1,000lb "Elec-tow’d” Electric Tow’d Dune Buggy Build Thread

I feel for you, so much meticulous effort and thought, only to be let down by a OEM unit ..that sucks !
Have you given up hope of any help from Kelly ? Or their agents ?
Maybe a bit of " Exposure" for them on social media and a few forums, may stirr them into action !
Start a separate thread on a few tec forums focussing on Kelly's crap products and support ?
Have you sought help or advice from over on the DIYEv car forum ?.. There is a lot of experience with these things over there.
I can understand you needing to "recharge" you own energy bank, at the moment.
Good luck, you know the solution is only a few amps away !
 
I love the project stay on it and finish it. Your so close to an all-some ride. Pictures are fun for all, but focus your the efforts on the product not picture's.
Complete the project and gloat as well deserved, and than post your picture's and videos. I am sure those interested will all be there to celebrate your victory.
I wish I had your skills and tenacity.
 
I finally got around to testing the zilla controller. Before connecting it to power, I wanted to inspect the internals for damage/heat/water etc. Case removal was straight forwards, IIRC 6 screws around the ends, and 6 more screws through the bottom plate.
.

View attachment 10

03IMG_5154.JPG

04IMG_5142.JPG

05IMG_5152.JPG

07IMG_5148.JPG

09IMG_5145.JPG

11IMG_5150.JPG

13IMG_5147.JPG

15IMG_5151.JPG

17IMG_5162.JPG

19IMG_5164.JPG
.

Click here for the Zilla 2K manual.

I setup some basic testing to see if the controller worked. When I power the hairball by connecting 12v(-) to #1, and 12v(+) to #2 and #3 and #4, the red error light comes on and stays on. 12v LEDs from #9 (chck engine) and #10 (batt) to 12v(-) stays on. Same behavior with or without cat5 to controller, and pot across pins #6 and #7.

Well, that means it is time to dump the Diagnostic Trouble Codes to see what is wrong. The "check engine" LED will blink out fault codes if you insert a special serial shorting plug. Being used, this controller did not come with a the short jumper, so I made an RJ11 6-pin connector that jumpers pin #3 (RX) to pin #5 (14v+), and tried the "4-second" insertion method of getting DTC, but the LED on pin #9 (Check Engine light) stayed on and never blinked.

The controller also did not come with a serial cable, so I fashioned a cable with the following pinouts, but hyperterminal at 9600/8/n/1 yields no response.


RJ11 ***** DB-9 connector
Pin 1 RX Pin #3
Pin 3 TX Pin #2
Pin 6 GND Pin#5

So I'm doing something wrong, or the Hairball has been hacked up; I sent an email off to Manzanita Micro, asking for their guidance...
 
Good News!  Rich at Manzanita tested my hairball interface, found nothing wrong with it, and pulled the error codes from it.  I was going to buy a new one for expediency, but he convinced me to send it in for repairs instead, and 1/2 hour of his time was way cheaper than a new one - thanks!

Of course, that means I am a spaz for not being able to pull the codes myself, I must have wired the cables incorrectly.  I ordered the factory serial cable and factory shorting plug to resolve that issue.  Hopefully it will be here in time for the weekend, so I can plug it into the power stage and see if I can spin up a small test motor with it.

If I can demonstrate that the power stage is operational too, then I will be eager to start Phase 2 of this build - the Zilla upgrade. (Tow'dZilla?)  Of course, I am all sucked into my 3D printer kit build, and both projects are huge fun, so it will be hard to choose which one to work on. (1st world problems)

-JD
 
Today is the 2-year anniversary of taking delivery of the Tow'd. :shock:

It's still running on the 500a(88a) Kelly controller so it doesn't get out much at all. I did some testing with the 2,000a Zilla2k that will replace the kelly, and found that the power stage was broken, so I did buy a defective controller. :oops: However, Rich at manzanita micro repaired and replaced the failed parts quickly and for a very reasonable price, so now I have a working Z2K for a fraction of MSRP. It is an odd feeling to have a domestic EV vendor I can depend on.

