Neato XV Signature got stuck in charging mode...

News and information about the Neato XV-11 Robotic Vacuum. All discussion and troubleshooting questions go here.

Re: Neato XV Signature got stuck in charging mode...

Postby Gordon » October 8th, 2014, 1:54 am

Lewiy:
Please accept my apology for my delayed response to your posting the data in which I had indicated interest. I was w/o internet connection for half of October 6th, and non-hobby matters occupied the rest of the day, as well as a good part of today.

It was yesterday that I realized I must have had a 'senior moment' on the day I suspected (and wrote about) your five-hours charging plot! I had commented that Neato was charging a fully charged battery (a statement that was solely based on the value of VBattV during high-rate charging). It was "17.4V", which corresponds nicely with a freshly charged 12S-NiMH pack.

My problem, at that time, was my failure to recognize that "VBattV" is not battery voltage while charging current is being forced through the battery. Under that condition, VBattV is voltage applied to battery terminals, a.k.a., E_Q14D, volts.

Anyone making use of Neato's GetChargerData to learn the value of their Neato's battery voltage, should also fetch VExtV, along with VBattV, so they may ensure VExtV ~= 24V (relaxed, thus indicating high-rate charging to be OFF), then, the value of VBattV should reflect battery voltage under light load.

The "light load" is Neato's system elex, which action I garner from devilmike's schematic diagram.

I appreciate both of your new Excel plots. Your "zoomed voltage" graph nicely shows battery voltage ramping upward as high-rate charging progresses; and later, when released from high-rate, we more clearly see discharging of the battery (which I think must be...) due to powering the system between those low-rate charging pulses.

Using data points on the "beginning of charging" graph, I plan to mess with the difference: VExtV minus VBattV, merge the dV(t) with D14+Q14(RdsON) resistance and see whether I come near the Neato-supplied Charger_mA(t).

I do believe that I am getting to like Neato's charging method quite a bit, especially the quasi trickle-charging tactic. Roombas and Scoobas blindly apply fixed level trickle charging current(s) to their batteries; and, without being assured that the trickle rate actually matches the sum of phantom-current plus self-discharge current. Whereas, with Neato (based on what I have so far seen) monitors VBattV & FuelPercent and injects charge on an as needed basis!

======== INFO ==========
On 5-Oct Lewiy wrote:...so when I connected the batteries, it exhibited FuelPercent; 14. It does so whenever the power is restored. Even if the battery is fully charged and not so cool. ...
Are you aware that member "neato" posted some words to the effect that: A Neato on mission is programmed to terminate cleaning when battery capacity has lowered to 15%? I think that number, (15), explains the "14" percentage points to which you refer, because charging starts with the controller "knowing" the battery contains the 15% unused capacity.
Gordon
Robot Master
 
Posts: 4304
Joined: April 6th, 2005, 2:02 am
Location: Santa Ynez, CA USA

Re: Neato XV Signature got stuck in charging mode...

Postby Lewiy » October 8th, 2014, 5:58 am

charging starts with the controller "knowing" the battery contains the 15% unused capacity

Perhaps the thing is that the controller "does not know" FuelPercent, when power is restored. It proceeds from the assumption that the batteries are fully discharged (conventionally FuelPercent is below 20). It can clarify the actual battery level only during the charging process.
In the graph below the batteries were fully charged, rested for a while in the fridge and were connected to the robot. The Controller had measured VBattV; 15,97. And immediately started its high charging. He received signal dT/dt in a few minutes. Only now the controller "knows" (it is absolutely "confident") that FuelPercent; 87. Following it switches to a low current.
Attachments
111.jpg
Lewiy
Robot Master
 
Posts: 152
Joined: September 27th, 2014, 3:33 am

Re: Neato XV Signature got stuck in charging mode...

Postby Lewiy » October 8th, 2014, 8:31 am

Neato ... monitors VBattV & FuelPercent and injects charge on an as needed basis!

