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 glnc222 » September 29th, 2014, 4:05 pm

charging plug test negative

An unconnected barrel plug was inserted in the on-vacuum jack, with the battery discharged from running and ready to charge. Off the dock. The charging indicator did not start blinking.
Same interrupting charging on the dock, pulling off the vacuum.
A separate jack was checked that the switch is opened inserting the plug.
There could be some question the plug size is wrong but not likely. It does not take much to open the switch.
The plug works to connect the power brick to the dock successfully; whether the dock uses the switch I do not know.

Plug size EIAJ-03, Digi-Key Part No. CP-046H-ND jack, CP-014-ND plug.
Jack datasheet http://www.cui.com/product/resource/pj-046h.pdf

I wonder if this problem was created by the board revisions v64, with the full LCD connector, compared to the older boards with much different layout of large IC's (see previously posted links above).

Also whether the on-vacuum jack was discontinued, covered up, because of this problem -- even if the jack is not the problem, more the incorrectly specified diodes and heat sinking. Maybe they missed the problem because never testing on 220v Euro power systems instead of American 120v A/C. Presumably those switching power supplies are not affected by the difference, as the output voltage is supposed to be the same. That voltage does need to be correct for battery charging. There were also revisions in the dock circuitry and charging software.
Old docks cannot charge newer vacuums, though docks backwards compatible with older vacuums.
Another reason for disabling the on-vacuum jack could be that dock compatibility, where the jack would bypass something needed in the newer setup. Or it just saves parts and board costs, being of little interest to customers.
glnc222
Robot Master
 
Posts: 4541
Joined: January 23rd, 2012, 9:19 pm
Location: North Carolina, U.S.

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

Postby Gordon » September 30th, 2014, 5:37 pm

glnc222 wrote:charging plug test negative
An unconnected barrel plug was inserted in the on-vacuum jack {J3}, with the battery discharged from running and ready to charge. {bot}Off the dock. The charging indicator did not start blinking.
Note: "{...}" demarks Gordon's clarifying inserts.
Why were you expecting a proper 'under-charge' response? With NRInc using J3's switch as a binary indication of charging-power available (at J3), all that is known is: the switch changed state from CLOSED to OPEN! Based on that tidbit of information the onus would be on U8's F/W to output a charging_ON instruction via U8-61, to then cause power FET Q14 to close its channel and connect voltage on the node (D14K, D15K, D11K, et al.) to the R62/R60 divider that feeds U8-15. U8-15 has been sampling +VBAT for some time, so all the F/W must do is note there to be significant increase in node voltage, for it to then decide to invoke a regular (blinking indicator) charging session. If there is no voltage increase, no further action is necessary. Although it would be nice for the robot to display one of the many battery-charging errors codes that it has in store!
...A separate jack was checked that the switch is opened inserting the plug. ...
A good thing to have done.
I wonder if this problem was created by the board revisions v64,...
It is not clear to me just what "problem" you may be pondering. Is it:
    a) The 'burned D14" problem;
    b) bob80's problem; OR...
    c) your "non-blinking" problem?
My current thinking is:
    1) Element (a) does not appear to have been reported in rr's Neato Robotic's forum before Lewiy's recent post. That "burned D14" issue may have been a fluke. I am guessing that way because there should have been damage to other, in-series, components that I see on NRInc's basePCA (built on "PCB Charge Base, Rev-3, #290-0033", ca. 120328). Those in-series components are: DPAK, power P-MOSFET DMP4051LK3 (seems a bit less capable than mobo's Q12 & Q14); power Schottky diode MBRA34OT3-D (which looks to be about the size of mobo's D15 (not the larger D14); and the SMD, R1, current shunt resistance of "R020", (0.020 ohms), 2010 chip. Point being: In the Neato system that suffered burning of D14 on its mainPCA, there should also have been signs of stressed components on the Charging Base's basePCA that had handled essentially the same magnitude of current as felt by D14. Nothing was said about that.
    2) bob80 Neato, element (b): I am willing to believe that J3's switch is only being used to alert Neato's F/W that someone has plugged into the jack, hence no action should be taken by Neato. But, IF D14 is damaged, and can conduct reverse current, a charging voltage coming in from J3 (that exceeds the normal range of +VBAT that is on the node between D14 & D15) would then reverse through duff D14 and reach the R101/R98 voltage divider, to then signal U8-8 port about presence of a valid charging voltage, thus indicating a false charging state (incited by an apparent Base input voltage). It will be interesting to read about bob80's findings.
    3) your 'problem' element (c): could use clarification.
Also whether the on-vacuum jack was discontinued, covered up, because of this problem --
Right! To some extent. I.e., maybe not due to problems in this thread, but related to NRInc wishing to avoid directly connecting its 24Vdc PSU to the mainPCA charging controller. The question is: With that intention in mind, why did NRInc keep mounting J3 hardware on the mainPCA? The jack, D15, and two passives could have been omitted, and the MCU port at U8-36 freed up for other use. What incentive might NRInc have for keeping that hardware on the board?
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 1st, 2014, 12:43 am

