Alps key switches - ultrasonic cleaning adventures

[XMIT]

tl;dr: Ultrasonic cleaning works wonders for Alps switches, making them feel new again. A tiny amount of Teflon oil makes them even better!

Alps key switches tend to suffer more than most when exposed to dirt, dust, and other contaminants. Dirt enters the switches easily through the tops. Due to close tolerances the dirt interferes with travel resulting in a scratchy, gritty, slow, mushy, or grinding feeling when keys are pressed. This is all too familiar to anyone with an old Alps board.

We had discussed ultrasonic cleaning in some other threads, most recently photos-f62/ultrasound-keycap-cleaning-t14413.html. The ultrasonic cleaner should be perfect for Alps switches with all their small hard to access spaces. There was some concern that the cavitation could harm the click leaves and switch plate leaves. Some research suggested that a fairly neutral pH should not damage thin metal parts in the sonic cleaner.

I experimented today with cleaning some Alps switches. The configuration was as follows:
Alps SKCM White and Cream Damped switches
iSonic P4820 ultrasonic cleaner
Solution consisting of two Efferdent denture tabs, about 1.5 tsp 7th Generation dishwasher fluid, and reverse osmosis water at about 60 C
Disassemble switches
Clean for 480 seconds
Remove, rinse in distilled water, dry, and assemble

Here is a photo of the disassembled switches:

Switches disassembled. Switch plates remain in lower housing. Upper housing, click leaf, slider, and spring are separate.

600_3264.jpg (264.24 KiB) Viewed 9472 times

Here are detail photos of the click leaf and the switch plate post cleaning showing no visible damage.

Tactile leaf detail post cleaning.

600_3267.jpg (167.56 KiB) Viewed 9472 times

Switch plate detail post cleaning.

600_3266.jpg (239.75 KiB) Viewed 9472 times

Here is a photo of the switch about to go back together.

Assembly post sonication.

600_3273.jpg (130.62 KiB) Viewed 9472 times

The switch feels brand new. I cannot detect really any grittiness or hesitation. Some switches still hesitate a tiny bit in response to aggressive off center key presses. There is a small chance that enough abrasion over time roughens the plastics themselves.

I did experiment with lubrication using a tiny amount of Super Lube 51010 synthetic oil with Teflon. I applied a tiny amount, as thin as possible, with a toothpick all over the slider. This improved key feel even further.

I'm looking forward to applying this technique to all the switches on a very gritty Northgate Omnikey 101 to get it back into working condition.

[Wodan]

Excellent work, so glad someone took the risk and threw some Alps into the ultrasound cleaner!

What confuses me is the reverse osmosis water. Where do you get that or how do you make that? I have plenty of distilled water but reverse osmosis is new to me.

[Chyros]

Excellent work, so glad someone took the risk and threw some Alps into the ultrasound cleaner!

What confuses me is the reverse osmosis water. Where do you get that or how do you make that? I have plenty of distilled water but reverse osmosis is new to me.

Basically deionised water, distilled water should work fine as well. EDIT: come to think of it, if you're using denture tabs, you can just use tap water for the first step anyway xD .

What's the role of the denture tablets and does this work without them?

[Khers]

Nice job, XMIT. I found something similar when ultra-sonicating the creams Alps in my AEK2. They went from pretty bad to pretty much as good as new. As I don't care for the tactility on offer in tactile Alps I removed the dummy-leaf and made them "linear" and added a dash of lube to finish the job.

[XMIT]

What confuses me is the reverse osmosis water. Where do you get that or how do you make that? I have plenty of distilled water but reverse osmosis is new to me.

I have a reverse osmosis filter in my kitchen for drinking water. Distilled water is better for this purpose but for me reverse osmosis water is less expensive. Either is fine. The idea is to make sure the water does not have any minerals dissolved in it, that could become a problem in the key switches when the water dries leaving the minerals behind.

[XMIT]

What's the role of the denture tablets and does this work without them?

The denture tablets are a combination of disinfectant and mild cleaner. This should work just fine without them.

I'll add that the tiny Alps springs slip through the gratings of the basket that I use in the cleaner. I fish them out with a magnet on a stick once cleaning is done.

[Chyros]

What's the role of the denture tablets and does this work without them?

