Last modified:
The presentation of this page should be looked upon not as an invitation to take apart your Scooba, but rather as a WARNING - YOU MUST EXPECT IRREPARABLE DAMAGE TO PLASTIC THREAD-FORMS AS A RESULT OF REMOVING AND REPLACING THREAD-FORMING SCREWS. THE AUTHOR OF THIS DOCUMENT RECOMMENDS THAT SCOOBA OWNERS AVOID THE TEMPTATION TO DISMANTLE THE ROBOT, UNLESS THERE IS NO OTHER RECOURSE. One reason for that warning is: Experience has shown that the first run of Scoobas is showing a high propensity for stripping plastic screw-threads during re-assembly of the bot! This subject has, therefore, been assigned as the number one section for this page. A second reason for the warning is hardly any opportunities for repairing Scooba have been identified.
After those additional warning-words are given, the simple instructions for separating the upper-chassis from the lower-chassis will be listed. Roger Crier has produced a video of that process which provides perfect accompaniment to all these words. In fact, the video can replace many of these words; however, it can't fully present precautions. For that reason, the link to his video is not given until a propitious moment!
This page concludes with another warning, one that cautions mechanics to have a plan in place for getting the final screw installed!
Do not intermix screws: While this advice is very critical for Roomba's screws, it is less so for the Scooba because thread-pitch is uniform at 1.4-mm per turn, in all positions measured. One needs to guard only against using the wrong diameter and / or length screw where improper.
Endeavor to align the screw's helix with the existing female helix: While putting screws back into Scooba's chassis, a different re-phasing characteristic may be sensed prior to getting the correct engagement, i.e., different than sensed on Roomba's hardware. A high percentage, 80% or more, of screw engagements followed this pattern:
While viewing the assembly portion of Roger's video, watch his handling of the screw-driver, and its back-twirling, as he senses proper engagement of the screw's helix, with the groove in the plastic.
Avoid over-tightening the screws: Perhaps it is unfair to extrapolate findings about a single Scooba to many others, so readers will have to judge for themselves whether the following observation should be applied to all Scoobas within a given production lot. The following data comes from a single specimen which must have been part of the earliest batch of Scoobas delivered to customers. The important observation is this: While loosening the Thread-Forming Screws, (TFS), that secure the upper-, and lower-chassis together, it was immediately noticed that none of the screws were tight ! Inside the Scooba, the results were variable, with the exception that the set of four TFS used to retain the lid over the main-PWB (CPU), were all stripped. Then, as Scooba was put back together, three screw-bosses became stripped -- even though the plan was to reproduce the light torques which the Chinese assemblers had applied! This weak system of screw-fits is presumed endemic throughout the lot(s) of Scoobas produced during 0512xx. Owners of these vintage machines would do well to not exceed screw-torques which were established by the Manufacturing group.
One of the three referent stripped-holes lost its threads because the pilot-hole was too large, and the screw's length could be only long enough to engage one-plus threads of the screw in the pilot-hole! Details of this event have been packaged in a separate file. If interested, take this link: Poorly Implemented Fastener Position
Do not loose any thread-forming screw: The long and the short of it is the TFS used on iRobot's floor cleaning products cannot be replaced by procuring them from any known source.
Prior to removing any chassis-screws described in the REMOVAL section, it would be wise for the mechanic to "torque-calibrate" himself.
If, at the very first time Scooba's screws get removed, it is sensed that the factory assemblers had not used the normal tightening torque (e.g., those torques generally experienced on the Roomba, Floor-Vac robots), it would also be wise for the disassembler to: a) memorize the sensation of screw-release torques, in a general sense, across all same-size-fasteners removed, and b) map each fastener to the hole it came from (so it can be put back in the same place).
Upon re-assembly of Scooba, restore all chassis fasteners to their 'own' screw-bosses, and a) use exemplary effort to phase the screw-threads with female threads before running the screw in more than one turn, b) use finger-tip rotation of a Phillips-bit, sans handle, to run the screw as far as possible, and c) finally, use a driver handle to apply a final torque that is a good estimate of that evaluated in the previous list-element.