That "burned D14" issue may have been a fluke.

Several people have reported on the Russian forum that they have eliminated the problem by replacing the diode D14. Perhaps the problem described here is typical.
why did NRInc keep mounting J3 hardware on the mainPCA?

Probably J3 is used to supply power in the debugging process during manufacturing. Charging the robot through this jack is blocked for users.
Lewiy
Robot Master
 
Posts: 143
Joined: September 27th, 2014, 3:33 am

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

Postby Gordon » October 1st, 2014, 3:16 am

Lewiy wrote:
Gordon wrote:That "burned D14" issue may have been a fluke.

Several people have reported on the Russian forum that they have eliminated the problem by replacing the diode D14.
I accept your input on that matter. Yet I also wonder why they are not reporting failure of a smaller diode, located in the Base-elex, that handles the very same current that burns D14!
Perhaps the problem described here is typical.
Quite right. There can be many instances that users 'solve' by calling NRInc's customer support number, thus those owners don't know what failed, and neither to we!
Gordon wrote:why did NRInc keep mounting J3 hardware on the mainPCA?
Probably J3 is used to supply power in the debugging process during manufacturing. ...
Yes, I can visualize that possibility. Thanks for the inputs, Lewiy.
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 shirgal » October 1st, 2014, 6:59 pm

Hi guys, can confirm i fixed 2 boards with same issue by replacing D14 , didn't have the time to run the bots but charging faults gone while undocked.
User avatar
shirgal
Robot Addict
 
Posts: 155
Joined: July 8th, 2013, 10:54 am

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

Postby Gordon » October 1st, 2014, 9:58 pm

shirgal wrote:Hi guys, can confirm i fixed 2 boards with same issue by replacing D14 , didn't have the time to run the bots but charging faults gone while undocked.
I guess all that I can say is Good Work shirgal!

FYI, I consulted with vic7767 as to whether his robot repair operation has been faced with replacement of duff D14 diodes in Neatos. He said: " I have not run into that particular issue with the Neatos running here nor have any customers reported such.". And that makes me think there has got to be some major difference in charging-elex inside the European Neato Bases (there is a little PCA between the AC adapter and Base contact-bars) that is not controlling current delivered to Neato's main_PCA-P19.

Would you please tell us all that you can about the Base's innards (data on the power brick, and base_PCA's meta-data); and about any 220/240VAC to 120VAC converter that may have been used (in association with D14 failures).

TIA, Gordon
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 2nd, 2014, 12:50 am

I will not vouch for everyone, but the Russians tend to buy the Neato on Ebay from the United States. Therefore they own models VX-11 and VX-21 (for the US market). There are available models VX-12, VX-15 in the European market. I guess they only differ in the power plug. Maybe a defective diode D14 is found in product batch from one seller who sends the Neato outside the US.
Lewiy
Robot Master
 
Posts: 143
Joined: September 27th, 2014, 3:33 am

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

Postby Gordon » October 2nd, 2014, 7:16 pm

Lewiy wrote:... the Russians tend to buy the Neato on Ebay from the United States. Therefore they own models VX-11 and VX-21 (for the US market). There are available models VX-12, VX-15 in the European market. I guess they only differ in the power plug.
You refer to the AC mains-"plug", I presume. I think one germane thing here, in this thread, is the elex module at the other end of the AC power cable. I am talking about the AC to DC adapter (a.k.a., the "charging_PSU", or the "power brick"). NRInc apparently put charging_PSUs having a rated DC output current of "2.5A" in 'docks' for early XV-11 robots. Since then, the Company has reduced that current rating to, first, "2.0A"; and, finally, to 1.67A for the current v64 system.

A second item is the Base's electronics module that is between the brick and Base contact bars.

Did you notice in glnc222's translation post , the Russian forum indicated a measured "3A" being handled by diode D14? Thus, the main question should be: "Why is the Base assembly delivering three amperes current to the robot's P19 interface?". It would be great if the Russian forum members, those that suffered burned D14s, would post meta data for their docks' power bricks, and for their internal base_PSAs (details such as PCB #290-0nnn, REV-#, and DWG#-Ennnnnn). They should also state whether some form of 240VAC to 120VAC converter had been in use (which is unnecessary if using the 1.67A output PSU (fyi, I am ignorant about the earlier ~2A PSUs, other than: all have 24Vdc outputs)).

I think, by looking at devilmike's schematic, it will be clear that the robot's charging controller is not designed to do anything about a too large current! Here, I had better display his schematic diagram:
VWks_devilmike_Markup.PNG
{well, I did add a few things to it, but changed none of its original content.}
After an investigator gets up to speed on what the three D-PAK-cased power MOSFETs do for Neato, and what service the two related voltage sensors may provide to Neato, s/he can come to the conclusion that Neato's high-rate-charging function has no sensible method for moderating an applied 3A, much too high, current!

The unexpected excess power simply begins heating all components in its path. An overheated weak link expires first! Viz., D14.

An interested reader might suggest: "Surely, the current can be reduced by pulse-width modulation, PWM, of Q14's gate voltage!". I would respond to that by saying: Yes, one may do that, and current would reduce; however, I must ask: How will the controller (MCU-U8) 'know' the correct duty cycle to apply? Notice, in devilmike's circuit, there are no means of detecting current. Thus, there is no way to close the feedback loop directing PWM application, hence, PWM is not practical, IMHO.

FYI, thanks to member 'poorredman', I have been studying his duff base_PCA, one that is built on "PCB, CHARGE BASE REV 3 290-0033 03/28/2012", and can claim that the device is designed to measure charging current and (I think) control that current to a specific value by reducing the charging_PSU's 24Vdc input voltage down to a level that slightly exceeds the dynamic battery voltage. If the base_PCA fails to reduce its output voltage, charging current can be expected to be much greater than three amps!
Maybe someone can tell me how renewing a burnt D14 on Neato's main_PCA can overcome the high-current issue (that has to be sourced off the main board) that leads to D14's death. (Heat sinking the new D14 does not reduce the high current, it can only accommodate the high current).
[Edit 141003: Handle of schematic's author IS: "devilmike", WAS: "Neatovtsy"]
Last edited by Gordon on October 3rd, 2014, 11:56 pm, edited 1 time in total.
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 Gordon » October 2nd, 2014, 7:37 pm

shirgal wrote:Hi guys, can confirm i fixed 2 boards with same issue by replacing D14 , didn't have the time to run the bots but charging faults gone while undocked.
Hi shirgal.
See if you can keep a finger pressed against D14 for more than ten seconds after high-rate charging has been going on for more than ~1/4th hour.
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 3rd, 2014, 3:51 am

the Russian forum indicated a measured "3A" being handled by diode D14?

