Window regulators upgrade

[Hugabuga]

The S2K regulator does not fit in the NSX door and the NSX window does not fit in the S2K regulator!
If you make a single calculation mistake in your retrofit the most sure is to brake the window during the opening or closing movement! (HAHAHA). Then there is another problem: the height of the S2000 window is smaller than the NSX, so you would not be able to fully open the NSX window with a S2000 regulator.
Finally the S2000 regulator track is made of steel ,while the NSX is made of aluminium :wink:, this would be a DOWNGRADE, not an UPGRADE!!:tongue:
The S2000 regulator weights actually 2,2Kg and the NSX about 1,6Kg

 

[drew]

The cable ends can be removed from the transport at the window connector...why could you not use that end point instead of breaking open the motor mechanism?

Since you can't put in the entire regulator, why not just retrofit in the S2000 cable system. I don't see any good reason to put in an extra 90-deg turn. I know you stated the cable is too short, but is it still too short if the extra loop is removed?

I also suspect that there might be good pulleys from a hardware store destined for sliding doors and screens. Did you look up that angle?



This is all GREAT WORK, I'm just offering up suggestions....I have a spare regulator lying around and I'd like to really make it work a lot better too.

 

[shawn110975]

I agree with this being the best first post.

I am going to take apart my windows this weekend and see if I can fix the problem another way. but he has pointed me in the right direction, looking at Honda's design it may have been the only flaw in the NSX. great post though

 

[Hugabuga]

So today I decided to take a close look at the motor, to find out why are the windows slowing down over the years, as I do not accept a "drying grease" for an answer:wink:

So here is the motor, after a complete petrol cleaning to remove all the grease.

and here is a close look at its interior:

The AXLE is completely DAMAGED where the roller sits and there is a clear evidence of abnormal wear of the case, showing that the roller was tilted and not aligned with the axle, so it will be slow down by the case!

In the 2nd picture of the motor interior, you can also clearly see that the base of the axle is showing wear that means the axle of the roller have becoming the base!
As the diameter of the base is much higher then the axle itself (25mm instead of 10mm), this would obviously result in DRAG increasing!

So this is my answer to why the windows are actually slowing down over the years: as the axle wears the roller will touch the case, until it produces more and more drag!

 

[drew]

Dood, most excellent. Looks like some speedi-sleeves are in order!

I called bogus on the dried grease a while back. I was unable to duplicate any abnormal friction when I took out the regulator....the glass could travel up and down slick as snot.

I installed a newer regulator and the window did ok doing nothing else. The old regulator track did have some friction points in it, but not the window tracks.

Drew

 

[Hugabuga]

I have placed the roller in the case, and you can see the HUGE gape there is between the axle and the roller, it is impressive, the axle is completely ruined!

So I decided to solve this problem!

I thought of drilling a 10mm hole in the case and replace it with some shaft ...but there is a problem: the axle is connected to the rotational spring, it will be complicated to fix the shaft so it could actualy to compress the rotational spring without turning.

So I thought of using a SLEEVE!
The axle has a diameter of 10mm.
In the market I found a 12mm steel tube with a 1mm thickness that sits PERFECTLY in the axle.

So the idea now was to increase the center hole of the roller from 10 to 12mm, so the "tube" axle could fit in it!

Carefully I drilled a 12 mm center hole in the rollers

here are the 3 parts of the roller with a larger 12mm diameter hole

and the roller parts sit in their new 12mm steel axle...

 

[Hugabuga]

Then finally I just cut the tube and placed it over the damaged plastic axle.

and voila!:biggrin:

Now all I had to do was to put new grease and place the roller and the cables.

Job is completed, I have a "new" electric motor, I think it might lasts longer now as the axle is made of steel instead of plastic!:wink:

 

[Hugabuga]

Here are my final conclusions:

the regulator is the one responsible for slow windows, not the tracks or died grease

To increase speed I found 3 solutions

1- release some cable spring tension, this will reduce drag, increase speed but will add some lag in movement
2- repair the axle with a 12mm sleeve as it is most probably already damaged
3- replace the 180º plastic part with a pulley to reduce cable drag.