The denture tablets are a combination of disinfectant and mild cleaner. This should work just fine without them.

Cool, good to know.

Tbh I don't think you need the deionised water in the denture tab step as it'll be liberating minerals anyway. But for sure you'll need the rinsing step with deionised water afterwards, no question.

Do the switchplates not retain the moisture?

[XMIT]

Do the switchplates not retain the moisture?

They most certainly do since there are so many small places where water can get stuck. I worked around this by using compressed air (either from a can or from a compressor) to blow out as much water as I could. I recommend doing this ASAP after washing so that rust doesn't form. You might also try a hair dryer.

As with any switch cleaning procedure I suggest testing every switch prior to re-assembly.

[XMIT]

Tbh I don't think you need the deionised water in the denture tab step as it'll be liberating minerals anyway.

Perhaps. But again, the calcium hardness and total alkalinity of our tap water here in Central Texas are so high that they interfere with the ability of soap to do its job. Softer water of any sort will result in less soap usage and more effective cleaning. This is important to me as I tend to reuse my ultrasonic cleaning solution for multiple key washes.

(Yes, a water softener might help. We don't have one.)

[Hypersphere]

I'm using a similar procedure, but for the final step (before air drying), I rinse with 70% (v/v) isopropyl alcohol instead of distilled water.

[XMIT]

I rinse with 70% (v/v) isopropyl alcohol instead of distilled water.

That would work, I guess I've just grown tired of inhaling isopropanol fumes. Also I can easily a gallon of distilled water for a dollar. If I'm lucky I can get about a quart or a liter of isopropanol for one or two dollars locally.

I tend to use 91% since it dries that much faster. Do you have any special reason for using the 70% concentration?

[Spaceman1200]

glad to see you tried this out! nice to see it had great results!

[Chyros]

I'm using a similar procedure, but for the final step (before air drying), I rinse with 70% (v/v) isopropyl alcohol instead of distilled water.

Yeah, that's exactly what I was thinking too, azeotrope it, much more reliable for drying the insides.

[Hypersphere]

I'm using a similar procedure, but for the final step (before air drying), I rinse with 70% (v/v) isopropyl alcohol instead of distilled water.

Yeah, that's exactly what I was thinking too, azeotrope it, much more reliable for drying the insides.

I think that the azeotrope of isopropanol with water is 91% isopropanol by volume or about 88% by weight. I just happen to use 70% (v/v) isopropanol, because it is more readily available in nearby shops than the 91% variety.

[XMIT]

I'm using a similar procedure, but for the final step (before air drying), I rinse with 70% (v/v) isopropyl alcohol instead of distilled water.

Yeah, that's exactly what I was thinking too, azeotrope it, much more reliable for drying the insides.

Heh, I hadn't thought of azeotropes since high school chemistry but for sure that would help with drying. I'll try that next time.

At least according to Wikipedia the azeotrope concentration of IPA in water is 87.9% by weight. I'll keep using my 91% concentration.

[XMIT]

I just happen to use 70% (v/v) isopropanol, because it is more readily available in nearby shops than the 91% variety.

(I wonder why this is - here in Texas getting 91% is super simple but in other places I've never seen it! There don't seem to be clear answers online.)

[Chyros]

I'm using a similar procedure, but for the final step (before air drying), I rinse with 70% (v/v) isopropyl alcohol instead of distilled water.

Yeah, that's exactly what I was thinking too, azeotrope it, much more reliable for drying the insides.

Heh, I hadn't thought of azeotropes since high school chemistry but for sure that would help with drying. I'll try that next time.

At least according to Wikipedia the azeotrope concentration of IPA in water is 87.9% by weight. I'll keep using my 91% concentration.

To be honest, the exact azeotropical concentration is a bit of a technicality, the two are sufficiently miscible that evaporation of one will also completely strip away the other. Calling it azeotrope is just a lazy way of describing this relationship that we use here xD . For a similar reason chemists use acetone to clean glassware, as acetone is extremely volatile and dries off all water in the process. Although I wouldn't advocate using acetone for cleaning switches as they will most likely completely decompose before your eyes.

[Firebolt1914]

I was thinking of doing this and then applying a dry lubricant such as molybdenum disulfide on white alps to attempt to see if it will feel like blue alps.

