Electric car conversion: Project ForkenSwift

[MetroMPG]

While tinkering on the potbox today, I got to thinking about the various sources for recycled parts used in the car so far.

A partial list...

• geriatric forklift (obviously)
• 1993 Suzuki Swift (various parts)
• dead inkjet printer (used nylon gears in the potbox)
• broken child's tricycle (used metal to make standoffs to mount the heatsink)
• used copper tubing (same as above)
• junked Canadian Tire SuperCycle 10-speed bike (metal for component platform frame)
• scrap piece of lexan thrown in the garbage at Ivan's old work (component platform)
• plastic top of a used 45 gallon barrel (12v accessory batt. tray)
• dead 12v cooler (heat sink)
• old bed frame (metal for battery racks - too hard... wrecked many drill bits)
• golf cart (controller)
• used EV floodies (from http://sparkyev.ca)
• used alarm system batteries (12v power)
• discarded industrial machine fencing (source of square & angle iron for the tow bar)
• used welding cable (battery/motor cables)

[MetroMPG]

Quick cable end-making tutorial for lemmiwinks, Bigmouse... I'll add pics when I make my next batch of ends.

Stuff you need:

- roll of non-lead content solder (lead solder has less conductivity and lower melting point = bad); flux
- piece of copper pipe, diameter slightly larger than your wire dia.
- torch (I used oxy-acetylene)
- hammer
- vice, vice grips
- grinder/files
- drill

---

1) Make a crucible from your piece of copper pipe. You'll fill this with solder and dip the wire into it



a) hammer/crimp one end closed;
b) overall depth should be slightly less than the desired length of your cable ends (solder will wick above neck of crucible)
c) I somewhat flattened the body of the crucible to hold my wire in the desired shape as I pushed it in

2) place crucible in vice; melt in solder, roughly 2/3 full .... EDIT: You don't necessarily have to pre-fill the solder pot/crucible. You can just feed solder into the side of the cable at the neck. It's heavy - it'll run down the strands as well as up.


3) strip cable: cut unsulation & slide down but not off the end (will help keep strands in order as you flatten the end)

4) flatten end of cable with a hammer; slide off piece of insulation



leaving the insulation on while flattening helps keep the strands in order

5) flux cable end

6) crucible may be cool enough to work with now - insert cable end into top 1/3, onto hardened solder

7) heat crucible and cable end while pushing down gently on the cable (to move it into the solder when it liquifies) - don't push too hard, or solder will squirt out the sides when it drops down

You may now need pliers/vice grips to handle the hot cable...

8) once cable drops to bottom of crucible, optionally keep adding heat & solder until it has wicked up as far as you want

9) remove cable from crucible before solder hardens (be quick - but careful when you pull it out - solder may drip/splash from end)



cable end after removing from crucible

10) wait for solder to just solidify, then gently hammer to flatten further if desired - make sure solder is hard first! I rushed one, and molten solder spit out of the cable end when I hit it!

11) wait for cable to cool, then grind/file flat surface, drill hole.

I have only tried this using fine strand cable. Coarser strand cable may not work - I think it'll be more likely the end may split when hammering in step 10, or drilling in 11.

[MetroMPG]

Added a few pics to the cable end DIY, above.

Made a few more cables this weekend, and noted that the forklift cables we have used so far are all 3/0, not 4/0 as previously thought. The lugs all say 4/0 on them (the source of my confusion), but the cable itself is 3/0.

Which makes me think that the welding cable we have is 2/0, not 3/0 as previously mentioned (I was guessing by size relative to the forklift cable).

Edit: how do I know if 2/0 is appropriate for our voltage/amperage/cable run lengths? To this point, I've just been assuming, "if it was good enough for a 19,000 lb 36v/500A forklift, it's good enough for the 48v/225A ForkenSwift".

I used about 12 feet of welding cable to make up one of the front/rear battery interconnects today, and it's installed in the conduit mounted under the car last weekend.

Ivan & I talked out a mechanical/electrical disconnect method, and I think we're going to use:

- the existing clutch cable (which I hadn't yet taken out)

- and an Anderson connector that came with the forklift (main batt pack / control panel connection)



The plan is to align & mount the connector inline with the clutch cable, with one end secured, and the other end connected to the clutch cable. We'll also use a spring to ensure separation after the pedal is pushed/released.