Not sure this is so actually. It seems to me that charging batteries during periods of inactivity of the robot is controlled by a timer. After the batteries are fully charged then 'FuelPercent' decreases linearly. When the level comes to 89% (after approximately 3.5 hours), the low current charging switched on. It lasts about an hour, but is interrupted by thermal monitoring (46/39 rule works). The batteries heated to 48-49 degrees several times a day and they never cool below 30 degrees. I think it seriously reduces their lifespan. Too high a price for its continued readiness to work. My first battery packs have become weak after 6 months of operation in this mode, and after 9 months completely worn out. Now, I usually keep my Neato in a cool place (without connecting to base station), in the intervals between cleaning sessions.
Attachments
112.jpg
Maintenance of battery charging for 21 hours
Lewiy
Robot Master
 
Posts: 152
Joined: September 27th, 2014, 3:33 am

Re: Neato XV Signature got stuck in charging mode...

Postby glnc222 » October 8th, 2014, 12:54 pm

The latest software release v3.4 September 2014 may have modified the charging algorithm, described as power management improvements, similar to remarks made on BotVac release. Not usable on oldest units with old system board types with direct charger connection jacks on the vacuum besides the docking contacts.

I cannot quite tell whether the charging method degrades the battery longevity lacking an adequate measure of the optimum cycle lifetime of the cells to begin with. Advertising for those cells often mentions only 250-300 cycle life, which can easily be exhausted by daily operation of the vacuum, especially with multiple charges for automatic repeated runs in large spaces.

Leaving the bot off the dock can be inconvenient if the interval is too long, with a around 100ma drain from the idling system. The Shut Down menu option may be needed. Yet when rebooted, the system always performs a charge.
Maybe all these things were addressed in recent software changes. Screenshot http://www.robotreviews.com/chat/viewtopic.php?p=126619#p126619
improved battery management resulting in improved battery life and longer cleaning cycles in some environments
glnc222
Robot Master
 
Posts: 4943
Joined: January 23rd, 2012, 9:19 pm
Location: North Carolina, U.S.

Re: Neato XV Signature got stuck in charging mode...

Postby Lewiy » October 8th, 2014, 2:19 pm

The Shut Down menu option may be needed. Yet when rebooted, the system always performs a charge.

I do ShutDown after cleaning but don't reboot the system. Self-discharge is about 10-15% for 2-3 days of inactivity.
The latest software release v3.4 September 2014 may have modified the charging algorithm

I have updated the software to v3.4. And I have not noticed any changes in the charging algorithm.
Lewiy
Robot Master
 
Posts: 152
Joined: September 27th, 2014, 3:33 am

Re: Neato XV Signature got stuck in charging mode...

Postby Gordon » October 8th, 2014, 4:16 pm

Lewiy wrote:
Gordon wrote:charging starts with the controller "knowing" the battery contains the 15% unused capacity
Perhaps the thing is that the controller "does not know" FuelPercent, when power is restored. It proceeds from the assumption that the batteries are fully discharged (conventionally FuelPercent is below 20). It can clarify the actual battery level only during the charging process.
In the graph below the batteries were fully charged, rested for a while in the fridge and were connected to the robot. The Controller had measured VBattV; 15,97. And immediately started its high charging. He received signal dT/dt in a few minutes. Only now the controller "knows" (it is absolutely "confident") that FuelPercent; 87. Following it switches to a low current.
Ack & accepted. Having learned more about Neato's charging behavior I have no problem with revising my above quotation, so that it becomes more encompassing of possible events. See how this plays:
As Neato begins re-charge of a battery that it has charged numerous times, the charging controller's F/W expects the battery's capacity to be on the order of 15%, and tags the FuelPercent value at 15% as high-rate charging switches ON.
However, should the battery have been: a) disconnected from the robot, or b) the robot's Power switch was set to OFF; that battery, or a different one, when connected to Neato, or Neato's Power switch is reset to ON, will be considered by the F/W to be an unknown entity that causes F/W to select an alternate means to assay the strange battery's capacity.

