[Reid_in_QC]

Hi everyone.

While my zero turn mower project patiently waits for me to find a perfect pair of worm reducers I have started on a much more practical project that enjoys much broader support here on our farm.

I am converting our Hefty-G cultivating tractor into a battery-EV from it's original gas-powered config, mostly so the damn thing will work when we need it to rather than pulling untimely stunts like it did most of last year. It's a pretty unusual tractor and not too terribly common so hopefully some of you will get a kick out of it.

The Hefty-G is a late 1970s copy/homage to the 1950s era Allis Chalmers model G which was, I think, the first small cultivating tractor in North America to move the engine aft of the operator, thus freeing up the space between front and rear axles for "belly-mounted" tools for precision work like seeding, weeding, hilling, etc. A brilliant re-think of the standard tractor configuration aimed primarily at vegetable growers before the advent of chemical weed control.

Fast forward 50+ years and these old tractors are being dug out of barns and put back into service on organic vegetable farms like ours where they continue to ably perform the tasks they were designed for. Well, with the possible exceptions of starting reliably and running smoothly without constant breakdowns. Luckily these problems are all easily resolved by throwing out the ICE and replacing with a suitable electric motor.

The Hefty-G is perhaps the easiest conversion project anyone could ever find. It was built out of pieces, almost zero integration, and everything just bolts together. They were turned out by a smallish concern in Wisconsin – the Holtan Axle Co. or Haxco – which built gearboxes and diffs and decided to try their hand at copying and updating the venerable Allis G. They did a great job in many respects but the rebadged Renault engine supplied by Continental Teledyne hasn't aged gracefully or made any friends at our farm of late. Nor are parts available. I guess I can thank the little bugger for motivating me to extract it this winter rather than trying to wait another year.

Okay. The build is pretty simple as you'll see in the pics. The gas engine shaft-drives the clutch and tranny and though it's not quite bolt-on easy I can't see how it's going to be too challenging to mount the electric motor in almost exactly the same spot.

The motor I found was rebuilt and shelved. Looks super clean and flawless and wasn't too expensive. AMD 203-02-4004 badged as Raymond, pulled from a lift, probably a standup 3-wheeler. It's a sepex/shunt motor which is fine because I found a big 600A General Electric sepex controller at a fair price on eBay. I may not even reverse the motor so it being sepex is not necessary but the price was right so there we are.

The AMD motor is 8" diameter, approx 14" long and is similar to one whic forum member Boekel used in a boat but now I can't find the thread. I'm planning to run it at 48V using a battery bank of Chevy Volt modules, probably 4P12S. I don't need tons of runtime or load as the tractor is primarily used for weeding, a couple hours at a time, and it's geared-down like crazy. It is not a tillage tractor and was ably powered by a wheezy old 20-odd HP gas engine with all of 800cc of displacement and a centrifugal governor. Good riddance to that particular piece of tech

The motor also has the advantage of having a tailshaft that should let me keep my original open center hydraulics with no modification (pump shown in mockup pics). I think I'll have two throttle pots in parallel if that's feasible, one foot and one hand throttle. I'll keep the clutch so that when I want to stop the tractor but keep spinning the pump for hydraulic flow I'll just depress the clutch with my left foot and spin up the motor with my right.

I'll post more photos and questions as I progress. There's a photo album here on Facebook but here are a few and a link to a walkaround
video on youtube here.

Don't be confused by the cylinder cradle and propane hoses on the right side fender. The tractor was not propane powered. That gear is all part of the flamethrower... sorry, the "flame weeder" mounted at the very front of the machine. It's a totally valid and accepted technique for killing certain weeds at certain times, even in organic systems... honest It's fun for lighting campfires too flame weeder demo on youtube.

Suggestions and words of wisdom are always welcome.

Anyone ever set up one of these GE IC3645 controllers? Mine is IC3645SR4W606N6. Should I worry about the controller map being properly suited to this particular motor? I'll have to figure out programming for the controller eventually.

