The other day I looked back through my EV email folder and was reminded of some of the bugs I had faced and solved in developing our EV. At the time some of these bugs were quite a worry, but looking back I feel very satisfied that I actually solved them. Many of them were way beyond my skill set at the time but won’t ever bother me again because of the skills learned along the way.
Root causes of the problems have been lack of skills (e.g. machining, welding, automotive), stupidity or bad judgment and in some cases plain bad luck. I know some people with the right mix of skills who have avoided many of these mistakes, but many more who have repeated some of them or had setbacks of their own. So I thought it might be useful to blog about these bugs as a learning experience for others.
When we experience problems I tell myself “the Wright Brothers broke a few wings”. Todays EV drivers are pioneers, and each EV is a one-off prototype. However it’s not a big step to imagine mass produced (or even mass conversions) that are far more reliable and lower in cost than todays Internal Combustion Engine (ICE) vehicles. For example the second EV I converted went much more smoothly.
Throughout I have had an amazing amount of help and support from the on-line (especially the EVDL) and South Australian EV community. My motor and battery suppliers have also been very supportive.
Most of the problems below happened during the final stages of construction and the first few months of driving. The only unresolved issue today is the best way to treat the lead acid batteries for maximum life and acceptable range. This has taken us a while to work out as we made a poor choice for the original battery pack and charging system (see below).
However the end result is a car we are very happy with. It gets used for 90% of our daily driving and has traveled 8500 electric km in the last 10 months of daily use.
My wife keeps telling me “I love my EV”!.
Optima Yellow Top D31A Batteries
When I started my conversion in mid 2007 Yellow Tops (YTs) were strongly recommended for EV use (Lithium batteries were not an option back then). The YTs were very expensive (AUD$5400 for 12 D31A rated at 75AH) but I thought worth the expense. However there were two problems.
Prior to installing in my EV I load tested each battery at 250A for 30 seconds and 50A until discharged down to 10.8V. These two currents model acceleration and cruise currents. When I tested my YTs I found that 3 out of 12 were duds – one couldn’t supply 250A and two had only around 20AH capacity at 50A. These were replaced under warranty. My experience was not unique (see the EVDL archive) – it seems that in mid 2007 Optima was sold and several quality problems emerged. In particular the deep cycle ability of the YTs seems to have been reduced. They are still a good starting battery, and OK for moderate deep cycle loads (like a fridge in your 4WD). But a poor choice for heavy deep cycle loads, like an EV.
How do you test a battery at 250A? I used a coat hanger as the resistor (0.04 ohms) – in a bucket of water. A 3kW electric kettle. The 50A load was some copper wire in a bucket of water, the type used for winding inductors for radios. A 1 hour/50A test would heat up a bucket of water nicely. I have since moved to an EV battery tester based on a Wifi router.
The second problem was capacity. When tested at 50A the range was 37 to 50AH for a battery rated at 75AH. The problem was Peukert effect – lead acid batteries deliver less AH at higher currents. At 60 km/hr my EV used around 45A at 120V. You are only as good as your weakest battery so we effectively had around 40AH available. This was not really adequate and meant we had to deeply cycle the batteries, reducing their lifetime. So the maximum range was initially 40 km in November 2008 but this drifted steadily down to less than 30km in July 2009.
In short the YTs were a poor choice for our EV – too small for the job and much too expensive. The timing (mid 2007) or my purchase was also unlucky. However even when they were of no further use to us they still were in pretty good shape. The reserve capacity (20 hour discharge rate) still met the YT specs and the cranking current was an impressive 1100A. So I sold them on ebay to some local rev-heads who like YTs as ICE starter batteries.
They money raised paid for a much cheaper new battery pack of 8 PG10312V 100AH Powersonic AGMs and we were back in business with a more appropriate battery!
Jaycar MB-3612 Chargers
After some reading I decided to use separate 12V chargers, one for each battery. They charge each AGM battery separately, avoiding overcharge of batteries in different condition that is possible with a series-string charger. They were also cheaper than an EV charger at the time, and I liked the idea of trying something a little different. You always learn something. The chargers have worked out OK, but in hind sight I would go for a proper EV charger next time (in fact I have one on order).
