500 series charging error 3 error3

The very latest news and updates for the iRobot Roomba robotic vacuum, the Scooba Robotic Washer and the Dirt Dog workshop sweeper. All discussion and troubleshooting questions go here.

500 series charging error 3 error3

Postby robertbible » October 8th, 2012, 12:07 am

I have read what I can find about this error. Here is my situation. We have and maintain 11 4xxxx roombas.. this is our first 'next generation' unit.

My neighbor gave me her 5xx. It was filled with cat hair. you wouldn't believe how much was packed under the front wheel.

After cleaning, I hooked up the charger and got error 3. I checked the charger, tried another charger, cleaned the battery terminals, cleared the fault etc... It would go amber for a couple of seconds and then pop up the error.
The battery measure about 3 volts.

We bought a new battery. Put it in. It beeped a pleasant tune and when hooked up to the charger proceeded to charge just fine.

We ran the roomba through a complete cleaning cycle and when we hooked it back up to the charger.. we had the error 3 again.

It obviously starts charging, detects too much current or draws the charger voltage down and then gives the error.

Since the new battery had a substantial charge when we first plugged it in, It didn't overdraw the charger..

anyone else have experience with this problem..

I just ordered new MOSFETs and will replace them when they come. It is simpler to change them out then to test them. But, I doubt that is the problem because it looks like it is trying to charge prior to popping up the error and if there was a bad MOSFET why did it work with a partially charged battery pack..

I am not overly impressed with this next generation so far... for instance you need to disassemble to get to the batteries.. and it sure loaded up with dust and cat hair...

Though to be fair the older ones would have stopped running long before you got that much junk accumulated.

robert
robertbible
 
Posts: 3
Joined: October 6th, 2012, 10:42 pm

Re: 500 series charging error 3 error3

Postby Gordon » October 8th, 2012, 3:04 am

robert, you may not fully understand the R3-Roomba Charging FETs Test, and the misunderstanding may be due to how the R3 Service Manual explains Err3. The explanation is both misleading and incomplete.

In the column "What It Means" in the table on page-39 the explanation is stated in two sentences:

R3's Service Manual wrote:Charging FET failed. The occurs when the FET test fails at the beginning of the charge cycle.
The first (quasi) sentence is misleading and incomplete. While the second sentence is truthful by saying it is the TEST that failed!

The first phrase is misleading because it blames a failed FET as the cause of the failed test; and, it is incomplete because the statement fails to take into account any problems with buffer transistors between the FETs and the MCU (which is the item commanding the FETs at all times), or any problem with the MCU itself.

Naturally, when some component in that circuit is not working, the above details are moot because the end result is usually the same -- replace the robot! However, in your case, my point is: replacing an assumed duff FET with a new FET will result in no change when another component may be the faulty culprit!

Now, regarding this notion:
robertbible wrote:I just ordered new MOSFETs and will replace them when they come. It is simpler to change them out then to test them.
Once you know how to power-check installed charging FETs I think you will retract that statement. Here is a link to a procedure by which you may verify that FETs U2 & U4 are controllable on an R2-Roomba's main_PCA: viewtopic.php?f=4&t=13298

You probable have some loose R2 main boards lying about to try that procedure on their FETs. Once familiar with the process it will be easy to adopt it to in situ testing of R3's charging FETs (Q17 & Q18).

If you find that Q17 & Q18 are controllable, it will be necessary to check the buffer xstrs that pull down the gates of Q17 & Q18; and, check other features related to driving those gates to a low voltage.
Gordon
Robot Master
 
Posts: 4304
Joined: April 6th, 2005, 2:02 am
Location: Santa Ynez, CA USA

Re: 500 series charging error 3 error3

Postby robertbible » October 21st, 2012, 6:59 pm

I replaced the fets and it fixed the problem for 4 charge /clean cycles. Then,,, it reappeared. coincidence??

Any suggestions..

does anyone have a circuit schematic.. and is there a self test mode on the 5xxx that you can put it in known charging states, to measure what is going on...

Also it charged the batteries in a 3 hours.. our older roombas always take overnight.






robert
robertbible
 
Posts: 3
Joined: October 6th, 2012, 10:42 pm

Re: 500 series charging error 3 error3

Postby vic7767 » October 21st, 2012, 7:50 pm

robertbible wrote:Any suggestions..