The combination of these 3 solutions is the ultimate UPGRADE I found for the NSX windows!

 

[greenberet]

Fantastic. Thank you for posting this!

 

[UnhuZ]

Hugo my man.... i got to meet with you so you explain me some stuff....i'm overwhelmed with all this excellent work of yours :wink:

 

[nsx_2k]

Very impressive work! One question, though. There is a post here on prime where some people have posted their window speeds (how long it takes to fully open & close the windows).

Is there any chance you could share with us the speed of your newly repaired windows so that we can quantify the difference?

 

[Hugabuga]

nsx_2K gave me the idea to make a movie comparing the speed of a modified regulator in the left (with pulley and sleeve), and a standard one in the right.
I don't know how to post a video, so here's the link from YOUTUBE:
http://www.youtube.com/watch?v=7t88j1BR19I

It takes about 4 sec to open and 3 sec to close (due to the rotational spring inside the motor to compensate glass weight), without any load!
The standard regulator is much slower in the opening movement as the cable tension over the 180º plastic is higher in this movement!

Closing times seems to be identical but installed in the car the standard one will slow down as we all know, while the modified one will be much faster.

I will post a film of the window closing in the car so with can time it

 

[ftuhy]

I also think your posting is fantastic information.

I want to look for an alternate source for the pulley. What is the diameter and thickness of the pulley you used from the S2000 regulator?

 

[Hugabuga]

Thank you all for your support!

The S2000 pulley has a diameter of 23mm and thickness of 5mm, passing cable has a diameter of 1,75mm.
The internet is full of Chinese plastic pulley suppliers.

Some of you have already PM me asking me to modify their regulators!

This post was to freely share my experience, I had no profit in mind this is why I posted it in the "Do-it-Yourself" section.

However by PM, I found that perhaps some members do not wish to follow my steps, a "Plug & Play Kit" consisting of a pulley and its support, sleeve and the drilled rollers would be a much better solution for them. I also found that 50$ including shipment for this "kit" could be a fair trading value.

If there is generic interest, I can try to source some pulleys locally but the roller will always have to be exchanged as I have no way to source them.
So every interested member will have to send me back its old roller after installing the kit, so I can drill it and send it to another interested member, and so on....

For those who do not wish to open the motor and manipulate the messy cables to apply the kit, we could also think of a complete regulator exchange with higher shipping (and customs tax perhaps) though.

[ADMIN NOTE: See for more information http://nsxprime.com/forums/showthread.php?t=136481 and please keep all business in that thread, leaving this one for DIY discussion]

 

[Hugabuga]

Regarding the pictures, I suddenly remember that not only I had slow windows but the closing movement was irregular. Now I think the reason is clear: due to a worn axle, the roller was turning in a tilted position in a way that the small top "appendix" part was touching the case between the 2 red arrow marks were there is an evidence of wear. This was acting as a BRAKING period during every rotation of the roller, making the closing movement irregular
This is in fact a good way to verify if the axle is worn or not, without any disassembling of the doors or motor!

 

[Hugabuga]

Here is the film of the window opening and closing, with the modified regulator installed in the door:

http://www.youtube.com/watch?v=JYv_LwY8BfQ

Opening takes about 2.5sec
Closing is 3.5sec

Windows are now faster then going from 0 to 100km/h (or 0 to 60mph) with the NSX :biggrin:

 

[710]

What wonderful information!!!!
Wow, this has to be the best first post in history, congratulations on your investigations and welcome to the site.

I have also been working on this problem (in between building a house, so very slow progress on the NSX window project).

I scanned the plastic part into 3D and designed an adapter plate to allow mounting a more powerful and faster motor. I was at the point of trying to source a cheaper electric motor. I also need to determine the design characteristics of the worm gear so it can be machined. I have a potentiometer to adjust the window speed.
I have a Hal sensor to allow auto-up on both right and left side windows. That was as far as I got up to now.

But my solution would not solve the friction and wear problems you have found to be the actual cause of slow windows! Everything can be solved using HugaBuga’s findings. He should get a medal.