 

[Yabert]

Nice project Reid
Hope the clutch will remain reliable as you choose to maintain it in place.
Can't wait to see how you will integrate 8 kWh of chevy Volt battery in this.

 

[electro wrks]

If you find you're using close to 1:1 on the in and out RPM of the transmission as the ideal reduction ratio, may be the transmission could be eliminated. If so, the motor could be located near the front of the differential, freeing up more space for batteries behind the seat.

 

[Reid_in_QC]

If you find you're using close to 1:1 on the in and out RPM of the transmission as the ideal reduction ratio, may be the transmission could be eliminated. If so, the motor could be located near the front of the differential, freeing up more space for batteries behind the seat.

That's a good point. I've never bothered to figure out the ratios when it was gas-powered but with the big final drives at each wheel it might be possible that the transmission could be eliminated. I'll keep it for now I think but worth figuring out if I could do without as I have a "spare" one of these tractors that should one day be converted too. Also, the clutch and transmission add drag and complexity and both are obsolete with zero parts available so being able to ditch them would be a plus in terms of efficiency and reliability. I'd have to figure out a different setup for my hydraulics rather than using the tailshaft.

Space isn't an issue for batteries or seat. I'll be able to put all the batteries, wiring, controller and accessories I want in the space above the motor and behind the operator seat without changing much at all. It wasn't tightly packed at all in original config.

Thanks for the ideas.

 

[Reid_in_QC]

Nice project Reid
Hope the clutch will remain reliable as you choose to maintain it in place.
Can't wait to see how you will integrate 8 kWh of chevy Volt battery in this.

Do you think I'll destroy the clutch?

I've never had problems with clutch slip using 20hp gas engine and often needed to ride the clutch to start in higher "road" gears because no synchromesh means always start from a standstill.

I hope it'll hold up for several years.

Did you delete the clutch from Yan Gordon's Hefty-G? If so maybe I can use his as spare in case I'm wrong

~ reid

 

[Reid_in_QC]

I made a bunch of progress on the Hefty-G-to-E project recently so thought I should at least post some pics.

I started assembling the case for the battery pack and have ordered some active balancing boards to act as a cheap BMS.

The battery is to be assembled from BAE modules built with A123 26650 m1a cells which I acquired in a consignment auction. They've seen a fair bit of use and seem to have capacities of about 65% of original spec but that's adequate for my needs and the price was right. The BAE modules are really well built and I happened to find a nice orange steel box in my shop that was kind enough to undergo some minor surgery to become the perfect host.

Battery layout is planned as 14 modules of 8p12s laid out in 2 layers. I want to run busbars across all parallel groups and connect top and bottom layers in parallel across groups as well. The result is 14x8 cylindrical cells in parallel for each step of the string with the idea that any weak or failing cells will have a negligible impact on the pack as a whole. Final layout is thus 112p12s for a 40v pack composed of 1344 little 26650 cells. The whole pack measures 32" long by 21" wide x about 18" high. Not sure of the weight yet.

I completed the new motor mounts recently and now I'm waiting to find time to finish rewiring my lathe to turn some adapter hubs to reassemble driveline. The pic of rear (brush) end of the motor shows the odd-shaped rear motor mount which will also support a hydraulic pump (just barely in the frame) which will mount to the tailshaft.

The last photo shows the clutch+sliding gear transmission (and a spare twin) that I pulled to see if I can replace some shaft seals and stop it leaking hydrocarbons all over my soon-to-be-electric tractor.

So far so good. More to come soon as the farming season winds down and I manage to find more time to work on this project.

 

[Reid_in_QC]

Progress on the drivetrain:

• Pretty complete cleanup/partial rebuild of original sliding gear transmission (new seals throughout, lighter synthetic oil) and clutch (cleanup and test, no repairs needed)
• dismantle and reassemble differential and half-shafts. I want to use the best of the two diffs I have (one from parts tractor) so that one can sit on a shelf as backup because these are obsolete unavailable parts from long-gone mfgr. When I get around to reviving the parts tractor as EV I hope to use a readily available automobile diff.
• hydraulic motor mounted to tailshaft with double-roller chain link and custom adapter hubs. My new-to-me lathe is starting to earn its keep!
• hydraulic freewheel loop to allow pump to spin the "wrong way" when drive motor reverses. Two poppet-style check valves and a couple tees + the inevitable transitions from JIC to NPT and back...