It was not measured. 'devilmike' had suggested that current is 3A. Another member of the forum 'DenRassk' (http://forum.ixbt.com/topic.cgi?id=47:10516:4725#4725) has measured charging currents using an ammeter and oscilloscope. So he got a high-rate current about 2A and low-rate current 0,34A.
Notice, in Neatovtsy's circuit, there are no means of detecting current

The processor "knows" the voltage drop, as well as the resistance between P19 and P13. So it calculates the value of the charging current (and lets us know about it - Charger_mAh in GetCharger). And it adjusts this current using the Q12 and the Q14. As you said - PWM. And we can look at its work on the chart, which was built based on data acquired via USB.
Attachments
Diagram.jpg
Lewiy
Robot Master
 
Posts: 143
Joined: September 27th, 2014, 3:33 am

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

Postby Gordon » October 3rd, 2014, 2:53 pm

Lewiy wrote:...It was not measured. 'devilmike' had suggested that current is 3A. ... 'DenRassk' ... has measured charging currents using an ammeter and oscilloscope. ... got a high-rate current about 2A and low-rate current 0,34A. ...
Thank you for those clarifications.
Gordon wrote:...Notice, in devilmike's circuit, there are no means of detecting current
Lewiy wrote:The processor "knows" the voltage drop, as well as the resistance between P19 and P13. So it calculates the value of the charging current (and lets us know about it - Charger_mAh in GetCharger). And it adjusts this current using the Q12 and the Q14.
The in-line "resistance" between P19 and P13 is primarily the sum of diode D14's dynamic resistance, plus FET Q14's RdsON value while high-rate charging is underway. The resistances of those semiconductors vary non-linearly with current, and with junction temperature. A quick spot-estimate of the total series resistance shows it varies by factor ~seven for I=2A, 25C=<Tj=<100C, and with each deltaE value obtained from respective data sheet plots. Seems to me that use of those junctions as a current-shunt would be a really tough design challenge.
As you said - PWM. And we can look at its work on the chart, which was built based on data acquired via USB.
Thank you very much for posting that data chart. I will write later what I have gleaned from every bit of it! BTW, my initial response is that those data appears to be from a nominal, well behaved Neato robot. Do you happen to know whether that robot was one of the Neatos that had suffered a burnt D14, and then repaired?
[Edit 141003: Handle of schematic's author IS: "devilmike", WAS: "Neatovtsy"]
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 4th, 2014, 3:14 am

those data appears to be from a nominal, well behaved Neato robot.

That's right, the data were obtained from my Neato. And it never suffered a burnt D14.
Lewiy
Robot Master
 
Posts: 143
Joined: September 27th, 2014, 3:33 am

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

Postby Gordon » October 4th, 2014, 3:57 pm

Referring to my mark-up of Lewiy's attached five-functions, GetChargerData chart:
* The robot's charging system appears to being a reasonably nice job.
* I have inserted a "MINUTES" time scale (to which I will refer while discussing events), to supplement the time-of-day abscissa.
* All variables, except "current", use the left-side ordinate. "Charge_mAH", should be considered as "Charge_mA", applicable to the right-side ordinate scale.
* At minutes = 0, we see Neato's temperature, prior to charge-start, was very close to the same value ('zero-C', that is)!
realChargingData141004.PNG

Onward, charging gets going:

Between 0<t<120, mins, I see:
    a) Charging current quickly set up to ~1.86A, continuous. There is no way to tell (other than suggestive 'hash') if Q14's gate is being PWM'ed to hold current to the value reported. "1.86A" falls between the standard 'brick' output currents of 2.0A and 1.67A, so, from data shown I don't know if a 2.0A brick's output is being pulled lower by the battery load, or the Base_PCA has reduced output current, or (less likely) is a 1.67A 'brick' misbehaving?

