IBM Model M Screw, Nut and Bolt Mod

[E3E]

Hey, totally oblivious E3E here. I work with my Dremel all the time, but what are these divots you guys are speaking of? Was the previous suggestion to literally burn holes through the plastic film or whatever?

I don't really work with IBM keyboards, so I am very much in the dark, but I have seen and heard of the bolt mod.

[emdude]

Yup, it's to help guide the drill. I've seen it done with a soldering iron as well as with a dremel with spherical burr, as fohat has done.

I wonder if there is any difference in accuracy between the two methods though. I've only done the former.

[fohat]

any difference in accuracy between the two methods

Night and day. The soldering iron mounds up barriers of resistance.

[emdude]

That makes sense; I would try a bolt mod with divots made with a dremel but unfortunately I only have access to a soldering iron at the moment. One of these days!

[phosphorglow]

Would you agree that the ability to accurately drill the holes is proportionate to the ability to confidently and safely hold the barrel frame in place while drilling, leaving both hands free for support and control...?

[emdude]

Certainly, but unfortunately, I do not have a vertical drilling press like the one in your old video. Right now, the best I can do is to trust in my own accuracy with a normal drill and have a friend help hold the barrel frame down on the jig while I do the drilling.

I was curious because I would think that having divots made with a soldering iron would be more accurate (assuming that you were accurate in making the divots themselves) because of a physical hole in which to sort of insert the drill.

[phosphorglow]

Alrighty! This is good stuff. I now have a slight revision to my guide I can work with. I've been trying to think in terms of drilling everything by hand like most people will do. My press is rather modified to specifically drill M barrel frames and I'm used to that mentality. But with patience and a steady hand, one can accomplish quite acceptable levels of accuracy.

Excluding an M15 I just did, it's been a number of M's since I've drilled one by hand. Slightly different story for regular M barrel frames than the M15 though... Either way, I know I can get the holes drilled very close to my "ideal" specifications by hand.

And the slight revision I mentioned, well, I'll be including it in my methodology from now on. It accomplishes the same result and saves me a step. Instead of the two pen marks I make as reference lines to drill between on my press, I can just use my awl like I'll be explaining in the guide... heh. I am nothing if not consistent.

(Y'know, it's been... over 100 M's since I've made a change to how I do things, and the last thing was a switch to a 1.7mm drill bit - a whole whopping .1mm difference.)

Let's back up a second...

Before I even considered offering my services, I wanted to come up with a method to consistently drill holes...:
- perpendicular to the barrel frame
- parallel in relation to the other holes in the same row
- within the confines of the original stud's circumference, biased lower (not higher)
- centered in relation to the holes in the steel plate (NOT dead center on each one; there's a gradual shift from center to sides)

The first two are primarily because I'm a little bit of a perfectionist. The third one is important for the integrity of the plastic (there's a happy 'meaty" zone). The fourth is quite important when using flat head screws with a conical shaped underside, which makes getting a finely tensioned and tuned M so much easier.

So! The divot I agree is incredibly important. What I'm not happy with is the size and location of the divot when a soldering iron is used (and ruining a good soldering iron tip!). A finer point to dial down the drill location in relation to the hole in the steel plate is what I want to stress about this subject. Such as the point of a nail, or this great cheap Harbor Freight special:

2016-03-21 23.01.34.jpg (59.89 KiB) Viewed 18152 times

Sort of difficult to take this picture, but that's pretty much exactly where I want that hole in relation to the plate. Drilling that hole accurately will be significantly easier with a proper work jig.

2016-03-21 23.09.00.jpg (52.27 KiB) Viewed 18152 times

As for that jig, I'm working on a printable PDF plan with instructions to build one like I posted in the FSSK thread. I've built three of them from my current "plans" and they're getting easier and more refined. Soon everyone can make their own!

