Voltron, the Electric RG!

Post up a thread for your EV. Progress pics, description and assorted alliteration
User avatar
jonescg
Senior Member
Posts: 3004
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

Voltron, the Electric RG!

Post by jonescg » Sun, 21 Nov 2010, 19:16

The two bolts on the controller that look really close together are the motor cables. Since both motors are in parallel, they can touch and it makes no difference. But yes, it's frigging tight. Not a lot of room on a motorcycle for all the goods. There are two round pin plugs with all the throttle inputs, thermistors, axillary 24V power supply and so on. Of course these are about 700 mm long, and I need about 900 mm.

Most annoying thing for me now is the weight of the bike. It's a full 30 kg heavier than the original weight without me on it. As a result, the forks bottom out when you rock it forward Image Might still be a while yet before she's on track Image
AEVA National Secretary, WA branch chair.

7circle
Senior Member
Posts: 540
Joined: Thu, 04 Feb 2010, 02:18
Real Name: Ken
Location: Melbourne

Voltron, the Electric RG!

Post by 7circle » Mon, 22 Nov 2010, 01:06

The "30Kg's Heavier for the original weight", was that with a full tank of fuel and oil etc. Specs often quote Dry weight. But you might have actually weighed it with scales on the front wheel and one on the back wheel.

The batteries will stay the same weight so at the end of your ride the acceleration will still be the same.

May be that's a good thing as you know how the bike will handle as the weight doesn't change.

Is the 24V supply isolated from the Motor and Battery circuits?

I guess having the 12/24 DC/DC provides it if you need it.

How have you wired the motors for parallel.

Do the M+ and M- go to one motor first then the other.
Or have you got M+ to one motor and M- to the other and then two links between.

Just wondering of they will share power/torque evenly.

Probably many other issues with wear and mechanicals that would make a special connection not worth while.

{edit: sloppy grammar ]
Last edited by 7circle on Sun, 21 Nov 2010, 15:59, edited 1 time in total.

User avatar
jonescg
Senior Member
Posts: 3004
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

Voltron, the Electric RG!

Post by jonescg » Mon, 22 Nov 2010, 04:37

The weight of the chassis was 60 kg (forks with oil, tyres, frame, brakes, fairings and chain). The batteries are almost 65 kg. The motors are about 23 kg. The controller and wires etc make up about 7 kg. So yeah, this combined with the fact that much more weight is located over the front wheel makes it bottom out without much effort.

The Kelly 1200 A controller needs a 20-30 V axillary power supply in order to function. I am providing this through a small sealed lead acid battery and a 12-24 V step up converter. The contactor is 12 V so I'm running this directly off 12 V.

The Motor lugs on the controller are both M-. So I have one going to one motor and one going to the other. The positives are connected in parallel right after the contactor, off the battery + terminal. I have a 600 A fuse in the middle of the pack and a shunt on Battery negative for the cycle analyst.

They will share power and torque just fine - they are fixed at the shaft. If one is working harder than the other, it self-regulates bu slowing it down.

Observe:

http://www.youtube.com/watch?v=Bc4xMWoxnBg

   Image
AEVA National Secretary, WA branch chair.

7circle
Senior Member
Posts: 540
Joined: Thu, 04 Feb 2010, 02:18
Real Name: Ken
Location: Melbourne

Voltron, the Electric RG!

Post by 7circle » Tue, 23 Nov 2010, 03:28

Thats a strange Controller with two M- terminals.

Which model is it.
I did a search in this topic and on Endless and could find which one your using.

The newKellyController web site KDC range
Last edited by 7circle on Mon, 22 Nov 2010, 17:52, edited 1 time in total.

User avatar
jonescg
Senior Member
Posts: 3004
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

Voltron, the Electric RG!

Post by jonescg » Tue, 23 Nov 2010, 17:27

It is the KDB12121 or some bloody thing. It has regen and is good for 1200 A. I don't think they make this particular model any more.

This is the replacement, again, with two motor terminals:
http://www.newkellycontroller.com/hpm12 ... p-801.html
Last edited by jonescg on Tue, 23 Nov 2010, 11:26, edited 1 time in total.
AEVA National Secretary, WA branch chair.

