If you love your Roomba, shim it now!

[Mavrik]

solid lubricants will quickly be pushed out of the load-zone

This can be said for liquid lubricants as well, unless there is a sealed reservoir and gears submersed.

While graphite has slightly conductive properties, just like dirt and dust, it seems to adhere to allot of surfaces. I applied some graphite to a piece of glossy white nylon and I couldn't rub it off with my fingers, it just smeared. A nylon to nylon application would just embed the graphite into the gears. This would result in permanently dry lubed gears. (Only a very small amount is needed.)

...tends to replenish a load-zone, by flow-back.

Liquid lubricants will flow to the lowest point of the casing by gravity (and eventually on carpet). Flow back can only occur if enough heat is generated by the gears to draw lube to it (ever try to weld something oily?) and has a lube supply to replenish what gravity removed. One reason car engines have an oil pumps. Add dirt into the equation, the dirt will absorb the oil and become the 'solid' lube you mention. (Ever use dirt to clean up an oil spill?)

Agreed. But, is it not the case that grit generally imbeds the softer material, then wears the harder material? If one signs up to that philosophy, it will be the metal-shim parts that wear.

The soft nylon gears is the concern here, they are the ones that will embed grit and wear out. You bring up a good point about the shim, since it is harder than the plastic counterparts it's effectively wearing the area out quicker by introducing hard material into the equation. Combine this and oil/dirt into the area, one has effectively lapping compound wearing out the area quicker.

[Gordon]

solid lubricants will quickly be pushed out of the load-zone

This can be said for liquid lubricants as well, unless there is a sealed reservoir and gears submersed.

You are not visualizing the macro-picture. The "reservoir" here, is the small gap that exists in the non-load region between hub and static material. Its volume is on the order of a liquid 'droplet'. Liquid lubricant is retained in that gap by capillary force. Question: What force is there, that retains particles of dry-lube which may be pushed aside, longitudinally out of the bearing?

Note that my rejection of dry lubricants originally applied to member "fugli's" question about using a Teflon-based dry lubricant. I presume you read his May 17th post, and subsequent posts, above.

While graphite has slightly conductive properties, just like dirt and dust, it seems to adhere to allot of surfaces. ...

I fail to see what conductivity has to do with retention of graphite, I'm quite certain the graphite platelets find homes in the pits, fissures, cavities, scratches (roughness) which most any commercially finished surface exhibits. As the plastic hub wears, those islands of reserve graphite will come into play and help reduce friction.

... I applied some graphite to a piece of glossy white nylon and I couldn't rub it off with my fingers, it just smeared.

That is not a very satisfying 'test'. You need to rapidly move the two surfaces past each other at the rate the Roomba gear-hub travels, and emulate a load-force equivalent to the floor pushing up on the brush while it revolves for 30 to 90 minutes.

A nylon to nylon application would just embed the graphite into the gears. This would result in permanently dry lubed gears. ...

Give it a try, for about 100 hours of run-time, and tell us about it.

I agree that graphite and moly-disulphide dry lubes have the best chance of working. However, I think many owners would not care for the black-mess experienced by applying either of those materials. Also, it seems to me there would be a bit of hassle injecting the powder into the thin annulus around gear-hubs; even with the usually provided applicator that is intended to service tumbler-type key locks.

I prefer to apply a drop or two of viscous oil just prior to each cleaning mission.

...tends to replenish a load-zone, by flow-back.

Liquid lubricants will flow to the lowest point of the casing by gravity ...

No, no -- that's not correct thinking for this topic! Capillary attraction is holding the reserve oil. If you experience gravity-flow, you have applied too much oil--or too thin! Granted, after a few cleaning missions, there may be some oil film surrounding the bearing area, but not enough to run down...

...eventually on carpet).

..., is it not the case that grit generally imbeds the softer material, then wears the harder material? ...

The soft nylon gears is the concern here, they are the ones that will embed grit and wear out. You bring up a good point about the shim, since it is harder than the plastic counterparts it's effectively wearing the area out quicker by introducing hard material into the equation. Combine this and oil/dirt into the area, one has effectively lapping compound wearing out the area quicker.

You almost got it right in your last sentence. The tail end of the sentence should read "...lapping compound wearing out the harder metal-shim"; and that would be all right, because the shim is renewable!

FYI: In a machine shop it is common practice to make laps from the soft metals lead or copper. The lap is easily machined to size and shape, and is then *charged* with fine silicon-carbide grit (or graded-size diamond dust), by pushing the grit into the soft material. The lap can then be used to polish very hard, tool-steel items while also finishing them to size.