However, between having our house repiped, knee surgery, and work issues, I haven't been able to carve out time to open the tow'd back up and install the Zilla. When I make progress again, I will post it here!

-JD
 
Awesome! I wait patiently for a video wth the zilla!!!!!
 
.
I never got around to publishing my last post, so this week you get (2) at once!


I've gotten a lot done since my last post, although much of it has been engineering a control system new wiring layout - which doesn't photograph well. Essentially, I had to re-wire the whole thing, figure out what stays and what goes, and it took forever. Three things are driving the do-over:

1) Moving the controller from the water-jetted mounting plate, to the trunk above, meant I needed to extend most of the wiring harness to the new location. I also had extend wiring for reverse switch on the tranny and other sensors etc.

2) The zilla manual and FAQ/Blog repeatedly mention its vulnerability to high-voltage noise, so I wanted to isolate the high-voltage CA wiring from everything else, so I moved them to a dedicated shielded cable. I don't think the tiny amp draw of the CA is really going to be an issue, but it is a lot easier to plan it now than change it later. Freeing up that cable meant moving a lot of things around, so most of the control wires have a different job now.

3) The Zilla Control scheme is much different from the Kelly's. For example, the zilla controls the main contactor, whereas the Kelly did not so I drove it with a dash switch. The Kelly monitored the motor's thermistor, but the zilla does not. The Kelly had (1) power lead, whereas the Zilla has (3) - "Always ON", "power on" , and "Start" - to better emulate an ICE experience, where the typical ignition switch has an Aux, On, and start positions.

To make this work, I connected the "always on" wire to pushbutton Dash Switch #2 so I can disable the controller power drain if I need to. The "power on" wire is connected to Dash switch #3 (note: the zilla also uses this power source to energize the coils on the main contactor). The Start wire connects to the starter terminal of the ignition switch; when it sees a momentary signal, the hairball precharges the power stage, then activates the main contactor to let power through. This is different from the Kelly too, which had to precharge through a resistor across the main contactor.

So the idea is to leave Dash switch #1 (Negative Contactor) and #2 (Zilla always on) powered up, so When I turn on the iginition key, the Negative side contactor will close, and the Zilla will power up. Then I'll push button #3 to turn on the controller or just leave that on too. Finally, turning the key to "Start" will activate the zilla.
.
00 IMG_9885.JPG

02 IMG_9939.JPG

03 IMG_9941.JPG

04 IMG_0078.JPG

09 IMG_0155.JPG

View attachment 5

11 IMG_0158.JPG

12 IMG_0157.JPG

View attachment 2

14 IMG_0168.JPG

15 IMG_0169.JPG
 
.
With the control wiring sorted, I started working on the Power Wiring. I intended to leave slack on the cables so I wouldn't have to disconnect them to lift the body, but after hours and hours of experimenting I changed my mind. It would be a challenge to fit the slack from an individual cable under the trunk, let alone 4 cables, let alone orient them in a manner to minimize battery loop inductance and avoid leaking high-voltage noise into the control wiring and wire chafe and road debris damaage.

Instead, I built the power wires to fit. Even with the tow'd body on, I can install/remove the (-) battery cable because the shunt is accessible. I can unbolt the (+) battery cable at the Zilla's power stage, and stuff it back out its hole to clear the body. The motor cables can disconnect at the Zilla's power stage or at the motor. It might take me an hour longer to lift the body, but it isn't a huge obstacle, and hopefully I won't need to do it often. Now that I have moved the controller and wiring terminations to the trunk instead of underneath it, I'd only need to lift the body to work on the main fuse, the contactors, or the shunt.

21 IMG_0234.JPG

22 IMG_0233.JPG

View attachment 15

24 IMG_0244.JPG

25 IMG_0251.JPG

View attachment 12

27 IMG_0255.JPG

28 IMG_0259.JPG

40 IMG_0245.JPG

View attachment 8

42 IMG_0278.JPG

43IMG_0280.JPG

45 IMG_0283.JPG

46 IMG_0286.JPG

61 IMG_0297.JPG

View attachment 2

63 IMG_0308.JPG

64 IMG_0305.JPG

Next up, mounting the zilla's power stage and hairball into the trunk. Since the trunk is fiberglass, I'd like to distribute the zilla's weight with a layer of 1/4" plywood over foam. The power stage mounts with bolts through the plywood, foam, and trunk; two of them will be eye-bolts, so I can run a short steel cable from them to the chassis, to hopefully retain the controller in the event of a crash.

