4Springs' Brumby

[woody]

2 days too early on your post

I regularly ride 40km to work - amazing what your body gets used to quickly.

[4Springs]

2 days too early on your post

Ok, it has officially taken me 5 days to work that out - better late than never!

I regularly ride 40km to work - amazing what your body gets used to quickly.

Yeah, my work is 15km away, and I can ride to work in my steel-capped boots, do a 10 hour day and ride back again without any kind of "bike fitness". That is, I can do this even if I haven't ridden a bike for a couple of months. Uses even less fuel than my EV!

[Thalass]

Wow those are great numbers on the lithium pack. I'm sure with their longer life and greater range you'll find they at least break even with the cost of buying several lead packs. I daresay they'll even be cheaper.

[4Springs]

Wow those are great numbers on the lithium pack. I'm sure with their longer life and greater range you'll find they at least break even with the cost of buying several lead packs. I daresay they'll even be cheaper.

They were about twice the price, so break-even time is about 3 years. If they last 10 years like some people seem to think then they'll be much cheaper! I'm not sure that I'm exactly molly-codling them though, I go further than 80% DOD more often than I like. Need more charging places!
I'm glad I started with lead though, since it let me work out exactly how large the lithium pack needed to be. If I'd gone straight to lithium then I would have bought a bigger pack to make sure I got the range I needed, and the car would have been that much heavier.
I'm using the ute much more now that I can go further. Just went past 10,000 km as an EV. Just sold our 3rd car too, so the Brumby needs to be our reliable 2nd car!

[4Springs]

Some photos of my LED decoration of the Brumby:


This shows the red LED I put behind the fuel cap door. This LED is powered by a 12V plug pack which was installed to turn on a contactor while charging. So the fuel cap glows red whenever plugged in to the 240V mains. I installed this in response to the feeling of never quite knowing if the 240V was on or not - especially in places other than at home.
With the lead-acid battery pack I had audio feedback in the form of a slight hum from a small water pump associated with the battery warming system. Could only hear that at home though, it couldn't be heard over the general traffic noise in a city.
Photo taken with a tripod using moonlight.




These two photos show the strip of blue LEDs I installed under the bonnet so that I can see what is going on under there. I put in a microswitch so that they turn off when the bonnet is closed, and there is another switch as well so they can be turned off when the bonnet is open.
Also on the microswitch is the backlighting for the BMS display. You can almost see that, it is behind the stop button and is green/yellow.
Other glowy bits that can be seen are:
A large red spot from the flashing red LED that denotes "150VDC is connected".
A small red dot on the charger, saying that it is charging.
Two red temperature readouts from the battery warming system.
Something that can't be seen is the small glow from the Zeva 12V low voltage cutout device.


And here is a photo showing the whole lot by moonlight. I had a tripod by this time, so I should have used a lower ISO setting.

[evric]

That's excellent Christopher, maybe get that red light to "throb" - even better! Eric

[Johny]

Veeery cute. It reminds me of that scene in Back to the Future where the Delorean ends up in the barn.

[weber]

I'm a sucker for coloured glows in the dark. Very nice.

[4Springs]

That's excellent Christopher, maybe get that red light to "throb" - even better! Eric

Ha!
The red light is to show that the power is on. Even better if it would show that the charger was operating. It could pulse quickly when the pack is near empty, then slow down as it charges up. Sort of like the car is panting...

[evric]

Yeah - do that!

[4Springs]

Kelly controller died this week. The report from the wife was that she turned the key on and the motor went flat-out in an uncontrolled manner! She did all the right things - she had stared in neutral and she turned the key off to stop the motor. Apparently it took an age to spin down - I wonder how fast it went?
So this Kelly lasted 15,000km and about 26 months. I've ordered a replacement, but I've gone for the next model up in the hope that a bigger controller might last a bit longer.
The controller now measures as a dead short across all terminals. It powers up under 12V and the fault LED flashes "I'm Stuffed" in Morse. It was waterproofed, which basically means filled up with epoxy resin, so I'm not too confident that I'll be able to get it apart. But I'll have a go. Any idea how to remove epoxy resin? First step is to unscrew the screws in the case, but even they are covered in resin. Burn it off perhaps?