FYI: In an optics' finishing lab it is common practice to make polishing laps for working glass-optics (or metal-optics) from firm-wax mixtures. The lap is warm-formed to conform to the shape of a reference surface, and then used to polish lens surfaces, or optical flats. A slurry of water and a variety of abrasives are applied to the lap while a special machine oscillates the lap across the part(s), which are held on a rotating spindle. In that process, the wax becomes charged with the abrasive particles, then essentially 'fine-grinds' the glass or hard metal-mirror.

[glo69]

I'm not going to get involved in the wet/dry lubricant discussion although I use a wet lubricant and am more comfortable with the wet lubricant idea. Scientifically I don't really know which is better although it seems that wet is the most common in all types of machinery.

What I am sure of is that using shims is the best way to go for several reasons.

1) I have been using shims longer than most due to mundt1 sending me a set when he first got involved with the shims. I have found that Gordon is correct. The shim material is wearing out while the brush hubs show no wear at all after several months use. I still have trouble getting the .002 shims into the gap around the hubs.

2) Mundt1's first post that started this thread demonstrated a 10% reduction in the friction at the brush hub.This gain not only saves on wear at the brush hub but the reduction in drag should also reduce the loading and wear on the brush hub drive gears by ~ the same amount. It should also reduce the strain on the brush gear drive motor and lengthen it's life. Battery life will also increase since less drain per cleaning cycle increases the number of cycles a battery will last.

3) Some of the brass dust will obviously be getting into the gear box but I think the reduction in total grit caused by the shims blocking the gap around the hub will at least even this out.

4) Finally there will be a big gain in the life of the brush drive gears later in the Roombas life. As the brush drive hub and bushing wear out the brush drive gears will not only be loaded by the drive pressure, they will also begin to carry some of the Roombas weight and the more the brush drive hub begins to bounce in the bushing the more weight that is transferred to the gears. So not only does the hub wear out but it will also accelerate the wear on the gears.

Just as an extra note I believe that mundt1's idea of Li-ion batteries will also lengthen Roomba life. The Li-ion battery weighs ~3/4 lb. less than an APS. Taking ~9% of the total Roomba weight off of the Roomba drive train should decrease wear on a lot of parts.

[geekgranny]

Thanks so much for your comment GLO69. It make sense and your manner of presenting the info makes it easy to understand for us "laypeople".

gg=alice

[Mavrik]

I mentioned conductivity in case some were worried about graphite getting on the circuits, but then again, it would simply be vacuumed. The reference to the glossy white nylon is to demonstrate that graphite sticks to nylon and embeds itself. No amount of run time is going to remove it completely.

I missed this before that it's recommend to lube before each use (sounds like a pain and messy). You concur then to capillary is subject to gravity and has no relevance since your adding oil before each use, not to mention oil loss from the gears via centrifugal force. Since your adding oil before each use, at some point you may want to open the gearbox and empty the oil soaked dirt before it does some damage.

"...lapping compound wearing out the harder metal-shim" "The shim material is wearing out while the brush hubs show no wear..."
This defies the basic laws of physics, soft material will wear faster than harder material in every case where friction is involved and lapping compound (house dirt and oil) increases this wear.

Your FYI's are polishing techniques, I don't know why your mentioning this.

[glo69]

Hello Mavrik

Just curious. Did you read my post above, particularly item #1.

[Gordon]

I mentioned conductivity in case some were worried about graphite getting on the circuits, ...

There should be no concern at all about that effect.

... graphite sticks to nylon and embeds itself. No amount of run time is going to remove it completely.

Graphite cannot imbed Roomba-gear plastics. Graphite is too soft and friable to imbed--I would say--any material found in the Roomba. Oops! It could be 'embedded' in the gear grease; and the optical-coupling grease used in the R/C Sensor!

... it's recommend to lube before each use ...You concur then to capillary is subject to gravity ...

In no way do I agree to that! Capillary attraction is due to the *wetting* force between liquid and solid materials.

...you may want to open the gearbox and empty the oil soaked dirt before it does some damage.

Yes, that should be done every five to ten years. ;-|

"...lapping compound wearing out the harder metal-shim" "The shim material is wearing out while the brush hubs show no wear..."
This defies the basic laws of physics, soft material will wear faster than harder material in every case ...

That is true; and your statement reveals I was remiss in making that exact point. Here is a more complete description of the gear-hub / brass-bearing / grit situation.

We have three levels of solid-material hardness in this scenario:
1) very-soft (the gear-hub material);
2) medium-soft (the brass-shim); and
3) hardest (the ambient grit - silica, alumino-silicates, etc.)
These relationships are established by determining how readily one material is able to scratch another.

