Roomba 630 no lights / dead

[ifcho]

Hello,

Today I took my Roomba 630 from the post office and I thing that I almost immediately killed it.

The Roomba is imported from the US (and I am in Europe), so it came with a 110v charger.

I thought that it would be safe to use a generic charger, but I was wrong.
The generic charger is 24v/1A output, but the multimeter shows 50V without load :-/

So while the Roomba was on the base both lights (home base and roomba) were blinking , along with a warning sign. I think my next action killed the roomba, as I plugged the charger directly to it.
After a while it just turned off all its lights.

I removed the battery a couple times (it gives out 15.3V), and tried to reset the robot, but nothing.

Do you have any guidelines or suggestions what have I burned and how to proceed?

The 630 looks a lot like the 555 model, has anyone disassembled one? Do they share parts?

[Gordon]

Hello, Today I took my Roomba 630 from the post office and I ... immediately killed it.

Yes you did!

...The generic charger I took, is 24v/1A output, but apparently, when I measured it with a multimeter it gives out 50V without load :-/

That is typical for inexpensive wall-wart power adapters that lack any type (voltage or current) regulation. The first Roombas (2XXX & 3XXX) were built to use a like power adapter having 24Vdc output when delivering 0.5A current. With current = zero, the output voltage was about 28 volts. Years ago I experimented with that 24V,0.5A unit by powering it with a variable voltage transformer, and with its dc-volts output fed straight to a Roomba battery (12S cells as in your 6XX battery) through an ammeter with 20A full scale range. As I ramped up AC voltage, heading to 120VAC, I witnessed current quickly exceed the adapter's 0.5A operating point; and upon raising input voltage even higher, yet far below the 120VAC goal, I saw output current climb up to several amperes (and quickly reduced the AC power)!

Your 630 probably experienced several amps of fault current(s) from two loads spaced apart in time. Your suspected TVS shunts will have reacted to the 50-volts surge, since they begin conducting around 30-volts. That would have been surge #1. A second surge will have occurred (and act to relieve TVS heating because adapter voltage will have been pulled lower than 30V) if Roomba actually started the charging operation.

...while the Roomba was on the home base both lights (home base and roomba) were blinking , along with a warning sign. I think my next action killed the roomba, as I moved the charger directly to it, it was still blinking with the warning sign on its display. After a while it just turned off all its lights.

That can be explained for charging mode if fuse F2 tripped open, resulting in de-powering the Roomba controller. And the likely thing to make that happen is for Roomba to have actually entered charging mode, which would have been assisted by two TVS diodes (acting in //) to load the troublesome adapter's output down to some level around 30V. Then, IF charging became authorized by the controller (thus switching into conduction the two charging FETs, a very high charging current would have flowed through Fuse F2. Current would have been on the order of: I_chrg ~ (30 - 15.8) / 0.1 = "142" amps! Nah, we know that can't happen, instead we have to presume a high enough charging current momentarily settles in to pull adapter-output down closer to battery voltage (plus the voltage drop across the battery's ESR -- which voltage drop I assigned as 0.5V (that was added above to the 15.3V battery level)).

Let's say that 24V, 1A adapter can output five amps with its dc-voltage lowered (due to its own internal impedance) loaded by the high battery charging current that results. Then we may estimate the batt's ESR, to see if we're in the ball park: ESR ~ dE_esr / I_chrg = 0.5V / 5A = 0.1 ohms; and that, to me, suggests the brief duration charging current could have been on the order of several to five amperes.

I don't have data for fuse F2 (P/N: R250,7167S) found on 5XX mobos, but do have data for a similar device (RTI Elex, Inc. #Lp60-250). Its datasheet claims the fuse will HOLD at 2.5A, 25C temperature, and a graph shows it TRIPs in about 15s when current is 5A.

I removed the battery a couple times (it gives out 15.3V), and tried to reset the robot, but still nothing.

Three things here:
1) Keep that battery disconnected from the 630, until you wish to connect it for testing the robot.
2) Did you try to put Roomba in Clean Mode, say, an hour after the 'accident"? The battery also contains a polyfuse (poly-switch) that might have tripped open during the possible high-current surge. vic7767 might be able to tell us about the 'fuse' that is built into a 5XX/6XX/7XX battery assembly.
3) Do periodically monitor battery voltage. When you find terminal voltage is approaching 13 to 12 volts, you should recharge it to an open circuit voltage in the range ~15 to 16 volts, and then begin the periodic checking all over again. You could easily fashion a make-shift charger out of that 24V, 1A adapter. Just figure out how you might introduce some series resistance (ballast) across which excess voltage can be dropped. Plan on starting at one ampere, and allow current to tail off naturally as battery voltage rises. But, you must be Johnny on the spot to interrupt charging as your selected finishing voltage approaches.

Do you have any guidelines or suggestions what could have happened?

Well, I have already done that above, to an introductory level. However, it is important for you to know that I have never touched a model 6XX Roomba, therefore anything I say about the 6XX circuitry is based on the assumption that 6XX circuits are identical to those in 5XX, "R3" Roombas. I have read about that consistency ( 6XX versus 5XX PCA) on this rr board.

I found a schematic sheet of the power board of an older Roomba model, but probably, much has changed since then.

Yes, generally changed for the better, and different enough that R2's schematic1 is of no value to your problem.

I see that blown up TVS circuits are common cause for such issues. Do the recent Roomba models still use this type of circuits, and do you know what would be the P/N for replacements?

Yes, iRobot continues to connect TVS devices with no means to protect the device from dc current (not that they are specified to snub dc power). TVS devices are carefully specified to absorb energy contained in a specific shape pulse-current vs. time curve.

TVS,SMD,ONsemi,1SMB26CAT3,bi-dir.common-K describes the TVS devices I have seen on some R3 mobos (main_PCAs).

Here are some reading materials for you:
* Read several pertinent posts on this page: viewtopic.php?p=81508
* Download and study the pdf diagram attachment in this post: viewtopic.php?p=98692#p98692
------------------------------------------------
After posting, the possibility of additional component damage came to me --
There is a master current shunt (very low value resistance) through which system operating current passes. The shunt is composed of two 0.1 ohm resistors // connected to give a nominal 0.05 ohm shunt resistance. I believe that if my alluded tripping of F2 required 15 seconds to switch open there could have been subsequent heating of the shunt's power resistors, with five amperes passing through the pair, that might have damaged them. Checking their // resistance for higher than nominal value is easy to do. This is what they (R235 & R257) look like, and where they are: