When I was reassembling my bare metal restoration, I probably spent close to 80 hours installing and adjusting body seals and gaskets. I wanted to get the car as air tight as possible to control cabin noise and leaks. I'm happy to say that my car seals tight enough to feel that pressure bubble when you close a door with top on and windows up like you get with a new modern car.

However, my one trouble spot was the top of the side triangle window. The 914Rubber main front targa seal is nice, but the rubber on mine was just a little too stiff to let the triangle frame to bed in for a proper seal. No amount of adjustment to move the frame forward helped. I even tried wrapping a ratchet strap around the top of the triangle window and windshield frame to pull the frame into the rubber and left it for two weeks to see if it would take a set. It didn't. There was this gap that let water trip in and created significant wind noise in an otherwise suprisingly quiet top on, windows up, cabin.

Click to view attachment

Here is how I fixed it.

1. Clean the main targa seal, triangle window end cap, and top edge of the triangle window thoroughly with isopropyl alcohol.

2. Smear the triangle window end cap and a few inches of the top edge of the window with a generous, but smooth coat of petroleum jelly. This will act as a release agent for the silicone caulk you are about to apply.

Click to view attachment

3. Apply a thick bead of black silicone on the targa seal in the area where the gap needs to be filled. Don't worry about getting it smooth at this point. Don't overdo it or you will have more mess to clean up in the next step.

Click to view attachment

4. Close the door tight and use a lint free cloth soaked with isopropyl alcohol to clean up any silicone that oozes out on the inside. Be thorough with the clean up on this side.

5. Use your finger on the outside to force the silicone into the gap and smooth it out. I had to squirt in a bit more silicone to make sure it was completely filled.

Click to view attachment

6. When you are satisfied the gap is completely filled with silicone and it looks nice and smooth, carefully go over inside and out with a clean alcohol soaked cloth to tidy up any stray silicone. Much easier at this stage than after silicone has cured. Careful not wipe the petroleum jelly off the end cap though.

7. Leave the car alone for at least 12 hours to let the silicone cure. After the silicone has cured, open the door and clean off the petroleum jelly with alcohol.

Click to view attachment

You should have a perfect seal and nearly invisible repair. This picture was taken with a flash to better show the cured silicone.

Click to view attachment

And here is without flash which is more representative of how it looks in natural light.

Click to view attachment

The small amount of jagged slag is only noticeable if you are looking for it, and disappears when the door is closed, but I'll let the silicone cure a few more days and try to smooth it with some fine sandpaper.

Click to view attachment

Happy drip and whistle free motoring!

Your ingenuity never fails to impress. Practical solutions to real-world issues. Really nice work. Again.

Nice one Brent! I may need to do this. Evil genius you.

Good job on a fix.

Would be far nicer if the targa seal fit properly in the 1st place. The seal material is too hard (not pliable enough) and doesn't have quite the right profile to make proper contact. Like you, I spent many hours adjusting the seal, the window triangle, and even nudging the aluminum track outboard as far as I could without resorting to re-slotting the aluminum tracks.

I hate to bang on 914Rubber since they do so much to support the community. I'm just thinking I need seals that fit more than I need plaid blankets from Scotland. Don't get me wrong I sincerely appreciate that there is a vendor supporting us with the rubber products that we all need. I just think that the 914 has appreciated in value to the point were I'd be willing to pay double if the seal actually replicated OEM form and function.

I actually am saving my OEM seal since it fit noticeably better sealing in that critical area both for water leaks and wind noise. Currently contemplating a hybrid. My OEM seal was perfect on the top but had torn on the driver side bottom where it interfaces with the door seal which is why I replaced it. I just traded one leak path for a worse leak path.

My thinking is that it may be better to use the OEM on the top and then splice in the lower straight sections where there is less going on with respect to all the interfacing surfaces. It will be a while before I do it but seems like an experiment worth doing.

Your fix looks like the much easier solution. Let us know how it holds up over time.

Mine fits the same way and just not happy with it. If I could find an oem one I would buy it. Perfect example of oem vs aftermarket I experienced this past week. Ordered some parts from. 928srus.com and In that order was a Porsche badge seal. It was 12 dollars. I have owned a dozen different porsches over the years including 944s, 911s, 964s etc. Well pull the seal out of the package and it fit the badge like none I have ever purchased before. It fit exactly Perfect. I mean Perfect Perfect.. in all my years I have never had one fit that good. I emailed Roger telling him how nice the seal fit. He responded thank you and said that is the reason why he only sells Porsche badge seals and not aftermarket. My little story just was to express how surprised I was in the difference in fitment.