Then, as Roomba goes about its work, some of the hardest material will wend its way into the oil-lubricated brass-shim / plastic gear-hub gap. Grit, now embedded in the oil, will get smashed between the gear-hub and brass on evey revolution of the gear as the oil+grit mixture passes through the *load-zone* {let me know if you don't grasp that term}.

That "smashing" operation is where the grit has opportunity to imbed either the soft hub, or into the medium-hard brass. Both may occur, but since the path into the softest material is more easily taken, the largest particles of grit will press into the plastic hub.

You now have a gear-hub that is emulating a *drum-sander*! As the gear rotates, the portions of imbedded grit that protrude may then scrape away at the softer brass, so you do have hard-grit scratching medium-hard brass!

Surely you can see that the softest material is now acting as a *hard-grit-carrier*, hence the plastic is basically removed from the wear picture!

Your FYI's are polishing techniques, I don't know why your mentioning this.

I guess I thought that you might learn something. BTW, while the end-goal is to achieve very smooth (highly polished) surfaces, the lapping operations all remove material by the process described above. Material removal is greatest at the start of lapping, where the largest and hardest grit particles are generally used. As lapping progresses, grit size will be reduced; and that requires use of a separate lap, since the beginning lap will still be charged with large particles. Several laps & grit-sizes and materials may be used to finish an optical surface.

[Mavrik]

Yes I did, but I'm sure the physics scholars of the world would interested to know how you changed the laws of physics.

I would agree that a shim would reduce friction because the shim is smooth, as long as it has enough clearance. This is subjected to how much play is in the gearbox gears.

Introducing brass or steel dust into the gearbox is the worse case scenario and will increase wear.

A shim improves alignment and slop, however what one basically has is a stone mill. Dirt settles on the shim and works it's way into the moving hub, subsequently grinding dirt. Add a dirt attractant, oil and increase the amount of dirt to the area. If a dynamic seal was engineered over this area it would increase life.

I don't see how the brush hub is subjected to Roombas weight, my Roombas brush carriage assembly pivots independently from the body.

[Mavrik]

I didn't see you latest post before posting above.

If capillary was relevant, you wouldn't be lubing before each use.

As far as the rest of your post, we are mostly on the same page!!

"Surely you can see that the softest material is now acting as a *hard-grit-carrier*, hence the plastic is basically removed from the wear picture!"

This is an interesting concept, however oil generally keeps dirt suspended. This is what I keep referencing as lapping compound (grit and oil).

[glo69]

Hello Mavrik

Mavrik wrote

Yes I did, but I'm sure the physics scholars of the world would interested to know how you changed the laws of physics.

That's strange. I don't see anything in my statement about changing the laws of physics. I just reported an observable fact. I was a very good electrical troubleshooter before I retired and one of the things I learned was that when an observable fact and electrical theory didn't seem to match I just had to dig deeper into the situation and determine what it was that I wasn't understanding about the situation. Observable facts don't go away by just childishly deciding thay don't fit my conception of what theory says should be happening. I believe that Gordon has very adequately explained what is actually happening in this situation.

Introducing brass or steel dust into the gearbox is the worse case scenario and will increase wear.

That's a strong statement for what is only a guess. I know that some grit, not dirt but grit, is harder than brass because I have seen that something is wearing out my brass shims and I know it isn't the plastic brush hub. Brass ( BTW I've never mentioned steel, that's your addition ) is obviously better to have in the gears than whatever it is that is wearing out my brass shims. Observable fact.

I don't see how the brush hub is subjected to Roombas weight, my Roombas brush carriage assembly pivots independently from the body.

My Roombas brush deck is independant of my Roomba on one end but it is connectd by pivot points on the other end. These pivot points allow rotation of the brush deck but are rigid in the vertival plane. You would have to do a vectoral analysis of the situation, including at least the distance of the brushes from the pivot point, the length and weight of the brush deck, and the friction inherent in the pivot point, to know how much weight is transferred. I'm sure that when the Roomba is new this weight transfer is small to insignificant. However in real life other factors will come into play. One will be the fact that as the Roomba ages the friction in the pivot point will increase, thus adding to the weight transfer. The second point is that the weight transfer is not static in an operating Roomba. As the Roomba bounces, which is especially bad on carpeting, the inertia of the deck and the Roomba itself will cause extra loading, so the weight transfer can be additive. At the same time the brush gears are the most heavily loaded, on carpet, the weight transfer will be at it's worst. Add in an aging Roomba that has worn gears, a worn brush hub bushing, and pivot points that are at their worst for drag and I think this weight transfer could become significant.