There's a bunch I feel I need to explain still. A lot of it is in a big text document with so much shorthand writing only I can understand it. And a number of things will be easier explained with visual references. So I've been working on reference material. In Adobe Illustrator 5. On my Macintosh SE/30. Because concise reference material should fit on a 9" black and white display. I'll post something relevant here in a bit... ;P

[phosphorglow]

holesinrelationtoplate.png (4.71 KiB) Viewed 18142 times

Hah! It's only 4.7KB - I don't have to feel guilty about using up deskthority bandwidth. Mhuhahahah.

Anyways, most of that should be self explanatory. At least that's the goal. I'm working on it.

(The screw diagram on the right at an angle is to show the merits of the minimal amount of contact area when flat head conical bottom screws are used. I actually took a few shapshots to turn into a gif. I'm having fun with this...)

[emdude]

Wow, this is fantastic. So you use a pick to make your divots? I might have to try that next time. I also did not consider making the divots relative to the holes on the steel plate itself.

It's good that you'll also be addressing tensioning when you make your future guide; I don't think I've seen a single bolt mod guide go into this important detail. Also looking forward to the printable plans for your work jig!

[phosphorglow]

Yup, a pick or awl - whatever you want to call it. Though, my method up until now has been to mark two lines with a pen, relative to the hole in the plate and drill in between those two. I have a magnifying glass on my drill press, which dials me in even closer. But with an awl/pick I just made my life one step easier by not having to use a paint pen on each location so the pen marks are visible. Heheheh.

2016-03-02 13.54.56.jpg (279.8 KiB) Viewed 18131 times

And yeah, tensioning. It's an art. It's taken me this long to finally get to a point where I feel I can probably convey how I do it. It sounds a little crazy, but the only reason I strive to be a better typist is to verify I can make a Model M feel as close to stock as I can after a full overhaul. A huge part of that is proper tensioning. If I can't comfortably get around 110WPM on a typing test, I'll tweak it or take it apart again and go for round two. But most of the time I get it spot on the first time.

[emdude]

Even though I have only bolt-modded a few Model Ms so far, I definitely agree tensioning is a matter of trying to get the right feel out of a Model M after it's put together again, messing with the screws as needed to accomplish that.

It makes me wonder how IBM's engineers managed to create a pretty great typing feel with melted plastic rivets.

[phosphorglow]

Yay, you gave me a reason to make another section of my 6 page ramble notes more coherent:

Clever IBM engineers knew that the barrel frame would naturally settle while cooling after being heat pressed. They knew the barrels had to be parallel and the same height and angle for consistent typing feel. My theory on the flex channels in the barrel frames was to primarily keep a lateral consistency, not just to help it curve to the steel plate. They also tried various plastic compositions. Each of them I've noticed has their own subtle character. Like, the pearl white frames tend to feel... “buttery”? And the harder plastic that was more of a brown/tan shade during '88-'89 felt “sharper”? I could, of course, be insane. UK plastic was also different (more rubbery). I'm also off topic.

Back on topic, all of the studs get heat pressed at once, it's not some individual process. So, stock, there's a natural push/pull dynamic. Replacing the studs/rivets with screws means we're now working primarily with the pull side of things. Also, different thickness plates had different thermal profiles, so some cooled down quicker than others, and some spots of the barrel frame cool at different rates. As they're cooling, the center (think from a side view) will have more of a tendency to separate. So, it's important to pay attention and “listen to”, to what the 30 year old plastic “wants”, and not just indiscriminately tighten everything at random expecting things to be better or sturdier. By the way, there's a reason my design for the work jig has a curve that's slightly less than the steel plate.

So, uh, the short answer to that last question: IBM engineers were clever.

[emdude]

I have handled more or less all of the different iterations of the barrel plates you mention, the original black one, the dark gray/tan one, the pearl white one, and the black Lexmark ones. Not sure if I have noticed the differences in the plastic compositions that you mention, perhaps I will as I handle more Model Ms. I wonder if any difference in feel can be contributed to age..