Prior to start of high-rate current, the only datum the F/W may use to give a hint of that power source's capacity is VBattV. We know that NiMH VBattV data can be divided into coarse capacity zones, such as:
    * (VBattV >15V) = "ought to run the robot for a while";
    * (VBattV ~ 14V) = "may or may not have operating capacity";
    * (11V < VBattV < 13V) = "depleted more than F/W permits"; and
    * (VBattV < 10V) = "not wise to re-charge it".
Your reconnected battery was assessed to have a VBattV in excess of 15V. Somehow, the F/W converted that voltage to a FuelPercent of about 19%, and charging was authorized for that 'strange battery'.

Seven to eight minutes later, the F/W read VBattV, == my VBattAppliedV, as ~19V, and cut off high-rate charging and immediately switched to low-rate. Based on the quickly acquired, apparent, battery voltage, the controller's F/W assigned an ~87% FuelPercent status to the strange battery.

Low-rate charging continued for about an hour, until a specific (your 46C) temperature was reached. (I think the slope, dT/dt, was not involved here). All charging was cut off.

About 0.7 Ah capacity increase had been made available to the stranger, and Neato's F/W deemed FuelPercent to be at about the 95% level.
Gordon
Robot Master
 
Posts: 4304
Joined: April 6th, 2005, 2:02 am
Location: Santa Ynez, CA USA

Re: Neato XV Signature got stuck in charging mode...

Postby glnc222 » October 8th, 2014, 5:40 pm

The theorized difference between charging an unknown battery and one previously charged, without disconnecting it, is consistent with observations made in alternative Lithium battery research (another thread).

There it was found there can be a problem connecting a nearly full battery the first time, as the over-charging behavior NiMh tolerates cannot be performed on Lithium; the charger can go to a voltage over 21v which damaged one system board component. When an initially empty battery is fully charged, subsequent charges when nearly full did not produce the same problem. It appeared the Neato had remembered something about the battery and modified its procedure.

Some Lithium installations can tolerate the over-charging behavior, depending on what bypass circuits are used to emulate the qualities of NiMh cells.
I have observed high temperatures on cells only when charging a nearly full battery, what I call an over-charging, turning all the current into heat instead of accumulated charge. When charging from a mostly empty condition, temperature did not increase more than 10C from room temperature around 22C.
Neato Robotics described the procedure as detecting a delta-T greater than 1C per minute.
glnc222
Robot Master
 
Posts: 4943
Joined: January 23rd, 2012, 9:19 pm
Location: North Carolina, U.S.

Re: Neato XV Signature got stuck in charging mode...

Postby bab80 » October 8th, 2014, 10:38 pm

okay, back from vacation. :( Had some time tonight to take a part the Neato I hope this is not going to hinder anybody but is not a signature model :help: , due to the fact that the majority of them are the same I'm hoping that's not going to be an issue. If it is I'm sorry but that's my mistake then. The PCB is a Binky Rev 5 dated 1/2/2012. Visual inspection of the board itself on every chip seems to be perfectly fine 0 burn marks or missing solder points. It looks like the board had seen some water damage but it seems to work fine other than not charging correctly. sector d14 and d15 are perfectly fine along with Q 14 & Q 12. I'm a little novice when it comes to voltage meters any idea what they should be put through and of course if that is the only possible problem on D14 where can I get the part? furthermore I have read the forums since my last post and I am super confused as to which direction I should take next. physical hardware/software BTW software has been updated to the latest firmware charging error occurred before update and after update. :thanks: Image
bab80
 
Posts: 13
Joined: September 26th, 2014, 10:56 pm

Re: Neato XV Signature got stuck in charging mode...

Postby glnc222 » October 8th, 2014, 10:51 pm

I cannot analyze boards, but you might want to be very specific and detailed about what a "charging error" is, exactly what was done and what happened, and at what stage of the process. There are many points of failure and every detail can be relevant (as Mr. Holmes would advise; but he's not available).
glnc222
Robot Master
 
Posts: 4943
Joined: January 23rd, 2012, 9:19 pm
Location: North Carolina, U.S.

Re: Neato XV Signature got stuck in charging mode...