Removal of the outer Bumper (basically a decorative shell) is the key to accessing six screws which are normally enveloped by the shell. Only one of the six may be seen while the Bumper is mounted.
Disconnecting the Bumper is easy; only two screws must be removed from points near the trailing ends of the shell. It will be necessary to show these screw attachments by viewing 'after-disassembly' photos; and the reason for that is the installed screws are not clearly seen until a Bumper end is sprung away from the robot's chassis! Keep that point in mind, because the mechanic will find, while attempting re-insertion of the second screw, more than two hands may be needed to maintain that same sprung condition while aligning the screw to the threaded hole and then turning it. A few more words will be given about this in the final section of this page.
Figure 1 shows the ends of the Inner-Bumper's curved casting, where clearance- holes for each screw can be seen. Note: The left-side hole is illuminated via a lamp in that sloping, black-tube (with white wires at its lower end). Another clarification may be needed for Figure 1, it shows the robot right-side up; however, as each of the two screws are being removed and replaced, (R&R), it will be necessary to position the bot upside down. In other words, if loose screws had been left in the bumper-holes, screw-threads would be visible above the bumper's upper-surface.
Figure 1. Inner-Bumper Has Clearance Hole in Each Trailing Tip
Those same 'upper-surfaces' contact the "nut-blocks" which are secured to the Bumper-Shell. Figure 2 (Bumper-Shell inverted) shows the RH-nut-block. Take a moment to consider the mechanical interface that exists between the Inner-Bumper and a nut-block. This is one of the few places where iRobot designers have not provided a self-registering interface -- you know, where two parts can be snapped together, and tend to stay together before any fastener has been inserted. It is that lack of 'fit-up' which makes re-mounting the Bumper- Shell a tough task.
Figure 2. RHS-Nut-Block, Inside Bumper-Shell
The above information should be adequate to locate the Bumper-Shell screws while Scooba is fully assembled; thus the instructions to accomplish Bumper-Shell removal are simply these:
Roger Crier's "Scooba 'How to' Covers off" video
Figure 3 shows the Bumper-Shell separated from the robot.
Figure 3. Bumper-Shell Dismounted
Store the Bumper-Shell away from the work space, and prepare to separate the chassis castings.
Next, right the robot and locate the six fasteners to be removed, by studying Figure 4.
Figure 4. Bumper-less Robot, Front View
Before removing more screws, it will be worth while paying attention to the ribbon-cable disconnect that Roger does at this point (which then minimizes the mechanical-manipulation health-threat to the cable as shown in this document). If you missed it, play the video again...
Ummmm, right! What does that mean? For technicians that work to ESD-Safe conditions everyday, they will know how to safely "ground" themselves and Scooba hardware to keep all at the same ES-potential; but, for most folks who have not been initiated into the ESD-Safe Club, the idea is to keep your body at the same electro-static potential as the electronics you are coming close to, or are getting ready to contact. One way to do that is to first grab onto Scooba's 'minus' battery-contact; and do that every time your body goes through some dynamics, such as leaving and returning to the bench, or turning and reaching for a tool that is a couple feet distant. Which of those battery contacts IS the minus-terminal? Check here, for "power-return" contact.
Don't ignore Reservation-2! Try to avoid excessive handling of the ribbon-cable. Do not bend it, then un-bend it without reason. The conductors may work-harden, suffer "metal-fatigue" and break. Any such break will be difficult to diagnose, even while performing an end-to-end continuity check on each conductor. Play it safe with that ribbon cable!
When ready (edge-connector has been de-mated), remove and store the following Thread-Forming Screws:
When ready, remove and store all Thread-Forming Screws from the bottom of the lower-chassis. Looking at Figure 5, that would be every screw except the 'machine-screw' which secures the black, sort-of-triangular, squeegee-retainer (positioned at the vertex of the leading pair of squeegees).