    b) I see evidence that the base_PCA's circuit has lowered Base output voltage such that it is only slightly (tenths of a volt) above battery-terminal voltage (noted: D14A vs Q14D volts). If the Base-elex is unable to accomplish that 'accommodation', there will be damaging current delivered to Neato's main_PCA elex.

    c) "E_Q14D voltage of "17.4V" suggests the battery was not in need of charging, yet it was high-rate charged for ~2 hours. I assume that excess was induced by being faced with charging a very cold battery! Note that VBAT(+) probably changed by only ~2V over the entire high-rate interval.

    d) Upon reaching the ~120th minute, Neato recognized that battery temperature rate (slope), dT/dt, had met a programmed slope at "~38.3C", so it shut off high-rate current (by, I think, raising Q14G voltage). Notice that 38C is only one degree above mean human body temperature--an extremely comfortable halting temperature, so far as NiMH chemistry is concerned!
At 120 mins, I see the local controller (main_PCA) quickly switched from high-rate to low-rate charging, to (apparently) sneak up on a preferred, higher halting temperature, of "~41.4C". Still, very comfortable! Q12 had been switched ON to provide that low-rate service. I estimated that current level to be "~385 mA" for that finishing event, and for subsequent intervals of low-rate charging. Battery temperature never gets close to even the 50C level. Beautiful!

At ~127 mins, I see battery temperature coasting slightly higher after charging current has halted. The up-coast is typical for all battery packs having only a single point of temperature sensing. Statistically, the hottest cell, or cells, will not be touching the sensor, hence remote cells will send some of their heat content to cooler cells until the battery is in thermal equilibrium.

Following minute ~127, there is no current passing through the battery for ~55 minutes. Temperature droops. Also, the "FuelPercent" datum droops.

At minute ~182, low-rate current is again switched ON.
    a) Battery temperature was between the earlier mentioned control points. Was the turn-on triggered by temperature?

    b) At that time, FuelPercent had fallen to a new low. Did that reduced Fuel Percent value trigger the turn on?

    c) Or, might NRInc have simply specified that a bit of trimming charge be applied following a period of cool down?
Of course, there are insufficient data to answer those queries.

By minute ~295, battery temperature as again fallen into the T_lo, T_hi band shown around minute-120, and may have caused turn-ON of the final pulse of low-rate current, and the mild impact it had on temperature droop. But, notice the wild increase in FuelPercent that appears to be associated! I think there is a bit of "magic" related to deriving "FuelPercent", and some time will pass before the magic becomes known.

Harking back, now, to minute 120: Notice these actions:
    a) Upon being released from high-rate duty, the base_PCA is able to pass into P19 the full 24Vdc brick output voltage; and, its level remains ~constant until minute ~370.

    b) Via close inspection of that light-blue curve, one may see (3) very slight dips in that voltage, each time Q12 is switched ON.

    c) Similarly, in response to the applied voltage, the purple, VBAT(+), curve rises during the Q12-ON interval.
Wonderful data!

That's about all that I can conjur about the chart at this time.

But, my issue with anyone using voltage drops across semiconductors (viz., D14's + Q14's RdsON for the high-rate shunt-resistance, and D14's + R68 + Q12's RdsON for the low-rate shunt-resistance) to determine charging current values remains unaddressed.
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 4th, 2014, 5:48 pm

Was any kind of script or program used to send multiple data requests to the Neato over USB over time?

[edit] improved schematic drawing for system board charging circuit shown above http://www.robotreviews.com/chat/viewtopic.php?p=132686#p132686
Last edited by glnc222 on September 11th, 2015, 6:48 pm, edited 1 time in total.
glnc222
Robot Master
 
Posts: 4541
Joined: January 23rd, 2012, 9:19 pm
Location: North Carolina, U.S.