Then it is a matter of running the control wiring pigtail through conduit to the hairball and punching down 15-20 wires, bolting the car back together, soldering on a CycleAnalyst connector, and then we will see what we see.
 
Thanks Nicobie!

Fechter, cooling is indeed the missing link in this picture, and I have dash switch and wiring run to power a pump. But I am eager to get it on the road, and since the car is so light, and since the manual says its OK, I want to see if it can handle moderate driving without cooling. Battery side amps will be low because of the 400a silicone fuses, and the Red dash LED will indicate degree of heating. Cooling is definitely next up on the project list, I've got some research done, looks like Otmar was using an Aquarium pump that is poorly made these days, do you have any suggestions? It would be elegant to fit a reservoir between the controller plate and the trunk bottom, otherwise it ends up in the trunk.

Anyhow, good progress this weekend. Since the trunk is fiberglass, I put a layer of 1/4" plywood down to help distribute the weight of the zilla's power stage. The fiberglass is rough and uneven, so I planned to put foam underneath as a conformal layer, but I used coraplast instead.

00 IMG_0473.JPG

02 IMG_0474.JPG

View attachment 12

04 IMG_0511.JPG

06 IMG_0513.JPG

07 IMG_0514.JPG

08 IMG_0516.JPG

09 IMG_0518.JPG

10 IMG_0521.JPG

11 IMG_0523.JPG

12 IMG_0525.JPG

13 IMG_0528.JPG

14 IMG_0526.JPG

15 IMG_0531.JPG

16 IMG_0549.JPG


Next up, I'll be playing with ideas for retaining the zilla's power stage in a crash. Maybe I am putting too much thought into this because the power stage weights about the same as the ICE battery that was in pretty much the same location, but I don't like the idea of either one flying at my head after a collision.

I think I will fit a smaller piece of 1/4" plywood underneath the trunk to distribute the load and keep the zilla bolts from pulling out. Then I will run a few strips of aluminum under that, to also keep the zilla bolts from coming out. Finally, I might replace a pair of the power stage mounting bolts with eye bolts, so I can connect them to the chassis with a wire strap.

After that, it is just a matter of bolting everything together, validating connections as I go, and then punching down wires at the hairball. Sounds simple (although it never is) but I am getting pretty close to having the tow'dzilla ready for a test drive!

-JD
 
Oatnet!

This is awesome!

Still Following your work! Very Nice Indeed!

Tommy L sends....
 
Thanks TommyL!

Recently I was musing on parts of the conversion I would do differently now that I know better. Once of the things I had the most angst about was the balance wires, which I brought out the back of the battery box onto the Kelly controller tray. This location exposes them to road conditions/debris, they could get tangled in anything from the clutch cable to the CV joints, and they complicate access to contactors\power wiring. They had become a big enough problem that I had a task on my list to "figure out what to do with the balance wires."

Since the CellLogs will plug into the balance wires, that project highlighted the problem again. When I was designing the route to get the CellLog wires to the balance connectors, I decided to come up from the Kelly controller plate, around the side of the battery box, then over it and out the gas tank filler hole between the seats. It struck me that since I used a tarp instead of a metal lid on top of the battery box (space requirements) if the balance wires had come out the top, the wiring would be much easier. It was too much to take on now, but I added a "Phase III" task to re-route the battery cables the next time I took the pack apart.

The body was already partially bolted down, so to install the CellLog conduit I had to snake it through the space. The space turned out to be too tight to snake, so I gave up after a half hour, unbolted and lifted the body back up. Now that I had access again, I took a peek inside the battery box at the balance wires - and realized that while they are hard to access, I might be able to extract them without dissecting the pack! It would be difficult, take a long time, and require a lot of concentration to do it safely, but still much easier than dissecting the pack.