I warned Ivan that we'll undoubtedly press the ForkenSwift's clutch pedal unintentionally (both of us drive manual shift ICE cars), which could create problems when the power cuts out (and requires opening the hood & reconnecting the Anderson to restore it). His reply was: you'll only do it once...

[MetroMPG]

Also, I made a cable end lug with coarser strand 3/0. It didn't turn out quite as well as the fine strand stuff - more prone to cracking when flattening/drilling as I suspected it would be. But it worked OK.

[MetroMPG]

Added a refinement to the end-making DIY:

You don't have to pre-fill the solder pot/crucible. You can just feed solder into the side of the cable at the neck. It's heavy - it'll run down the strands as well as up.

[MetroMPG]

Time to start thinking about this again:


---

I mounted the pack negative contactor this evening and wired it to the 12v keyswitch relay (recycled the fuel injection relay).

Man, is it ever LOUD when it pulls in. I mean LOUD. Like... !!CLACK!!

Now, I mounted it more or less on the passenger side shock tower, so it's telegraphing directly through the unibody, but still. I took it over to show Ivan, and it's got us reconsidering whether we should go ahead with the plan to put the other contactor on the go pedal to pull in/out pack +ve every time the pedal is pushed/released.

One of the joys of this little car is how utterly silent it is ghosting around at low speed. If the pack +ve pedal contactor can't be physically isolated to make it quiet, I definitely don't want it clacking away under the hood. It'll drive me nuts to have to listen to it every time I move my foot. The single keyswitch CLACK I can live with. It kind of adds drama to the "start up" process ... "Contact!"

The reason for the pedal-actuated contactor is: in case of controller failure, the driver's first instinct is to release the go pedal, theoretically cutting power pretty much instantly (assuming it doesn't weld). If we omit it, then a split second is lost, before jamming down the clutch pedal to mechanically shut things down.

Given that this is a low power machine, that split second is perhaps not an extreme problem. For someone with a lot more voltage & power in their pack, it's a problem.

[MetroMPG]

skewbe:

is this a problem? I've done a bit of EV reading (likely not as much as you) and I haven't come across a "WOT" controller failure (yet).

You know, in the year and a half I've been following the EVDL, 2 Cursit failure stories jump to mind (there are undoubtedly more) - and they were "fully closed throttle" failures.

I haven't actually read about a WOT controller failure, though I'm sure if you ask on the EVDL, you'll get some.

Still, everyone insists controllers will fail fully on, and they're basically ticking time bombs...

[MetroMPG]

Note re. controller failures: DC controllers may fail ON. Apparently this is one of the other benefits of AC, which fail OFF.

---

Clack, clack, clack! I hooked up a second, identical contactor, this one not bolted to the unibody, and it's much more reasonable in terms of noisyness. Still louder than nothing, but nowhere near the volume of the shock tower mounted one ... probably acceptable.

So it looks OK if we can soft mount the pack +ve contactor and run it off the low pedal potbox microswitch. I'll probably still try to insulate it to keep it as quiet as possible...

[MetroMPG]

People keep asking me how much it costs to drive (recharge) the ForkenSwift. So without further ado...

After 20+ km of driving, the pack takes about 24 hours to recharge, using the slow, "semi-smart" (voltage sensing, auto shut-off) 24vdc/10A charger. That's splitting the pack, 12 hours on the back 4 batts, then 12 hrs on the front 4.

Yes, 10A is that slow.

3.6 amps (24v charger) at 120 volts AC =
432 watts total

24 hours per charge for 25 km
10368 watt hours per charge

@ 6 cents per kwh (standard Ontario energy mix rate - Ivan's house)

0.62208 $ per charge
25 km/charge
0.024883 $ per km
2.5 cents/km

@ 9 cents per kwh (100% renewable energy rate - my house)

0.93312 $ per charge
25 km/charge
0.037325 $ per km
3.7 cents/km

Compare to gasoline cost for the Blackfly @ 75 mpg (US) & $1 / L ...

3.14 L/100 km (3.14 = 75 mpg (US))
0.0314 L/km
1 $/L
0.0314 $/km
3.1 cents/km

[MetroMPG]

Just realized I can calculate the car's energy efficiency from these figures...

(10368 wh / 25 km) / 1.61 km/mi = 257.59 wh/mile

And that's based on charging. Discount the charging inefficiencies (as I believe is standard practice in talking about wh/mi), and it's better than that. I'm not sure what the charging efficiency is though.