The reason for the "hydraulic freewheel" is that I want to use the motor controller for reverse and stop using the transmission's reverse. It's a Sepex motor+controller so it seems a shame not to use the inherent excellence of this setup for reversing. However, the hydraulic pump doesn't want to spin in reverse, or at least we don't want fluid circulating the wrong way in the hydraulic circuit. I looked at integrating a sprag clutch or roller bearing one-way clutch in the coupler hub but didn't find a compact and trustworthy design. A friend suggested a check-valve regulated shunt loop to allow the pump to spin in reverse without screwing up oil flow in the circuit beyond this loop.

 

[Reid_in_QC]

Since last entry I finished fettling all the bits of the driveline. Clutch, transmission and diff were all dismantled, cleaned and rebuilt with new bearings and seals wherever appropriate. brakes and output shafts to the final drives too while it was in parts. Finally some steering repairs to replace an utterly pooched U-joint and inspect and rebuild the steering gear box with new seals and lube.

Then I finally finished the driveline and was able to actually test run it on 12V! It's handy having a design that retains the clutch at this point because I can actually drive the tractor around the shop - a real improvement over pushing it around which is what I've been doing for about a year.

Having finally set up my lathe in the fall, learned at least the rudiments of turning and acquired a handful of tools, I was able to finish making a drive hub to fit my motor's tapered keyed output shaft. It came out well and seemed to fit the motor perfectly but now that I spin it up I can see a tiny bit of runout and I'm not sure where it comes from. I'm pretty sure it'll be within the tolerances of the rubber donut drives in the first part of the output shaft so I'm not going to sweat it for now. One of the pics shows how I cut a keyway on the lathe after turning the inside tapered bore. I set up a 6mm wide HSS parting tool on it's side in the toolholder and then ran it in and out with the compound rest slide, gradually scratching a thou at a time. It probably took 15 minutes but that's fine with me. The hub part is cast iron and it turned out really nice as far as I'm concerned.

The output shaft from motor to clutch had to be shortened and re-welded but otherwise required no modification as put the drive motor shaft on the same centreline as where the ICE had been.

Overall the results are both pleasing and functional. The sepex motor spins up nice and quiet at 12V and has enough torque to get the tractor moving without trouble. The field winding was shunted for these tests. Hydraulics also worked perfectly with the pump mounted on the tailshaft.

All told I'm pretty pleased.

Videos here for those interested https://www.youtube.com/channel/UCMf6w_UWjICkvjODMkpufPA/videos?view_as=subscriber

 

[Yabert]

Nice work Reid!
I can't wait to see your work on the battery of this awesome electric tractor.

 

[DMA]

Amazing work, Reid! I'm looking forward to progress, with anticipation.

 

[Reid_in_QC]

Time for another update as the cold winter weather finally draws to a close and I can more often make progress on the tractor in my unheated shop.

Attached pics:
1) battery box support frame
2,3) placement of battery box (orange) and control enclosure. Also SB350 battery disconnect.
4) motor brush cover to force air intake at drive end. Thanks Boekel for confirming that this is right for this motor.
5) enlarged footrest with added pedal-pot assembly (forklift scavenge)
6) configuring the various controls and instruments around the operator.

The grab handle/supports rising from the front of footrests are new and make it safer and easier to get on and off while also protecting accelerator pedal and stiffening the larger footrest. It also looks like a good place for the fire extinguisher and maybe the torch-lighting torch (this tractor has a flame weeder mounted up front and needs tools for fire-ON and fire-OFF)

I'm almost done the fabrication tasks, with a few notable exceptions, and am working on layout of instruments and controls. The operator station is actually quite small and needs careful planning to get everything set up.