You might consider using the beefier TO-220 MOSFETs. The ones I use are rated at 60 volts 10 amps. I've never has one fail but I have not installed them into a 5XX Roomba, only the Dirt Dog and other 4XXX R2 Roombas.
Roomba and Neato Mods, come visit: http://www.vic7767.com/
User avatar
vic7767
Robot Master
 
Posts: 14944
Joined: January 14th, 2006, 8:31 pm
Location: Haughton Louisiana - USA

Re: 500 series charging error 3 error3

Postby Gordon » October 22nd, 2012, 1:12 am

robertbible wrote:I replaced the fets and it fixed the problem for 4 charge /clean cycles. Then,,, it reappeared. coincidence??
You won't know until you discover the actual cause.
...does anyone have a circuit schematic..
Nothing complete in that area. I did post what I call a Power Forms diagram that shows the high current paths through the charging FETs, but none of the gates' circuitry shows on that diagram. Its pdf is available in this post: viewtopic.php?p=98692#p98692 ; and there was cause to upload a JPG of it in this one: viewtopic.php?p=98797#p98797
...and is there a self test mode on the 5xxx that you can put it in known charging states, to measure what is going on...
Yes. Refer to Appendix A of the 500 Series Service Manual. There are tests that check for three levels of charging current.
Also it charged the batteries in a 3 hours..
That could be a reasonable period of time.
our older roombas always take overnight.
Not much can be learned by that statement, since the Roomba models, their PSU characteristics, or the battery capacities and age are not detailed. The longer time could be due to using the "Standard" R2 PSU with large capacity batteries.
Gordon
Robot Master
 
Posts: 4304
Joined: April 6th, 2005, 2:02 am
Location: Santa Ynez, CA USA

Re: 500 series charging error 3 error3

Postby robertbible » October 22nd, 2012, 2:45 pm

Thanks for the schematic. I guess that I will pull the top and trouble shoot it tonight...

In the self test mode, does it maintain a given charge state?? eg. trickle test stays in trickle.

How is the high charge current regulated if the FETS are always on. Typically you would expect them to be operating as a pwm current source.

Do you know where the current sense amplifier is?

robert
robertbible
 
Posts: 3
Joined: October 6th, 2012, 10:42 pm

Re: 500 series charging error 3 error3

Postby Gordon » October 22nd, 2012, 3:15 pm

robert wrote:Thanks for the schematic.
Welcome, you are.
...In the self test mode, does it maintain a given charge state?? eg. trickle test stays in trickle.
Yes, as long as BiTs are being "Manually stepped", so enter the BiTs mode carefully (else abort if Auto-stepping happens & then begin anew).
How is the high charge current regulated if the FETS are always on.
The external switching mode PSU not only voltage regulates the 22.5Vdc, but it current limits by rolling off voltage if more than 1.25A (NOM) is drawn. I have not run a test on the R3 PSU, but the R2-PSU (only one unit tested) did not exceed 1.4A as its output was loaded enough to reduce voltage way below a healthy battery's expected voltage.
Typically you would expect them to be operating as a pwm current source.
Yes, I made that assumption / mistake in 2004 and held onto it longer that I like to advertise!
But, they are only always ON during high-rate (1.25A, NOM) charging. All other charging states (Recovery, Trickle, ramp-up to high-rate and down to off, and, for the FETs' Test) apply PWM gate drive(s).
Do you know where the current sense amplifier is? ...
Yes. The high-rate control limit is in the external SM-PSU; however, the robot has a master current shunt through which all peripheral loads return their load currents (whether charging {treated as positive current}, or running cleaning peripherals {treated as negative current 'flow'}). That means the MCU and its F/W has the option to take action (warnings, error announcements, or power shut off) when system current exceeds some limit, if it has been trained to do so.

Here is a picture (provided in a post by member red_sword_fish) that shows most top side components of his R3-PSU:
r_s_f-top-sideVw.jpg

If you study the right side of that image (its the low-voltage end of the PCA) you can spot two vertically mounted power resistors -- looks like R12 & R20. Inspection will confirm they are parallel connected to form a current shunt having a few ohms resistance.
Gordon
Robot Master
 
Posts: 4304
Joined: April 6th, 2005, 2:02 am
Location: Santa Ynez, CA USA

Re: 500 series charging error 3 error3

Postby jc777 » May 11th, 2013, 12:37 pm

hi robert,

did you manage to repair the charging error 3 (err3) problem eventually? how did you do it?

i am having charging err3 after plugging in the charger, with or without battery connected.

some background, bought 560 towards end of 07, was using it a few times weekly until early 09. left it in storage until i took it out recently to try charging it.