I have used a product called Shoe Goo. On things other than shoes.

I comes in black or clear. I think it would work in this application, maybe better than silicone.

In full disclosure I'm biased since I used to work for an OEM.

I can't tell you the number of hours of design, development and validation that go into something as seemingly simple as seals.

Water test booths that simulates monsoon rain as well as high pressure wash testing. Dust intrusion testing. Door slam testing. Windnoise / Windtunnel testing. Testing in extreme hot and extreme cold. The list goes on and on.

There are most certainly significant differences between OEM and the aftermarket for almost every type of part (sheetmetal, brake pads, tires, seals, etc.) There is a reason OEM service parts cost more and it's not just mark up (though there is a good bit of that too!).

914Rubber has a very hard job and does a pretty good job trying to balance that tight rope between trying to balance tooling costs vs. making a fair profit. Don't want to bash them unnecessarily. I don't envy what they do. But yet, I would glady give an arm and a leg for OEM quality seals that fit.

I'm pretty sure you can adjust the angle of the chrome channel forward to get a better initial seal. I remember doing this on mine many years ago.

You SHOULD be able to adjust the channel forward and, in fact, you can. The problem is that the rubber in the seal is stiff enough that it just pushes the channel back instead of letting it bed in for a tight seal. There's just a lot of leverage at the top of the channel for the stiff rubber seal to push against.

Yup, triangle adjusted all the way forward, and inboard as far as the bolt on the bottom down at the door will allow. Seal just pushes it all around without deforming the rubber bulb of the seal.

I even considered re-slotting the door and/or extending the bolts -- but that is not acceptable. Should not be necessary and will only drive up door closing effort.

Interesting. Used clear but never seen black before.

When I was unsuccessfully trying to get mine to fit, I removed the plastic end cap on the channel and moved it around through the window frame seal to see if there was a spot where it fit the contour of the seal. There was, but a little lower. I've wondered if a little deeper channel in the bottom of the triangle seal would allow the whole triangle window assembly to sit low enough for the corner to nestle tightly into the targa seal and still have everything else line up. I wasn't about to start carving on the seal to find out though.

I looked at the Shoe Goo in my adhesives draw when I was deciding what to try, but I only have clear. My recollection is that Shoe Goo sets up more rigid than silicone but can't remember for sure. I knew silicone would stay soft and flexible which I figure would be good, but I think Shoe Goo might be less likely to peel off the main seal over time. Will be interesting to see how long term this work around is, but it looks to be adhered tightly at this point.

Man! This is one issue I have on both doors. I spent countless hours trying to get the top of the triangle glass to set properly in the Targa seal. I thought this was on me. I felt like I had screwed it up somehow. I personally think the bottom part of the triangle window seal is too thick and prevents the full adjustment but I'm not sure.

If you bought the 914Rubber window triangle seals, they are indeed too thick in several places.

Again, I'm not trying to beat down 914Rubber -- I'm happy to have access to the triangle seals that would otherwise be NLA. I was able to rework the part which is better than having no part at all. But like the targa seal, would pay more for a part that simply fit exactly like OEM and didn't need rework.

I ended up only using the driver side. The passenger side OEM part was still in decent shape. After the amount of rework I had to do with the driver side triangle seal, I decided to pass on doing the passenger side for now.

I ended up having to carve down the bottom V-ribs on the bottom to allow the seal to sit deep enough in the door sheet metal pocket.

Click to view attachment

OEM (top) vs. 914Rubber (bottom) - picture doesn't quite do it justice. V ribs were a little taller, and seemed to be less compressible. Some of that could simply be that OEM seal was installed for so long that it had taken a set. However, that "set" is where the triangle needed to be to have a chance of mating with the targa seal.

Click to view attachment


I used carbide in a Dremmel to deepen the pocket that the triangle glass sits in to get the triangle back down to the same position as the OEM seal allowed.

My part is off a 73' and has a pocket to fit the metal blade that bolts to the window channel that helps hold the window triangle glass. In PM's with 914Rubber, it sounds like that pocket was eliminated in later 75' & 76'. So to use with 73' window channel, I also had to carve out the square pocket. I think the alternative was to remove the metal blade from my window channel. I opted not to do that since I think that blades helps stabilize and retain the triangle glass. Personally, I've only owned 73's so I don't know how well later cars fit without the pocket and/or the channel bracket my car has.

I didn't know later cars omitted that metal tab.

I forgot about having to carve out for the blades, but I did the same. I'm not a fan of trying to outsmart the engineers by cutting off parts that clearly had a purpose.