I _really_ wanted to re-route the balance wires, but I didn't want to get caught up in a the downward spiral of tangent projects, resolving each minute issue but never finishing the project. The cellLogs were already a "tangent" project that was delaying the road test for a couple days, so re-routing the balance wires for them was a "tangent" to the "tangent".

However, it would make the cellLog installation a lot easier, so time invested relocating the balance wires could translate to time saved on the CellLogs. The new location of the balance connectors would make the cellLogs easy to remove instead of permanent - nice, since I only need them for deep discharges and the occassional baselining. Now that I knew a better way, a substandard installation would bug me more than adding a few days of work to the project. I decided to take it on.

The plan was to pull all the balance wires from their current path, bundle them up at the top of the pack, and build a pair of "extension cords" (with 23 andersons on each side) so they could reach the gas tank filler hole between the seats.



01 IMG_0878.JPG

02 IMG_0882.JPG

03 IMG_0881.JPG

04 IMG_0885.JPG

View attachment 23

View attachment 22

11 IMG_0851.JPG

12 IMG_0858.JPG

View attachment 19

View attachment 18

15 IMG_0902.JPG

16 IMG_1007.JPG

17 IMG_1009.JPG

18 IMG_1043.JPG

file.php


Here they are, (6) cellogs in a row, 7 cells each, covering the whole 42 cell pack.
.


21 IMG_1065.JPG

22 IMG_1105.JPG

25 IMG_1180.JPG

26 IMG_1107.JPG

View attachment 8

28 IMG_1248.JPG

View attachment 6

30 IMG_0795.JPG

31 IMG_0793.JPG

32 IMG_0800.JPG

33 IMG_1237.JPG

View attachment 1



So - It's Alive!

While the performance is much better than with the Kelly, it was still far short of what I expected, and did not feel like 465 amps to me. Before I fixed the amps display, the CA reported a vMin to 133v which is @ 3.16v cell, which is more like a 2c-3c discharge, 120a-180a not 465a.

The next day, I had a few minutes after work for another test drive. I found that in second gear, it started a bit slower but pulled much harder as I got up to speed. Since torque (amps) are limited by backEMF, I thought maybe first gear RPMs come up faster than the zilla's throttle curve, backEMF restricting amps before the zilla can deliver them. Got the burny smell again, and the motor/controller were still cool to the touch.

The next day, I was able to fix the CycleAnalyst. After the Kelly build, the CA would display high amps at standstill, and I sorted out that I had reversed the "S+" and S-" leads. I wasn't sure if I had fixed this at the shunt or the CA connector... Since the problem recurred when I move the CA wiring to a new bundle and connector, I knew I had to swap pins #3 and #4 on the connector again, and voila I have a working amps display.

This time I tried launching in third gear, it was slightly slower than 2nd gear to launch, but really pulled as speed increased. I thought I saw 370a on the display, but the CA said aMax was 230a. After seeing white Zombie launch with his zilla, I am suprised I don't have full amps at launch, decreasing as backEMF goes up. However, just like with the kelly, I see low amps at launch, which increase as I go faster. If I take my foot off the pedal at speed, amps jump up as soon as I step on it again, so it seems to be RPM related - but I haven't hooked up a tach to the zilla yet, and the kelly had no way of sensing RPMs that I was aware of. Maybe I need a better understanding of how DC controllers behave, maybe I have something else going on (throttle wiring?), but I look forward to experimenting with gearing and longer runs to see if I can get amps up higher.

I did 4 hard launches on that last test, and traveled maybe 2/3 of a mile, and the CA said I used whopping 225wh... I have been hoping that Rickard's theory that an EV will use 100wh/m for every 1,000lbs would apply, so I could get 100wh/m. Maybe if I drove tamely, or didn't have frequent stops, that number would apply. However, my chunk of suburbia has lots of stops signs and streetlights and I am more interested in what wh/m I get without hypermiling. With ebikes, for example, I know I am pretty much gonna get 35/40 wh/m (which fits with Rickards formula, huh!) - I look forward to experimenting with the Tow'd to get better data and expand my design vocabulary.

