[m38mike]

I'm new to this site, just discovered it Christmas day. I've been interested in EV's for a while and when diesel hit $5/g I decided to get invested in an EV. I rebuild old Willys Jeeps as a hobby, so I found one that had all the right things wrong with it, and started the conversion process.

I've been working with a shop near my home that converted ICE cars to all electric. Unfortunately the 3 owners each had different ideas and the company folded. So now I've got a complete conversion that performs like a golf cart. I went to them with these requirements. I have a 35 mile daily commute - round trip - so I wanted to be able to drive 40-50 miles on one charge. I don't have the option of charging it at work. Most of that drive is 2-lane 55mph highway. I don't want to be a speed-bump on that road.

Here's what is in the 1952 Willys Jeep right now:
Netgain WarP-11 motor
Kelly controller - KDH14650B
18 GC8 Napa 8v batteries
Zivan PFC1500 charger
500 amp gauge

I understand the basic concepts of how all this should work, but I'm not very smart - technically - in this stuff. So I'm posting this to see if any of you can help me with what I need to diagnose and what I might need to change to make this little jeep get up and go.

My top speed in 3rd gear is about 25mph. I'm pretty sure that the Warp-11 can push that jeep faster than that. The max reading I've seen on the amp gauge is about 110-115 amps.

What is holding me back???

Mike

 

[TheSGC]

I think your Kelly Settings need to be changed through a serial port and their software. Particularly the Current Mode and the Low Voltage Dectection, and probably a few other things. Also is your Kelly on a heatsink?

I have a 96 Volt Civic EV using a Kelly KDH09401 and I was getting 80 AMPs max before adjusting my settings and now I get 180+ AMPs from the batteries. I set my Control Mode to Speed, all the Current/Speed Settings to 100%, the Throttle Ramp to 3 and my Low Voltage detection to 77 volts for my 96 volt system. The Low Voltage Detection (shutdown) starts at 110% of the value you set it, so most likely yours is set pretty high and the batteries sag low enough under load to activate the slowdown/shutdown.

I would suggest setting your Low Voltage to 118 volts for your 144 volt system.

 

[dtbaker]

you might want to also check all your battery cable connections, make sure they are tight, and after a few minutes of driving make sure none are warm (indicating resistance).

 

[tj4fa]

I learned how to drive on my Dad's 52 Willy's. He and my uncle pulled out the Flathead 4 and put a 327 in it. That's back when ICE/gas was popular.

I WANT TO SEE PICS DAMMIT!!

 

[m38mike]

Thanks TheSGC, we've reset many of the parameters on the Kelly trying to get better performance, but I don't think we reset the low voltage shutdown that low. But in thinking about what you said, it makes sense that the shutdown setting may be what's cutting us down. We have a computer fan wired in to blow cool air onto the controller to keep it cool. In checking the air stream, we haven't noticed any real heat. It's been warm, but not hot.

dtbaker, thanks for the idea. Having repaired lots of very old electrical connections while rebuilding 50-60 year old Willys, that was one thing we checked regularly. At 144v if one connection is just a little loose or too dirty we'd have a big impact on what we could get from the pack. So far every connection has remained cool to touch.

tj4fa, I'll see about getting some photos posted by this evening. It really turned out to be a nice looking set up. You'll see soon.

m38mike

 

[Guest]

A small computer fan may not be enough after you change your parameters and begin to draw more current and for longer periods of time. My Kelly heated up to where I could not touch it within 3 miles. It is not installed at this time. We will be installing the motor and controller into a Dune Buggy for further testing soon. My Kelly is different in that it's a SepEx controller and it worked from the start. However I never installed a proper heat sink or fan. Have a look at this setup:

http://www.evalbum.com/1798

Thanks TheSGC, we've reset many of the parameters on the Kelly trying to get better performance, but I don't think we reset the low voltage shutdown that low. But in thinking about what you said, it makes sense that the shutdown setting may be what's cutting us down. We have a computer fan wired in to blow cool air onto the controller to keep it cool. In checking the air stream, we haven't noticed any real heat. It's been warm, but not hot.

dtbaker, thanks for the idea. Having repaired lots of very old electrical connections while rebuilding 50-60 year old Willys, that was one thing we checked regularly. At 144v if one connection is just a little loose or too dirty we'd have a big impact on what we could get from the pack. So far every connection has remained cool to touch.

tj4fa, I'll see about getting some photos posted by this evening. It really turned out to be a nice looking set up. You'll see soon.

m38mike

 

[tj4fa]

...Have a look at this setup:

http://www.evalbum.com/1798

Very Nice EV!!