The charger fans have all failed and have been replaced with high quality fans. During installation the amount of wiring (one pair of wires for each battery) was a lot of work. The charger starts at about 12A, which tapers off to 1A at the end of the charge (it then switches to float). This means the charge time is fairly slow compared to a charger that has a bulk current phase, however we haven’t found charge time to be a problem in practice. Out of the box they switch to float at 14.1V which is a little low. There is a little pot inside that you can tweak to make it switch over at 14.7V.
The chargers do not do equalisation which I think has contributed to my battery problems. Each 12V battery has 6 x 2V cells which can get into different states of charge, a periodic overcharge is required to make sure they all have the same capacity.
Having said all the above the chargers have been bouncing around in the back of my card for a year in temperatures up to 40C and keep doing their job. Not a bad charger, just perhaps not appropriate for repeated deep cycle applications.
They do look pretty when charging at night, though. All those glowing LEDs.
My EV made a sound like a creaking door at low speeds, coming from the bell housing. After some investigation (and lots of help from the EVDL and Sean, a local mechanical engineer) I figured out it was a mis-alignment of the motor and gearbox. This means the centers of the motor shaft and gearbox shaft were out by a few mm. The clutch plate would slip against the pressure plate making the creaky sound.
Somehow I came up with the idea of a gauge that shows when the adaptor plate is centered:
After working out a procedure to align everything I put the car back together the creaking was gone and I was no longer in danger of killing my gearbox. Must admit I still can’t believe I worked out how to fixed this one! Lesson learned was get the adaptor plate aligned right when machining.
I had problems with my clutch slipping under acceleration. My electric motor had twice the torque of the original motor so I had a custom clutch built up for $300. On reflection I suspect this problem was at least partially due to the adaptor plate mis-alignment, however clutch upgrades are common for EVs. After some driving experience we settled on 3rd gear for all our driving so I now feel a clutch isn’t really necessary for our sort of driving (60 km/hr around town). On EV #2 we went clutchless which worked just fine.
A clip that retains one of the electric motor brushes fell off inside the motor! I found out as the loose clip introduced a variable resistance connection between the battery pack and the chassis. No idea how this one happened, but easily fixed. Out came the motor again……
Over-revving and Over-temperature of Electric Motor
My EV had an odd problem of slipping out of third gear into neutral, especially when accelerating. For inspection I needed to get a brake test done. This meant accelerating to 60 km/hr, braking hard, and repeating this test many times. In the final test the motor was quite hot and it slipped out of gear. We didn’t notice for a moment due to our speed (you can’t hear the motor, only road and transmission noise in an electric vehicle). Suddenly the over-temp light came on. Afterwards the motor was making a noise like a sewing machine. At low revs I could see the brushes moving up and down as the motor rotated. Oh dear.
The combination of a hot motor and over revving had “raised the commutator bars”. Fortunately mine was a mild case – a local electric motor guy machined the commutator and its seems to be OK now. It usually means the end of the motor. Heat and over revving are the enemy of an electric motor. Apart from that they last virtually forever.
A few months later I looked inside another Charade (same car as mine). I noticed a small bracket securing the gearbox to the chassis (the gearbox is also supported by two rubber engine mounts, but they are flexible and allow twisting under torque). DOH! Hunting around my garage I found the same bracket – I had forgot to re-attach it! This bracket restricts the movement of the gearbox under torque, the reason why I was popping our of third gear. Chalk this one up to stupidity – although in general a rev limiter for EVs is a good idea!
Inspection and Registration
The Department of Transport inspection felt harrowing at the time but when I think about it I will not be worried about the next one. There is a degree of inconsistency about the whole inspection process, as the people inspecting have often never seen an EV and the rules are open to some interpretation. For example I was requested to “seal my motor against water and dust ingress”. This would effectively destroy the motor due to over heating. Eventually my engineer argued that this would make the car unsafe as the motor would become a fire hazard, and DoT accepted that. Mainly the inspection was about road worthiness, and sometimes you just have to swallow your pride and say “Yes Sir, No Sir, three bags full……”.