When One Mistake Can Kill

This article was first published in 2005.

Car won’t start.

Hmm: flat battery.

Answer’s easy, huh? Just charge it.

But what do you do when the battery has over 280 volts in it, and no battery charger exists? Welcome to the world of modifying petrol/electric hybrid cars....

WARNING!! Overcharging a Prius High Voltage battery can result in the battery pack catching fire.  If overcharged, the battery pack can also release toxic, corrosive fumes.  Charging currents over 0.5 amps should be undertaken with extreme caution and for short durations only.  Being electrocuted is a likely outcome if you do not know what you are doing with high voltage AC mains power systems and high voltage DC battery systems.

I’d had my ‘99 Toyota Prius off the road for three weeks, while I removed the supercharger and installed a turbo. Like the blower installation, the turbo job was completely custom: a newly fabricated exhaust manifold, new oil and water lines, and new intercooler plumbing. So explaining the three weeks. During that time I’d had the low voltage (12 volt) battery disconnected, but the 288 volt battery pack that fills the complete width of the car behind the back seat had been left connected. Not that it would have mattered if it was connected or not: there’s no current draw on it when the car is turned off. Or is there? In any event, now, with the turbo installed, the car wouldn’t start.

I recharged the 12V battery (it lives in the boot) and turned the key. But the engine wouldn’t even crank: not because there wasn’t enough power available to turn it over, but because one of the many Electronic Control Units in the car was preventing the start occurring. That could be determined because the ‘READY’ indication wouldn’t appear on the digital dash. And without a READY, the car won’t start.

I checked and rechecked under the bonnet, making sure I hadn’t done anything silly like leaving off the airflow meter plug or something like that, but could find nothing wrong. I entered the diagnostics mode but could see nothing unusual in the displayed fault codes. However, without the proper service tool, the faults for only one of the ECUs can be seen – the engine control. The much more sophisticated Hybrid Control Unit (an ECU perhaps three times as big as the engine ECU) could not have its fault codes sighted.

But another fault display screen gave an indication of what was wrong. This screen came up with two problem areas – and one was the 288V battery. On that basis it seemed likely that the high voltage battery had dropped below a level at which the ECUs permitted starting. But how to charge it?

Connections

Normally, in the rare situation where a Prius high voltage battery drops below the starting threshold, the car is taken to a Toyota dealer where apparently a special charger is used. However, firstly, my car is a grey import and is the earliest model Prius, which was not sold here in Australia . That meant that the local dealer’s charger (if in fact they have one, which I doubt) probably wouldn’t work. And secondly, even if the dealer did have the charger, I can just imagine what they’d say when they saw how non-standard the car is. “It won’t start cos of what you’ve done to it, mate.”

So any battery charging would be up to me – and those people I could drag in to help.

My first call was to John Clarke, an electronics engineer who works for the magazine, Silicon Chip. I contribute material to that magazine and have also written a book in conjunction with John. He’s a man unfazed by nearly any electronic problem: I knew when I stated the requirements, he wouldn’t throw his hands up in the air but instead carefully consider a workable solution.

But what actually was needed? In this Prius model, the high voltage battery consists of nickel metal hydride D-cells. Yep, D cells. In fact, there are 240 of them wired in series, each (as normal with Ni-MH cells) having a voltage of 1.2V. That totals up to 288V. However, the battery pack doesn’t just consist of these batteries; instead much of the space inside the box is taken up with the battery ECU, relays, a current sensor, and a large thermostatically-controlled cooling fan.

In other words, the charger – the one that didn’t yet exist – wasn’t going to be just connected to the two bright orange cables coming out of the box.

Or could it be?

Perhaps the cables could be disconnected at the engine end, the 12V ignition turned on (which would hopefully click over all those relays and power-up the ECU in the box), and then the battery pack charged via these cables?

I wanted to find out – but first a word about the volts. Two hundred and eighty eight volts DC from a high capacity battery is lethal. As in, you grab those cables with bare hands and you die. To prevent this nasty occurrence, I’d invested in special High Voltage Live Working gloves. These are good for 1000 volts and cost only about AUD$70 – more with the protective leather over-gloves that I also bought.

There is also a safety breaker located in the boot. This is electrically located in the middle of the series battery pack, so pulling the breaker results in a maximum nominal voltage across any part of the system of 144V – still dangerous, but less dangerous!

So with the breaker pulled and my high voltage gloves on, I undid the two high voltage connections in the engine bay and used a high safety rating multimeter to measure what voltage was available across these cables. Firstly, with the ignition off and the breaker pulled, then with the breaker in position, then with the breaker in position and the ignition on. The answer in all cases was: nil. The battery ECU could detect that the cables weren’t connected and had switched off the power.

So there was no convenient external way of connecting the charger. Instead, I’d have to delve inside the high voltage battery box itself.