The burny smell is be waning, and the brush path is cleaner now, so the grime might be gone and the needed commutator patina may be building up.

I did some more research today, and yeah it looks like I should see a full 465a on launch, so while the car is powerful enough to be streetable, something is wrong. I sent an email to Manzanita to see if they can give me any ideas...

-JD
 

Attachments

  • 19 IMG_1051_1024.JPG
    19 IMG_1051_1024.JPG
    40.5 KB · Views: 2,865
Hi Etrike!

Sounds like things are going well with that Zap!

The smell did not appear on my last ride, and the commutator looks much better, I think the griome wore off and a patina is starting to build. The zilla had not been installed when the brush pictures were taken, so if that is damage it happened with the kelly.

144v is typical voltage, 72v is probably about right for the 6"? motor in your zap, but the 9" motor is significantly larger and rated at 144v. Here is a 144v factory motor chart for the 9" ADC FB1 in the car:

ADC FB1-4001a_800.jpg

Otmar (the guy who invented the zilla) has a great quote on the zillaFaq:

"For a typical street conversion running a 8? or 9? motor, I run the motor current limit between 500-600 Amps. Higher than that often results in a slipping clutch and increases motor heating as well. I have tested Advanced DC motors in 8? and 9? sizes in standard and XP models and have found that with perfectly broken in brushes in race situations the motor voltage can be set as high as 170 Volts without causing flashover. In street applications I suggest limiting them at 150 Volts."

Rich at Manzanita Micro (who supports the zilla) said I should be spinning tires off the line at my current settings, but
chided me for only running 465a and not taking advantage of the zilla. He reports he is running 800a on his smaller ADC 8", and I should set my motor voltage at 160v.

So I thought about that... 800a x 160v / 745 = 171HP, and 200+ lb/ft. The 2015 Chevy corvette ZO6 weighs 3.5x the tow'd, and has 600+ hp and 600+lb/ft of torque. The math shows the tow'd could have a better torque-to-weight ratio than the ZO6, although the HP falls short after factoring in motor/controller efficiency. I look forward to probing those limits, but will run tamer ones while I sort things out and find out what the chassis can handle.

Anyhow, I had noticed that the copper strap external to the motor was hot after a recent ride, which later made me wonder if that was not passing current well, so I replaced it with a 4/0 cable, but that did not improve performance.

IMG_1519 640.JPG

IMG_1521 640.JPG

I got zillaview going, and was able to record a run. I cut out most of the parts where throttle was not 100%. The second one has even more clipped out, but is at 25% slomo so I can read the dials better:

[youtube]7atERS61I44[/youtube]

[youtube]h292wlojBgQ[/youtube]

It is nice that this confirms that I am getting 100% throttle, so that is not a problem, and controller temps stay under 34c with this light load. Rich reports that the cooling and lack of encoder will not cut rate, so there go all my lead suspects for the performance problem.

One thing that is odd is that it shows my motor amp limit as being 1,000a - but I have that configured as 470a. Another thing that is odd is that amps seem to peak at 235a, 50% of the 470a limit. But the really odd thing is that Motor volts are sitting so low. Maybe a motor problem has a resistance problem that is pulling the voltage from 140v down to 36v, but I don't see the massive motor heat I'd expect that to generate. Maybe the power stage needs a little more fixing. Maybe it is something I don't have the perspective to understand yet.

I really have to give thanks and Kudo's to claudio@evgear.com.au for creating zillaview, I am 100 steps further down the troubleshooting path thanks to the dashboard he created and shared.

-JD
 
oatnet said:
One thing that is odd is that it shows my motor amp limit as being 1,000a - but I have that configured as 470a. Another thing that is odd is that amps seem to peak at 235a, 50% of the 470a limit. But the really odd thing is that Motor volts are sitting so low. Maybe a motor problem has a resistance problem that is pulling the voltage from 140v down to 36v, but I don't see the massive motor heat I'd expect that to generate. Maybe the power stage needs a little more fixing. Maybe it is something I don't have the perspective to understand yet.

