Scooba's Tank Assembly Details

Last modified:

Features found on and in Scooba's Tank Assembly are reviewed in this document.

  1. For servicing, the Tank-Assembly (TA) totally disconnects from the main-bot.
  2. Topside external features of the TA include: a) the (metal), Handle & Tank-Release at front, b) two, tethered, rubber-stopper Caps at the rear, and c) opaque plastic top and side surfaces, all colored light-gray with blue-tint. [Apologies to iRobot, regarding the lack of color fidelity in these photos.]

    Figure 1-a. Scooba's Tank Assembly -------- Figure 1-b. Tank Flipped Over

  3. As shown in Figure 1-b, the underside is of complex geometry, probably to maximize tank volume and accommodate inlets and outlets (which are further discussed in later list elements). Overall, what is noticed is the rear portion is constructed of the same plastic used on the sides and top, while the forward portion (slightly more than half the area) is a transparent, blue-colored plastic. Four open ports are visible, along with two smallest ports that either have a check-valve, or are a "check-valve". In addition to those air and moisture I/O ports, the Handle / Release is seen front and center, and to the RHS (Scooba's) of the Handle / Release, there is in an inlaid, external part (same plastic as on top) which provides two functions that are described in the next numbered element. One other thing... the TA has its own bar-coded pedigree number, on a bottom-side label.
  4. On the front wall of the TA, just to right of center, there is a light-gray el-shaped fitting (ca., 11 to 16-mm wide x 40-mm tall) which projects a small ledge forward to engage a spring-loaded slider as the Tank is latched into the robot; and that spring-force later serves to slightly elevate the front end of the TA just after pressing the Release.

    Figure 2. Tank Ejector Block And Position-Sensor Magnet

    Roger was perhaps the first owner to notice this de-centered force causes the Tank to not register flush to the robot's right-front, while the left side appears fine. Hopefully, this will explain the cause to any other astute owner that notices this slight imperfection.
    Notice, also, the fitting contains a small bar-magnet. When the TA is latched down, that magnet actuates a reed-switch (sensor), which then reports to the robot's CPU the presence of a Tank Assembly. (The mating reed-SW can be found on the Control-Panel module's PWB, immediately to the left of the serial-I/O connector ).

  5. After a bit of inspection of the TA, it becomes clear that the transparent-tank portion is that which contains the feed cleaning solution -- the Clean Tank; and the opaque fraction, which is essentially 'below' the Clean Tank, collects all materials taken off the floor, is therefore the Dirty Tank. "Clean-Tank" and "Dirty-Tank" will, generally, be abbreviated to CT and DT in the remainder of this document.
  6. Of two obvious fittings that penetrate the CT's wall, one is the feed-port that passes cleaning solution downward to the pumping system in the robot. A spring-loaded check-valve is visible at the top end of this female fitting, and there appears to be (inside the CT) a small filter-screen at one side of the port (see Figure 3). The screen, which is not resolved in the picture presented, is bounded by the square-frame, lower-left of image-center. A rubber, sphincter-seal (of about 5-mm ID) penetrates the tank-wall; but there is apparently a rigid-plastic enclosure / fitting that is used to mount this port & valve assembly to the interior surface, because two screws hold it to the wall.

    Figure 3. Clean-Solution Feed-Coupling To The Robot's Pump

    If the Scooba fails to squirt out cleaning solution, one of the first things to check is whether the screen above this connector is clogged. A method of testing for that condition is given here.

    Note: Figure-4, a bottom-side view of the Tank, has been marked with the list-element (6) and with numbers of the next five elements, to assist with grasping relative sizes and locations of the features discussed.

    Figure 4. Partial Bottom-View With Certain Features Correlated To Text

  7. To the left (Scooba's 'left') of tank-center and close to the forward edge, but still on the underside of the CT, there is a rubber, air-bleed (apparently) valve that also penetrates the CT-wall. Its OD is ca., 10-mm, and port-diameter is roughly 3-mm. Air could pass into the CT via this port, but air or liquid may not exit.
  8. At the TA's LHS, and fore & aft centered, there is a major air-duct. Dimensions are ca., 20 to 30-mm wide x 60-mm long, with long-sides curved. This is the air-outlet from the DT, but it is also the 'inlet' to the Blower in the robot. See footnote 1.
  9. Diametrically opposite the duct described in (8), the "Prep"-collected debris' passes into the DT via an elliptically shaped input port (which has approximate diameters of 15 and 17-mm). See footnote 1.
  10. Inset from the two ports just described, there are left and right oval-shaped intake-ducts that accept dirty water from the plenum of the Squeegee-Vacuum that trails the Main Brush. These two ports are approx. 6 x 18-mm oval openings. See footnote 1.
  11. Dirty-Tank's High-Limit, level-sensor: Located just inboard of the RH-intake from the squeegee, there is a two-contact electrical connector (male-type) which engages a mating receptacle in the robot. It is likely that these tank-contacts go to an 'impedance' type of transducer that can signal the controller when the collect dirty liquid begins to impede air-flow through the DT. Here is a close-up view of the male connector:

    Figure 5. Electrical-Connector To High-Limit Sensor In The Dirty-Tank

  12. CT-volume: In various iRobot documents the capacity of the CT is given as one liter. The CT has been found to accept only about 0.95-liter -- and at that level the CT was so full, the water-meniscus was half way up the stopper's bung-hole! At that level, upon closing the Cap, excess liquid will be squirted on the operator! However, (giving considering the "one-liter" claim) it was noticed that additional volume remained (i.e., an air-space) in the CT. This was noticed by viewing the TA's underside after reorienting the TA to be in its 'lifting orientation' ('Release' at the 12 O'clock position). An air-space was quite noticeable -- possibly amounting to 0.05-liter. To get the idea of quantity, envision the meniscus of the water to be approx. 13 to 15-mm below the center-line (being held horizontal) through the two Latch-Retainer's screws (which make a handy datum). It would seem recommendable that users fill the CT using a total (including the added volume of Clorox-concentrate) of only 0.9-liter. Note: We might later find that the volume quoted by iRobot is actually the theoretical-volume, a value computed by the CAD-program that was used to design the Tank Assembly; and no one ever did a volumetric-test to confirm!
  13. Rear-Window: The DT has a view-port installed in the rear wall of the TA. The level of the collected dirty solution may be seen at any time via that window.
  14. DT-mechanism: The DT has been found to contain some movable items; items which output a clunking sound whenever the tank is inverted. The sounds occur no matter which way the TA is rotated. A little has been learned by probing inside the DT with finger-tips. Crude, imaginary pictures of the mechanical elements have been formed by such probing. There is a rigid-plastic sphere, with diameter about that of a ping-pong ball, but it seems heavier than a ping-pong ball. The ball is permitted some lateral, both fore and aft, freedom (an amount that is roughly half the ball's radius) to travel; and it can move a small distance vertically (i.e., with the TA in its floor-cleaning orientation). Surrounding the ball, there is a hinged-plate, swinging about a horizontal axis, an axis that is parallel to the Main Brush's axis. That plate is perhaps 5-mm thick on edge, and it seems to cage the ball via an integral 'box' that exists above the ball. If one tries to view this item by looking into the DT's dump-port, the scene in Figure 6 shows that only a very small portion may be seen.

    Figure 6. Viewing Into The DT Via It's Dump-Port

    Note that only the aqua-colored edge of the part visible in Figure 6, and one wall of that "box" can be touched, so imagining three more sides, along with a top surface on it, is a bit of a 'stretch'! In any event, some 'testing' with the DT partially filled and totally filled with water, has basically revealed the device to be an 'anti-spill-out valve'.

    The valve's closure is not total, i.e., 'perfect', but it can prevent major spills. Here are the results of testing its action:

    1. Approx. 0.4 liter of water was added to the TA's DT (while the TA was held in roughly a horizontal orientation), and the stopper closed.
    2. The TA was then lifted by the Handle to rotate the TA's disc-shape to be vertical (as shown in the Owner's-Manual). No water spilled from any of the open ducts that are located on what is normally the TA's down-side. Low level clunking could be heard while re-positioning the TA.
    3. The TA was then returned to its former orientation, as in (a), and a small amount of water spilled from two of the small-size intake-ducts ('intake' to the DT), but the flow was brief.
    4. The TA's orientation was restored to the vertical, as in (b). Then the TA was rotated about a horizontal axis, one that was also square to the plane of the disc-like shape of the TA. As the rotation angle exceeded about 30-degrees, some 'dirty-water' (of greater volume than before) spilled from either the largest duct, or from the smaller debris-intake on the opposite side of the TA, depending, of course, on which way the TA was rotated. Here too, the flow was stemmed -- presumably by the clunking-valve.

  15. DT's volume: The Dirty-Tank portion can hold, ca., 1.5 liters (TA held horizontal; and, filled to the stopper-bung's lower-neck transition). No volumetric test has been done, but most likely the high-limit sensor will halt Scooba before dirty solution begins to choke-off air flowing through the DT.

    This concludes the Tank-Assembly discussion.

    Footnotes:
    [1] For clarification, this port does penetrate the wall of the CT, but an interior duct separates the dirty path from the clean-solution; and the duct can be seen to veer toward the middle of the TA (presumably to a connection with the DT).

    To return to the Hydro-Pneumatic-Paths document, click return to Hydro-Pneumatic-Paths

    or, to the main document, click Scooba Technical Comments

    This page is currently maintained by G. Plews