I also have noticed differences in the construction of the frame with various iterations. That is, the rubber tubes around the edges of the frame that were present in the early frames, but not the later ones, as well as the extra strip/bar of plastic that ran down the flex channels in the Lexmark-produced Model Ms. Anyhow, as I worked on those boards, I wondered if they were a cost-cutting measure or did something to improve how the barrel plate handled the flex or the feel of the board, etc.

On the topic of the heat-pressed studs, I remember a member here mentioning how early Model Ms had unevenly melted rivets and were manually melted. I suppose that isn't the case then.

[phosphorglow]

Heheheh, there's another section of my rambles that amuses me. I won't work on it right now but basically it's comparing the various generations of M's to Sneetches. All of the Sneetches are good, star bellied or not, but I'm biased towards the eldest Sneetches. Younger generations had their genetics modified for the survival of their race.

So, yeah, a lot of it was streamlining of the manufacturing process and subsequent cost cutting. IMO, they nailed it the first time, rubber (silicone?) tubes, separate metal stabilizer bar plastic doodads, thicker traces on the membrane silkscreen, all that jazz.

The drainage channels are a great idea, but do you know how many boards I've seen where the liquid spilled ends up draining INSIDE the subassembly because the space bar stabilizer risers have holes that lead right inside? A lot. (Edit: I'm not giving them enough credit for their usefulness - they do drain the bulk majority of the liquid from the keyboard in that event instead of letting it sit inside the case. The holes in the risers are the weak point.)

[fohat]

I do not have a drill press. The awl was my preferred method before I started using the tiny burr.

The work is exacting, but I set the object in a sturdy frame, put on a strong pair of reading glasses and lean in close, and use both hands holding my tool (with my elbows sticking out in opposite directions) to do my work.

Keeping tools sharp and fresh and clean is always important.

[zuglufttier]

I just did the screw mod on my SSK, I wanted to do that ever since I got it... About half of the rivets were already gone

I think, the mod could even be more easily done without removing all rivets in the beginning. Just remove some of them so that everything still fits tightly together and drill the few holes and put in the screws. Then remove more rivets and replace them with screws and so on. Sure, that wouldn't leave clean holes and all...

That's why I didn't do that But, man, the screw mod really made a big difference on my SSK!

[dblarssen]

Hey Guys, I'm normally not so in to posting and reacting to posts but I THINK I CAN ADD SOMETHING IMPORTANT TO THE NUT AND BOLT THING !!!

As I bought an M from Clickykeyboards (and not so satisfied, seen the transportation costs to Belgium + This..... the thing works OK, but I opened it up to see what they had done, and they only replaced 3 rivets by screws... a lot more plastic rivets broken off and loose in the case. So, to bring this KB back to life with the intention of using it full-time, I say there's only 1 solution, and that's to do the WHOLE MODDING, and not like Clicky says 'to just replace the broken rivets' ....because the next week others will break, and a month later some more... and more! I can also say the rivets didn't come off by using the KB, because I didn't!!). But that's not what I wanna talk about!

THIS IS THE IMPORTANT PART : There seems to be 2 camps, promoting OR bolts, OR screws..... the problem with the bolts being de metal plate getting lifted by the thickness of the protruding part and nut of the bolt (as I see it, there's about 1mm extra space there, from the lips on the lower shell, under which the keys-assembly goes... so with most bolts that's not enough, and the bottom of the case gets deformed a bit; the lips on the lower shell get overstressed a bit, and might brake, being old plastic). So I studied the problem and have some remarks and a solution :

a) The extra 1mm is good enough for the screw method, with the flat and sunken heads (proven OK).

b) I think the upper 5 rows (that's at least the upper half, probably more) of rivets don't touch the bottom shell, and can be bolted without any problems.

c) There's NO REASON why BOLTS should go in from the top, protruding the metal plate, and putting the nut on the metal shell side (with screws everyone does the opposite). Bolts with exact the same flat sunken head (as screws) exist ...nothing exotic! So just turn the bolt arround... the heads then only sticking out a mm or less... and UNDER THE KEYS THERE'S AT LEAST 4 OR 5mm SPACE FOR THE PROTRUDING BOLTS!!

