Mik's blog

Mik's picture

Auxiliary Battery Cooling

The writer of this blog accepts no responsibility or liability resulting from attempts to repeat or perform the procedures described in this thread. This information may be used at your own risk. Click for more details.
.
.

I have been working on additional battery cooling systems to extend battery life for my Vectux.

Keeping NiMH batteries cool is one of the most important parts of looking after them.

There is every chance that charging efficiency, range and battery longevity will be positively influenced by removal of heat from the batteries when they are more than 25°C warm.

.........................................................................................
The Auxiliary Battery Cooling System (ABCool) consists of the standard Vectrix Battery Cooling Fans inside the battery housing and additional 240V-->12V power supply with it's own cooling fan and a double timer with (supposed) surge protection.
Some pictures of the fans inside the battery cover: (Click image to enlarge)

.

Relays are used to ensure that one, and only one 12V supply powers the fans at a time.
It plugs into the connector under the front seat and allows the original charger electronics to function completely unchanged when the ABCool is not plugged into the grid.
Additional cooling periods before and after charging can be programmed via one of the two timers; the second timer turns the on board charger on and off.
...................................................................................
Follow the links to more detailed descriptions of the device and the background.
Click on photos to enlarge them.

...................................................................................
...................................................................................
I the process of observing the latest "Recharge after BaLPoR" I discovered a design fault in the ABCool 12V power supply which causes two potential problems.

The design fault is not only easily fixed, but also makes it half as complicated and cheaper to build. KISS!

Here is a schematic of the inferior design, which is currently installed in the Vectux:
Photobucket
This design is unnecessarily complicated, but works well in winter....
.

And this is the new, simpler and better design (but not built or tested, yet...):
Photobucket

The first problem
with the first design is that the Vectux charger activates the DTDP relay, which then switches the 4PDT relay on, so that the Vectux charger electronics provide power to the fans, no matter if the ABCool is plugged in, or not. This can cause unnecessary battery depletion in very hot conditions.
The second problem is that if one of the two relays fail to switch and the ABCool is not on, then the charger would run without cooling fans. This would not cause immediate severe damage, because once the battery temperature climbs to 46°C the charger will pause. This was involuntarily tested by siai47.
.
The new, improved setup requires only one relay.
The Vectux charger is supplying the fans with 12V by default, even if the relay becomes faulty and fails to switch years later.
When everything is still working and the Auxiliary power supply is turned on, then the relay switches so that the auxiliary power source supplies the fans with 12V power, no matter if the Vectux charger is also trying to run the fans, or not. So mains power is used whenever available for cooling the batteries.
.
.
The new design has 2 advantages over the current design:

Advantage 1): If any of the relays fail to switch, then the Vectux will function like a normal Vectrix. Currently a relay failure would cause loss of battery cooling during charging (if the ABCool is not plugged in) because the "default" position of the relays is to have the Auxiliary 12V supply switched on.

Advantage 2): Even in very hot conditions (battery and/or ambient temperature more than about 35°C (= the trigger temperature for the switching on of the battery cooling fans by the Vectux electronics), the power for the fans will be provided by the ABCool when it is connected to the mains power supply.
.
.
With the first design the following scenarios are possible:
Scenario 1):
The Vectux battery is hot after riding in 35°C ambient temperature; the battery fans will therefore run spontaneously when the Vectux is being parked. Even if I plug in the first design ABCool, the power for the fans will continue to be provided by the battery, not by the ABCool. This leads to further depletion of the battery and further heating of the battery. Not much, but totally unnecessarily.

Scenario 2): After charging has finished in a hot environment the battery will be so hot that the fans remain switched on by the Vectux electronics, using power from the hot battery, even if the ABCool would otherwise run the fans by using the grid without depleting and heating the freshly charged battery.
.
.
In summary (it's really the same stuff for a forth time, sorry to be a bit repetitive here...):

The first ABCool design will not allow additional cooling to be powered from the grid at those times when the Vectux electronics have decided that battery cooling is needed.
These are of course the times when cooling is needed the most and for the longest periods, and the power for this will come out of the battery, not the grid.

All I need to do to "upgrade" to the new version is solder connectors to a new 4PDT relay, and click it in to replace the double-relay in this photo:

Mr. Mik

Mik's picture

Vectux = No Warranty, No Worries Vectrix!

The writer of this blog accepts no responsibility or liability resulting from attempts to repeat or perform the procedures described in this blog. This information may be used at your own risk. Click for more details.

.....................................
.....................................

As of today my scooter has done more Km's since I renamed it Vectux than as a Vectrix!
Photobucket Click here for a higher resolution picture.
And there has been a lot less trouble since then.
That makes this penguin very happy !
Photobucket

4999km total. (Plus about 1400km on the first Vectrix.)

2511km as Vectux, since the warranty was voided by Vectrix Australia.

2176km since the second, proper repair to the motor controller board.

