Solving Faulty Roomba Button Problem

[THX-1138]

Here is a nice tip from our fellow Forum member, Pierre Collet, that asked to post this for all of us:

Btw, an information that may interest some of you:

On my parents in law's Disco, the CLEAN button was not working any more, which was a pain in the neck to start the test procedure.

So, I removed the screws that were holding the panel, and opened it, just to discover that the buttons actually push on a kind of metallic cup, that plops out when you release the button. Some dirt had sneaked in (the cups are "isolated" from the outside with a bit of sticky tape, that had a fold in it, that allowed dust to sneak in...).

So, the solution to cure dead button is simple: remove the sticky tape, clean the area under the cup, and put the tape back.

A tip for people who want to keep such contact corrosion free: add a tiny teeny droplet of oil for Diesel engines. Diesel engine oil both prevents corrosion, and is also slightly abrasive, which helps to keep nice contacts for a long time.

Hope this helped.

[Gordon]

Attached, you will find a composite image that gives graphic support to P's cleaning process, AND, shows how to seal off the portal through which dust migrates into the switches' chamber.

The top picture shows a portion of the assembly that is directly below the user-interface buttons which you push to make Roomba do its thing.

A lot of dust can be seen around the rectangular array of solder-connections, which are just inside the dust-access portal. Plenty of dust can be seen to have migrated across the white board, and has gathered around the dome-shaped switch elements--some grit appears to have gone under the sticky-tape that holds the elements in position during mechanical assembly. It is *that* tape that needs to be lifted, and discarded to clean each contaminated switch.

New swatches of Mylar, or PVC packing tape can be cut to replace the discarded pieces. Make them somewhat circular, and large enough to seal all the way around their perimeter, and the switches will be happier, longer, than life iRobot provides for them!

The bottom picture shows how you may apply cloth-backed tape (duct, gaffer's, etc.) around the ribbon-cable entry to the switches' chamber to effectively shut off the dirt-flow.

If anyone needs help to R&R Roomba's Cover (Bumper-shell with it), guidance is available at 'mysite', via the sig below).

Continue to the addendum posted on 080217.

[Pierre Collet]

Hi THX-1138, and Gordon,

Thanks for forwarding my "tip" on the correct forum.

Just to be precise, concerning diesel engine oil, I meant "detergent" rather than "abrasive". As far as contacts are concerned, the result is quite the same: it cure permanently most contact problems you may have. Just don't put too much (the easiest way is to drop the tip of an open paper clip in a glass of oil, and just touch the problematic contact with the oily tip to leave on it an ever so tiny droplet of oil).

[davem]

As Arte Johnson would say "very interesting"

Those dome switches look like recycled Atari joystick button technology.
If you actually break one of them, you can buy replacements from the gamers on the net and eBay. Here is a site with pictures, scroll down.
http://www.best-electronics-ca.com/joystick.htm

Replacing the tape to re-center the dome is a common fix.
As Gordon suggests, cover the whole thing well.

In my Radio Shack contact cleaner spray, mineral oil is the contact lubricant.

Dave.

[Guest 1]

Gordon, do you have any pictures on how to take Romba Scheduler apart to clean?
thanks

[Gordon]

Gordon, do you have any pictures on how to take Romba Scheduler apart to clean?
thanks

Follow the instructions given at THIS PAGE, and you should get safely under the hood. Your Scheduler should be of the *Handle on Bumper* construction.

I have not taken time to complete associate web-pp of other invasions, such as into a wheel-assembly, or brush-deck, but have numerous pix of each. If you tell me where you are going, I may be able to post a few key images.

You should also explore http://www.roombaexchange.com/ {REMOVED the comma} where Chris has informative, dynamic explosions of hardware which may be enough to get you there and back.

And, you might also ask the other Chris, Fraggboy that is, about cleaning the optical bump-switches. I know he is chafing at the bit to post his images about that subject.

[ProTech Robotics]

The link above has a "," after it making it invalid so here is the correct link www.roombaexchange.com/roombatd.php

[Gordon]

This post is an addendum to my August 20, 2007 post, above.

The following provides guidance for checking User-Interface (UI) switch performance, cleaning switches as necessary, and sealing them to prolong the next time of cleaning.

Preliminary Check of SW-contacts' Resistance:

If you would like to confirm the necessity of cleaning the User-Interface, UI, switches prior to dismounting the switches assembly, you may do so by using the steps found below under the heading "Verify switching action". When operating properly, closed switch-contacts should show a very low resistance reading, say 0.2 ohm, or less.

Dome-element removal {assumes top Cover has already been dismounted}:

a) Remove three screws (#1 Phillips driver) securing the UI assembly to Roomba's top Cover.

b) Carefully cut loose any hot-melt-glue which bonds the two ribbon cables to the top Cover, then snake the cable through its via so the Cover may be set aside.

c) Peel off each piece of tape which holds down a switch spring contact (dome).
The attached panorama of opened switches represents what will be seen during this step.

d) De-laminate tape from domes. Lifting domes away from the PCB is easy, but since they come away still stuck to the adhesive-tape, caution should be exercised while peeling tape off of the domes--you don't want to distort their basic shape! If necessary, a dip in an acetone bath should help release the tape, and dissolve adhesive residue.

Cleaning the domes:

Use an "acid brush", with bristles slightly shortened to stiffen the bundle, along with clean acetone or IPA (isopropyl alcohol) to flush both sides of each dome.