One final point. I was going to stay out of the discussion between you and Gordon. I only wrote my entry to reassure people in a simple way that they are still better off using brass shims. However I thought it was perfectly clear that both of Gordon's FYIs were examples of what we have happening in this situation. If you reread his FYIs you will see that both his examples, not of theory but of actual industrial practice, are where a softer material is used, with a hard grit, to lap a much harder material. I can see by your last comment to him that you still do not understand the concept. Dirt suspends in oil, grit imbeds in the softest material.That he is correct is proved both by the fact that his examples were real, not guesses based in an imperfectly understood theory, and by an observable fact. The brass shims are wearing faster than the plastic hubs.

[roddenshaw]

Well, my Roomba now has shims. They spin a little, but for the most part they remain stationary whilst the brushes spin. I made them out of aluminium so it will be interesting to see how well they last.

[Henk Poley]

Someone here mentioned using teflon sprockets/gears. Would it already help to use teflon foil as shims?

[glo69]

I think any material inserted between the hub and the case would help keep the embedded grit from wearing out the Roomba. I'm just not sure how long something as soft as teflon will last.

As for the aluminum, I will be very interested in hearing a report on how well that works.

[Jeff Carver]

Having just read through all (or most) of the discussion here about shimming, I just decided to try it on my Discovery Sage, which is showing considerable wear, especially around the beater bar socket. Following the path of least resistance, I cut up a beer can for aluminum strips. They're just a hair shy of going 360 degrees around the shaft, and I oiled with 3-in-1 oil.

I haven't done a full-up test yet, but on short tests, I encountered two problems. First, the Roomba started running haltingly. I took out the larger shim (which had less clearance, though it still fit easily), and the behavior went away. I trimmed it a bit more and reinstalled, and then it seemed to run okay. But when I opened it up after a short test, I found the shim (again on the larger shaft, which was a little more snug) was working its way out of the opening. That's not good.

Those of you who have more experience with shimming--do you know what I might be doing wrong, that the shim doesn't want to stay in place? Too narrow, maybe? (I didn't measure, but the strips are probably about 2mm. They push in far enough to be flush with the outer surface.)

I'll probably do the ball-bearing mod later, but right now I just want to slow the deterioration.

[glo69]

Hello Jeff

There was a discussion a while back about that problem, but I can't find it right now. The one common denominator though, between your problem and the other one is the use of beer can aluminum for shims.

I siggested at that time that everyone who has become shim proponents are using brass shims. I don't know if the problems have to do with properties of the aluminum, correct thickness, or ?.

Small sheets of brass, usually in 3 thicknesses, such as .002, .005, and .008 are in one package and can be found at most art or craft shops. Use the thickest shim you can get into the opening, and I don't think you will have any more problems. Obviously, if you use the thickest sheet, you will have to avoid the problem of overlapping the shim ends.

If you read toward the end of this thread http://www.robotreviews.com/chat/viewtopic.php?p=23913 you will get a little more information on making shims.

[RoombaHolic]

. . . I'll probably do the ball-bearing mod later, but right now I just want to slow the deterioration.

I'd skip right to the ball-bearing mod if it were me. (In fact . . . it was, and I did!!!)

I haven't had any wear and tear on the gears of my re-manufactured Scheduler since replacing the brush deck and gears and doing the ball-bearing mod.

In fact, I just received the 3rd replacement for the new Pet Scheduler I purchased, (Two of which suffered from stripped gears within the first few uses.), and I'm considering going ahead with the bearing mod on that unit so it won't suffer the same gear failure like two out of three of the others did. I guess that would void the remainder of the 12 month warranty, (About 9 months remaining.), but at the rate it's been going I can't afford to keep shelling out money to return a defective Roomba every couple of weeks or so because of the very poor design of the brush/roller bar gear housing.

[Jeff Carver]

I'd skip right to the ball-bearing mod if it were me.

Well, I'd like to. But right now I don't have the time to spend on it. And my 535 just went back to irobot for replacement, so I need the Sage up and running.

[VFM]

Hello,

When I first shimed my Roomba 550 I used aluminium from a can of Coke. But it did not work very well. Aluminium had a tendency to went out of place. No matter how I cut the shims or if I put bike oil or grease. So I change to brass with good results.

But ball bearing mod is a far better option. (I bought a set of ball bearings from Vic7767). The mod have 3 main advantadges:

1. Stops dirty or dust or hair from entering to gear case.
2. Stops hub wear.
3. No more lubing needed on hub zone. So is easier to clean.

(1) i (2) are in part covered by shiming. But ball bearing is a more efective fix (and more durable). I think that Irobot should install this bearings in a redesigned deck.

Best,
Ventura.