Defective rotary encoders

[Wolf0 31]

Hello all.

This post is a minor hijack but I need advice on a subject and maybe someone here has gone through this very same problem.

I have a MCP Combo II - Airbus FCU which is starting to show it's age. The rotary encoders are full of noise (electronic) and are mostly useless now. I have opened the unit multiple times and sprayed contact cleaner on the encoders but the improvement is residual and it fades out quickly.

I've decided that I'll have to de-solder the original encoders and apply new ones. Which brings me to my question: which ones should I buy (that will hopefully last longer) and from which vendor?

Thanks in advance.

Pedro Carvalho

[BIGSKY 321]

I have had good luck with these, https://www.digikey.com/en/products/detail/tt-electronics-bi/6127V1A180L-5FS/2620661

[Wolf0 31]

Thank you for your help @BIGSKY. I'll be posting an order soon!

[Wolf0 31]

Hello again.

I've now completely solved my problem with the defective encoders by replacing them with compatible new ones.

This post is aiming to point any other user with the same problem to a possible solution which was the one I found.

First of all, for reference, the installed rotary encoders on my VRInsight MCP Combo II - Airbus FCU were ALPS ENCODER part number EC11E15204A3.

These are rated for an operating life of 15.000 cycles, which is the root of the problem. VRInsight chose low end encoders for the hardware, with a very low expected life. On my hardware, with low-moderate use they lasted 4/5 years without problems and after that electronic noise started creeping in and they became progressively useless.

These encoders have 5 signal carrying pins. 2 on upper half (for the push button action) and 3 on the bottom (for the left/right/left fast/right fast action). Additionally there are 2 latches on the sides which are there just for stability. Overall there are 7 metal pieces contacting (and soldered to) the mainboard (5 pins + 2 latches) which must be de-soldered (by hot air or a very tiny soldering iron tip). I draw your attention to performing this de-soldering step with care, as the mainboard tracks where the pins connect are brittle and if the button is still connected one risks lifting a track together with the button (this is not a deal-breaker, but will involve soldering a tiny electric wire between the encoder contact and the mainboard track origin (been there, done that two times.....).

The replacement encoders that I was able to order are also from "ALPS ENCODER" (which, as far as I understand are no longer in active production). I was able to get them via a electronic store supplier which still had stock. I chose the part number EC11K1524406, and ordered them via this store:

https://lcsc.com/search?q=EC11K1524406

The reason why I've chosen this model is simple: they are almost identical to the original ones (5 electronic contacts and same base size and orientation plastic pin), but they offer 100.000 operating life cycles! Nearly 10x as long as the original ones.

They do come with some minor disadvantages:

1) They lack the side latches to the mainboard, so they stay connected only by the 5 tiny electric signal pins. This is not enough to prevent the encoders from dethatching from the board when the user pulls the black plastic caps for a new disassembly, or even for regular use. I've circumvented this issue by creating two solder bridges from the old mainboard landing plates for the latches to the sides of the rotary encoder. It looks terrible but helps securing the encoders to the board.

2) The new encoder has a shaft slightly wider at the base. This will prevent the metal front panel of the VRInsight device to be fully screwed to the back cover. The round holes drilled at the factory where the encoders protrude will scrap the sides of the new ones. This is circumvented by using a power drill to widen the holes by a small bit. This looks scary at first but actually is really easy. The holes are already made, so the drill will self-center. And the metal is not terribly stiff, so it is easily drilled.

The good news is that I now have a working device ready for a lifetime with "minimal" effort, and the new encoders add a very nice rotating quality feel to the unit.

I hope someday will benefit from this information someday.

Pedro

 

Edited September 21, 2022 by Wolf0

[gboz 157]

On 9/22/2022 at 7:06 AM, Wolf0 said:

I hope someday will benefit from this information someday.

Pedro

 

What did you think of Bigsky's Digikey linked ones, (from their specs)?
They seem to have only three contacts?