Postby bab80 » October 8th, 2014, 11:10 pm

The following is the process of which I get the unit to fully charge: push unit to dock,green light turns on plays its opening song then either throws error code 0003/0004 unit asks to be reset. Pull unit away from dock, press the reset button, and it turns on but charging light indicator now flashing from amber to green. move unit back to dock. unit will act like its charging, and actually does charge to full solid green light. turn unit on,and again to clean. unit will proceed to move away from the dock by itself with solid green light still illuminating and proceeded to go through a cleaning process with no error messages. But if you pull the dirt bin, pick it up or switch the on an off switch while its in its cleaning function. the unit will go back to the home screen and act like its charging again. I would then put the unit back on the dock till I get solid green light and start the cleaning process again. Also if it helps to know at the end of its cleaning process when it's going back to the base it never finishes the dock process unit gets within 2 feet of dock and indicates that it's charging but never plays the I am done song :)also I might add once the indicator light is solid green the unit can be pulled away and put wherever to start a cleaning process or the unit can be shut down through the menu screen process. But if green amber indicator light is on unit cannot be shut down unless switched off at the on off switch "message appears saying cannot shut down while charging"
bab80
 
Posts: 13
Joined: September 26th, 2014, 10:56 pm

Re: Neato XV Signature got stuck in charging mode...

Postby glnc222 » October 9th, 2014, 12:27 am

To get more info on this you could connect a PC via USB and run a repeated Get Charger command as illustrated above when the unit is placed on the dock in the way you obtain the 0003/4 error messages on the LCD.

What I can surmise from the description is the cpu inputs of data from the charging system on battery voltage and current is screwy and highly irregular, so the cpu cannot perform meaningful calculations. The error message to reset means the system detected a failure in which some data it is using is incorrect and a program must be started over.
I had a broken system board which would do this and could never be used to charge, and had to be replaced (I do not do component level repairs, though others do). There could always be a fault in the processor components, memory etc., but what you would look for is faulty components in the sensing components used to feed the processor.
Whether the schematic posted above locates these and how to check them I do not know.
Possibly USB data might point to a particular measurement which is faulty and narrow down points to measure with instruments.

Over USB the system also sends separate error notifications out USB in addition to replies to request commands.
There is no full repair manual published by the company to interpret all the messages.
glnc222
Robot Master
 
Posts: 4943
Joined: January 23rd, 2012, 9:19 pm
Location: North Carolina, U.S.

Re: Neato XV Signature got stuck in charging mode...

Postby Gordon » October 9th, 2014, 12:35 am

bab80 wrote:okay, back from vacation. ... the Neato ... PCB is a Binky Rev 5 dated 1/2/2012.
Perfect! That is the one that matches the partial schematic diagram that Lewiy piped into this thread!
Visual inspection of the board itself on every chip seems to be perfectly fine 0 burn marks or missing solder points.
No burn mark is good, for the reason: Your Charging Base elex may still be healthy. Semiconductors can go duff w/o leaving any change in appearance.
It looks like the board had seen some water damage but it seems to work fine other than not charging correctly.
If apparent wetting of the PCA is not on the port/aft quadrant of the board, let us set that concern aside while investigating health of D14.
... I'm a little novice when it comes to voltage meters ...
I hope your "voltage meter" is part of a Digital Multimeter, because the first meter function you will need is low resistance continuity and / or a diode check function. Measurements will begin with all power sources disconnected from the robot.
Pls tell us about your electrical test-equipment.
... on D14 where can I get the part?
Let's find the duff part first, and then find a source for it.
... furthermore I have read the forums since my last post and I am super confused as to which direction I should take next. ...
Granted, this thread may appear to have gone off topic, but no, it has not. Practically everything given to us by Lewiy will play into being able to answer why you are seeing your Neato do odd things.
Gordon
Robot Master
 
Posts: 4304
Joined: April 6th, 2005, 2:02 am
Location: Santa Ynez, CA USA

Re: Neato XV Signature got stuck in charging mode...

Postby Lewiy » October 9th, 2014, 12:41 am

bab80
It would be nice to measure the resistance of the diode D14 (do you have an ohmmeter?). If the resistance is close to zero in both directions, then you have to replace the diode. It is MBRS540T3G.
Lewiy
Robot Master
 
Posts: 152
Joined: September 27th, 2014, 3:33 am

Re: Neato XV Signature got stuck in charging mode...