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

Postby Lewiy » October 5th, 2014, 7:30 am

Was any kind of script or program used to send multiple data requests to the Neato over USB over time?

Here is my command line: "C:\Program Files (x86)\BEL\Realterm\realterm.exe" baud=115200 port=12 TIMESTAMP=4 CAPTURE=C:\Users\Ali\Desktop\NEATO\ChargNeato.txt SENDFILE=C:\Users\Ali\Desktop\NEATO\GetCh.txt SENDDLY=3000 SENDREP=7200 SENDQUIT. The Realterm sends 'GetCh' every 3 seconds and record responses to the text file. After 6 hours, the file is ready. I process it in Excel to get the graphs and statistics of the charge process.
Perhaps some comments to the chart above will be useful. The batteries were fully discharged (each cell separately to 0,9V using MAHA MH-C9000). Then I cooled them in the fridge. This is in order to reduce charging time. The robot did not know these circumstances, 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. High current charging starts after the connection to the base station. It ends when the signal dT/dt arrives. Low charge current follows on. It is paused when the temperature on one of the thermistors is growing up to 46 degrees Celsius, this point is here (I cut superfluous):
"12:09:10";Getcharger
"12:09:13";FuelPercent;88
"12:09:13";BattTempCAvg[0];42
"12:09:13";BattTempCAvg[1];45
"12:09:13";Charger_mAH;383
"12:09:13";
"12:09:13";Getcharger
"12:09:16";FuelPercent;88
"12:09:16";BattTempCAvg[0];42
"12:09:16";BattTempCAvg[1];46
"12:09:16";Charger_mAH;0

and resumed after cooling to 39. Here it is:
"13:03:11";Getcharger
"13:03:14";FuelPercent;85
"13:03:14";BattTempCAvg[0];39
"13:03:14";BattTempCAvg[1];40
"13:03:14";Charger_mAH;0
"13:03:14";
"13:03:14";Getcharger
"13:03:17";FuelPercent;85
"13:03:17";BattTempCAvg[0];39
"13:03:17";BattTempCAvg[1];39
"13:03:17";VBattV;16,80
"13:03:17";VExtV;23,86
"13:03:17";Charger_mAH;0
"13:03:17";
"13:03:17";Getcharger
"13:03:20";FuelPercent;85
"13:03:20";BattTempCAvg[0];39
"13:03:20";BattTempCAvg[1];39
"13:03:20";Charger_mAH;405

The total duration of this phase (net of pauses) is always for about an hour. Last steps:
"15:06:46";Getcharger
"15:06:49";FuelPercent;89
"15:06:49";BattTempCAvg[0];39
"15:06:49";BattTempCAvg[1];37
"15:06:49";Charger_mAH;376
"15:06:49";
"15:06:49";Getcharger
"15:06:52";FuelPercent;100
"15:06:52";BattTempCAvg[0];39
"15:06:52";BattTempCAvg[1];37
"15:06:52";Charger_mAH;373
"15:06:52";
"15:06:52";Getcharger
"15:06:55";FuelPercent;99
"15:06:55";BattTempCAvg[0];39
"15:06:55";BattTempCAvg[1];37
"15:06:55";Charger_mAH;0

From this moment the robot considers that the battery is full.
Lewiy
Robot Master
 
Posts: 143
Joined: September 27th, 2014, 3:33 am

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

Postby glnc222 » October 6th, 2014, 1:19 am

Very Informative. I was not aware terminal emulator programs had a repetition feature.
Realterm is open source at SourceForge.net, major resource. http://realterm.sourceforge.net/
I had not seen this one mentioned before, but am not into all the Windows software.

Extracting data from text files and formatting for input into other programs is not too difficult with the right string functions, found even in BASIC. With only one job easier just to do it manually.
I haven't looked for the best graphics program though. One spreadsheet was not too impressive.
[edit] Actually, you might be able to read that Neato output with some options in spreadsheet input functions.
glnc222
Robot Master
 
Posts: 4541
Joined: January 23rd, 2012, 9:19 pm
Location: North Carolina, U.S.