7circle
Senior Member
Posts: 540
Joined: Thu, 04 Feb 2010, 02:18
Real Name: Ken
Location: Melbourne

Voltron, the Electric RG!

Post by 7circle » Wed, 24 Nov 2010, 02:57

Do you have all the documentation you need on your controller.

Thanks for the the pointer to the replacement controller in the HPM range.

You mentioned it has "again, with two motor terminals".

The manual for the HPM describes them as M+ and M- and uses them to get reverse.

They describe controller as "Kelly HPM full bridge or 4 quadrant motor controller provides fast and reliable electronic direction control."

So your controller having two M- terminals still appears very unusual.
Your motor would be two quadrant to get forward and regen.

I thought they made some dual motor controllers too that could use two M- terminals.
You also mentioned it can balance the power between the motors.

Sounds interesting I'm sure your keen to get it's weight balanced and suspension tuned so you can then start tuning the controller parameters.

Its looks lots of fun



Ken

User avatar
jonescg
Senior Member
Posts: 3004
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

Voltron, the Electric RG!

Post by jonescg » Wed, 24 Nov 2010, 18:04

At this stage I'm inclined to set the regen to zero. If just cracking the throttle mid-turn caused regen, I don't want a bar of it!

About the controller. It does not divvy up the power - it just delivers power to the motor. Since the motors are fixed at the shaft, overpowering one will simply cause the other to speed up and compensate.

Image
AEVA National Secretary, WA branch chair.

7circle
Senior Member
Posts: 540
Joined: Thu, 04 Feb 2010, 02:18
Real Name: Ken
Location: Melbourne

Voltron, the Electric RG!

Post by 7circle » Thu, 25 Nov 2010, 05:37

jonescg wrote: At this stage I'm inclined to set the regen to zero. If just cracking the throttle mid-turn caused regen, I don't want a bar of it!

About the controller. It does not divvy up the power - it just delivers power to the motor. Since the motors are fixed at the shaft, overpowering one will simply cause the other to speed up and compensate.


Thanks for posting the side view of the connectors from the manual.
I looked at the KDHB series but they don't list the 121 models and the 141 look different. It's a benefit that you have an extra terminal to wire to. and the extra internal bus bars are a benefit too. (Don't know why my search for kelly in endless under your topic came up with nothing)

If you where backing off throttle on the ICE in 2nd or third gear how much engine braking would you expect. Once you have all your 235kg (=60+65+(11.5+11.5)+7+80) zipping along the motor regen may be fine.
Its all about current control of the motor and the resulting torque.
If the controller only has a switch FET between the M- and the B- they usually have a Diode that allows free wheeling to the B+. This may not allow you to have zero motor load when you back off on the throttle. The motor Bemf will keep the current flowing through the diode and cause a braking load on the motor. This won't be charging the batteries so its not regen. Its just motor braking. But maybe your controller has a FET also bewteen the B+ and the M- terminals. So the regen can be controlled.

Also if the motors don't have the same Kv due to diffences in the Permanent magnets have different fields (over temp can degrade the permanent magnets) they can force current between each other. a bit like connecting a fully charged battery to flat battery. The current will flow between them. When they are in series they can do this. So when at low RPM's when there is very little Bemf from either motor its not a problem so going to Twin Series Motor connection wont fix it. (But the problem's not there yet so just keep it in mind for the future)

But don't take my words as gospel as your way ahead of me.

What version of their software are you using?

If you can show some screen shots of the parameters that would be ace!

Have you got a clamp meter that can give measurements above the PWM ripple.

I saw some new cheap ($130) jaycar/electus clamp meters with true RMS, its quoted for only 10kHz.

To know what is really happening in the circuit will be very helpful.

The twin motors might be on the same shaft but that doesn't mean they will share the load evenly.

The brush advance will be critical in power/torque sharing.

So measuring the actual current through each motor will tell you if they are sharing the torque. But its abit hard with out a dyno. Having two current sensors while your riding that record average or peak could be helpful.

The other of test is to touch it (or use a thermometer) if they are the same temp then that's what you want.

The IR non-contact temperature sensors that can see in between the grill to measure the core temps. The Closer you can run to the max temperature of the motor the more torqe/power you can get through the motor.

The controllers only allow for one thermocouple. Unless your model has an extra one, which could be possible.
May be a switch between them could work.