Postby Gordon » October 9th, 2014, 1:22 am

bab80 wrote:The following is the process of which I get the unit to fully charge: push unit to dock,green light turns on plays its opening song then either throws error code 0003/0004 unit asks to be reset. Pull unit away from dock, press the reset button, and it turns on but charging light indicator now flashing from amber to green. move unit back to dock. unit will act like its charging, and actually does charge to full solid green light.
As Lewiy correctly pointed out in his earliest posts: a faulty D14 may conduct electricity both ways, assuming it has not fused open circuit. Passing current into the charging control section (Q14, in particular) will occur while Neato receives power from its Base; but, when disconnected from the Base, it is battery voltage that establishes a small level of current flow from P13, through D11 (and through the // parasitic diode inside Q14) up to duff D14, where the current passes backwards (cathode to anode) through it and goes to GND through resistors R101 & R98. Notice the tap point between those two resistors. That tapped signal voltage heads off to MCU-U8, pin 8. When that tapped voltage reaches that U8 device it reports charging voltage is on the Base input connector P19. That is why the system believes the Base to be connected to Neato!
... turn unit on,and again to clean. unit will proceed to move away from the dock by itself with solid green light still illuminating and proceeded to go through a cleaning process with no error messages. But if you pull the dirt bin, pick it up or switch the on an off switch while its in its cleaning function. the unit will go back to the home screen and act like its charging again.
By removing battery power from Neato (via sliding the switch, or by de-mating one of both battery connectors), you have purged all knowledge that Neato had acquired about that battery. When you then close the battery switch, or re-install the same battery, or install another battery into Neato, the robot will commence charging the battery in the mannner shown in one of Lewiy's latest graphs, this one:Image . Neato apparently attempts to determine whether the battery is worthy.
I would then put the unit back on the dock till I get solid green light and start the cleaning process again. Also if it helps to know at the end of its cleaning process when it's going back to the base it never finishes the dock process unit gets within 2 feet of dock and indicates that it's charging but never plays the I am done song :)...
I have no answer for that behavior. Hopefully, it will cease once the charging elex is functioning properly.
also I might add once the indicator light is solid green the unit can be pulled away and put wherever to start a cleaning process or the unit can be shut down through the menu screen process. ...
That is a 'soft-OFF'. The system continues to be powered by the battery. When you open the battery switch, nothing remains powered in the Neato.
------------------------------------------
So, while I was babbling on here, I see Lewiy is up for the day, and has submitted a brief & to the point directive to you! May as well do what he says!
Gordon
Robot Master
 
Posts: 4304
Joined: April 6th, 2005, 2:02 am
Location: Santa Ynez, CA USA

Re: Neato XV Signature got stuck in charging mode...

Postby glnc222 » October 9th, 2014, 2:58 am

D14 part# MBRS540T3G shows distributor stock on Octoparts.com only for large quantity orders. No stock in single units. However, they are listed in small quantities from China on Ebay. Around fifty cents each.
glnc222
Robot Master
 
Posts: 4943
Joined: January 23rd, 2012, 9:19 pm
Location: North Carolina, U.S.

Re: Neato XV Signature got stuck in charging mode...

Postby Gordon » October 10th, 2014, 2:38 am

bab80: Too bad about the S540T3G being N/A at this time.
I hunted for alternates today. http://www.mouser.com normally stocks that rectifier. The site directed me to a page indicating there to be a 24 weeks lead time for that device.
At http://www.digikey, I found a very similar (major spec values are ==) Schottky rectifier, the VISHAY SSC54, that stands a fair chance of acting OK in place of the On-Semiconductor part. DigiKey's stock number is: SSC54-E3/9ATGICT-ND; and price is $0.63 each.When the time comes, you might wish to buy several of them just to have back-ups for soldering mishaps. The SSC54's data-sheet suggests the device will probably run a bit warmer during high-rate charging than does the original S540T3G at position D14.
Gordon
Robot Master
 
Posts: 4304
Joined: April 6th, 2005, 2:02 am
Location: Santa Ynez, CA USA

Re: Neato XV Signature got stuck in charging mode...