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

Postby Gordon » October 6th, 2014, 2:10 am

Really nice work, Lewiy, regarding fetching charging data and doing the MSExcel plot-ops! I'm guessing that a few members are right now transforming your command line to run on their systems so they can fetch their own Neato's charging data. Time will tell for that to pan out.

I'd also like to repeat thanking you for introducing us to devilmike's schematic diagram for Neato's v64main_PCA charging-controller, and battery-power distribution FET switch. Collectively, your posts contain critical information that this particular (stuck in charging) thread requires. I think that rr is in a very much improved position to address the subject matter.

I have a few items, relative to your today's introductory comments, but I will limit to the subject of VBattV and VExtV data for this post.
on 141005 Lewiy wrote:...The batteries were fully discharged (each cell separately to 0,9V ...
Yes, quite a nice starting point. If those 12 cells had each retained their 0.9V discharge level the 12S battery would have exhibited 10.8V terminal voltage. Yet in the compressed representation of the few tens of seconds following time 10:02:38 on your graph, I am only able to read VBatV as "~16V". My question is: When you look in your raw data txt-file. did VBattV quickly ramp up to about the 16-volts level in the first few minutes of charging? I'm well aware of NiMH cell voltage quickly rebounding from a depleted-charge level, but typically the rebound level for 12S NiMH fails to rise much above ~14V; and it is that background that prompted me write this about (the 0->120mins) VBattV:
Gordon wrote:"E_Q14D voltage of "17.4V" suggests the battery was not in need of charging, yet it was high-rate charged for ~2 hours. ...
I'm also curious about the differential voltage between VBattV and VExtV over the beginning several minutes of high-rate charging. As you have said, that will be the continuing basis upon which Charger_mA is computed.
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 6th, 2014, 4:27 am

Yet in the compressed representation of the few tens of seconds following time 10:02:38 on your graph, I am only able to read VBatV as "~16V"

Here is the the beginning of charging on a large scale,
And a quote from a data file:
10:03:46 Getcharger
10:03:49 VBattV 13,19
10:03:49 VExtV 0
10:03:49 Charger_mAH 0
10:03:49
10:03:49 Getcharger
10:03:52 VBattV 13,15
10:03:52 VExtV 23,35
10:03:52 Charger_mAH 0
10:03:52
10:03:52 Getcharger
10:03:55 VBattV 15,31
10:03:55 VExtV 15,65
10:03:55 Charger_mAH 1930
Attachments
111.jpg
Lewiy
Robot Master
 
Posts: 143
Joined: September 27th, 2014, 3:33 am

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

Postby Lewiy » October 6th, 2014, 4:33 am

I'm also curious about the differential voltage between VBattV and VExtV over the beginning several minutes of high-rate charging

I zoomed voltage here:
Attachments
112.jpg
Lewiy
Robot Master
 
Posts: 143
Joined: September 27th, 2014, 3:33 am

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

Postby glnc222 » October 6th, 2014, 1:05 pm

Additional charging profile data was graphed in the lithium battery thread using external data logging instruments.
A 24 hour profile of battery terminal voltage and current
http://www.robotreviews.com/chat/viewtopic.php?p=125034#p125034

The charging profile of lithium ion batteries with heat generated in balancing circuits instead of in the cells (which do not get hot with that chemistry)
http://www.robotreviews.com/chat/viewtopic.php?p=124053#p124053

Old system board operation pre-V64.
glnc222
Robot Master
 
Posts: 4541
Joined: January 23rd, 2012, 9:19 pm
Location: North Carolina, U.S.

PreviousNext

Return to Neato Robotics

Who is online

Users browsing this forum: Brett, CherkyB, Franzl, moradi2, susantx3, Wing Nut and 340 guests