No pistons, no fuel, no worries! Toyota Sera EV Conversion begins…

[greywolf]

Did you weigh the bike pre and post conversion ?

[VA61DETT]

Ah forgot to get a pre conversion weight,
But post is 15Kg Front / 20 KG Rear

So yeah, A little heavy!

-A

[greywolf]

I can't imagine you'd want to be carrying it too often, but handy to know if you run out of battery and have to start using leg power alone.

What's the charge times like ?

[egamirrorim]

When you said 'bike' I thought a motorcycle style bike lol!

[brett_celicacoupe]

hey mate,

Do you have the ZEVA fuel gauge driver? i thought i read earlier in this thread about the drifting problems. i was thinking of getting one but am worried about the same thing.

I was thinking that there is a reasonably straight forward way to deal with this. you could setup a transistor-zener diode circuit triggered by a high battery pack voltage which drives a relay in series with a 10000uF cap. this should cycle the relay on/off for a second or so and reset the fuel gauge driver as the battery pack voltage passes through a trigger point.

you would need to block this circuit with a NC relay driven off the 240AC charge circuit. otherwise you would get the condition where the batteries arent naturally above the 'full charge' point but the charger has brought them above it. this may also need a capacitor delay to account for the slow battery voltage drop after the charger is removed.

As long as the circuit is consistent, the fuel gauge driver could be calibrated to suit.

thoughts?

[VA61DETT]

-brett_celicacoupe

I am using the TBS E-pro meter, I believe my charge state drift is now due to a defective cell.
One of the units is coming up to voltage way before the rest of the cells & causing the charge to terminate earlier then expected (only about 1-2%).
By separately charge cycling the individual cell we managed to get the meter to trigger a 'full' state, though I think the cell will need to be replaced eventually.

I briefly looked into the zeva gauge, it would be nice to have the fuel sender working as a basic state of charge gauge, that way its a no-brainer for anyone driving it.

I am not too sure what you are getting at with the capacitor trigger stuff?
Any good battery gauge (zeva included) will actually count the amp hours in & out of the battery in order to determine the state of charge.
Simply looking at voltage is not really accurate enough. *Though for lead acid cells the 'open cell voltage' is apparently a good indicator of SOC.

-greywolf
Charge times ~4 hours for full charge with 200W charger,
I have tested the range a little over 42km with a moderate ~%25 pedal power input.
Though when the battery shuts down (around 32v) it becomes a massive drag to pedal due to the motor cogging.

[greywolf]

No clutch to disengage the motor for free pedalling ? Or is that only adding to the weight ?

[VA61DETT]

-greywolf

Would be nice but not really feasible with the hub motor set-up.
Its the price you pay when having the regen capability.

[greywolf]

I suppose after testing is out of the way and you know your range, it's all about planning your trips accordingly.

[4lex]

Just spent an hour reading all of this. Love your work!

What's the news, three years in?

[VA61DETT]

Hi,

Funny you mention that.
Just in the process or upgrading & documenting a few of the systems on this.
Have ran about 20,000 Electric K's
The only major issue was with the motor controller failure due to heat stress.
Since installing the upgraded cooling system it has not overheated once.
We have had one cell go a little faulty, it would cause early charge termination & SOC reporting issues.

Was out at the TECCWA C-red dyno day last weekend:


Whopping 51 Hp!
Which Was a little disappointing, considering it was pulling around 430A @ 150v (~64kw).
Doing the math reveals about 30% power loss from battery to rollers.
Though this is probably about right for the efficiency of the motor & drive-line losses .

Here is a clip of the least noisy run of the day...


Currently in the works:

1. Upgrade of brake vacuum pump:
Going to a ford OEM hybrid part I picked up surplus. Quieter operation & far less power draw.

2. Addressing persistent battery sync errors:
Still have issues with the E-pro meter not correctly syncing up on charge, running some tests to try and iron this out once and for all.

3. Adding Yaris electronic assist power steering:
This will be a interesting experiment, had a friend run one of these units in his Starlet & it works like a champ.
Currently sourcing some parts from the UK

4. Upgrading the 12v systems:
Adding a secondary DC-DC converter for the steering assist & cleaning up a few of the 12v systems.

5. Upgrade TPS:
The old 5MGE throttle position sensor is getting a little long in the tooth, going to investigate into using a hall-effect or industrial pedal mech.

Watch this space!

[VA61DETT]

Ok,

Just finished installing & testing out the installation of a 2007 model Yaris EPS module into the 'Leccy car.

Here is some nifty info:

EPS controller circuit


Steering column comparison


Steering drive shafts


Shaft machining & joining


lower servo mount


Joined columns


Modified UNI joints


EPS Wire diagram


Testing!


There you go, easy as pie!

[Ae95/4agze]

So with 07 eps ecu with abs is it a can communication? How are you getting it to provide assit? Speed signal input or just providing power to the d31 pin 6 connector?

[VA61DETT]

Just give it power & the ignition on signal.
When you move the wheel it will be stiff at first, but after a few seconds the assistance will kick in.
It appears to remain active once it have been 'woken up'.

From reading the service manual, it appears that if communications are lost from ABS module or ECU it will default to normal assistance.

You can tweak the feedback level by interrupting the torque sensor voltages.
There is a good article on the 'auto speed' website on how to do this.

-A

[Scuzzy]

Had you installed a non-power steering rack to remove all the hydraulic lines etc from the front?