But better to monitor the peak/max temp of either with a schotkey diode with 0.35V drop from each thermistor resistor divider. This would allow the thermistor at the higher resistance to dominate the voltage feeding the controller sensor circuit.
The diode adds a predicable offset that you can adjust for n the temperature setting.

As long as the controller has a much higher input impedance than 100k (like 1Mohm) other wise the resistor values or other software parameters would need to be adjusted from the default for the KTY83-122 sensor.
(that's if you want to bother with temperature monitoring)

Have you measured the weight on the front and back wheels.

Most bathroom scales can go to 200Kgs. You might need to but a chock under the other wheel not being measured to keep the bike wheels at the same height when measuring the mass if you only have one scale and swap it from wheel to wheel.

You could also see the change with you on the bike in riding position. If you lean against a wall at a very small lean angel so your feet are on the pedals of get someone to hold you still.

Hopefully the suspension people can use this to setup the right springs for you.

Go Jonescg your nearly their.


Ken

User avatar
jonescg
Senior Member
Posts: 3004
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

Voltron, the Electric RG!

Post by jonescg » Mon, 06 Dec 2010, 19:19

Hi all,

It was a slow Sunday for progress, but I did manage to set everything up for charging. Made up some bus-bar connections and wired them all in series, plugged in all of the BMS leads and set up the master control unit. My charge leads are good, with an Anderson clip at one end. I soldered these suckers with the help of a mini butane torch Image These things are SO cool!
Image

But since I didn't have a 16 mm2 ring-lug I had to bodgy something up with a few washers. I had the Elcon 10 A charger ready to go, but it refused to acknowledge the battery. Turned out I had the charger cut-off lead wired into the wrong ports :oops: . Anyways, it seems to be working.

Image

When plugged in, they all glow green indicating they are above 2.5 volts, or some low voltage point... or just indicate they are simply plugged in :) Switch the charger on and watch the pack voltage rise! And rise... and rise until it gets to 124 V
Image

At this point, the red lights are coming on indicating they are above 3.7 V and are shunting current into the copper end of the BMS units. Man do they get hot:!   Some got to 4.0 V, while others, still green, are just getting up to 3.4 V. Rather than let it do damage I just turn the charger off and wait. Eventually they cool down and the red light fades. I suspect the voltage drop between the BMS unit and the pack is enough to potentially do damage, so I really need to shorten them up as much as possible.

So really, I need to wait until the dedicated 118 V (maximum!) charger is ready. Oh, and to check the failsafes, I had it charging and unplugged one of the packs from the BMS board. Sure enough, the alarm sounds, red light on the master unit comes on, and the charger is switched off. Brilliant Image
Image
AEVA National Secretary, WA branch chair.

User avatar
EV2Go
Senior Member
Posts: 2059
Joined: Wed, 16 Jul 2008, 00:21
Real Name: Paul
Location: Brisbane 1963

Voltron, the Electric RG!

Post by EV2Go » Mon, 06 Dec 2010, 19:31

You have to love when things work with minimal correctional effort (better yet none at all) Image

Can't wait to hear how this thing goes it is one classy conversion.
Last edited by EV2Go on Mon, 06 Dec 2010, 08:32, edited 1 time in total.

User avatar
jonescg
Senior Member
Posts: 3004
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

Voltron, the Electric RG!

Post by jonescg » Fri, 31 Dec 2010, 16:39

Thinking about the ignition wiring... The off-ACC-on switch may or may not keep the ACC circuit on. Not that it matters for pre-charging the caps, because they will hold their charge for as long as it takes to break the precharge circuit and complete the contactor circuit. But still, I'm wondering if the precharge relay is necessary. I'm worried if I'm not careful I'll have the might of 106 V worth of A123 coming back up to meet me at the switch.
What do you guys reckon?

Image
AEVA National Secretary, WA branch chair.

User avatar
evric
Site Admin
Posts: 500
Joined: Sun, 20 Jul 2008, 01:57
Real Name: Eric
Location: Adelaide SA
Contact:

Voltron, the Electric RG!

Post by evric » Fri, 31 Dec 2010, 17:23

This arrangement looks good, even though I've never seen an ignition switch switch the negative before. As long as you pause on the ACC position for a few seconds this will work well. What's to stop someone going through ACC quickly and straight to ON ? Training ?
Prius Plug-in Conversion: http://www.evplus.com.au ...Holden Barina EV: http://www.evric.kestar.com.au

User avatar
jonescg
Senior Member
Posts: 3004
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

Voltron, the Electric RG!