Same goes for the tabs in the aluminum channels at the top of the windshield frames. I believe 914R wants us to lop those off to install their seals. I refused as they look important for holding the seal in place and probably prevent creep from the umpteen heat cycles the seal will experience over its life. I did (reluctantly) trim mine a bit shorter to make the job of slotting out the new seal a bit easier, but those tabs are now in my seals doing their job.

SO, to comment on the production of these seals. 1st rubber will shrink over time and we are always playing guess the coefficient. It just so happens that this particular part didn't shrink as much as others did. We replicated the sample that I was able to obtain off my 75, unaware there were variations.
As to the pocket in the bottom of the channel we did mold it most of the way and it can be cut open, but we made it this way for 2 reasons.

1- the limitation of the tool we made.

2- no part fits into the "drain" created. It would allow any water to drain into the door.

Giver #1 we saw no reason to replicate that feature exactly when the end user can choose this option.

These are a difficult piece to install to begin with, this makes them a little more so.

The net result was yes they can be fitted to work properly, but will require additional pressure to do so. It makes the job easier if they are trimmed slightly.

These are the 2nd generation of these parts we have made.

We will be retooling on these to make them easier to install. We must complete our interior series 1st. In order, the pillars, the new dash top, the lower dash.

I will put these next on the list. As you know we typically don't raise prices on the parts as we gain efficiencies, maybe a bad business decision, but I think we owe it to our customers so pay back their support to allow us to make the things we do.

@Mikey914

Thanks for this Mark. As others have said, we couldn't restore these cars to the level we do without you. None of this is intended as a dig, just being honest about the realities of installing these aftermarket parts. It is understood that there is a lot to juggle to hit the sweet spot with these.

I'm good on seals. Next on your list I'd like to see that subwoofer box. I need that fat bottom.

Thanks,
Yes this winds up being about more than making parts. We are currently in the final phase of implementing and a Enterprise Resource Platform, that will really help us to juggle a lot more balls, more efficiently. We've been at this almost 18 months now. One of the reasons we had slowed parts development, it has been a major undertaking for us and I will be happy when we can start using many of the features it has.

I'll take a look and see what I can do on the sub speaker pod. I may be able to figure something out as this is partially started.

I think I've been pretty clear that I appreciate what 914Rubber is doing to support the community.

In my opinion, I'd rather have a part to rework than nothing at all. But having said that, personally, I'm willing to pay more for parts that fit and function exactly like OEM than having to spend time reworking parts which then is time I could have better spent elsewhere.

With respect to the business model, If you gain efficiencies that rightfully should go right to your bottom line. You should adjust your pricing as material prices, tooling, labor costs, shipping, etc., all all going to be going up with inflation. I'd much rather that you make quality parts, cover your costs, make a profit, and stay in business rather than seeing you try to win the race to the bottom that cheap pricing forces.

The flip side of it is the customer experience. I'm not shy about reworking a piece to make it suit my needs. However, there are many others that are feeling frustrated that parts don't fit properly and they either assume they have done something wrong or don't have the tools and skills to rework them. Similarly, as 914's get more appreciation and attention in the restoration trade, shops don't want to be spending $50-$100/hr shop rate to rework parts to fit and then potentially disappointing a high dollar customer when fit & finish isn't 100% correct.

I know intimately what it takes to make these parts, and I don't envy what you do and the tightrope you have to walk trying to make it a viable business model. Just in case it wasn't clear, I appreciate and applaud you for making NLA parts available.

@Mikey914

If you try this method for closing up the gap, I would recommend staying with the silicone.

Shoe Goo or Amazing Automotive Goop both have the same basic plastic, Styrene Butadiene polymer. Shoe goo uses toluene as solvent that has to evaporate. Amazing Goop uses perchloroethylene (chlorinated brake cleaner) for the same task.

Both these solvents will dissolve petroleum jelly so it may end up sticking to everything, and the release agent won't work and the sealant may not cure with petroleum jelly dissolved in the goo(p). The solvents can also cause the rubber to swell and once evaporated the rubber may shrink or be permanently deformed.

Interesting side note... A factory trained Ducati master mechanic showed me a neat use for black automotive goop. He uses it as a removable thread locker. Cleans off with brake cleaner when you need it completely gone. It does work for that purpose... I used it to hold down the nuts on the air cleaner tins for the Weber carbs. Tried to undo the nut and the rubbery stuff prevented the nut from unscrewing. It actually tried to tighten the nut until the rubber gave. Kind of like using nail polish to set a trim potentiometer from moving once positioned. Better than relying on that tiny lock washer. I shudder at the thought of that nut going down the carb's throat during a run...