[jonescg]

Yikes! One of the joys of an AC motor is that can't happen. Good thing the start in neutral policy was adopted...

[4Springs]

Good thing the start in neutral policy was adopted...

Can you imagine what would have happened in the garage? The brake booster wouldn't have had time to build up vacuum, and I doubt the brakes would have stopped her anyway...
Our other car will not let you start the engine unless your foot is on the clutch. A good habit to get into, although it took a bit of getting used to since I used to drive a tractor that would not start if your foot *was* on the clutch.

There are two separate fail-safes to stop the motor runaway happening, but neither of them work if the controller is stuffed. In this situation you have three options to stop it:
1 - turn the key back to ACC (turns off the drive contactor)
2 - turn the key to OFF (turns off the other contactor)
3 - hit the big red panic button (breaks the battery in a third place)

My wife did the first two, and was ready to do the third if they didn't work. So I'm proud of her and pleased that my planning (to make things like this intuitive to a driver) worked in this case!

[evric]

Christopher, have you got a good precharge circuit in place?

[coulomb]

Any idea how to remove epoxy resin?

There are so many types of goo that they could be using.

There are two different types in the TC (Elcon) chargers that I'm getting to know. Both are softened by heat. The black soft material seems to respond best to a chopstick: flattened at one end and sharpened in a pencil sharpener at the other end.

The yellow gunk is tougher, and seems respond better to pliers. I have a small needle nosed set that close really well (very little gap when closed).

For heat, I use an Atten hot air station. These were about AU$65 when I bought mine years ago. They are for reflow soldering of surface mount components, but mine is used more often for work on solar panels (removing the backing sheet, EVA, and the other gunk), working on chargers, and shrinking heat shrink tubing.

Here is one: Atten hot air gun Ebay listing.

The resin sounds like it might be a bit tougher than what I've been faced with. I read somewhere that someone had success with putting a controller into an oven. Obviously you need to be very careful both of melting the electronics (even though it's broken now it might be repairable), and burning yourself.

Good luck!

[4Springs]

Christopher, have you got a good precharge circuit in place?

Well...
My precharge circuit consists of two high wattage resistors to the value recommended by Kelly (can't remember the values off the top of my head). When I first put it together I used a multimeter to see how long it took to charge up the capacitors. I found a guideline somewhere that said that it needs to get to (I think) about 80% of full voltage before you should turn on the drive contactor. I timed this as about 20 seconds.
I rely on the driver to wait this amount of time. I have given them a light as a guide. So the process is:
1. Turn key to ACC (and on to ON if you like). This pulls in the main contactor (not the drive contactor, just the one that connects the traction battery). A blue light comes on to remind you to wait.
2. Once the blue light goes out (30 seconds) you can turn key to START to bring the drive contactor in.
There is no interlock to stop you turning to START prematurely.

We both drive the car a lot, and so we are both used to this process. I made a point to ask my wife if she did this this time, and she said that yes, she turned to ACC, then loaded the dog in before she got in herself. (this is the type of thing we now do out of habit because 30 seconds seems like an age sometimes!)

It does sound like a precharge problem doesn't it? It didn't overload during a period of high stress, it was the first time it was used that day. That blue light is very bright though, I can't imagine that she could have not noticed it.

I did look at the resistors and they are clean and white. I'll test them though to make sure they are still the correct resistance.

