Karcher RC3000: Battery Load Test

[robocleaner]

This is a battery load test for the Karcher RC3000. It is identical to Vic7767's load test for the Roomba battery (see http://www.vic7767.com/how-do-i/).

UPDATED 29 MARCH 2012 - see "CYCLING UPDATE" below

Do not try this yourself unless you're confident in what you're doing - shorting batteries is dangerous


Background to this test

The need for this test arose due to a discussion regarding the impossible length of time Karcher RC3000 batteries seem to last. There are at least 6 members on this forum who have Karchers in excess of 6 years old, that are/were still working on original batteries. My oldest Karcher is between 6-7 years old, and cleans min 80sqm 2-3 times a week. Given that the Karcher returns to base to empty/recharge typically every 20-40 minutes, that's an awful lot of cycles over 6+ years... many, many more than the accepted 500-1000 cycles for a NiMh battery (up to 1200 according to Panasonic).

A member rightly pointed out that, logically, NiMh batteries couldn't possibly last that long... and he's right, in theory at least! So I promised to load-test my 6+ year old ones to see what they were still capable of...

Karcher batteries made in EU by BMZ Germany

Test set-up

The Karcher was fully charged on its own base before the batteries were removed. Karcher uses 2 x 6v NiMh batteries at 1700mAh (10 x 1.2v cells) wired in series to give 12v at 1700mAh total capacity. To make this as accurate as Vic7767's accepted Roomba test, the batteries are wired in series and the load (lamp) is identical to Vic7767's - an MR16 12v lamp drawing 20watts (the Karcher draws 18w, so like the Roomba, its near identical to actual consumption of the robot itself). A digital volt meter is wired in parallel to monitor the voltage. The test cut-off I determined as 10 volts to make the test similar to Vics (although I'd usually cut-off at 0.9v per cell).

Load test under-way

The pliers are holding an insulated holder and not shorting the bulbs terminals!

Test results

At full charge, after 20 minutes rest, the batteries showed a combined total (series) voltage of 14.07v. The test ran until the voltage showed 10.00v. The time elapsed was 43 minutes. In theory, this means that the Karcher is still capable of running up to 43 minutes on 6+ year old batteries! Well, it impressed the heck out of me since my Karcher normally returns to base within 20-30 minutes, presumably to empty rather than charge! I have no logical explanation as to why these batteries can last that long after all these years, but the test clearly shows that they can (and frequently do).


Cycling to (hopefully) improve performance

Samsung quote their Navibot SR8855 battery to have a servicable life of 1 year. After 18 months, my orginal 14.4v/2000mAh Navibot battery was giving a poor 26 minutes maximum run-time. Cycling that battery increased that life to a consistent 67 minutes! So here I'm going to cycle these old Karcher batteries, then repeat Vic7767's load test and update the results - to see if these batteries performance/run-time is/can be improved. As the Karcher would charge these batteries in series, I'm going to cycle them in that manner too although better results would be likely if cycled individually.

Karcher batteries set-up for cycling

CYCLING UPDATE 29 March 2012: Normally when cycling a battery you’d expect to see a progressive increase in the amount of charge being taken into and discharged from the battery pack with each cycle. That didn’t happen here. Over 4 consecutive discharge/charge cycles, these 6+ year old Karcher batteries returned a remarkably consistent 1470-1473mAh over each cycle - quite a high proportion of their original 1700mAh capacity. Repeating vic7767’s load test showed a marginal increase to 51 minutes (the maximum theoretical run-time of the robot). It’s not the dramatically improved performance you’d normally expect and usually obtained when cycling, for example, the Samsung Navibot battery which showed a more than doubling of run-time.

The Karcher does not drain its batteries once its finished cleaning as other robots do, but neither does it "max out" its batteries on every cleaning cycle. These unexpected cycling results lead me to suspect these batteries are indeed being “hyper-cycled” – a theory first suggested on this forum by Spirit Force. This is a methodology used by electrically assisted cars such as the Toyota Prius - frequent cycling in shorter bursts using just a small percentage of the batteries total capacity to prolong life well beyond the norm. Spirit Force has suggested a method to test/measure this theory which I will try, and update results here when I can: It may hopefully help explain why these batteries last so long! But for now I need my charger back to cycle my work tool pack batteries...

[Fraggboy]

WOW!!

Thank you for sharing that with us. I am shocked that the batteries are still giving that much run-time. But, then again. Nothing drains the battery while it's parked. After it's done charging, it parks. Then, when you want to run it again, it then empties out the dust bin (Unnecessary), and charges once more.

I can't wait to see if the 'run-time' increases after the cycling.

[mfortuna]

That is very interesting. One of the things I wonder about is if Karcher charges the packs as two x 6 cells or one x 12 cells. Based on my experience with 6 cell RC packs you can get a lot of life out of them if you don't over discharge cells. Over discharge is a lot easier with 12 cells than with 6 cells.