[fohat]

I wonder why this is - here in Texas getting 91% is super simple but in other places I've never seen it!

Wal-Mart started selling 91% in the pharmacy department a couple of years ago so that probably flooded the market if nothing else. They also sell 70% at a cheaper price, but I have never compared prices to see if it was 70/91 as much money.

[duynguyenle]

Would the concerns about corrosion and moisture retention be alleviated by disconnecting the switchplate from the bottom housing before cleaning? (I do realise doing this is a huge pain, but I have managed to remove switchplates before without causing any damage)

[Hypersphere]

Today I put Northgate Omnikey 101(SKCM white Alps) #3 back together after using ultrasonic cleaning on the top housings and sliders. For the springs and click leaves, I just rinsed with 70% (v/v)) isopropyl alcohol.

In addition to cleaning, I lubed all the switch rails with Superlube 51010 oil using a small paintbrush.

This Northgate was the cleanest of the lot thus far, and the top plate was almost corrosion-free. Nevertheless, this is the one that had the most noticeable binding when keys were pressed off-center.

After cleaning and lubing, the difference is astounding. The switches feel smoother than the blue Alps in my LE DC-2014 (which I have not cleaned or lubed). Overall, this board sounds and feels incredible. Part of me doesn't ever want to experience a NOS blue Alps board, because this one with cleaned and lubed white Alps feels good enough (for now).

While I had the switches opened, I tried an experiment with sliders from brand new Matias Quiet switches. I temporarily fitted the switches in the Northgate with unmodified Matias Quiet sliders and with Matias Quiet sliders that had the bottom dampers clipped. I also tried these with click-modded tactile leaves from Matias Quiet switches. I preferred the native white Alps configuration to all of these other combinations, with one exception -- I ended up putting a fully damped Matias Quiet slider in the spacebar switch. This works wonders for producing a quiet spacebar -- highly recommended!

As for the other slider combinations, the fully damped Matias slider feels too mushy for me for the regular keys (but it is great for the spacebar). The hemi-damped Matias slider just didn't sound or feel right, although it accomplished what I had set out to do, which is to damp the return stroke but not the down-stroke. The click-modded tactile leaf resulted in a louder click and heavier-feeling switch than the stock white Alps, and I did't particularly like the effect.

Thanks to XMIT for recommending the iSonic P4820 ultrasonic cleaner (with basket)! I got a brand new one for a very reasonable price through Amazon.

This rejuvenated Northgate is just amazing to type on! It's as if all the best characteristics of IBM Model F and Topre switches were somehow merged into a new species of keyboard. (I suppose it would be known as an Alps keyboard! But if this cleaned and lubed board approximates the sound and feel of a brand-new Alps board, it is a wonder that they stopped making them).

[XMIT]

Would the concerns about corrosion and moisture retention be alleviated by disconnecting the switchplate from the bottom housing before cleaning? (I do realise doing this is a huge pain, but I have managed to remove switchplates before without causing any damage)

I don't believe so. The water retention concerns come from places with really tight clearances and a tendency for strong capillary action. Namely, the switch plate internals. I am *not* about to take switch plates apart for this.

The switch plates are tough to separate from the lower housing because my desoldering station leaves enough solder on the legs to keep them from pulling through the holes in the lower housing. It would be quite a lot more work to pull them out for very little benefit.

[Chyros]

To be honest I'd be very surprised if the cleanliness of the SWITCHPLATE made a difference Oo . The contact leaf maybe, but really, certainly not the switchplate itself. I think even if someone reported that I wouldn't believe it :p .

[XMIT]

I could believe that dirt/dust gets caught in the switch plate itself which leads to poor switch behavior. I've definitely un-stuck some switches by pressing them enough times to work some grit free back there. So, my procedure will include cleaning the switch plates for now.

[pyrelink]

Very nice work XMIT. I was always under the impression that a decent ultrasonic was closer to the $4-500 range, but very promising that you are getting good results out of just a $90 unit. I might put one on my shopping list.

How bad were the switches before you put them through the ultrasonic? Would you have considered them too scratched to save before you tried the ultrasonic?

[Chyros]

I could believe that dirt/dust gets caught in the switch plate itself which leads to poor switch behavior. I've definitely un-stuck some switches by pressing them enough times to work some grit free back there. So, my procedure will include cleaning the switch plates for now.