Inside the Battery Pack

In addition to the car, I also own a half-cut Prius, ie the front half of the car imported from Japan and bought for spare parts. The battery pack is in the back of the car - and so normally wouldn’t be included with a half-cut - but I’d made sure that it was supplied as well. That gave me a guinea pig battery pack to examine before I opened-up the battery pack in the car.

As supplied, the cover was already off this pack so inspection was just a case of looking. Or more accurately, fitting the high voltage gloves and taking off an internal plastic cover before looking. What I was trying to find were the two ends of the battery string – if I could directly access these, then I could perhaps charge the battery while bypassing all the relays and current monitoring stuff. On paper it sounds easy: after all, the high voltage cables are all in that brilliant orange and all I needed to do was to find two thick ones that went to each end of the battery. But there’s a helluva lot of stuff in this box...

However, after about 20 minutes of inspection and wire tracing, I was pretty confident I had access to the two ends of the battery series. I measured the voltage across these two points and found just under 300 volts. So this battery pack, which hadn’t been charged for at least 6 months, had a higher voltage in it than the nominal 288V it was supposed to have! However, given that Ni-MH rechargeables rise to above 1.2V each when charged, that kinda made sense. After all, 300 volts divided by 240 cells is only 1.25V per cell.

OK, so I now had the charging points on this pack – but how to charge it?

The Charger Circuit

“Hmm”, said John. “You probably only a need a bridge rectifier and a resistor.”

He paused, thinking hard.

“Yes, that’s it. Say, a 400V 8 amp bridge rectifier across the mains, and a 10 ohm resistor. Course, the resistor will have to dissipate a lot of power, so perhaps use a 100-watt light bulb.”

In Australia the AC mains supply voltage is 240V; by the time a bridge rectifier is used to turn it into DC, there will be peaks in the fluctuating voltage of about 330 volts. That’s high enough to charge a 288V battery without problems. The resistor limits the current flow – the greater the current flow, the more it resists.

But John sounded concerned.

“It’s all very dangerous,” he said. “Do you have an earth circuit leakage safety breaker on your switchboard?”

I’d been going to install a safety switch for years, but hadn’t yet got around to it.

“In that case, buy one of those portable ones and use that. And put a low current fuse in the circuit,” he added.

Having sketched out the very simple circuit, I then rang my father, a long retired electronics engineer and physicist. I recounted the circuit John had suggested, and found my father insouciant about any dangers.

“Nah, you won’t electrocute yourself,” he happily said. “I am more concerned about this resistor value. I think it should be much lower – perhaps less than 1 ohm.”

The theory of sizing the resistor is surprisingly complex, primarily because the unsmoothed DC voltage coming from the bridge rectifier is bouncing up and down all the time. The upshot was that it was easiest to start with a high value resistor, connect the charger to the battery pack, and then measure the actual current flow. If it was too low, reduce the value of the resistor.

I built the charger on a scrap of chipboard, using the 500 watt, 3 ohm (cold resistance) floodlight bulb shown here as the resistor. Double checking all my work (hey, this would be a good battery charger to connect up with the reversed polarity....not) I wired it to the half-cut battery pack and switched on. The result was positive – a charge rate of half an amp.

That was good – I was charging the high voltage battery – but at a current flow of half an amp, it would take a while to charge.

I could reduce the value of the resistor, which would increase the charge rate, but what about the car’s battery pack? If that was lower in voltage than the half-cut’s pack, the current flow might be too high, even with the 3 ohm resistor in place. Best to now move the system to the car.

Heart in my mouth, I connected the system up to the car battery pack, inserted the circuit breaker and switched on the charger. But again, the charge rate was only half an amp or so. Over the next hour, I reduced the value of the resistor and watched the charging rate go up a little each time. (Of course, each resistor change required pulling the circuit breaker, donning the high voltage gloves, undoing the connections, unsoldering the resistor and soldering-in the new one, then.... It was not something to rush or do without a lot of care!)

The current flow into the battery kept on rising, but it never got very high – about an amp being the maximum I read on the meter. With the resistor value by now absurdly low, I pulled it out of the circuit completely – now the bridge rectifier across 240 volts AC was charging the battery directly, with only a 2.5 amp fuse and the portable safety switch for protection.

But even in this configuration, the current flow kept on dropping. It looked as if a 240V AC source wasn’t going to be high enough to get the charging results desired. Oh bugger, bugger, bugger.

I rang my father to tell him the bad news, and he had a brilliant idea. Why not pull the circuit breaker and then charge the battery in two 144V halves? That way, starting with 240V AC would be plenty high enough.

By this time it was late at night and I’d been doing this tricky and dangerous work on the car for more than 12 hours. Best to stop and start again the next day. However, out of curiosity, I used the multimeter to measure the voltage of the battery. Would the charge I had been able to give it have increased its voltage at all? Amazed, I looked down at the multimeter to read: 318V. Before charging it had been about 275....

In fact, was it even possible that the car would now start? I disconnected the 12V battery for a few minutes to reset (some) of the fault codes, then turned the key.

The engine started.....