any idea could it be due to the battery is old? will replacing it with brand new battery work?
jc777
 
Posts: 12
Joined: May 11th, 2013, 11:14 am

Re: 500 series charging error 3 error3

Postby vic7767 » May 11th, 2013, 3:04 pm

Measure the battery voltage. If under 10 volts do not install it into your Roomba. Use an alternate method to charge the battery to at least 12 vdc then install it and test.
Roomba and Neato Mods, come visit: http://www.vic7767.com/
User avatar
vic7767
Robot Master
 
Posts: 14944
Joined: January 14th, 2006, 8:31 pm
Location: Haughton Louisiana - USA

Re: 500 series charging error 3 error3

Postby jc777 » May 13th, 2013, 12:01 pm

What other alternative methods are there to charge the battery? Does replacing MOSFET Q17 & Q18 work?
vic7767 wrote:Measure the battery voltage. If under 10 volts do not install it into your Roomba. Use an alternate method to charge the battery to at least 12 vdc then install it and test.
jc777
 
Posts: 12
Joined: May 11th, 2013, 11:14 am

Re: 500 series charging error 3 error3

Postby vic7767 » May 13th, 2013, 12:07 pm

You can use any NiMH/NiCD smart charger that will support the 14.4vdc configuration.

There may be no need to replace the charging MOSFETs in the newer R3 Roomba due to a better more robust charging system.
Roomba and Neato Mods, come visit: http://www.vic7767.com/
User avatar
vic7767
Robot Master
 
Posts: 14944
Joined: January 14th, 2006, 8:31 pm
Location: Haughton Louisiana - USA

Re: 500 series charging error 3 error3

Postby jc777 » May 13th, 2013, 12:16 pm

Can use the original charger that comes with the roomba to charge? Just connect (+) from charger to (-) of battery and (-) from charger to (+) of battery, can?
vic7767 wrote:You can use any NiMH/NiCD smart charger that will support the 14.4vdc configuration.

There may be no need to replace the charging MOSFETs in the newer R3 Roomba due to a better more robust charging system.
jc777
 
Posts: 12
Joined: May 11th, 2013, 11:14 am

Re: 500 series charging error 3 error3

Postby vic7767 » May 13th, 2013, 2:03 pm

The stock power supply from iRobot is not suggested to be similar to a smart charger and should not be used as such. The charging smarts are contained in the Roomba firmware not the power supply.
Roomba and Neato Mods, come visit: http://www.vic7767.com/
User avatar
vic7767
Robot Master
 
Posts: 14944
Joined: January 14th, 2006, 8:31 pm
Location: Haughton Louisiana - USA

Re: 500 series charging error 3 error3

Postby Gordon » May 13th, 2013, 2:49 pm

jc777 wrote:Can use the original charger that comes with the roomba to charge? Just connect (+) from charger to (-) of battery and (-) from charger to (+) of battery, can? ...
Those connections will probably ruin something! Instead do:
PSU(+)==>Batt(+)
PSU(-)==>Batt(-),
but LIMIT THAT TYPE OF TEST CHARGING TO ~15 MINUTES, then disconnect and measure battery voltage to confirm it has increased.

Until you know what you are doing, don't plan on doing a total battery re-charge by that process.
Gordon
Robot Master
 
Posts: 4304
Joined: April 6th, 2005, 2:02 am
Location: Santa Ynez, CA USA

Re: 500 series charging error 3 error3

Postby jc777 » May 14th, 2013, 1:59 pm

Any example of stand alone charger that can be bought online (e.g. ebay) that can be used to charge the battery directly?
vic7767 wrote:The stock power supply from iRobot is not suggested to be similar to a smart charger and should not be used as such. The charging smarts are contained in the Roomba firmware not the power supply.
jc777
 
Posts: 12
Joined: May 11th, 2013, 11:14 am

Re: 500 series charging error 3 error3

Postby vic7767 » May 14th, 2013, 2:45 pm

Here's the one I use to charge the 4XX, 4XXX, 5XX, 6XX, and 7XX Roomba battery pack:

http://www.onlybatteries.com/showitem.a ... 2&uid=1041
It's a smart charger with a temperature sensor probe.
Roomba and Neato Mods, come visit: http://www.vic7767.com/
User avatar
vic7767
Robot Master
 
Posts: 14944
Joined: January 14th, 2006, 8:31 pm
Location: Haughton Louisiana - USA

Re: 500 series charging error 3 error3

Postby jc777 » May 23rd, 2013, 12:04 pm

by the way, what is the voltage / current that the battery supply to the roomba?

