Speak EV - Electric Car Forums banner

Ampera Door Pushbutton Failure Analysis

8.4K views 35 replies 13 participants last post by  GJO  
#1 ·
@GJO had kindly donated me a failed door module, so I've now taken it apart & photographed the inside. Despite clicking the switch, it remains open circuit. Basically the pushbutton mechanism is a very cheap maybe-nickel-plated buckling-plate thing, and the central raised point makes contact with a faintly gold-plated bit of etched circuit board. A combination of the impact from the spring-steel plate and whatever current is passed has worn away this thin gold plated section until there's nothing left to make contact. I've met these cheapo buckling plate switches before, typically found on very cheap calculators years ago. And I had to strip & clean those to keep them working.

Top view of failed switch.
121071


121072


121073

Underside with potting rubbery compound removed. Damage to blue wire is my scalpel attack.
121074

Self-adhesive white layer shows no sign of water ingress to me.
121075


121077


Minor damage visible where the 4 outer prongs of the steel plate rest, but the central part is the problem.
121078
 
#2 ·
Stripping the plunger-side of the switch reveals a one-piece black rubbery button + membrane item which appears intact with no splits in. There's a thin white layer with a clear window attached, I suspect this is a moisture barrier, as well as the clear piece being sticky to retain the steel click-plate in place when it's all put together.

121079
 
#4 ·
Way ahead of you, sunshine ! :)
Just been looking into this! RS & CPC are useless; Farnell have a decent low-profile, gold contact pushbutton switch that's rated >3 million cycles at 50 mA 24V dc. ÂŁ1.40 each. But it's not going to be a direct swap, as you're looking at a surface-mount item, not a PCB assembly like this is. So I plan to order a couple of these to play with, and I'll 3D print a housing to hold it, reusing the black rubbery plunger item above. Then I hope to give the thing back to Graham to test - means he'll have to yank the door panels off again!

This button will have a slightly harder press to operate, and a slightly shorter travel, but I doubt the difference will matter tbh.

Assuming this works, and I see no reason why not, I'm prepared to offer a repair service. Drive down to Winchester, and I'll rebuild the switch in-situ. By the sound of it, it's easier to snip the wires off to remove just the switch, and then replace it with ?longer? wires soldered on. Needs a looking at - I'm not familiar with the insides of my door!

I see no reason why I can't offer a diy repair kit for anyone to do this themselves, providing they're happy soldering the replacement wires in & heat-shrinking them to seal against moisture. Not everyone's within easy driving distance!
 
#6 ·
+1 for a DIY/exchange based solution; Winchester is a 145 mile round trip for me.

The one working button (driver-side rear) on my Ampera died about 2 weeks after I got it. I've managed to get a handle off (albeit still connected by wire) but I couldn't fathom how to get the handle itself apart to get at the switch inside.
 
#7 ·
The outer cover springs off from the small end. To do this effectively, it is necessary to remove the complete handle and mount it in a vise - holding it by the internal lugs so that it is just possible to get a robust flat bladed screwdriver under the edge of the narrow end of the cover. Using this leverage, the cover springs off to to reveal the inner handle. Note the wider end of the cover is secured by two robust lugs - this is why it is necessary to free the narrow end first and then lift/slide the cover off.
 
#9 ·
The problem looks like chemical corrosion between the two surfaces somehow...

It's also poorly designed as the via (the "pipe" between the top and bottom surfaces of the PCB to take the signal away) hasn't been filled and capped, which means when it's manufactured, some of the gold plating would be drawn down into it, making the plating nearby thinner.

If the problem is that there is a spot on the pad with no contact on it any more (the black area in the gold disc above) then could you simply put a blob of solder across the whole pad? Alternatively you can get conductive paint and just fill it in, like you would a paint chip?
 
  • Like
Reactions: GJO
#12 ·
Well done @HandyAndy. I keep meaning to sort mine out although both rears work fine. Just waded through some really stupid adverts from the powerflush company to get here and reminded myself to install an ad blocker on my moby.
 
#14 ·
Unfortunately at least the Omron IP67 protected surface mount switches I used to replace the failed ones also fail before long. Those were the best I could find, but if you manage to find better ones, I'm all ears. While the switches should have been resistant to water etc. we also have the cold weather that takes it's toll. But the good thing is that at least the Omron B3SL and similar ones have long enough button that it's possible to just drill a hole for it to the center of the round gold plate and glue the switch to the underside. Makes it more DIY for us without 3D printers.. :)
 
#26 ·
Unfortunately at least the Omron IP67 protected surface mount switches I used to replace the failed ones also fail before long. Those were the best I could find, but if you manage to find better ones, I'm all ears... :)
Yes, the Omron B3SL is yet another of the "cheapo" designs out there, 50p/switch appx. It's:
silver-plated contacts, 100,000 operations, 50 mA 12V.
I'm eying up the MEC 6CJ1NOPR instead, ÂŁ1.40. It's:
gold-plated contacts, 3,000,000 operations, 50mA 24V. Should be a d*mn sight more reliable.

... While the switches should have been resistant to water etc. we also have the cold weather that takes it's toll...
Can't do anything about that!