Hi oatnet,

It looks to me like you are limited to the 235A. The duty cycle stays around 25% even at 100 throttle. So that's why the motor voltage only hits about 1/4th the battery voltage. Confirmed by the battery current being so much lower than motor current.

Have you got it in valet mode? Also, do you have any indication of motor RPM?

major
 
.
.
.
So I was in the middle of writing this reply to Major's post...

Hi Major!

Clearly you know zilla.   :D   I don't have valet mode setup yet, but I have been considering using a momentary switch like a "nitrous" button - higher amp/volt settings than I typically run on the street.  Does the valet mode come on the moment you press the button, or do you have to release the throttle/repower the controller?

Thanks for pointing that out, I got so lost in volts and amps and temps that I didn't process duty cycle, but that is a big piece of evidence, which is making me think something is up with the power stage.

I reset the hairball to factory defaults with the D) option, then turned off option p) to set scaling to Z2K - and noted that programmed amps doubled as a result.  I set them back to what I wanted, but the test ride was same as before. 

Thanks to the duty cycle, I know the controller is delivering max of 235a, exactly half the 470a I programmed.
Max motor 1,000a on zillaView is not the 470a  I programmed, but it is half of Z2K's max.
Manzanita recently repaired the power stage and replaced PCB cards - is it possible that some factory setting is on Z1K instead of Z2K?

I guess I could double the amps, and expect to get

... when I ran downstairs and did just that.  Setup the camera on a notebook running zillaview, and went for a spin on the standard course in 2nd gear.  Leaving the garage, I was very cautious with the throttle, low speed control was smooth and easy but clearly there was more power there.  I rolled onto the street, conservatively applied 50% pedal and 

omg Omg OMG      O M G       O M G !

the seat slapped me on the back and it was pulling so hard I let off the throttle.   I rolled at 25mph for a moment, worked up my courage, and stomped the accelerator all the way down.  The nose came up sharply then the back tires broke loose with a whiiiirrrr until I let off the throttle.  I tried this a few more times, startled and amazed.

HOLY CRAP! Going through my mind, over and over.   When I got back, my EV grin was more like a rictus of fear, but in a good way.  Despite the raging fury, the controller, motor and new motor jumper cable were all cool to the touch, the only burning smell was the tires.

I am so happy to feel that the original vision for this project has finally been attained.  My heart sank when I realized the kelly was giving everything it could, so maybe 500a was not the power I expected...  Wondering "Have I been fooling myself all along, blind to some trick of logic?"  And sank again, when the zilla was not doing much better - thinking, "is that all there is?  Is that what 500a feels like in a car?"   But now, I know the raging beast borne of lightweight construction and excessive power, long imagined, is now real and in my garage waiting for me to learn how to control it.  

I still have the issue of why the power stage is only delivering half the amps I program, but that is OK the Z2K has double the amps I can use... A topic that will be the subject of a post later this week.

Anyhow, the highest motor amps I see on the Zillaview footage are 355a/6c at 126v, which equals 3.00v/cell.  The CA registered a Vmin of 122v and an Amax of 768a, which was probably a short spike before the controller reigned it in.  Although I covered the standard short course in terror the CA says I used only 195wh, vs 245?wh with the old settings, waiting for acceleration to build.  

Well shoot, now that I have the driveline sorted, I have a 1,000 little tasks to clean up the overall build, and make some videos...

-JD
 
rtz said:
Definitely hook up some water cooling to that poor Zilla.

I'll definitely consider that when my testing cycles start making it even slightly warm to the touch. It starts cutting back at 55c, but per the zillaview today's wild ride took the zilla from 29c to 33c. Pushing a 1,000lb dune buggy instead of 3,500lb pickup has its benefits! :D I need to try launching in 3rd and then 4th gear and see if they warm it up.

Some pics from today's run:

IMG_1573.JPG

IMG_1574.JPG

IMG_1575.JPG

IMG_1576.JPG

IMG_1577.JPG
 
Yay, a new page that won't take forever to load.