I also managed to find a tachometer/encoder that I pulled from an industrial motor ages ago so I'm going to try to install it on the tailshaft end and put a tach in the "dash" for fun. Maybe I can even train employees to optimize motor performance for max runtime by reading a tach and instantaneous current draw? Probably not but a tach will be fun anyway and the one I bought includes and hourmeter so I don't need a separate one of those.

So, still lots to do and when I finally get it all running I need to do a major overhaul of the implement lift to change the geometry and install new weeding tools. The clock is ticking as the snow melts away and my wife and farm partners remind me that this MUST be done and working by en of April.

wish me luck.

 

[Reid_in_QC]

Here's some detail on the roller fairlead and retractable charger cord I set up on the tractor.

I found a photo online of a nice little roller fairlead made with plastic parts but I had a better selection of metal on hand so that's what I made mine out of. A piece of square tube, 4 carriage bolts and some smallish round steel pipe. Rollers were bored and dimensioned on the lathe but with a better selection of materials this job could have been done entirely with hand tools.

I'm really happy with the result and love that this tractor will need only be parked within 40' of a 15A 120VAC outlet and then it will take mere seconds to set it up to charge. Again mere seconds for putting away the cord before departing for the fields the next day.

The cord reel hidden in the fender is a robust industrial unit that I picked up used somewhere - as compared to cheap import crap one might not want to hide inside a fender and plan to use twice daily.

 

[Reid_in_QC]

Good progress on controls and instrumentation – at least new dash panels and roughing in the controls. So some satisfyingly colourful photos are ready to share.

The lower dash sits atop the transmission and holds tach/hourmeter plus pilot lamps for precharge delay (orange), ready (green) and stop/fault (red) as well as emergency stop mushroom switch.

The upper dash below the seat includes key switch, switches for various 12V accessories such as lights and flame weeder, a USB port for phone charging and an illuminated switch which will become one of 3 things.

The clear-top waterproof enclosure with sunshade that is finally mounted in the last two photos is for LCD displays of battery SOC and monitor for active balancer modules.

 

[Reid_in_QC]

A bit more progress but still seemingly avoiding tackling the battery pack assembly – mostly because I think i ought to make a punch & die for copper bus-bar strips and not yet sure how to do that.

So instead I dug through some boxes where I thought I once stashed an encoder of industrial origin. Amazingly I found it and thus was compelled to install. The toothed ring (reluctor?) went on the tailshaft hub that serves as adapter to connect the hydraulic pump. My new best friend Jimmy the lathe helped resize everything to cooperative dimensions and even helped me turn up all the necessary spacers and bits to get it all very much more concentric than I used to manage.

The encoder feeds a signal to a little chinese tach/hourmeter so that I can geek out about motor speed and possibly provide some useful instruction to employees about which gear to use at what speed for max eff and adequate air cooling to motor. Mostly I just want a tach for geek value but I think there's almost an argument to be made for including it.

I've also made some progress starting to install the control equipment into the main control enclosure. Cooling fans and vents are done as shown in pics but also starting to arrange components in the main cavity and working to finalize my wiring diagrams so I don't forget too much. So for just the controller, current shunt and main contactor are mounted but others will surely follow and space will get tighter.

 

[Reid_in_QC]

A few more pics for the Aux 12V battery install.

the 12V battery for accessories and logic is a 8P4S module of LiFePO cells same as the main 8P12S modules for the main 40V traction battery - but smaller by two orders of magnitude. It fits up inside the lefthand fender and I don't want to have to extract it regularly so I wired in a balance and power port so it can be monitored and balanced without having to remove the panel and battery.

The balance connector is a $5 5-pin "aviation"-style connector is from the Chinese eBay vanguard (who are hopefully not involved in aviation at all) and the power connector is an Amass XT-90 with a tidy little (hidden in pics) bulkhead mounting accessory I also found on eBay.

Otherwise it's really just an accessory battery and probably way overkill for the purposes. My DC-DC from main battery is cheap Chinese 400W and max load for 12V circuits is way under that but you never know. What if I install a killer stereo on 12V and a couple big subwoofers?

Next up is assembling the main battery pack. honest. I really can't mince around that elephant in the room any longer. Or can I?