Yeah, maybe with regards to switch actuation, but the actual movement is both orthogonal and tiny, I can't see how even a really crummy switchplate would affect key travel tbh Oo .

[duynguyenle]

Would the concerns about corrosion and moisture retention be alleviated by disconnecting the switchplate from the bottom housing before cleaning? (I do realise doing this is a huge pain, but I have managed to remove switchplates before without causing any damage)

I don't believe so. The water retention concerns come from places with really tight clearances and a tendency for strong capillary action. Namely, the switch plate internals. I am *not* about to take switch plates apart for this.

The switch plates are tough to separate from the lower housing because my desoldering station leaves enough solder on the legs to keep them from pulling through the holes in the lower housing. It would be quite a lot more work to pull them out for very little benefit.

I meant more in that you can take the switch plate out and just wash the housing and slider, but I see what you're saying about leftover solder on the switch pins. They do make it extra hard to pull the switch plate out

[XMIT]

How bad were the switches before you put them through the ultrasonic? Would you have considered them too scratched to save before you tried the ultrasonic?

They were pretty bad. Not the worst I've tried but not far from it. The cleaned ones feel as close to new as anything I've tried.

[Hypersphere]

FWIW, the white Alps switches in my Omni 101 feel incredibly smooth after cleaning and lubing. The only parts that I sonicated were the top housings and sliders. The springs and contact leaves were simply swirled in 70% isopropyl alcohol and air dried. Superlube 51010 was applied to the slider rails in the housings and to the underside of the slider that contacts the contact leaf. I can recommend this procedure for those who wish to work on the switches in situ without desoldering.

Perhaps if the switches are in really bad shape and/or if you have already desoldered them, then you could clean the switch plates as well using a sonicator.

Notes on keycaps: I tried out keycaps from several sets on the alpha keys of the home row before deciding which set(s) to use in populating the entire board. Here is a summary:

AEK dye-sub PBT: profile does not match the modifiers (blank black ABS from Matias), don't like Italic (Oblique) font on keycaps, don't like the homing keys on D and K instead of F and J.

Dell AT101 Old Logo -- not dye-sub PBT on this model, but very nice caps nevertheless: Okay, but didn't sound as good as the stock caps.

IBM 5140 dye-sub PBT: My favorite with respect to appearance, but the caps did not bottom out correctly resulting in unfavorable sound quality. In addition, I do not have matching keycaps to use elsewhere on a full-size keyboard, such as on the NumPad.

Northgate Omnikey 101 doubleshot ABS: I ended up going with these based on profile, fit, appearance, and sound. Usually I prefer PBT, but I am willing to use doubleshot ABS because of all the other good qualities of these caps.

SGI Granite dye-sub PBT: Good fit and sound, but I don't like an Italic font on keycaps

Wang 724 dye-sub PBT: These caps sat slightly lower than the modifiers (blank blank ABS from Matias).

For modifiers in the main typing area and the spacebar, I used blank black ABS from Matias. For Esc, F5-F8, Nav Island, and NumPad keys other than the number keys, I used WOB doubleshots from a Tai-Hao set.

Despite my negative comments about these various keycap sets with respect to putting them on the Northgate Omnikdy 101, I intend to hold onto them for use in other projects.

The top case is painted black. The board looks good in my estimation. Moreover, after cleaning and lubing the switches, the board feels and sounds great. The switches are remarkably smooth (yes, I know I've said this before -- I just can't get over how good the switches feel!).

Thanks again to XMIT for recommending the sonicator and to XMIT and Fohat for other advice on cleaning and painting keyboards.

[Chyros]

Damn. Looks like I might need to give this a try. I've got plenty of comparison material xD .

I've had to halt my QKT video series for the moment due to extreme demands on my time with regards to my thesis, but I'll be picking this up again in the future. I want to do another Alps switch restoration video as well. Hopefully you guys will continue to gain more ground in this process so that when I do the video we'll have a consistently working, detailed protocol handy, with a specific cleaning procedure, optimal lubricant, etc.

Has anyone done this with first-gen Alps btw? I suspect this procedure would interfere with the lubricant already present on them. It may even be necessary to wipe the lubricant off completely.