This might be a time to look into a more insistent precharge procedure. I did have a relay attached to the blue light timer circuit at one stage but I found it too restrictive. The relay meant that you could not pull the drive contactor in until the 30 seconds were up. There were two problems with this. First was that when plugged into 240V the main contactor is pulled in, and therefore the circuit is fully charged no matter where the key is. So in that case I know that I don't have to wait for the blue light. With the relay I would get frustrated because I couldn't start even though I knew it was precharged. The straw that broke the camel's back though was when I had a loose wire one day at work. I had to find a screwdriver, pull out a section of the dash, find the relay in the dark and hotwire it to get going!
I could re-instate that relay, but also put in an override pushbutton to satisfy my desire to be in complete control. Or I could see what other people do. Any suggestions?

[evric]

I would suggest to put it back so that you have to wait, but change the value of the resistor so that the time is only about 7 seconds (like mine). This amount of time is about how long it takes to get your seat belt on. If you have an override, it is too easy to make a mistake. Mine is wired so that if I turn it to Start, this latches up and if the high voltage is present, in my case via the circuit breaker, (in your case the main contactor,)the 7 second timerstarts. Then the drive contactor is actuated. My circuit actually checks for the presence of the 130 Volts. Nothing can go wrong, nothing can go wrong, noth...

[weber]

So that's your second Kelly controller failure? It pretty much has to be death by voltage spike in this case. Faulty precharge could do that.

You should disconnect the faulty controller and prove that your precharge is still working, to eliminate precharge from suspicion.

What model of Kelly controller was it? What voltage rating? How many LiFePO4 cells are you running?

The following thread is somewhat old. Have you seen it:
http://www.diyelectriccar.com/forums/sh ... 22300.html

By the way, one rule for precharge, to protect the precharge bypass contactor from welding, is that the voltage across its contacts (i.e. the voltage across the precharge resistor) should be less than 28 volts when it closes. So in the case of a 140 V system, 80% charged would be OK. But with a 360 V system the controller has to be 93% charged, and at 720 V, 97% charged.

But it's up to the controller manufacturer to say what the precharge requirements of their controllers are.

A classic mistake by vehicle converters is having other loads such as a DC-DC converter, in parallel with the controller, which load the precharge resistor and prevent it from _ever_ charging the controller caps high enough.

Since you're allowing 30 seconds for precharge I'm guessing you have something like a 2 kR precharge resistor. It would only take a 20 mA load to prevent the voltage across such a precharge resistor from ever dropping below 40 volts. But I'm guessing you've checked that in the past.

[Johny]

I also think your precharge is way too long. I would find that really frustrating to drive. My industrial controller precharges in less than two seconds so you can have it a lot faster.

Edit: What value resistor do Kelly recommend?

[4Springs]

Ok, I set up the precharge over two years ago and have not looked at it since. Let's look up some old schematics.

This is the circuit diagram from the Kelly manual. It specifies a 2k precharge resistor.

This is the diagram I have for the Brumby. I have made some changes since this diagram, but the precharge section is the same.

weber has correctly deduced the 2kness of the precharge resistor. This is why my precharge takes so long, I implemented the resistor recommended by the manufacturer. They didn't say how long to precharge for, so that is where my experiments came in.

After a while I replaced the lead-acid battery with a lithium battery. 48 LFP cells. Between the charger and the BMS the pack voltage will not exceed 175.5V, and will normally be 173V when fully charged. Specification from the Kelly manual (for the nominal 144V controller) is for a maximum voltage of 180V. On the day in question the vehicle had been sitting for a few hours since charging, so the pack voltage was probably around 165V.
I did not re-calculate the precharge delay with the new battery. The voltage is not a great deal different, so it was probably fine. I will do the test again though with the new controller, using weber's "28V rule".