I thought I'd capture some burnouts today, but it didn't work out - i bumped my camera into "time lapse" mode. :twisted: When I went out for another run, the clutch started slipping and stinking on the first launch and that was it - although it took me a minute to figure out what was going on. I had wanted to try launching in 3rd/4th too, but it will be dark before the clutch cools. At least I know where to mount the camera (on the motor!) for a good angle, on my next try. And I have officially found the first weak link - my brand-new clutch and pressure plate. :shock:

Sorry it isn't better, but here is the fail compilation:

[youtube]LMkQkPQTjnU[/youtube]

-JD
 
WOW! This is awesome Oatnet!!!!!!

I'm so happy that the 4 ga jumper is cold and the TOW'D is ALIVE!!!!

Tommy L sends.... :) :) GRIN!!!!
 
Thanks Tommy!

I have a bit of backlog of updates I have written but haven't posted:

Unposted writeup after clutch failure
=====================================
It looks like I'll be pulling the motor this weekend. I took the tow'd out for another test yesterday, and yeah my burnout runs broke the clutch. It holds fine at normal power levels, but it immediately slips/stinks when I stomp on the pedal. The clutch lever only goes about 1/3 down, but does not actually release the clutch. I can't really picture what failures would cause this, maybe some snapped fingers on the pressure plate? Anyhow, if things are breaking under load I know I am doing something right, power wise. A co-worker recommended I look at a Kennedy Engineering Clutch/pressure plate that can handle the torque loads, they did a good job with my flywheel so I will probably go that route. Once the clutch is squared away, I wonder if the tranny or the axles will be the next to go...

I ordered the Cobalt MJ-1200 submersible pump that Otmar recommends for the cooling system. Once an industry standard, I found many reports of those pumps failing, underperforming, vibrating and being noisy after production was moved to China. Luckily, I found that "Cobalt Aquatics" sources the original version of the MJ-1200 from the original italian manufacturer, I hope it lives up to the glowing reviews. The chinese version is half the price, but you get what you paid for.

The MJ-1200 uses 20 watts to pump 295gph with 69" of head pressure. I was tempted to use the MJ-900 instead, which uses only 8.5w to pump a respectable 230GPH - double the zilla's requirements - with 46" of head pressure. I decided the 135% power consumption penalty was worth the 50% increase in head pressure - better too much than too little.

I used to have an inverter at the dash that powered fans for the kelly. That inverter will be remounted in the trunk next to the (still-to-be-determined) reservoir/overflow tanks, energized by the same switch that powered the fans. No idea what radiator I am going to use, or where it will go, but they will exist.



Unposted writeup of Z2K vs Z1K
==============================
Everything I learned about DC motors came from experience with Brushless/BLDC on ebikes, so one of the challenges has been learning the different parameters around brushed series motors. I am going to spew out an overview of what I think I understand today, a little of it comes from experience but most of is it regurgitating what I have read so there will be flaws.

For example, I thought that Series motors do not have back EMF, because without load they can quickly spin up to bearing-killing speeds. However, back-EMF does exist. BackEMF increases with RPMs and restricts the amps, which decreases torque until it meets current load and cannot increase RPMs any more.

Amps define torque, it will remain constant no matter what the voltage is. Since a portion of amps is converted to heat, increasing amps requires more motor cooling. This is a particular problem at launch, when efficiency is low and amps are generating more heat than power and can quickly saturate the motor. The brush/commutator interface has a limit to how many amps it can handle before a flashover event.

As RPMs increase so does efficiency, so less amps are wasted as heat. However, BackEMF also increases until the motor can't take in enough amps to increase RPMs. There are two ways to decrease BackEMF - increase the voltage, or run a smaller motor. However, overvolting is highly limited by the brush/commutator interface as well as motor timing. Better brushes, better brush surfacing, and better patina on the commutator will let you increase voltage a little, but not double it like I could with a BLDC motor. Changing the timing also helps the motor run more effectively at a higher voltage, but push the volts too hard and a flashover can destroy your motor.

Smaller motors generate less backEMF and less torque/amp, but deliver torque at higher RPMs, but the smaller commutator can handle less voltage. The larger motor will deliver more torque/amp, and can run at a higher voltage, but backEMF restricts amps earlier.