1894km since I installed a Bussman fuse to replace the third blown Littelfuse. That is the longest distance I have ever covered on a Vectrix with one fuse....
I do not believe that riding up steep hills with a low battery had anything to do with the fuse failures as suggested by Vectrix Australia. The RETAMPI knows better...

............................

The RETAMPI continues to work fine, no problems with water entry so far.
It still displays the percentage of max. amps going in and out of the battery, peaking at 100 if the battery is full and warm, between 65km/h and 75km/h.

............................

The ABCool 12V power supply works beautifully and is into it's sixth version of charging / cooling pattern:

This pattern is suitable for when ambient temperatures are fluctuating daily between 10°C- 25°C, which happens in winter on the Australian Gold Coast. It does get a bit colder, but by the time I ride off to work it's usually above 10°C again.
It covers demanding, hilly and fast commuting, 20km x 2, including always charging at work.
I would not be able to get to work and back without a deep cycle of a fully charged battery or charging at work.

Timer 1 = Vectux charger:
2000 to 2130 Mo - Su
0900 to 1000 Mo - Fri

Timer 2 = AuxiliaryBatteryCooling:
1630 to 2000 Mo - Su
2130 to 1000 Mo - Su
0750 to 0900 Mo - Fri
1000 to 1100 Mo - Fri

These settings are supposed to achieve this:

A) Charge level remains between 25% and 90% most of the time.
B) It avoids unnecessary battery aging (due to avoidance of heating up during the later parts of charging and due to avoidance of deep cycling.)
C) Aims at stopping charging when "CP" has just finished.
D) Battery is cooled to about 20°C to 24°C before charging begins.
E) Battery is cooled after the charger has been turned off.
F) Energy wastage due to the chargers standby current consumption of 0.5A is avoided.
G) It might also increase the life span of the charger because the charger heats up if left plugged in once charging is complete, particularly in countries with higher grid voltages.
H) It reduces overall power consumption and increases the "Km/kWh" rate, because the later parts of the charging cycle, which are less efficient and produce a lot of heat rather than a lot of usable battery charge levels, are avoided.

A full recharge including "CC" and "tr" parts of charging to equalize cells will probably be done once every week or second week, when the full range is needed for some reason. I simply plug the Vectux "straight" into the grid instead of through the ABCool timers.

A BaLPoR will probably happen every few months, more or less when it is unavoidable.


I have only been testing this charging pattern for a few days so far, it might hold unexpected surprises....

Small changes to the charging times, to keep the battery levels within the wanted range, will probably be needed.
And the cooling times will need to go up once the weather gets warmer again.

.................
.................

The Vectux had new tyres fitted today.

The rear tyre could have gone for another 500-1000km before it would have become illegal, the front tyre a little longer, maybe 2000km. So I kept them, good spares for emergencies.
Whilst they wait for their comeback I'll grow fruit trees in them!

Photobucket

I replaced them with the same Pirelly tyres as the originals, about AU$140.- fitted.

This is how the Vectux was supported to get the front wheel out:
Photobucket Click here for a higher resolution picture.
....
Photobucket Click here for a higher resolution picture.

The front wheel is fixed to the fork in a standard way. It required 10g lead to balance, the same amount in the same spot as originally. That's good according to the tyre fitter.

The rear wheel, however, is a problem insofar that it does not fit onto a wheel balancing machine.
A custom Vectrix part would be needed to balance the wheel. This part would need to be perfectly balanced itself, and needs to attach perfectly centered to some or all of the six holes in the inner rim of the wheel. Not an easy feat to make that yourself!
So far I have not noticed any undue vibrations or wobble.
I made sure that the tyre fitter installed the wheel in the exact same position on the gearbox as before, by marking them. I was a bit worried that any small change could worsen the gearbox noise again, but so far it remained acceptable.

..........
..........

So far unsolved problems:

Photobucket

It remains to be seen what causes the "Disappearing 5 bars Syndrome".
A detailed computerised battery analysis is in progress and reported here.
So far it has shown that there are multiple cells with reduced capacity, which have also bulged. More about the uneven distribution here.

Almost all the low capacity cells are located in the lowest layer of 3 layers of cells. I believe the reduced capacity is due to radiant heat from hot road surfaces being transmitted to the lower cells.
I will probably paint the underside of the battery compartment with heat reflecting paint to prevent this from happening in the future.
Re-conditioning by controlled deep discharging of single cells shows most promising results so far, over 40% capacity increase for the weakest cell at 20A discharge rate.
.

The static thrust is limited too severely, much better acceleration should be possible.

.

A higher top speed could increase safety in some situations.
.

Software changes might be "all" that is needed to achieve these last two improvements.

.
Added on 2008-08-21:
A warning light to alert the rider to a rear/brake light failure needs to be added.
More details here.

Edited 2009-02-12: The hazard lights option has now been added to increase safety during low-battery crawls and other emergencies. It does not work during main fuse blow-outs, yet.

Mr. Mik

Short URL

Customize This