 

[Reid_in_QC]

Warm weather allowed me to do some (sloppy and quick) painting to make the battery box look less shabby and now I'm actually starting to assemble the main battery pack.

See attached pics for battery config. Cells are LiFePO4 from A123 model anr26650m1-a. The 8P12S Modules were built by BAE Systems and I stripped their BMS boards due to complete lack of comprehension as to where to even start with them. I've opted to make a massively parallel setup at module nominal 40V output. Using 14 modules (7 top & 7 bottom) final pack config is 112P12S because top and bottom layers shown in photos will also be paralleled once installed in battery box.

Because these batteries came to me used and ostensibly at "end of svc life" I've tested them thoroughly and chosen the best. I've discharged and recharged on a 12S balance charger and have eliminated any modules with faulty cells.

Original spec on these BAE modules is 723Wh so about [email protected] My best ones come in around 15Ah and the worst in the pack are down around 11.5. Given what I paid for them I'm pretty happy with about 70% of original capacity and energy density isn't a crucial issue for me (this is a tractor, not a motorcycle or dragster).

So far, so good. I couldn't think of a better way to make the parallel connections so I made some slender copper busbar material by drawing flexible copper tubing (maybe 1/8" nom.?) through a slip roll until it was collapsed, compressed and drawn out to proper thickness. Liberal application of elbow grease lets you draw a full 33% more length from your copper tubing as you collapse and flatten, even on a cheap little Chinese slip roll. Copper is so cool.

Next step is more elbow grease to collapse and draw some bigger copper tubing to make power connection busbars for the ends of the packs and then wrestle it all into the box without shorting anything.

 

[Reid_in_QC]

Power connections assembled on battery pack. Again I made busbars by rolling/drawing copper tubing through a modified slip roll until I got something flat and thin and uniform enough to convince me. There's lots of contact and really I'm sure it's all overkill but it's going to be hard to get to at any of these connections once the battery is installed in the box so i'd rather overdo it while it's accessible.

Putting half of the pack on the scale gives 61kg (nearly legible on the scale) so the whole pack will probably weigh close to 150kg when fully assembled in its steel case. This is fine for my application but doesn't yield very promising energy density figures.

If we go with my tested results for these second hand packs which yielded an average of around 13 Ah @ 40V or 520Wh per pack. With 14 packs in parallel I may have 7.2kWh max capacity and at 150kg that's only 48Wh/kg (barely better than book values for lead :-(.

In their original condition these modules, without case, should have yielded 84Wh/kg and I guess I could exclude my case weight for a more comparable 61Wh/kg which is fine given the price I paid.

All the same, this pack, even built of used A123 cells, should outlast lead by a significant margin and I can charge and discharge it much faster, should I need to. Also it cost me far less than a lead pack of similar capacity and does weigh less, or give me more runtime with the same allowable weight.

Really the big unanswered question is how much runtime for what kind of tasks will I get from this rather bulky battery made from second-hand lithium packs. Tune in a month or so from now for the answers.

~ reid

 

[Yabert]

Really cool battery and busbars.
Like it.

 

[gipetto]

That's a cool project and I would like to attempt similar, though I would have picked something with a 3 point linkage and pto, like an early fordson major or single cylinder steyr( smaller, and many succumbed to frost and crankshaft damage)
Those old non live drive tractors were an easy conversion compared to the modern stuff.

 

[Reid_in_QC]

I would have picked something with a 3 point linkage and pto.

This is not a utility/chore or general purpose tractor at all. Never was, never will be. It's a high precision tool carrier for seeding and weeding tools that are belly-mounted. It's a specialized high-precision tool carrier for high value crops.

I'm an organic vegetable farmer. We have 3 other "normal" diesel tractors and one more of these Hefty-G tractors awaiting restoration and conversion to electric.

The whole point of these tractors is to put the tools directly in view of the operator and between the four wheels to allow precise steering correction to follow crop rows with mechanical weeding tools.

And it does have a 3-pt hitch, it just happens to be in the middle of the tractor

~ reid