With Brushes limiting volts, and backEMF limiting amps there is an upper limit on how many watts a given series motor can handle, and for how long. Otmar (the creator of the zilla) recommended a street conversion with 8"/9" motor be run 500a-600a and 150-160v. I have read that the 9" Netgain Warp motor will survive 1,000a for 10 seconds, and Netgain rates their Warp 9 at 170v with Helwig brushes, and some drag racers mention running 196v. So with compromises, short runs, well-detailed equipment and good cooling, a bit more power can be had. A larger 11" or 13" motor can also handle more power too.

However, the 1,600a/400v that the Z2K's "EHV" version can supply is way more than typical series motors can handle, so what is the point? I think it is to run dual motors. The Zilla can drive contactors that switch dual motors between Parallel and Serial configurations in 0.3 seconds. In parallel mode, dual motors split the amps, so the 1,600a translates to a manageble 800a each. In series mode, the dual motors split the volts, so 400v translates to a manageble 200v each.

Battery Sag is another reason to run high voltage, especially on the Lead-Acid cells common when the zilla was developed. The zilla is not a boost converter, so Motor Volts will never exceed Battery Volts. If your 150v pack sags to 130v, the motor will see that 130v, not the 150v you programmed. However, the zilla _IS_ a buck converter that can reduce pack voltage, so if you have a 300v pack and program motor volts to 150v, the motor will see 150v. If your battery sags to 200v, the motor will still see 150v.

That said, the Z2K is clearly way more controller than the tow'd will ever be able to use. The smaller Z1K-HV would be more appropriate - capable of the 1,000a that would toast the motor in 10 seconds, and at 300v it has plenty of headroom to accomodate pack sag. Having too much controller means it can tolerate greater abuse, and it provides greater thermal mass so it can run longer without proper cooling. However, the z2k weighs and costs twice as much as the Z1K, so I could reduce weight and sunk cost with a Z1K.


000 IMG_1584.JPG


 
View attachment 19

file.php

I like my donuts glazed, my clutch disk not so much. The disk and pressure plate were brand new, from a cheap clutch kit I bought back when I had the VW bus. I think the pressure plate might have failed - when it was in the car I could not get it to release with the clutch pedal, so I thought I'd find a chunk of snapped part blocking travel. It was clean inside the bell housing, and I can't budge the pressure plate fingers with anything, so I think it locked up, but not quite all the way closed, hence the slip - Weird.

Anyhow, I bought replacements from Kennedy Engineered products. I bought their HD clutch disk, and their Stage 1 pressure plate which is rated to 180lb/ft. I could easily ramp up amps to exceed 180lb/ft so the Stage II/III/IV pressure plate may have been more appropriate. However, that would just tee up the axles or the tranny as the next failure. I could replace the tranny with a built Rancho Unit with LSD, and beef up the axles, but that would tee up the frame as the next failure. I could weld on braces to tie the horns of the transmission mount to the frame, but then the "loop" of the frame would twist, and I would have to weld a structure to brace the frame... But this project has always been about achiving performance by leveraging low weight, not shoehorning the ulitmate power into the chassis, so for now, I will restrict power to what the chassis can handle.
.


04 IMG_1803.JPG

05 IMG_1793.JPG

06 IMG_1767.JPG

View attachment 14

View attachment 13

View attachment 12

10 IMG_1834.JPG

20 IMG_1999.JPG

21 IMG_2004.JPG

22 IMG_2165.JPG

23 IMG_2166.JPG

24 IMG_2093.JPG

View attachment 5

26 IMG_2005.JPG

View attachment 3

33 IMG_1931.JPG

file.php

The red paint has been coming off the motor mount, so I decided to throw some paint on while I had the motor out. A smart man would have taken off the flywheel, hub, keystock, and motor mount, so the painting would be straightforward. Apparently, that is not me, I am the lazy man who painted it with the motor bolted on.
.


36 IMG_2208.JPG








Please ignore these duplicate pictures:
 

Attachments

  • 02 IMG_1823.JPG
    02 IMG_1823.JPG
    129.9 KB · Views: 5,941
Back
Top