My more powerful and slightly faster motor might even break the plastic guides, I don’t know yet. The existing wire might not be thick enough to carry the increased amperes.
We already know the plastic cable locking block must be replaced with an aluminium one ( the well known ”window thingy”) because it breaks. Something else might break if I ever got my more powerful motor in there.

So what is shown here is that the whole assembly is designed right down to the limit, and is full of faults. It is inadequate in almost every way.

The solution to our problems is one of the following:
1) We could use another, newer, better engineered mechanism, comprising parts from another car, either Honda or other. This would mean someone producing the kit from scratch. Would probably cost us about $500 per door minimum, but I bet a lot would be sold/bought. Like several thousand.

2) We could do the modifications ourselves, following HugaBuga’s very good advice. I believe few people would dare do this.

3) We could use a repair-kit to repair the existing mechanism, as in HugaBuga’s photos, cost would be lower, but, again, many (most) people wouldn’t dare do the job themselves, especially considering the risk of messing up the thing, thus needing a new regulator assembly costing new $800 each or second hand about $400.

4) We could use a complete regulator exchange kit built by HugaBuga or another person and marketed as a replaceable unit. I can imagine HugBuga or others having trouble with getting some of the parts for cores, unless they buy a brand new one and cannibalise it. HugaBuga may need to resort to asking a very high core cost, so if someone doesn’t send their old parts in to him, he won’t run out of parts.

Anyway, it doesn’t look cheap or easy (even though HugaBuha’s very excellent write up of the process looks do-able).

Just though I would summarise…because my windows suck.

 

[Juice]

I'm still in shock from finding out they used a plastic axle!

 

[tbromley]

Amazing post - I may not ever want to take this on but it is really explantory. I've been on several DIY sites for the 964 and the 993 Porsche and I don't believe I've ever seen a better job done with any project. The pictures were excellent - this is truly an a wonderful job - congratulations to you Hugabuga. If you ever get to a point where you are selling these upgrades I would like to know. You really are ingenious with the way you felt so easy with your mods to the motor and dilegence to understand the motor itself.

What do you do for a living? Are you a mechanic? Or a mechanical engineer?

Thanks for this. I really enjoyed the whole post and the videos - althogh the lighting wasn't too good for the video. I got the idea though. Much better and faster than mine and mine has had the upgrades.

 

[Hugabuga]

Thank you Sir, and thank you for the medal! :biggrin:

I must say it gave me a lot of fun posting all my work, and a real-time feedback from NSX owners all over the world turns everything even more interesting!

This is really a great Forum of an amazing car, thank you NSXPrime and its "patron"!

My only advice to all the members is: get your hands in grease and speed up your windows!

Yes, I am a robotics mechanical engineer

 

[Zishan]

Yes, I am a robotics mechanical engineer

That explains it! :smile:


Seriously, I don't think this is something I could do.....but if you have the time and patience, you should show us how to remove the regulators, and start charging to have these rebuilt by you.

Excellent post.

Thanks,

 

[Hugabuga]

Don't fool yourselves, anyone can cut a tube and fix a pulley!
Don't you all remember how to replace a broken bicycle brake cable anymore? it is as simple as that!:wink:

I will try to make a film showing all the process of how to remove the regulators from the door and how to modify them, in a 5min film :biggrin:
For now, just follow the manual to remove the regulators but remember to tape the glass in its fully closed position to maintain the original window adjustment.

I have no problem rebuilding the regulators, in fact some of you have already post them to me!

 

[OLDMNSX]

Hugabuga,
WAY TO GO MAN!!!!!! I am not able to open your pictures over here, but will when I get home. If I wasn't in Afghanistan I would be working on my regulators RIGHT NOW......anyway December isn't that far away......is it?
Brad

 

[nh_nsx]

Just wanted to add another "Thank you for the great work"!!
I guess plastic must be lighter than metal or Honda NEVER would have done this to the NSX!! Less weight=more speed!

 

[trainingrevolution]

OUTSTANDING WORK!

I did notice a difference on my car after removing the window tracks, cleaning them, and regreasing them with the Honda grease but not nearly as much as I would have expected. Can't wait to tackle this project! Thanks for the excellent write-up!