Might I recommend you add a couple wind generators or maybe a magnet motor to get some free energy out of that thing.

 

[JRP3]

To add to what others have said it sounds as if the controller is really holding you back. Many people have had trouble getting a controller to work well with the Warp11. Only the unavailable Zilla seems to be reliable with them but hopefully you can get your Kelly to play nice.

 

[m38mike]

Here are a few photos of my EV so far.

This photo barely shows the 10 battery rack in the back seat area. Each cable was custom built for the specific location where it's used. The weight load is well balanced from front to back, and still within the GVW capacity for this 56 year old Jeep. When it's ready to commute, it'll have a full hard top covering the seats and rear rack in winter, this is the summer setup.

This shows the 8 battery front rack above the motor. This was the original 8 inch motor that was installed. When the performance failed to live up to our expectations, we swapped this one out for a WarP-11. Up to now, there has been no appreciable improvement. Knowing what a WarP-11 has powered, that tells me that we have a choke point somewhere else in the system.

We put all of the controls under the driver's seat where the gas tank used to be. Lots of room, good ventilation, protection from the weather, and easy access for adjustments. The contoller is inside the green ammo can on the left of the charger. We added a 100cfm fan to the top of the ammo can to move air around the controller. After reading about heat sinks on this site, I may add one of them too.

This shows the original 8 inch motor under the fron battery rack. The rack bolts to the frame and can be removed as a complete unit in less than 10 minutes. We did it a couple times when replacing motors or making other adjustments. You can see the front drive shaft and differential. We kept the 4x4 because of the snow we get here in this Colorado valley.

The motor shaft is bolted to an adapter plate that slips over the original transmission shaft. No clutch, so it's direct to the transmission.

 

[tj4fa]

Sweet...

A little GIJoeJeep!!

WAY COOL!!

 

[m38mike]

gottdi,
yea contoller temp is something we'll be watching very closely. The 100 cfm muffin fan has done a good job cooling it so far, but I think a heat sink to increas surface area will be a good future add on. I like the setup you linked. Very neat work.

JRP3,
It sounds like the controller IS my choke point. We're going to try adjusting the settings again to see if that helps. I wish I could get my hands on a Zilla. I've heard good things about them. In the mean time, one of the former owners of the conversion shop that I was working with is also an EE, and would like to try building a controller for me. He has lots of parts from a Viper conversion they started but never finished before the company broke up. (poor Viper!) He would like to start building controllers commercially similar to the Zilla, but he said a bit more robust. He likes the jeep as a test bed since everything is so easy to get to.

tj4fa,
yup, it's a 52 M38. Made by Willys for the Army during the Korean war. I've got 4 of them in different stages of rebuilt. I like it for the electric conversion because it is so simple mechanically. Everything is easy to get to, and it has a very strong frame to carry the extra weight. The major drawback to it is the aerodynamics. It has all the grace and curvature of a brick!

Mike

 

[WSJ]

Great job Mike!

Wayne

http://public.fotki.com/WAYNEWANG/electric-car-beetle-ev-/
http://www.evalbum.com/1798

 

[m38mike]

Well, we've been playing with and adjusting the Kelly GDHB controller for 2 weeks now, and we still are not passing the amperage we should be. We did reset the timing on the WarP11 to zero, and that made a big difference in our performance, for the good.

With all the possible combinations in the Kelly GDHB 14650 controller I'm not sure what we need to do next.

The batteries we have are Deka GC8V. 18 - 8v wet cell golf car batts for 144v. The 20 AH rate on them is 165. The minutes discharged at 75 amps is 85 minutes. The Deka web site didn't list a 6 AH rate for these batteries. Should I be able to pull 300 amps for 30 minutes out of these batteries???

 

[dtbaker]

Should I be able to pull 300 amps for 30 minutes out of these batteries???

I doubt it....

You'll have to find curves from the battery manuf, but when you start pulling big amps like that you will run into Peukerts Law pretty quick with Lead, then voltage sag, with 'performance loss'.