The Kelly model is a KDH14800D - the "800A" model. Someone mentioned that this is my second one - that is correct. The first one was faulty when I got it. The story goes that I installed it and realised that the case was not isolated from the HAZV. I emailed back and forth with Kelly, but they assured me that it was ok. When I tested it by driving around the house it blew up in a spectacular way. They replaced this under warranty even though it was out of its warranty period due to me taking a long time between buying it and using it. The replacement was fully isolated as expected.
The KDH14800D seemed a good match for my motor and weight of vehicle. A nice amount of acceleration without ever getting very warm. I have it on a vertical aluminium plate with an airgap underneath, and there are two 12V fans bolted on top.
I have ordered a KDH14100D - it was shipped yesterday. This is the 1000A model "Racing Controller", and it uses the same case so I should be able to just drop it in.

Long term I'd like to build my own controller. Note that I say "build", not "design". My electronics knowledge is way too basic to design one of these things, so I've been looking at the OpenRevolt/Ecomodder project. If nothing else is salvagable from the Kelly I might be able to use the nice aluminium casing for that project.

[evric]

The circuit shows the contactor in the centre of your pack running off ACC. In my car ACC is turned off while I go to START. It should be wired to IGN side of the switch. Does your radio go off when you switch to START? If it does this maybe the cause of your problems.

[4Springs]

The circuit shows the contactor in the centre of your pack running off ACC. In my car ACC is turned off while I go to START. It should be wired to IGN side of the switch. Does your radio go off when you switch to START? If it does this maybe the cause of your problems.

Well spotted Eric, I didn't know that when I drew this circuit, but I found out once I built it! So I put in a diode or two so that the contactor (labelled BB in the diagram for Battery Box) is powered by ACC, ON, START and 240V. That contactor is therefore only rarely turned off since most of the time I am parked I am plugged into 240V. I bought a more expensive Gigavac contactor for that reason - the Gigavac has a low current usage while on compared to the cheaper one I used for the drive contactor.

[evric]

As the circuit is now, if one of the Emergency stops is left open and then after trying to start is suddenly closed this will put full voltage straight on the controller. I suggest a HV sensing circuit, which via a resistor and the input of an optocoupler senses the HV and then a transistor output enabling the start circuit/drive contactor after the preset delay(timer cct). Automatic with no way to hurt the controller. Just a little more circuitry in the green section.

[4Springs]

As the circuit is now, if one of the Emergency stops is left open and then after trying to start is suddenly closed this will put full voltage straight on the controller.

Yes, you are correct.
But an even more likely scenario is that I switch to ACC, then ON, and forget to go to START. I try to drive by putting my foot on the accelerator, and the controller tries to oblige by driving the motor. What I feel is a small nudge from the motor. My reaction is to go "Doh" and hit my head, then turn the key to START. But if that nudge drained the capacitors, what I *should* do is start the whole precharge process again.

So I agree, I should put in something to react to the actual voltages. This would guard against both those scenarios and potentially others that we haven't thought of.
weber's 28V rule has got me thinking - could I make a simple circuit based on the voltage over the contactor terminals?

So this circuit goes into the one in the previous post. The blue is HAZV, with the controller and its capacitors off the bottom. The green isn't just 12V, it still connects to all the other interlocks to stop the contactor being closed until ready.
The idea is that when HAZV is applied the capacitors look like a dead short. So the zener bypasses enough current to open the normally closed relay down the bottom. While in this state the drive contactor cannot be closed.
As the capacitors charge up, the voltage across the zener reduces. Eventually it gets to less than 28V, and it stops bypassing current. The relay drops out enabling the drive contactor.

I don't have enough electronics knowledge to know if this could be made to work. My next step would be to build it and play around with values. Issues I can think of are:
1. The relay needs to work over a wide voltage range. At some point it would drop out, and it may chatter. Wouldn't want chattering to drive chattering on the contactor.
2. The relay/zener part would supply extra current to the controller. I would need to revisit the value of the precharge resistor.

As I said, I'm not sure that this could be made to work. Is a standard way of sensing voltage? Either of the traction pack itself (as Eric suggested) or of that contactor?

[evric]

You already have too many contactors and relays...
Ian Hooper (www.zeva.com.au) makes a precharge circuit. Just buy and install one of them.