So, to the guys with their excellent guides (Phosphorglow, Imgur, Madhias, I do this from memory so don't be insulted if I mention wrong or don't mention ...but all know whom I talk about) :

IF YOU THINK THIS IS INTERESTING AND IMPORTANT ENOUGH FOR ADEPTS WORLDWIDE? PLEASE ADD THIS PART TO YOUR GUIDE.... I WON'T WRITE ANOTHER ONE, AS IF I INVENTED THE HOT WATER AGAIN You could (but don't have to) mention 'HARRY from Belgium' for this addenum. Please use the information, and put it in correct English, using better names for the parts or positions I mention!

Greetz,
Harry.

[dblarssen]

Oh... forgot something!

d) About the material of the Screws/Bolts I also need to add something : If you use 2 different kinds of metal together, sometimes they can react (rust). This goes for Steel and Inox.... probably the metal baseplate (normally Electrolythically treated steel) will get scratched in the hole and on top, by the Screw/Bolt and therefor loose a bit of protection on that spot. Theoretically it's better to use treated steel Screws/Bolts than Inox because the Inox may cause oxidation where the contact is (in fact also the Inox will rust a bit by this reaction... but the steel plate will react the most).

Not pleasant to hear, but I'm the same : If possible I go for the best (so Inox) but I've learnt a bit by now

Harry.

[taylorswiftttttt]

removed

[POTV]

ssk 3.JPG (1.46 MiB) Viewed 17687 times

First of all I want to thank you guys for doing the tutorials on how to bolt mod and screw mod the IBM Model M. I rarely contribute much on DT, but visits quite often to see what’s happening.

Now to my mod: A few years ago a seller on eBay had a load of SSK’s for sale. I bought 3, one ISO French/Canadian and 2 ANSI – all made from between july and September 1987 and in very good condition. The ISO SSK has been my daily driver since – and I have quite often thought about having the 2 other ANSI converted by a skilled person. Sending them to Fohat or Phosphor Glow seemed a bit costly, since I live in Denmark. And I don’t know others that do it nearby. So I finally decided to do it myself. Got hand of a cheap vertical drill, Einhell BT-BD 401 for approx. 70 USD on Amazon and the other tools needed.

The first screw mod went according to plan, nice holes in the plastic plate thanks to a slow pace, and the final assembly went almost without any issues. Then I discovered that I has misplaced a spring, damn it. Open again and reassemble. The test afterwards showed, that 5-6 keys didn’t react. A time consuming and frustrating fine tuning went on with the screws (different pressure on the plates and membranes) and solved the problem 99 percent. But I wasn’t happy with the project over all. In my world things work 100 percent or don’t. So I almost gave up on modding the second SSK. Especially after I opened the case and noticed, that only 1 rivet had broken. A practically new keyboard specimen. Anyway, I started and learned from my first mistake, so it only took about 4 hours from start to finish. The test: All keys worked perfectly, and crisp and sharp key feel (Which the first modded SSK also had).

I wondered, why was the first SSK so screw pressure sensitive compared to the second?. I knew I had been very careful with the tightness. Suddenly I might have the explanation. On the first SSK I reused the original black rubber like blanket – on the second I used a new white silicone replacement blanket from Unicomp. I decided that had to try to use a new blanket on number one. And that did the trick. All keys responded perfectly afterwards.

ssk blanket.JPG (663.39 KiB) Viewed 17687 times

Lesson learned, for me at least. It is not that difficult to do a screw mod with the right tools. And do spend 10 USD on a new blanket from the very helpful people at Unicomp. When I talked to them, I also ordered some media keys – in light blue, which is not possible on their website but doable. (And thanks to Orihalcon for the mapping adapter)

Regarding the benefits of a screw mod in itself… The difference is not that big to my fingers, I would say 8 to 10 percent more precision in key feel compared to the almost new ANSI. If the keyboard is well used, and serves the owner on a daily basis, I would do the mod. I can’t judge on how to choose between a bolt mod and a screw mod, but the screw mod should be less time comsuming, and I’m pretty demanding and perfectly happy with my results.