[Wolf0 31]

4 hours ago, gboz said:

What did you think of Bigsky's Digikey linked ones, (from their specs)?
They seem to have only three contacts?

They appear to be high quality but they have some compatibility issues:

1) The shape is round and won't fit the mainboard without major adaptations.

2) The contacts are bellow the encoder and the mainboard is expecting a flat base with side contacts

3) There are only 3 contacts, which means the push action will be lost. The original encoders have 5 contacts to allow for two different actions: 2 for the push function and 3 for the rotary function.

Overall it appears that they could be used in a contingency scenario with additional workload for the modifications, but the rotary encoders that I ended up choosing are much more compatible (albeit with their own issues) and require less work.

[gboz 157]

On 9/23/2022 at 8:56 PM, Wolf0 said:

They appear to be high quality but they have some compatibility issues:

1) The shape is round and won't fit the mainboard without major adaptations.

2) The contacts are bellow the encoder and the mainboard is expecting a flat base with side contacts

3) There are only 3 contacts, which means the push action will be lost. The original encoders have 5 contacts to allow for two different actions: 2 for the push function and 3 for the rotary function.

Overall it appears that they could be used in a contingency scenario with additional workload for the modifications, but the rotary encoders that I ended up choosing are much more compatible (albeit with their own issues) and require less work.

Ok, thanks.
How do you get the VRI MCP grey plastic knobs of the shafts?
Mine refuse to come off despite a really hard pull.

[Wolf0 31]

Pull harder... They are not bonded by glue or any slot.

The shaft is a polished cilinder and I expected less friction, but they resist a lot initially to the pull.

Maybe fit a flat head screwdriver bellow the black caps to help leverage.

[gboz 157]

Right oh, thanks again.

[zetato 11]

Hi all,

sorry if I revive this topic, but my MCP COMBO II Boeing version started having the same problems.

I find this type of encoders https://www.farnell.com/datasheets/1564006.pdf, in particular the EC11j1524413 

What is not clear to me, is the difference in the two type of output wave and relative detent stability point, shown in page two and what this can change in the sim.

Can I have your opinion?

Thank you.

Mauro

 

Edited May 22 by zetato

[SergeyPe 96]

11 hours ago, zetato said:

What is not clear to me, is the difference in the two type of output wave and relative detent stability point, shown in page two and what this can change in the sim.

Hi Mauro,

EC11K15XXXXX and EC11J15XXXXX are identical other than the PCB mounting method (pins vs surface-mount). Also the difference with "E" (original model) is minor in terms of interfacing with the unit's controller, which is proved by the fact that the "K" version is working fine in Wolf0's system. It means that the "J" will be OK as well. Great encoders, by the way- I'm using the K's in a DIY B737 EFIS module: a very good tactile feeling with a nice click

[zetato 11]

Thank you SergeyPe,

I'll try with the J version to have the mechanical cage soldered to the mainboard.

Will post here the results in future!

Mauro

[SergeyPe 96]

Hi Mauro,

Almost forgot to mention another option- it's Bourns PEC11H series. 100 K cycles lifetime and the footprint is very similar to ALPS. 

Regards,

Sergey

[zetato 11]

Hi guys, thanks everyone, sorry @SergeyPe, sometimes I don't receive any notifications from the forum (maybe it's my fault in the Gmail configuration) and therefore I didn't read your last message. Anyway, good news, I replaced the encoders with EC11j1524413 and they work very well! I needed to enlarge the holes in the front mask a little, but that's easy, and the encoders are a little longer (maybe 5mm compared to the original, but for me it's not a problem to have the knob a little further out from the panel. To remove the encoders I used a hot air gun set at 260°C, carefully masking all the small components around it using aluminum tape, and for soldering I used the classic soldering iron, not exactly simple, but with the help of a very expert friend of mine we made it! I hope this information can help other guys in the same situation with this excellent piece of hardware!

Mauro

 

Edited May 31 by zetato