A pretty good explanation of this is at
http://www.caravanandmotorhomebooks.com/articles/peukerts_law.htm

As a rough number I would guess you will be unlikely to pull more than 100 or so amps for any 'sustained' length of time from typical Lead ...

 

[JRP3]

I'm sort of surprised the neutral timing made an improvement, exactly how so? I think you'd get more torque out of it by retarding the timing but I also think you're risking arcing in the motor which could damage it if you're running 144 volts. Also, do you really draw 300 amps for 30 minutes? That seems like a lot.

 

[WSJ]

Battery life is shortened as the average current increases.
An 8v 165 AH golf cart battery should not exceed 165 amps
for 5 seconds and 300 amps for 2.75 seconds.

Wayne

http://public.fotki.com/WAYNEWANG/electric-car-beetle-ev-/
http://www.evalbum.com/1798

 

[m38mike]

I think i've been thinking about the amperage available from my batteries all wrong. The batteries have a 20 AH rating of 165. That means that over 20 hours, each one could provide 8.25 AH on average per hour, for a total of 165 A. So cumulatively, 18 batteries discharging at this rate could provide 148.5 Amps per hour for 20 hours. Have I got that right?

If it is right, that means I should be able to draw perhaps twice as many amps for perhaps one quarter of the time. I guess that would depend on the power curves for this type of battery. And that's where Peukert's law comes into effect. The faster i draw energy, the less potential energy there is in each battery. So if I used the energy in the pack lightly over 20 hours I could get about 2,970A total. If I use the energy over one hour I may only get 1,000A because of Peukert's law. Does that sound right?

By resetting the motor timing to zero I think we did get much more torque out of it. I know that with my current gearing, if the vehicle is going 55mph then the motor is spinning at 3,200rpm. I guess we need to find out what the optimum advance setting is for some point in between 0 and 3,200 rpm. And to find out if the Warp11 can sustain 3,200 rpm without burning out. And can my battery pack provide the energy to keep the motor spinning that fast if I'm going up a 2-3% grade for 2 miles.

 

[m38mike]

Yesterday I talked with one of the guys at Netgain about my WarP11 motor. On setting the timing, they said there are 3 possible settings: advance ccw; neutral; and advance cw (clockwise). The shop owner I'm working with on this conversion told me that they had reset the timing to neutral. The Netgain tech said that was a sure way to burn up the motor. He added that in my application it needed to be advance ccw. I talked with the shop mechanic doing the work and it turns out they had set the motor to neutral and it ran poorly so they set it back to advance ccw and it ran much better. I hope they didn't damage the motor when it was set back to neutral.

The tech also advised me to change brand on my controller. We've been struggling with the Kelly to get anything but mediocre performance. I think it's time for a change. The tech also mentioned that Zillas are supposed to start production again by this summer. I don't know if I need a Zilla for my application but that sounds like good news.

I'm thinking about buying a Logisystem 550 controller. based on the motor power curves, I'll need about 175 amps to spin the motor at 3200 rpm to keep the jeep at 55 mph. Will a Logisystem 550 give me this?

 

[WSJ]

When batteries are connected in parallel, the capacity of each battery is additive. With 18 batteries in parallel your pack would provide 148.5 Amps per hour for 20 hours at 8 Volts or 1164 kWh at 77° F. When batteries are connected in series, the capacity of each battery is not additive, because the current flows through each battery. With 18 batteries in series your pack would provide 8.25 Amps per hour for 20 hours at 144 Volts or 1164 kWh at 77° F.

The capacity of the US BATTERY, US 8V GC is 94 AH @ 75 A @ 77° F, various discharge rates are specified at the following link;
http://www.usbattery.com/usb_images/USB%20Capacity%20Chart.pdf

The LINK-10 displays battery pack Voltage, Current, AH or kWh and % Charge Remaining. It’s a must have instrument in an EV.
http://www.evalbum.com/tech/e-mtrpdf.pdf

I have a GE motor which takes about 75 A of battery current at 55 MPH. Have you taken some battery current measurements at various speeds?

 

[dtbaker]

I'll need about 175 amps to spin the motor at 3200 rpm to keep the jeep at 55 mph. Will a Logisystem 550 give me this?

without Mozilla, it seems the next best bulletproof controller may well be the Curtis line.