Ouch. Went through TWO "Enterprise" platform implementations over the years. THe last company I worked for thought about it, but realized (a bit due to my input...) the task, complexity, time, and cost for a $4-6B/year company. Don't think the company would have been up to the task anyway. Glad I'm retired now.

We were "lucky" that there was a grant, and an Oregon company that wanted to do a build out for us. We started allocating time to it immediately building BOM's and build orders, as well as cleaning up our vendors lists and putting many other things in place. I have 2 consultants that are working to bring this online hopefully very soon. We have been very lucky that they wanted to build out a manufacturing and sales company as they know there would be many challenges, and in doing so they would be able to more easily handle both. We did find that the accounting platform was a bit weak, but they are rebuilding it and hopefully it will be everything we need and more.

I used Porsche factory (recent production) triangle window seals, and a 914 Rubber main targa seal. And I have the same gap problem at the top, with the channel adjusted all the way forward and inward, and I tried for hours to optimize the fit.

It might be the rubber block that caps the channel. I tried both my new 914 Rubber caps, and the original caps.

Me too. Mark is justly revered here. Still, i think the part is just too hard and perhaps not quite right. I can see that is not an easy problem for the rubber man to solve, but suggest that the old school approach would be best: new and improved. The good news is that we have a company that listens and a community willing to be crabby. If you make a better part I will buy it.

I drove 100 miles from Butte to home @ 80mph with top on and windows up last night. I'm happy to report that the wind noise inside the cabin is signifiantly reduced with those gaps sealed. Kicking myself for not taking before/after sound measurements.

I keep looking at this and something doesn't look right. Given that the blocks are identical to the factory ones in shape and durometer, what would the solution be here?

There appears to be a 1/2" section of rubber not fully seated, Is it holding it out?

Could the amount of butyl behind the "chrome" be holding it down slightly? I don't think that would be the whole story.

The triangle window looks like the top is seated deeper than the cap. Almost like if the top of the triangle window was shimmed out more it would push the bottom of the cap out more to "close the gap". This appears to be where the failure is. I've seen many of the caps worn and look slightly deformed from the top of the glass pushing it out, but not here. It looks like it's barely engaged.

Why it would sit lower is beyond me. The initial complaint was it was too tight, which would hold it up.

Thoughts??

I think going a little softer on the vertical seal on the windshield frame would accomplish two things. It would allow the seal to better follow the bend of the channel. As you note, the block of the seal is pretty stiff trying to make that bend so straightens out around the tightest part of the bend and pulls as far as it can away from the channel. It is in the channel, but not tightly following the contour. I actually think that is the bulk of the issue. Second, with a softer bulb, the triangle channel wouldn't be pushed back from the forward adjustment.

I always think about the American muscle car seal market. All of those seals are softer and the reason is the cars fit and finish was a bit sloppy by nature. Well hate to say it but the 914s were as well.

@Mikey914

Of the various replacement seals I've installed, I've found the new seals attachment block (base of the seal) to be significantly larger than what came off the car. As a result, they fit much tighter into the aluminum channels and have less flexibility than what came off the car. I apologize I don't have a picture handy of my car and/or my main targa seal.

Here is a comparison of the base of the seal for the upper window seal that goes into the targa.
Click to view attachment

In several cases, the OEM seal can be pulled though the aluminum channel with only moderate force. I believe the fact that seals could be moved with only moderate force is the reason the factory "staked" the aluminum channels once they had the seal in place. The OEM seals can sort of the "inch wormed" into and out of the aluminum channel. I found this to be true on my 1st car way back when I painted that one in about 1987. In the case of all the replacement seals on my 2nd car, I've had to install the seal by rolling it into the aluminum channel with a trim stick. Once the seal is in the aluminum channel, it becomes very difficult to move it and it really can't be "inch wormed" very much at all. Likewise, I haven't seen the need to stake the replacement seals to keep them from moving since they are so tight in the channel.

Having a bigger base stuffed into the aluminum channel will affect the overall stiffness of the bulb seal that that window interacts with.

As you stated previously, rubber shrinks and hardens as it ages and leads to a fair degree of guesswork.

I fully understand that engineering rubber parts is 1/2 engineering and 1/2 black art. I cant understate for the community how difficult a job it is for 914Rubber to reverse engineer 50 year old parts. Even more so when you don't have a budget for mass production tooling and a consistent production process that is running day in and day out. Worse yet when you potentially have to make a batch of seals for a minimum buy and then they sit an a self (months, years, multiple years?) and harden in the warehouse. I honestly have the utmost appreciation for what 914Rubber is doing for the community by making these seals -- It isn't an easy job and most don't appreciate just how difficult it is. .