is it safe to connect an adapter with the same ratings to the battery terminals on the roomba to test if the charging error still exist?

as i read from one of gordon's post that the MOSFETs require a certain minimum threshold voltage to 'put up' the state and 'turn it on.

so can safely connect the roomba (or any other with the correct ratings) adapter directly to he battery terminals on roomba to check for charging error?
jc777
 
Posts: 12
Joined: May 11th, 2013, 11:14 am

Re: 500 series charging error 3 error3

Postby Gordon » May 24th, 2013, 2:28 am

jc777 wrote:by the way, what is the voltage / current that the battery supply to the roomba?
Seems to me that you did not proof read your question. I can't answer it as written ("...battery supply to the roomba?"), but if you had asked: "...Charging PSU supply to the roomba?" I could answer by saying:
    a) Current will limit to about 1.4A as a nominally depleted battery begins its recharge, and voltage applied at the battery terminals will be only a fraction of a volt greater than battery voltage.
    b) Battery voltage will increase as charging continues, and charging current will proportionally diminish.
    c) When charging current has reduced to the nominal PSU current rating of 1.25A, the PSU's nominal output voltage rating of 22.5V (R3-Roomba) will be restored (during the (a)-->(b) period that output voltage will have been automatically reduced by the PSU's internal current regulation circuit).
is it safe to connect an adapter with the same ratings to the battery terminals on the roomba to test if the charging error still exist?
In general, NO; and the reason is: You will find it very difficult to locate a 3rd party PSU that emulates the current regulation feature (unstated on the Roomba PSU's label) of the various R2, and R3-Roomba PSUs.
...so can safely connect the roomba (or any other with the correct ratings) adapter directly to he battery terminals on roomba to check for charging error?
No again, but for more reasons. The simplest way to say it is: The Roomba mobo will not be in charging mode, and the charging current (which would be your personal responsibility to hold to reasonable levels) coursing through the battery would not be sensed by anything on the mobo, hence you would not be testing what you set out to test.
Gordon
Robot Master
 
Posts: 4304
Joined: April 6th, 2005, 2:02 am
Location: Santa Ynez, CA USA

Re: 500 series charging error 3 error3

Postby jc777 » May 24th, 2013, 12:42 pm

Gordon wrote:
jc777 wrote:by the way, what is the voltage / current that the battery supply to the roomba?
Seems to me that you did not proof read your question. I can't answer it as written ("...battery supply to the roomba?"), but if you had asked: "...Charging PSU supply to the roomba?" I could answer by saying:
    a) Current will limit to about 1.4A as a nominally depleted battery begins its recharge, and voltage applied at the battery terminals will be only a fraction of a volt greater than battery voltage.
    b) Battery voltage will increase as charging continues, and charging current will proportionally diminish.
    c) When charging current has reduced to the nominal PSU current rating of 1.25A, the PSU's nominal output voltage rating of 22.5V (R3-Roomba) will be restored (during the (a)-->(b) period that output voltage will have been automatically reduced by the PSU's internal current regulation circuit).
No, you might have misunderstood the question. Really means, when the roomba is switch on, what is the voltage / current that the battery supplies.

Had asked this because the roomba was not used for a few years, and currently displaying err3 when plugged the adapter in.

Trying to determine if it is the battery totally flat and spoilt, or the main board (Q17 / Q18 / any other component) is spoilt.

If it is the battery spoilt, do not mind changing to new one. If main board Q17 or Q18 or another other replaceable component spoilt, do not mind replacing as well. However, do not want to change both the battery and main board, which cost almost half the roomba price.

Read somewhere that the battery's condition might affect the initial 'boot-up' testing that would cause err3. Meaning if the voltage across the battery is below a certain threshold, or no battery inserted, err3 will show up.

Now trying to simulate the 'battery' by connecting a power adapter (need not be roomba's adapter), could be a generic adapter, that would supply the right voltage / current to test whether err3 still exist. If err3 does not exist, means it is really battery fault. If err3 still exist, means it is main board fault.

But since you mentioned "I will be very surprised if your 5XX-Roomba has suffered failed charging FETs -- there are several design features that, when combined, make failure of Q17 & Q18 much less likely than we saw for the 4XXX series.", so which other component could be faulty if not Q17 & Q18?