Postby bab80 » October 10th, 2014, 10:27 am

yes I know it's kind of crazy. I have not had a chance to hook up my ohms meter cuz of work but I hope to be able to do that sometime tonight. I was getting kind of getting desperate may purchase the 5 lot deal that are on eBay but they seem to be really really really overpriced. I'm going to be smart though and test it first before taking the plunge
bab80
 
Posts: 13
Joined: September 26th, 2014, 10:56 pm

Re: Neato XV Signature got stuck in charging mode...

Postby Gordon » October 10th, 2014, 11:40 am

bab80 wrote:... I was getting kind of getting desperate may purchase the 5 lot deal that are on eBay but they seem to be really really really overpriced. ...
The lot size that I just looked at is 10 pieces! So 29 cents per rectifier is plenty cheap!
That seller's photo shows the Assembly Location as "C", which I think to mean China; and the one Neato used (on the PCA that I have) is also made at the same general location.
The seller also says delivery of the 10 pcs will be two to five weeks distant. Thus, for the minimal expenditure of ~$3+$2 total, you may as well order today.

When you get ready to check D14's resistance, forwards and backwards, be sure to shut off power to the mainPCA via battery disconnect, or battery-switch to OFF; and no active charging connection. Same with "soldering" ops.

Now, regarding soldering/de-soldering of SMD components, have you had any such experience? And, do you have the elements of a decent soldering station at hand? If your answers are NO, then it would be wise for you to google for information to read, and to watch you-tube vids.

Be sure to arm yourself with soldering station equipment that you lack. Do some practice work on any trash elex, with mounted SMDs. The journey will then become more pleasant.
Gordon
Robot Master
 
Posts: 4304
Joined: April 6th, 2005, 2:02 am
Location: Santa Ynez, CA USA

Re: Neato XV Signature got stuck in charging mode...

Postby glnc222 » October 10th, 2014, 3:01 pm

Temperature controlled soldering stations cost more than a new system board (e.g. a v64 board with full LCD plug http://www.ebay.com/itm/Neato-Vacuum-XV-11-Working-Main-PCB-board-model-Binky-64-USED-parts-/111481243801?pt=LH_DefaultDomain_0&hash=item19f4ccc899.

One member used this station which happens to also supply a heat gun for shrink wraps etc., clever http://www.ebay.com/itm/New-YiHUA-2-IN-1-Smart-Hot-Air-Rework-Soldering-Iron-Station-898D-/191271295833?pt=LH_DefaultDomain_0&hash=item2c88a85b59

I have used a cheap 15 watt iron from Radio Shack for PC boards, but am leary of damaging a system board. It may take higher temps to desolder, compared to soldering. A basic technique of surface mount I read is applying separate flux. Large SMD parts can be handled manually fairly easily but some are so small they really require machines, with sub-millimeter lead spacing. Smaller the better making cell phones etc.
glnc222
Robot Master
 
Posts: 4943
Joined: January 23rd, 2012, 9:19 pm
Location: North Carolina, U.S.

Re: Neato XV Signature got stuck in charging mode...

Postby glnc222 » October 10th, 2014, 3:23 pm

charger revisions?

More questions than facts:
I just wonder if this problem has anything to do with the revision made to the charger and system boards, lowering the brick PSU capacity, and adding parts to the little board inside the dock -- to correct the problem more than create it. There were some posts indicating the problem is unique to Euro 220v line power. While the switching power supply brick may supply the same voltage, what if the current limiting properties are different and the higher line voltage supports higher current? Pure speculation, but what if the old charging system relied somehow on current limiting properties of the PSU, and results in some excess current through D14 compared to American 110v line power usage?
glnc222
Robot Master
 
Posts: 4943
Joined: January 23rd, 2012, 9:19 pm
Location: North Carolina, U.S.

PreviousNext

Return to Neato Robotics

Who is online

Users browsing this forum: BrettS, digira, someoneclever, sondejmarek and 384 guests