[robocleaner]

It's 2 x 5 cells Mike, but good question; the charger/base outputs 24v with no load and the bot runs at 12v - it suggests the two 5-cell packs are simply in series. I'd need to open the bot and check the circuit to be sure though.

[mfortuna]

Oops, I assumed it was 12 cells total when I saw something about 14V. Regardless of 2x5 or 2x6 that is very good battery life.

[robocleaner]

For those interested, I have just updated my original post above with the results of cycling the Karcher's batteries.

[Fraggboy]

Thank you for the update. From what I gather and understand (From your information), the Karcher is very efficient using the batteries over a longer period of time, thus the batteries are lasting longer. Am I correct to say that?

[robocleaner]

I think it safe to say that the Karchers battery management is very well considered; the robot itself uses (almost) as much power as a Roomba whilst it's actually running. To me, the results of the cycle test were totally unexpected: These old batteries are in much better shape than anyone could reasonably expect for their age, and even when cycled they had nothing else more to give - it demonstrates to me that how they're being used/charged in daily use is excellent. Spirit Force's suggestion/theory/logic (by PM) is the only plausible explanation as to how the Karcher manages it...

The basis of the theory is here on the Toyota Prius chat forum: http://priuschat.com/forums/knowledge-b ... -life.html. Spirit Force PM'd me some Karcher run-time/capacity based calculations that suggests this could logically make sense. So I've got to test that out, haven't I?

[Fraggboy]

Spirit Force PM'd me some Karcher run-time/capacity based calculations that suggests this could logically make sense. So I've got to test that out, haven't I?

If you have time, sure!!

[TechGuy]

Thank you for the information.

According to the graph on the PriusChat forum, if we want to extend the Roomba battery life span, we should limit each mission to no more than 25 minutes. Stop using lighthouse and give your battery much needed rest.

Roomba stops a mission when the loaded terminal voltage falls below 12V. What is the % depth of discharge when a cell falls below 1V?

[robocleaner]

I don't think the theory quite works like that: It doesn't revolve around a fixed cut-off voltage, but a variable percentage of total battery capacity - the less capacity you use per outing, the longer the cells last. Using this methodology, Toyota offers a 10-year warranty on their NiMH Prius cells. The Karcher's chosen method of operation appears very similar from initial calculations, but I want to test that theory in the real world to either prove or disprove it.

I'm not a Roomba user (although Spirit Force is, and this is his suggestion/theory!) so I'm not familiar with Roomba's battery capacity/usage patterns. All I can say with a reasonable degree of certainty is that my Samsung Navibot equally hammers it's battery to death on each outing, then charges really fast and cooks the battery, and then continues to trash it by simultaneous draining and trickle charging on dock indefinitely thereafter... which is why Samsung only warrant it for 6 months and suggest replacement after 1 year... and why cycling it after 18 months extends the run-time from 26 minutes to 67! The Karcher avoids all of this "bad practice".

I'm not sure you could adapt either the Roomba or Navibot easily to change it's battery usage patterns; all I'm concerned with here is trying to find out why Karcher batteries do seem to last an incredibly long time when others don't. As yet I don't know what the depth of discharge/cut-off voltage of the Karcher is... those are two of the parameters I'll need to measure for the "hyper-cycling" test!

[robocleaner]

Testing Spirit Forces' Hypercycling Theory

I did a more comprehensive response to this yesterday, but it got lost when the server crashed. In short, Spirit Force seems to be right.

The make date of this Karcher is 15th September 2005, so these original under-test batteries are 7 years 2 months old and still running well 2 to 3 times a week (comparable performance to my brand new Karcher). Voltage I measured by cribbing vic7767's Neato volt-meter idea using a lightweight/cheap voltmeter off ebay - accuracy checked against my Fluke, and in my case just blu-tacked to the Karcher since I'll be removing it. It's wired to the Karcher's Service socket VE+/- pins.

Using vic7767's Neato Volt-Meter idea to test Spirit Forces' hyper-cycling theory.

The Karcher charges it's 12 volt pack to 15.5 volts. At 12.4 volts, it keeps cleaning, but hunts the base to empty and recharge - it'll return to base if it crosses the IR beam which projects up to 20 meters from the base... if it doesn't cross the beam, it'll keep on cleaning until it does. At 11.4 volts, it ALWAYS stops cleaning and just hunts the base, which, as the batteries are still highly charged, it can (reportedly) do for an hour or more (Note: the running load equates to a 0.6-volt drop with vacuum and all motors driven, so even in this state, the actual voltage of the cells is still actually 12 volts). Remember, the Karcher will also return to base to empty and recharge when the small 0.2l bin is full, so in practice the point at which the Karcher recharges its' cells is variable, yet the batteries never (hardly ever) fall below their full rated 12-volts (still 1.2 volts per cell).

This contrasts to (for example) the Roomba 14.4 volt pack, which consistently runs until the pack is depleated (1 volt per cell - technically empty). So yes, it does appear that light drain and regular recharging does lengthen the life of these NiMh cells.