Cleaning the PCB-contact surfaces:

Use an acid-brush, with bristles cut short to stiffen the bundle, and ONLY clean IPA to flush the areas around each set of SW-contacts. Flush (at least) a 20 mm x 20 mm area around each switch position. Switch-contacts on the PCB should look like this example:



Replacing the domes:

a) Prepare three pieces of transparent packing-tape of size 15 mm square, and one rectangle 10 mm x (15 to 30) mm. Plan to use the large piece over the Power SW-dome (since opposing LEDs require the 10 mm width be fitted between them).

b) As each dome is laid onto the PCB, orient the dome so its two long legs fall into the locating holes provided.

c) At each switch position, fit one edge of the tape in alignment with the silk-screened boundary of the adjacent LED(s), and centered about the dome in the orthogonal direction, then lower the tape around the dome until it contacts the PCB. If the LED overhangs its marked rectangle, slip the tape under the LED.

d) Using one or more sizes of smooth, rounded rod(s), press perimeters of all tape-pieces into firm contact with the PCB, all around each dome.

Sealing secondary dust-leaks:

Each of the dome's guide-holes go clear through the PCB, hence blown-dust may egress through them, directly into a sealed-switch. It is prudent to tape over those eight holes (applying patches of tape to the far side of the PCB).

Sealing tertiary dust-leaks:

Clearance gaps around each push-button provide paths for ambient dust to sift through and reach the array of sealed switches. While this is not viewed as a serious threat, it is very easy during re-assembly to insert a plastic membrane between the user's push-buttons and the white PWB-assembly, as a means to minimize dust reaching the switches' area.

From a roll of product such as GladWrap, rip off a strip at least four inches wide and lay the strip across the white PWB-asm before mating it with the push-button molding and its bezel. Trim off excess film all around the perimeter, then mate that sandwich with Roomba's top Cover so the three retaining screws may be re-installed.

Verify switching action:

Perform switch-continuity tests as follows.

NOTE: This HIROSE-type connector expects pins of nominal diameter 0.5 mm to be inserted. Larger diameters may stretch the sockets and impair future use.

NOTE: "P10" is the plug at the end of the top Cover's ribbon-pigtail which engages J10 on the main_PWB.

NOTE: The numbers within "[ ]" are row and column numbers (as used in matrix notation) which orderly assign contact identification. For details on how to decipher this numbering scheme, refer to this document. See "REFERENCE INFORMATION", at a point about 1/3rd down page.

a) P10[1,8] is common to all four center SW-contacts, so connect one lead of an ohmmeter to UI-connector P10[1,8].

b) P10[1,3] connects to Spot's outer-contact, so connect the second ohmmeter lead to P10[1,3].

c) With ohmmeter powered, press the Spot SW and verify a very low resistance reading, say 0.2 ohm, or less.

d) Move the second ohmmeter-lead to P10[1,4], then press Clean's SW-dome to verify a very low resistance reading.

e) Move the second ohmmeter-lead to P10[2,1], then press Power's SW-dome to verify a very low resistance reading.

f) Finally, move the second ohmmeter-lead to P10[2,3], then press Max's SW-dome to verify a very low resistance reading.

[aqua]

Hi, I like I guess many others had my Discovery 4210 (round) buttons start to fail after only about a month, starting with the MAX button. I opened the machine up but couldnt see what the heck was wrong except the design looked like something of a sinclair zx81 keyboard - not good.

Anyway i tried to find a non surgical way around the problem as my buttons were not corroded, and i found if you press the button on one side (it rocks over kinda like a see saw) and then sweep finger to other side of button whilst pressing like a sweep from left to right or vice versa the button which previously didnt work engages about 90% of the time.

Hope this helps someone as pretty much all of my buttons have failed in this way, you may find occassionally that the 'button sweep' technique will cause the button to register two pushes, but i can live with that.

Aqua

[Gordon]

Hi, I like I guess many others had my Discovery 4210 ... buttons start to fail after only about a month, starting with the MAX button.

The Max switch is closest to blown dust / grit's point-of-entry, so it will be the first switch to be affected. 'Clean' is next, then Spot, and finally the Power-SW goes off line. Two years ago, I don't hink we realized how harmful air-blasting under the bonnet could be--only damage to the Speaker was suspected at that point in history.

I opened the machine up ... my buttons were not corroded, ...

Nor should they have been, since the spring-contact is made from corrosion resistant steel, and the stationary contacts are bright hard gold.

... i found if you press the button on one side (it rocks over kinda like a see saw)... then ... a sweep from left to right or vice versa the button which previously didnt work engages about 90% of the time. ...

Taken to the extreme, you have the elements of a mortar & pestle grinding grit into finer grit, and destroying the hard gold center-contact of each switch.

Once the gold is abraded off, there is (probably) a thin layer of nickel to wear through, and then you are into the copper foil on the circuit-board. The copper can oxidize, or tarnish by reacting with sulfur in the air (resulting in poor electrical contact).

That action is somewhat like flogging a dying horse, to get a little more work out of it.

[Jessi]

Can you tell me how schedule the Roomba http://irobotroombareviews.com/irobot-r ... ning-robot to clean specific rooms on different days?

[vic7767]

That is not an option for the Roomba. You can set a schedule to start cleaning one time per day. Place the Roomba into the room you want to clean.