... the Omron B3SL and similar ones have long enough button that it's possible to just drill a hole for it to the center of the round gold plate and glue the switch to the underside. Makes it more DIY for us without 3D printers.. :)
I can't make your sizes add up. The OEM switch protrudes 0.6mm above the 2.0 mm thick circuit board, and the rubber plunger is 7.4mm on top of the switch. So if you mount a new switch+plunger on top of the circuit board, it needs to be 8.0mm tall in total. Mount it below the board, and it needs to have 8mm plunger sticking up from the top face. B3SL-1022P has 2.6mm of plunger sticking up from a 2.5mm thick body. Or the low-profile version BSL-1002P has shorter plunger sticking up 0.9mm.
 
#20 ·
Three of my four door buttons were replaced under warranty not long after I bought the car.

Did anyone read my suggestion about simply using some conductive paint to cover the worn-off part? Seems far simpler than bodging in a different type of switch...
 
#22 ·
I suggest checking out the other thread too: https://www.speakev.com/threads/has-anyone-changed-a-door-handle.139352/

As I mentioned there, some Finnish people have reported that the handles from gen 2 Volt and others using the same one have a better switch that can be implanted to gen 1 Volt/Ampera handle:

I haven't done it because I don't like the idea of buying the whole handle for just the switch, but at least that's one option for a robust switch.
 
#23 ·
My experience of conductive (silver loaded) paint like the Electrocure stuff is its fine to repair a crack in a circuit board or a rear screen heater but it has very little resistance to rub so I doubt it would work long term. Tin and tin/lead solder both slowly corrode, tin may well be longer lasting than leaded solder as its less likely to flow.

These miserable dome switches are in everything now, When I used to service machinery (copiers and office equipment in the 80's) these were trouble then. No real repair, just replace the part.

I still use a 1990 IBM model M AT keyboard which has never missed a beat in 30 years and left a pile of other keyboards used by the rest of the family in its wake. It proves its possible to make good switches, however they were and are expensive (see Ebay).
 
  • Like
Reactions: Richard Gledhill
#25 ·
Mine had the drivers door one changed in warranty at just under 3 years old along with the alloys and the aquariums at the back. Just in time,any more fail and its pay time :( as I'm the 2nd owner.
 
  • Like
Reactions: Richard Gledhill
#28 ·
@HandyAndy If you haven't come across it, though I imagine you have , "Sugru" may work. 3D a mould and press the stuff into it.

Sort of silicone rubber I think, feels like it anyway.
 
#30 ·
... "Sugru" may work. 3D a mould and press the stuff into it. Sort of silicone rubber ...
Yup, seems to be a silicone rubber. But it's not cheap!!! ÂŁ7.50 for a 3-pack, with only 5 grams in a pack!
So instead I've just ordered a small qty appx 100g of 3D-printable flexy polyurethane filament to have a play with. Cost ÂŁ8.94. Should be good for at least a 100 switches...
 
#31 ·
There defo seems to be a lot of interest in a DIY repair kit! So I'll go ahead with getting decent switches in, and a replacement module made available. It will come as a switch looking v like the OEM one, same size & shape, with 2 flying leads & a couple of splicing connectors. I might change my own driver's door switch, working now but for how long, and take a series of photos as a guide. I think ÂŁ5 + postage would be a reasonable price to cover the cost of one switch. Replacing the switch will be the user's responsibility; if it doesn't work, or you shatter a door panel etc, tough luck, maybe the radio bit died.
 
#32 ·
re my proposed "Door Switch Repair Kit" :

Anyone interested, plz PM me a message to say how many door switches you want. This will give me an idea of how big a batch I need to start with, thx.

I have no idea if the same one is used in the tailgate; I expect it is, but anyone had one of these apart?
 
#33 ·
I suspect the tailgate is different , Never heard of a switch failure and unlike the door switches I use the tailgate daily often several times and I suspect others do as well.
It is protected from the elements though.
 
#34 ·
I have just replaced the third door handle - I have used new handle kits in each case. The downside of this approach is that in order to undo the electrical connection, it is necessary to remove the door trim - even then the connection is an absolute pig to undo and each time I have managed to break something. Most recently I snapped part of a plastic lug that secures the cable that connects the inside door handle to the door lock. I managed to effect a bit of a repair with a couple of cable ties but to do the job properly, I would have to replace the whole door trim because the damaged lug is an integral part of the plastic moulding. The point of this post is that, if it were possible to just remove the outside handle (which is easily released) and fit a new button switch with an new electrical connection made inside the handle, it would not be necessary to remove the door trim. This would save a significant amount of time and it may also protect the door trim and its components from the otherwise unavoidable damage.
 
#35 · (Edited)
Ok, things are progressing, if a bit slower than I'd like! I've made up my first-ever replacement door switch that I'm going to test on my passenger door. I plan to take pics of step-by-step attack & see what happens. Here's a couple of pics of the switch to compare with the OEM original.

I think the pressure required to operate will be slightly higher than the OEM ones, as the switch itself has a few grams weight greated actuating force, and my switch has a 0.2mm thick ABS membrane internally compared to the original rubbery item, so that's a bit stiffer. But it feels ok to me.

Dimensionally it's as identical as I can get to the original, similar tapers on the body & button, so am hoping it will fit well. Main difference is my button is hard ABS, while original was slightly rubbery stuff, so mine may be a tight fit in the handle; if so I'll have to adjust the design a bit. It's actually turned out to be a very fiddly job wiring it up, as the contacts for the 2 poles are almost touching, and are duplicated down each side of the button. Hence the use of the green wirewrap solid-core Mylar-coated wire you can just see,

121401


121402