Figure 5. Underneath the Robot
Nine of the TFS are 8-mm long, and one is 10-mm LG. All are #2 Phillips, pan-head types, of size TFS3x1.4. The 10-mm screw is located about 1.5-inches aft, and 0.6-inch inboard, relative to the machine-screw's position.
The mechanic might find it wise the hold off on removing the screws surrounding the two drive-wheels, until last. Reason: The internal, wheel-drop springs will force apart the two chassis assemblies sooner than one might like.
NOTE: The remainder of this document illustrates an alternate disassembly path, one which preceded Roger's discovery of the edge-connector in the Control Panel assembly. If the connector has been de-mated, then ignore references to "tethered assemblies".
Once the necessary screws have been removed, orient the robot right-side up. As indicated, internal springs will force the two sections apart to some degree. It is then only necessary to work loose the remaining interleaved screw-bosses, to make it possible to lift the upper assembly off. Some insight as to what will happen while doing the lifting, and what should be done while lifting, may be in order.
Figure 6. Scooba Split-Open
Figure 7. Race-way Caps
NOTE: iRobot engineers must have read this section, then made changes on the Scooba production line. Recently manufactured Scoobas have been found with these race-way caps securely bonded to the race-way walls. No more loose caps! When wire bundles must be lifted out of a race-way, any interfering cap(s) must be either broken off, or carefully slotted to a width which permits several wires to be worked through the slot(s).
Following that concept of protecting the ribbon-cable from damage, one might hope that the large mass of the upper-chassis casting could be taken away, so subsequent work on elements in the lower-chassis could be more handily tackled.
And so it is, there are two ways to obtain that relief! One method gets rid of that large casting, but leaves a smaller dongle, that is called the Control Panel Assembly in this document; while the second process leaves only the ribbon-cable in place. Hark back to Figure 4, for a moment. The Control Panel is all of that blue structure at top / front of the robot.
Figure 8. Control-Panel Attachment Points
Figure 9. Ribbon-Cable Disconnect Opportunity
Well now, the robot is opened up for study, or for some sort of repair or maintenance work. Before heading back to the main-page, you should take a peek at the one aggravating aspect of completing re-assembly of Scooba -- see the final section. Until you re-assemble to the point of inserting the last Bumper-Shell screw, all re-assembly work is basically a reversal of the work discussed to this point.
The "warning" part is a little anticlimactic here, in this section, since it must come prior to removing a screw from a spring-loaded assembly. Thus, the warning did get discussed near the very top of this file. As announced there, its importance was somewhat suppressed by the thread-stripping notice, hence repeating it is worthwhile. Bear in mind, the utility in providing an early warning was so the disassembler could conjure a plan for dealing with this awkward assembly step.
Upon first performing the final assembly step, most mechanics will probably be very frustrated by the non-relaxed access to the Bumper-Shell attachment screw. The reason for that 'non-relaxed condition' is: installed screws cannot be clearly seen until a Bumper end is sprung away from the robot's chassis! Hence, to unscrew it, one must have some sight-line to it; and to make that possible, one must deflect the Bumper-Assembly away from its relaxed position. Upon backing the screw out of the nut-block, the previously aligned holes then become misaligned.
Later, while attempting re-insertion of the second screw, more than two hands may be needed to replicate that same sprung condition while also aligning the screw to the nut-block's threaded hole, and then turning the screw! (Roger says he can get that Bumper installed in less than two minutes -- by himself! I need extra hands.) Here are a couple tips:
Figure 10. Stiff-Wax is Useful for Holding Non-Magnetic Screws
This concludes the section about going inside Scooba.
If you wish to return to the main (Scooba Technical Comments) document, click Main
Otherwise you may reverse travel the link that brought you here by pressing the 'Backspace' key.
This page is currently maintained by G. Plews