor because they didn't want to pay for a remanufactured transmission so there is precedence. Will they go with a $125 HTN-CF distance sensor rather than a $2k unit make of aluminum? I'm thinking they will, particularly in older cars, but who knows. Maybe there's a demand for both versions. As I mentioned even in small quantities a machined version is still way cheaper than a $2k aftermarket unit (but at maybe $700 much more expensive than a HTN-CF version). Lots to think about but I really appreciate your feedback.

PV997

Here's a visual representation of what I was talking about in comment #25. These are the material costs for the entire distance sensor comparing printed HTN-CH to machined aluminum. This does not include labor which is probably about 2 hours per unit. Note that this also does not include the cost of a 3D printer as I can get small quantities done for material cost as we know someone who has one. In larger quantities the 3D printer + accessories cost (~$1500) would likely need to be amortized in.

Prairiedawg

I'm not an engineer, but I joke with people that I fix the things that engineers F up. I'm a Journeyman commercial HVAC service technician by trade and see the direct results of ideas that looked good on paper but in reality were destined to fail whether by equipment manufacturers, installation problems, or engineering and system design flaws. Many times there are no manuals to follow or fixes that are readily available so it takes some ingenuity and "field engineering" to make repairs and keep systems online. Now more than ever with our proverbial "supply shortages", we must get even more and more creative. Simple things like fuses are........unavailable, lead time is.......x months. interesting times we live in.

Because I live and work in the world where the consequences of going cheap or "innovative" are real and costly, I always opt for durability. I also subscribe to the "if it ain't broke, don't fix it" mentality. The distance sensor is broken, but not the aluminum part. The reality likely is that the HTN-CF part will be fine and have a long life and work as intended, but I wouldn't put it in my transmission. At least not right away and be the Guinea pig. maybe after 5-10 years when it's proven. I also don't buy the first MY car either but that's my level of comfort. I've seen too many things in my field with all our modern plastics not live up to what they were supposed to do when subjected to extreme temperatures or even swings plus UV etc.

I look at it this way. What you and others are doing is this wonderful egalitarian and altruistic thing with the best intentions. I would hate to see it go sideways over something simple like this. To me, the aluminum casing is a no brainer, an easy lay up. Take the win with the low cost boards and electronics. You're still giving an extreme value to those that would purchase the sensors and may have said they would purchase them preemptively, I don't think getting 100 would be that difficult. You have one of the most read threads on Rennlist with members with all makes and models following and pulling for you. Start a GoFundMe page, Hell, I'll chip in. We're all Porsche owners who have all paid the Porsche tax in one way or another. Why wouldn't we support the one person who almost singlehandedly broke the PDK monopoly? Everyone with a PDK around the world owes you a deep debt of gratitude. This is a way they can show it.

jbkusa

Can you reuse the aluminum cases of bad sensors? Think of it like a core fee?

Wayne Smith

Part of the idea behind a group buy was to get to the number that makes aluminum possible. I don't think the case is a good place to bring in untested material. A simple risk assessment combined with foreseeable (installer) use and misuse tells me aluminum is the right path. And you won't need to press in those spacers.

Add a couple hours labor and we are at least $200 each. If it's more so you and your group take home some financial rewards I'm all for that too. The GoFundMe option when we figure the various Forums should get 50 people at least. Especially when this goes beyond just the 997.2 cars. I'm in.

But from what I can see, things are probably about 3 months to product verification. Do you want us in before or after?

And to the moderators, can we do a general Forum announcement?

Noah Fect

I wish I understood CNC costs better. The quantity/price curve that you show is more like what I'd expect from injection molding or some other process with actual NRE. By comparison, CNC machining has no NRE other than the original CAD work. You feed the machine a G-code file and a block of aluminum, walk away, and come back when your part's done.

With CNC, the 1000th part is no cheaper than the 1st, as they both take the same amount of machine time and raw materials. I guess small-scale orders are machined on smaller mills and take (much) longer than a bulk order that can be cranked out in a hurry on machines designed for mass production. Is that the core issue behind the apparently huge economy of scale? Or are there others that I'm overlooking?

My thinking is, if $1200 will buy a 3D printer capable of making the plastic housings, would $12000 buy a CNC mill that could make similar quantities of aluminum ones?

Because like prairiedawg, I'd feel a lot more comfortable with aluminum. It's good to drive cost out of your processes, but it can be false economy to drive too much cost out.