Post by jonescg » Fri, 31 Dec 2010, 19:54

It could easily be on the positive too. Just how I drew it late at night with MS Paint Image. Yes, only training will stop a noob from belting right through to on. I can't think how else you might stop this, but it's better than any previous circuits I've come up with so far. I think the second, smaller precharge relay is necessary, since without it I would have full line voltage coming up to the switch, and potentially trying to charge the auxiliary battery! Unless a diode would work?
AEVA National Secretary, WA branch chair.

7circle
Senior Member
Posts: 540
Joined: Thu, 04 Feb 2010, 02:18
Real Name: Ken
Location: Melbourne

Voltron, the Electric RG!

Post by 7circle » Fri, 31 Dec 2010, 21:57

You can use a timer interlock like a relay that has a resistor and large capacitor that closes after 5 seconds.

Or an opto coupler that turns on when the voltage reaches 70% on the Buss caps.

Its worth considering interlock to stop the two steps.

I was looking at waht could be done with the Start postion on most cars/bikes.

So you need to hold the Key in the start pos for 5 seconds until the precharge is reached before the ON will keep the main contactor ON.

If no one already has a sample schematic I might draw one up.

7C

User avatar
jonescg
Senior Member
Posts: 3004
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

Voltron, the Electric RG!

Post by jonescg » Tue, 04 Jan 2011, 05:22

Hi all,
I have come up with this schematic for turning on the CA at the same time as the BMS and currently, the 12-24 V converter. Unfortunately the 3 position ignition switch probably only completes the precharge circuit while it's in that position, but when it's moved to ON (contactor circuit) it would turn the precharge circuit off.

At least the 12/24 V converter and controller don't draw much of a current. And besides, the BMS master unit should be on whenever you are charging or discharging. Sitting idle, all it does is flatten the battery.

What do you think of this one? The box including the CA is my dashboard and it has a key switch for the 12-24 V converter and controller circuit, as well as the BMS. The relays are dual circuit ones, but unless I can be sure that the ignition key will keep both circuits complete I don't think I can piggy-back on one relay. Points a and b are potential sites for the charge leads. I will potentially be able to monitor charge into the battery through the cycle analyst.

Image

I still haven't put much thought into a failsafe for trying to throttle off in precharge mode, but I'm not real good at electronics Image
Last edited by jonescg on Mon, 03 Jan 2011, 18:25, edited 1 time in total.
AEVA National Secretary, WA branch chair.

User avatar
evric
Site Admin
Posts: 500
Joined: Sun, 20 Jul 2008, 01:57
Real Name: Eric
Location: Adelaide SA
Contact:

Voltron, the Electric RG!

Post by evric » Tue, 04 Jan 2011, 13:47

I often try to drive off in precharge before I "Start", so I am redesigning my start sequence electronics. I am going to use the "Start" position to enable the contactor and controller but only after the precharge delay. So mine will precharge on the "ON" position.
Prius Plug-in Conversion: http://www.evplus.com.au ...Holden Barina EV: http://www.evric.kestar.com.au

7circle
Senior Member
Posts: 540
Joined: Thu, 04 Feb 2010, 02:18
Real Name: Ken
Location: Melbourne

Voltron, the Electric RG!

Post by 7circle » Tue, 04 Jan 2011, 14:35

Don't mean to be pedantic, I would suggest showing the fuses in the schematic.

Sometimes the controllers have a output that says they are on.
This could be used to allow KEY-POS-3 to turn the main contactor on.

Maybe you can flip the KILL-SW so it doesn't cover the return wire.

If you can show on the Controller what model it is "KDCxxxxx ..."

The manual KDHxxxxB has functions on the J1 and J2 connectors that can be used.

Like J1-pin3 is Main Relay control. (Don't forget the free-wheeling diode)
Their schematics don't show a small relay on the Pre-charge.
But I think you need to use one.

Not sure what you KDH connector functions are.

Image

User avatar
jonescg
Senior Member
Posts: 3004
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

Voltron, the Electric RG!

