HP 46/81 idler gear
10-28-2020, 09:56 AM (This post was last modified: 10-28-2020 10:01 AM by Duane Hess.)
Post: #11
 Duane Hess Member Posts: 170 Joined: Mar 2014
RE: HP 46/81 idler gear
Hi RPZ:

Not sure my post has a point; you've probably know most info in it. But do you recall in your past after asking a person questions they stop and look at you for a moment? Much to your suspicion of their pondering "Do you really want to be punished?" was true. Woe to the person who does not ponder whom can communicate succinctly. Fortunately, you never asked me a question so... the punishment, err, my elaboration will be minimal. (and there is a suggestion at the end)

Used Sadsilence's solution myself; worked well. There have been numerous posts on the 46/81 printer. Believe many revolved around alignment the printer itself as opposed to gear specifications. If you're curious about the printer you might try:Epson 102
to see pictures with the covers off.

As you already know, the gear is a speed reduction idler combination gear of 17/68 teeth Mod 0.4. From my measurements with an old mechanical slide micrometer:
- small gear 7.5mm diameter, 3.9-4mm "tall" (thick)
= as odd # of teeth, one side of micrometer is setting on the corners of 2 teeth and other jaw is on the point of a single tooth
= if I understand correct [(#teeth + 2)*module] produces a very good value for outer gear diameter. 19*.4 = 7.6mm
= calculated pressure angle 27-28 degrees (thrust angle?) (described later)
- large gear 27.8mm diameter, 2.9-3mm thick
= micrometer mispositioned: each jaw was sitting on corners of 2 adjacent teeth. should have been tip to tip of opposite teeth
= (68+2)*.4= 28mm
= calculated pressure angle 18-20 degrees (described later)
- the gears have a 3mm bore. The bridge shaft measured 2.9mm. "Bridge" is how HP/Seiki printer documentation refer to the removable plate and shaft supporting the gear.
- overall gear thickness 7mm
** it should be noted I have no background using a micrometer. The first time using one was to measure this gear. The micrometer supposedly has a +/- accuracy of 0.1mm. Who knows what my accuracy is.

For your reference the printer in a 46 & 81 is a Shinshu Seiki EP-102. You may find various manufacturer names as several companies (owned by various family relatives) evolved names & formed subdivisions. Basically Suwa Seikosa -> Shinshu Seiki -> (some products+printers) -> Seiko; Shinshu Seiki -> (printers especially) Epson. The part number for the gear (hoping someone will find a stash somewhere):
Shinshu Seiki: B-2-11 "Idler Gear Wheel"
HP: 1535-3243
Seiko: 1240120

The "Gear Comparison" jpeg in the "HP 97 Printer Gear RPZ rev2" zip shows key geometry consistent with the 17 & 68 combo gear:
- rounded trough bottom (essentially or is circular)
- side extends up at an angle
- approx. top 1/2 of tooth angle changes again to match pressure/thrust angle
- end of the tooth is blunt (snub nose--flat)

I was curious what material PLA (yellow gear) you refer to is. Along with what material comprises the Shapeways fine detail & clear gears. Are they nylon, delrin, polyacetal? My interest centers around gear's relative hardness.

Wondering, as my opinion of the 46/81 combo idler (spur?) gear and the 97 printer idler gear is they were intentionally made of "sacrificial" material. Unfortunately, the material falls apart also. Sacrificial meaning made of a softer material as to incur most of the gear train wear. Hence, lessening wear on the gears connected to. In the EP-102 printer its obvious replacing the motor gear or drum gear would be much more difficult/consequential than replacing the combo idler gear. The fact the combo idler gear is an a separate removable plate ("bridge") underscores that design philosophy, in my opinion.

Is the issue of a softer sacrificial gear actually that important? Likely debatable . But I would not use a metal gear unless no other alternatives exist. Further, I'm more a hobbiest than anything, so my minimal use wouldn't underscore any vailidty to the concern irrespective.

However, consider automotive comparisons; granted high-force and extended duty cycles. Car people know you replace geared components in pairs. For example, you don't replace metal timing chains without a full timing kit (gears it runs on). Likewise you move the gear from an old replaced distributor to the new one. If the new distributor gear is used, replace the cam shaft gear also. People with rear-wheel drive cars likely have seen the differential gears. You'll notice a wear "pattern" between gears. The pattern is random & unique with each gear set. This is why you replace gears in pairs, particularly in large force and duty cycle situations.

Imagine your hands being two gears originally with nice and squarely aligned teeth. Now, years later, make one hand cup shaped and the other a fist. Put them together and pretend they are worn gears. Although worn, maybe with vibration, etc. the fist and cup fully seat. To make the point, replace the fist with a square of cardboard. Slide the cardboard along the cup shape. Contact between the two will be only in a few points. Small contact surface creates massive wear at the same points. Hence, you replace gears in pairs. Practical for cars, but not really for these printers in my opinion. The EP-102 motor & drum gears would be a tedious issue.

What's the point? I would like to know the relative softness between the gears at Shapeways and PLA (whatever it is). Personally I would like to share some info. I have minimal ways to examine the original gears and newer replacements--the 17 & 68 individual gears you can use to create a combo gear.

I made images on a flatbed scanner of the remnant pieces of the original broken gear. (gee 3D doesn't work well on a flatbed scanner) Results weren't that bad. Picked up the scanner (combo printer, copier, scanner, etc.) at a yard sale for about \$10. Scanned the large and small gears at 800DPI. Cut & pasted the gear only from the 8-1/2x11 inch image into a jpeg of its own. Portrait printed the gear on letter sized paper. So ended up with a 10 in wide gear. Unfortunately the only ink I had came with this 10+ year old device, only printed pink & ran out quick; ended up with crappy printouts. However, using a ruler and doing some trig am pretty sure of the following:
- the thrust angle of the large gear is somewhere between 18-20 degrees
- the angle for the small gear appears more like 26-30 degrees (probably 28); the small gear was tedious since the ink ran out
Off the shelf gears generally have a 20 degree angle.

Now I know why I keep a bunch of little boxes/vials full of broken chunks of all sorts of stuff that comes out of equipment. You never know when its useful!

If you are interested (and will return my gear chunks) I would be happy to send them for your examination (my expense). Should you be interested in comparison with your gear. Magnifying glass examination shows the thrust angle between the 17 & 68 tooth gears are different (or look that way). I don't want anything in return (other than returned parts unmodified) but would appreciate microscope images of my gear pieces. A flatbed scanner doesn't work well. Naturally the hope is you would verify and modify, if needed, your 3D printer definitions so we all can obtain gears from Shapeways. Or that's how I took your posts, you created 3D definitions and were allowing the rest of us to purchaes parts from Shapeways (or you?).

I understand you have a life, etc., so I don't want my gear parts returned in a week. A few months is fine. (a year would annoy me). Maybe you're not interested, but took it as your familiarity with the gear is relatively limited. I have the full 68 tooth gear and 17 tooth gear (separated), but about 1/3 of the teeth on the small gear are chipped.

Please let me know your interest and we can make shipping arrangements. Will include my horrid printouts. For you reference on how I came up with the thrust angles. Actually for my reference, to see whether I have a clue. Don't know crap about gears other than once upon a time owning 70's era cars. And what I've read on the Internet, as to have a clue to talk about it.