[OldIsNew]

If you have a Dremel that fits it, I would suggest considering the Dremel 220-01 Rotary Tool Drill Press Work Station:

dremel2.jpg (135.21 KiB) Viewed 17667 times

I paid $32 for mine, though I think it's going for about $10 more at most places now. It's NOT a high precision drill press, but overall it works well. I'm not an expert on screw mods but I've done five, including two M122s, with this and it really made quick work of the drilling with very good results. I've also used it for a lot of other things as a stand for the Dremel since it can be rotated 90 degrees. I've had mine for a year now and use it on a frequent basis and it hasn't broken yet (lots of plastic parts so don't crank on it). Again, if your Dremel fits it and you can get one for a reasonable price I'd recommend it.

[JP!]

That looks to be a nice little kit. I already have that same Dremel.

[002]

Huge thread necro but just wanted to say thanks for this guide, Madhias.
I did my first ever screw mod over the weekend and it went very well. I used these screws which are probably cheese-grade Chinesium given the price but they did the job.
I also used a 1/16 imperial drill bit (1.59mm) which felt absolutely DISGUSTING to buy given that I am a metric-boi but sadly I couldn't find a 1.7mm. I use my cheapo Ryobi rotary tool with snake attachment which made the drilling process very easy.

[fohat]

I also used a 1/16 imperial drill bit

In spite of the miscegenation I have found that an Imperial 1/16" hole is perfect for an M2 screw because the screw will cut its way into the plastic and make a wonderful threaded socket for itself without need for a nut, or can forced in a little harder to function as a regular pilot hole.

[002]

In spite of the miscegenation I have found that an Imperial 1/16" hole is perfect for an M2 screw because the screw will cut its way into the plastic and make a wonderful threaded socket for itself without need for a nut, or can forced in a little harder to function as a regular pilot hole.

Exactly what I found to be the case even with the M2.2 that I went with However, I suspect that the screws I bought are not quite the quoted width -- in the picture on the eBay listing, the threaded part looks quite sharp and pronounced yet in reality they are dull. Also I didn't bother with making pilot divots before drilling. I think the speed of the rotary tool combined with the snake extension gave me enough control that it wasn't necessary. Sending the screws home felt so satisfying -- I'd hate to think of how finicky everything is with tiny nuts and bolts to contend with.

[fohat]

I don't know whether you are using some variation of the Sandy55 method, but wouldn't a self-tapping or machine screw with a shaft thicker than 2mm damage the "crescents" on either side of the rivet shaft that are supposed to space the plates?

[002]

Ah...is that what those things are for? I thought they were for holding the layers in place so that nothing shifted sideways.
Anyway, I was pretty careless when I took the keyboard apart quite some time ago. I cut away most of those crescents but in the end everything still works and feels great. Your advice in that thread was applied; specifically, not locking everything down super-tight.

[elecplus]

In the end i spent about €100 or €150 total for different screws & bolts

Necro thread, but I just bought 48 screws for under $3!

[fohat]

Too bad you couldn't wait 2-3 weeks for shipping from China.

https://www.ebay.com/itm/M2-x-8mm-304-S ... rk:41:pf:0

https://www.ebay.com/itm/100X-M2-M2-5-M ... 2749.l2649

https://www.ebay.com/itm/50X-304-Stainl ... 2749.l2649

[John Doe]

If I remember correctly, it's the very thread that brought me to the IBM Model M/F world. Thanks.