Post by jonescg » Tue, 04 Jan 2011, 15:55

That's a good thought Ken, I don't know what all of the options are for the Kelly controller, but if it cuts the throttle signal until the caps are properly charged it might be worth using the in-build system instead.
AEVA National Secretary, WA branch chair.

Catavolt
Groupie
Posts: 170
Joined: Mon, 15 Feb 2010, 14:57
Real Name: Jon Eggenhuizen
Location: Newcastle
Contact:

Voltron, the Electric RG!

Post by Catavolt » Tue, 04 Jan 2011, 19:34

Be mindful of how you wire the dual motors to use equal cable length to balance the load.
Image

User avatar
jonescg
Senior Member
Posts: 3004
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

Voltron, the Electric RG!

Post by jonescg » Tue, 04 Jan 2011, 19:53

M- cables are identical. M+ cables vary by about 30 cm. Hope this doesn't make a huge difference. I guess I won't know till I can put a current clamp on each side.
AEVA National Secretary, WA branch chair.

User avatar
jonescg
Senior Member
Posts: 3004
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

Voltron, the Electric RG!

Post by jonescg » Wed, 05 Jan 2011, 01:02

Here is the new and improved ignition diagram. It leaves the precharge resistor in place, since it probably only draws a significant load as the capacitors charge up, but not a lot thereafter. It also allows for monitoring of the current in with the CA (which is turned on by a switch on the dashboard, along with the BMS master unit). Charge leads would go on positions A and B as indicated. The main ignition switch now turns on the 12/24 V converter, which in turn fires up the controller and by means of the leads from the controller, the contactor.

Image

What do you think?


(I know, I think I should get back to writing my grant application Image)
AEVA National Secretary, WA branch chair.

7circle
Senior Member
Posts: 540
Joined: Thu, 04 Feb 2010, 02:18
Real Name: Ken
Location: Melbourne

Voltron, the Electric RG!

Post by 7circle » Sat, 08 Jan 2011, 17:10

Is J2 Lead 2 "24Vout" a current flowing in or out of the KDH12121B?

You might want to use a spare relay to disable the pre-charge resistor to stop leaching loads when the Bike is sitting unused for a while.
Sounds like your aware of the issues.

Can you post your file as a zip.
I can then send you some changes via PM if you like.

User avatar
jonescg
Senior Member
Posts: 3004
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

Voltron, the Electric RG!

Post by jonescg » Sat, 08 Jan 2011, 17:32

J2 lead 2 is -ve 24 V into the controller. The 24 V out is referring to the 12/24 V converter. I have thought long and hard about it, but the risk of doing something silly like attempting to ride off in precharge mode, or something like it is too high. The controller effectively opens the relay, which is great. However, I seem to have done something to the controller itself in that it won't spin up any more. I thought I'd wired it up wrong but when I changed the throttle sensor lead to presumably the right one, it gave a throttle sensor error - in Chinglish: The throttle signal is higher than configured dead zone at power on. I'm frigging around with it today, but hopefully it has nothing to do with the massive spark while soldering the precharge resistor on Image
AEVA National Secretary, WA branch chair.

User avatar
evric
Site Admin
Posts: 500
Joined: Sun, 20 Jul 2008, 01:57
Real Name: Eric
Location: Adelaide SA
Contact:

Voltron, the Electric RG!

Post by evric » Sat, 08 Jan 2011, 18:40

This question may a little late but...Why are you soldering while it's all connected up?
Prius Plug-in Conversion: http://www.evplus.com.au ...Holden Barina EV: http://www.evric.kestar.com.au

User avatar
jonescg
Senior Member
Posts: 3004
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

Voltron, the Electric RG!

Post by jonescg » Sat, 08 Jan 2011, 20:00

Hehe, it's cause disconnecting the battery +ve is a royal PITA. No, the spark happened while doing some testing. All good - the motors spin up like they are supposed to. I had to change a few settings on the Controller. All good now Image

However, I'm finding that my Dodgytech current clamp is giving strange results. The Cycle Analyst tells me that when I wind them up to almost full speed they are pulling 30 amps (no chain attached). Yet the current clamp is giving different results depending on whether it's AC od DC, and whether it's on the + or - battery leads, or the motor leads... Who do I trust?
AEVA National Secretary, WA branch chair.

Post Reply