Regardless of that fact. I've done a lot of development in automotive around bushings and seals. Each time a supplier changes, parts usually have to be changed or tweaked to get equivalent performance. With rubber parts, It's not as simple as handing a drawing to two suppliers and getting EXACTLY the same part back from each supplier.

I can't tell you the number of times I've seen a major OEM purchasing department burned by attempting this. They will take a part that was engineered, developed, and produced by Supplier A - send that print to supplier B who promises a cheaper piece price. Parts come back from supplier B but yet, they don't perform as they should. Now purchasing is mad a engineering because they say the drawings didn't contain all the critical characteristics. Engineering is mad at purchasing for resourcing a part that was working fine and manpower is needed to fix it with Supplier B. Supplier A is mad since they are having business taken away to a cheaper competitor. Supplier B sits there looking stupid saying I produced what the print called for - not my problem. Supplier B often knows that they will recoup on the "cheaper" part by the time they get paid for "engineering changes" to the print to now make the "cheaper" part work within their production process. Now I remember why I'm retired!

Simply copying the OEM design and durometer doesn't guarantee an equivalent result when rubber parts are involved

Rubber formulations are not the same from supplier to supplier. One may use a bit more sulfur and a little bit different vulcanization time than the other. Drifts in durometer that are seeming irrelevant sometimes need to the offset by changes to the section of the part. Sometimes the part is the same durometer but ends up with different elasticity characteristics and that affects function. A seemingly small section thickness difference (within print tolerance) between the parts can leads to the need for the some tweaking and/or re-engineering to get the part to perform the desired function when changing from supplier A to supplier B.

This is a long way to say, that simply copying the OEM durometer and dimensions won't necessarily yield the same part, and the same functionality as the OEM part. As the others have noted, I've noticed that what came off the car is softer and more deformable than the replacement part.

@Mikey914

I had the same thing happen on my car in the same location. They seal was seated.

What was happening is that the larger base block was so tight in the aluminum channel it was forcing the seal to buckle as it bent around that radius.

In my case, I was able to minimize it but aggressively "stretching" the seal though the channel so as to thin it ever so slightly helping it better conform in that tight radius the same way the OEM seal did. Even when I did get that area of the seal to better conform to the radius, there is still an inability of that window triangle cap and the triangle glass to force the seal to conform to it because the seal is too stiff in that critical area.

Again, may be a combination of part dimensions, durometer, or elasticity differences that lead to that inability of the seal to properly follow the radius of that windshield to header transition.

No surprise when this came up for sale, i jumped on it.
http://www.914world.com/bbs2/index.php?sho...p;hl=targa+seal

Wow! I'm surpised they were still making that seal up though 2007? You gotta wonder if that was just the pick-n-pull date or if they really were running batches of seals in 2007

Ok,
Lets have a conversation on the process.

#1- we want to be as close to OEM as possible unless there is a specific reason. An example was the use of stainless or aluminum for internal clips. We were one of the 1st to take on this added expense that has been adopted by others too.

As far as hardness and size there is a bit more to consider. Rubber will shrink over time, you can clearly see that in the new / old comparison. We look to ward the fits to parts and try to extrapolate what the original size was using a fomula that estimates it pretty closely. This takes multiple samples, and we can actually see when we have an original vs an OEM that was replaced. This is ideal as it helps confirm the side is correct.

As for hardness, there are some standard harnesses that were used, and we can confirm the original while we are working on size. It's easy to say make it softer, but it's actually much more complicated and can actually create more problems. The hardness of the corner blocks are greater than that of the extrusions. The greater the difference, the more likely of a bonding issue. If the material is softer it may not retain as well in the tracks. These are just a few things that must be addressed. When the factory did something there is usually a reason, and small changes can have unintended consequences. I have been educated on this a few times, it's one of the reasons it has taken us more time than expected in bringing some products to market.

There is always the installation technique that can be a variable. I prefer to seat the corner blocks (attach and screw into the frame), then engage the inner edge and push the seal up while seating the outer. It requires working your way down the edge on the outside while keeping the inner edge seated.

I do think it's time for us to put together a formal install video so we can see a technique that works well for us.

I did see that the top of the triangle window glass seems to sit a little lower. It almost looks like if it was 1/8" higher in the retainer the edge would come in and fill the gap.

I did start a thread help us help you to try to deconstruct this specifically, as I would like to address specifically this issue.

You can look for more from us in the future on this.

Thanks,
mark