Tuesday, November 22, 2016

Discommodulated Battery

No, it's not a crime under Napoleonic Code, it is what we managed to do last week, and it is much easier than posting rotated photos on Blogger....

One LEAF battery pack has been disassembled into sub-packs.  We identified the components, taking special care to disconnect the busbars that were easiest to remove first, as those mid-pack breaks greatly reduce the voltage present and potential for shock.  We learned to snap back the insulating covers after removing the busbar fasteners, as the busbars can be removed without exposing any bare copper.  We double-checked the disconnection of all small sense leads and made sure there would be no snags or shorts in the wiring harness, and then lifted out each of the sub-packs.

Next step is to measure these sub-packs and confirm they won't be a ready fit into the Ranger Sarcophagus, and then we'll advance to disassembly down to the module level, taking careful note of how the sub-packs were assembled, as we'll need to reuse the end plates and hopefully many of the busbars and the BMS wiring harness as we reconfigure the modules into a series string of 'buddy pairs', and restack them using the original hardware, but with a different number of modules in each subpack.

Disassembly without Destruction

The LEAF battery packs are a wonderfully compact kit, with an interesting mix of fragile and overly beefy componentry.  We've now broken down the subpacks into individual battery modules, without a single broken wiring connector and only one stripped screw, out of dozens apiece.
At 8.2V apiece, the individual modules don't pose a shock hazard to bare skin, but the subpacks did, so extra care was employed when exposing the network of interconnected busbars and removing those bolts.
Arrg, can't get it to stay flipped, but you get the picture!
Now all we need to do is fill the big box with what we pulled out of the little boxes.  Seems simple enough, but so does posting photos to a blog!

48 modules, once again in an antigravity orientation...

Thursday, November 10, 2016

We're Lovin' LEAF

Lovin' Life with LEAF salvage, that is.  The two collision-salvage battery packs arrived this week, as did the prototype Wolftronix LEAF Pack Sniffer.  This amazing little device turns on the BMS and reports back the pack voltage, average cell V, # of bars of remaining capacity, identifies the high and low voltage cells, and then scrolls through all the individual cell voltages.  One of the packs was sitting at 396V with four bars of capacity, the other at 383V and 10 bars of capacity.  Voltage as a state of charge indicator is mainly valid at the upper and lower end, and these packs were built two years apart.  I think they will balance out nicely, as individual cells voltages were tight after months of sleep, with only 6 thousandths of a volt difference between the highest and lowest cell.  Most were within 3 thou..  Here's a video of the sniffer in action!
Leaf Pack Sniffin'

The packs arrived on a rollback wrecker, so without a forklift we very carefully set the back edge of the pallets on the ground and then drove out from under them!


It may seem scary to sic a student with an air chisel on a highly charged battery pack, but research revealed the safety of this technique, as the sealed surface to be separated is outside of a one-inch protective lip of steel (and by another air gap and metal module covers on the inside) meaning one would have to overshoot the target by quite a distance and cut through multiple layers of steel after the target of a single layer of sealant! The air chisel made short, but very noisy work of this job, much to the students' delight!  The students quickly learned to distinguish by both sound and feel whether the chisel was hitting sealant or steel!

Wednesday, November 09, 2016

Vindication Monday!

'twas disappointing to see a short in the motor after rebuild, but perhaps that could be explained.  The controller label still wasn't visible, but a document search said that the Sand Star E Plus offered regenerative braking.  That could only mean one thing on a brushed motor with four terminals labeled A1, A2, S1, S2..... SHUNT WINDINGS.  The shunt wound motor is another old workhorse, but rarely seen on golf carts.  It's prime characteristic is that speed is controlled by varying the voltage on the 'shunt' that energizes the electromagnetic fields.  At a certain shunt voltage, the motor will give its all to maintain a particular speed, until the load is too much to bear.  Conversely, if a shunt wound motor is 'overhauled', or forced to turn at a faster RPM than the shunt dictates (rolling down a hill), it will generate current back into the battery.   This revelation comes courtesy of a little consulting and measurement by Dan Bentler, who walked me through the diagnosis.  The motor couldn't go through a full bench test, as the output shaft bearing is integral to the transaxle, but it turned just as expected once bolted back on.  Reconnect the controller, and a dead short appears.  So we have a bad controller, and that's not so hard to replace.  Removal of the controller confirmed that it was an uncommon but affordable shunt motor controller.

On the Grumman project, we got the motor mated back up to the transaxle, but discovered a different bolt spacing on the end bell that didn't match the other motor, but it will match if we simply remove and 're-index' the adaptor plate (rotating the motor housing 90 degrees).  We need practice installing motors anyway, right?

Friday, November 04, 2016

Setback Friday

Today was hoped to be a milestone, as we were all set to install the replacement motor in the Grumman, and also completed re-soldering of a broken field wire in the Smithco trike.  Alas, it was not to be.

Prior to installation, we did a test spin of the whole assembly (motor with mounted hub, flywheel, clutch and pressure plate).  It wobbled.  A lot.  It wasn't expected to be a perfectly balanced assembly, as the clutch centering tool isn't that accurate and the heavy Rabbit MK1 flywheel with a scallop cut out for clearancing obviously wasn't high-speed balanced at the factory.   Here's the video (With apologies for the authentic mechanic's shop color commentary.)

So, as a team we discussed whether to cross our fingers and run with it, or investigate further.   Out came the dial indicator.  We all learned how to use it, and found 20 thousandth's of an inch in runout. A coupla' thou would be okay, but not 20.  So off comes everything except the hub, and we dial it up again.  Still at 20 Grand.  SIGH, ...despair briefly raises its ugly head.  One team member even said "Well what did you expect, bringing the truck to a bunch of dumb high school kids.  If you wanted it fixed, you shoulda gone someplace else!"  

We popped off the hub, which was easy because we hadn't painted the thing with locktite like the Texans did waay back during the Carter Administration when this thing was first assembled...  BUT, there wasn't a setscrew in the keyway either.   Matter of fact, the setscrew hole wasn't even tapped.  WHAT IF?
So, out came the taps, and a bolt was quickly installed.  WELL WHADDA YA KNOW!!
Upon reassembly, the hub spun true, at 1.5 Thousandths of runout.  The threads weren't tapped exactly straight on axis, and our key was fractionally shorter than the slot, so this opportunity to correct both of those situations was seized upon, and we'll reassemble it all and install on Monday.

Similarly, two teamsters had tackled a very challenging soldering job in the Smithco trike motor and managed to get the motor fields intact again.   However, it was still a dead short when we tried to do a test spin.  I now think that the field wire snapped during disassembly, as it wasn't scorched like a short should be, and the heavier-gauge motor leads showed a lot of heat...  Back to the drawing board for this motor, but we eliminated a variable.