Hope you can understand what I am refering to, and advise on how to repair / revive the roomba. Thank you.
jc777
 
Posts: 12
Joined: May 11th, 2013, 11:14 am

Re: 500 series charging error 3 error3

Postby Gordon » May 24th, 2013, 5:44 pm

jc777 wrote:...by the way, what is the voltage / current that the battery supply to the roomba? ... No, you might have misunderstood the question. Really means, when the roomba is switch on, what is the voltage / current that the battery supplies.
I'll grant you that I did interpret those words correctly, but decided not to respond to either of a couple meanings I could see because this thread pertains to a charging issue, not to a Clean Mode problem.
Had asked this because the roomba was not used for a few years, and currently displaying err3 when plugged the adapter in.
Certainly, that old battery that resided in a Roomba in storage for nearly two years is as dead (i.e., fully depleted of charge) a battery as possible. Within a week or two of disuse that Roomba's battery had discharged enough of its energy into Roomba's circuits to pull is voltage so low as to (likely) not be recoverable by Roomba's built in charging tactics.

As the months of storage racked to over more than a year of inattention the Roomba continued to (lightly) load the battery, thus slowly pulling all 12 battery cells down to zero volts. I don't recall that you ever measured the battery's (no-load) voltage and reported the value. A battery in that extreme state of discharge will express an internal resistance that remains at a very low value, but greater than for a healthy battery. When you decided to simply task the Roomba to recharge its dead battery the unsuspecting charging system entered its FETs' test by applying, say, about 20-volts (via a series of fast ON/OFF pulses) to what appeared to be nearly a dead short to the charging power supply. Therefore none of those pulses could actually rise to as high as '20V' because the current roll-off regulator within the PSU acted each time to reduce the PSU's output voltage enough to limit current through the dead short to less than 1.5A (in the manner described in my previous post). But, here is the 'kicker':

Unbeknownst to you the robot's MCU is powered by the Charging Power Supply, not by the battery to be charged. Therefore if the voltage peak of each attempted FETs'-test pulse is prevented (by the PSU's current regulation) from reaching a satisfactory level the robot's charging control may be impacted by one or more of the following actions:
    1) Loss of Charging Mode lock.
    2) Instability of buck-conversion of V_charge to +5VREG.
    3) Expected pulse currents measured via the system's current shunt resistance exceed the programmed maximum.
    4) MCU becomes unstable, via (1) & (2) requiring re-boots and starting FET-testing over & over.
We are not told (even by the 500 Series Service Manual) about the possibility that any of the actions just outlined may be permitted to pirate the charging-error-3 and express error-3 as a result of a really dead ("flat and spoilt") battery. But I would not be surprised if that turned out to be true.
Trying to determine if it is the battery totally flat and spoilt, or the main board (Q17 / Q18 / any other component) is spoilt.
Do simple tests first, and that means load test an uncertain battery. If the battery is shown to be useable, then migrate to tougher testing. BTW, you have not listed your stock of electronic test equipment (meters, clip leads, power-resistor stock (or incandescent test lamps) elementary stuff).

You begin a battery's load-test by fully charging the battery (charge by Roomba, as possible), or manually charge it in a special bench charging setup. The latter is required for your battery since to avoid polarity reversal of any cell it would be wise to confirm correct polarity of voltage across each cell as charging current is slowly increased.
...Now trying to simulate the 'battery' by connecting a power adapter (need not be roomba's adapter), could be a generic adapter, that would supply the right voltage / current to test whether err3 still exist. If err3 does not exist, means it is really battery fault. If err3 still exist, means it is main board fault. ...
You can't use a power supply to simulate a battery since the charging process would attempt to force current through that PSU in a direction reverse to its normal output current flow!

It would be possible to simulate a Roomba battery by another battery of 12 NiMH (or NiCd) cells. The alternate battery would be one for which there is no expectation of powering Roomba in clean mode. The alternate battery could be a used 4XXX- or 5XX-series battery that accepts charge (passes the FET's check, but has weak cells that hold very little charge). It could be had for shipping costs. Or, you start out by connecting 12 AA or AAA-size NiMH cells in series (somehow find very cheap cells to avoid paying too much). Any alternate test-battery would be (carefully) clip-lead connected to the 5XX battery contacts. Also clip leaded to the robot's thermistor contacts would be a 10k resistor.
Gordon
Robot Master
 
Posts: 4304
Joined: April 6th, 2005, 2:02 am
Location: Santa Ynez, CA USA

Next

Return to iRobot Roomba and Scooba Chat

Who is online

Users browsing this forum: im1kissfan, KeldBroe, mofan, ps0001, RockinRay, someoneclever and 378 guests