Sunday, December 27, 2009

Cleaning the Battery Posts

Since aluminum oxide is quite non-conductive I decided to clean all of the battery posts on my new TS-LFP100AHA batteries. You can see the surface oxidation on each of the battery posts in the sample photo below. The post on the left is the negative post, and on the right is the positive.Even though copper oxide is relatively conductive I decided to clean it as well. I preparation to do the cleaning I had ordered a stainless steel wire brush, fine single cut file, and a bottoming spiral M8x1.25 tap for the post threads. I first used the wire brush to clean off each post. I could have stopped there but I wanted to make sure I had a smooth surface for the copper straps to rest against. I put a thin layer of NOALOX on each post and then used my file to smooth off the top of each post. The NOALOX helped keep the aluminum from sticking to hard in the file grooves and also put a thin layer back immediately on the aluminum to minimize oxidation. Using the fine turned out to be a good thing. I found a battery where the center post didn't stick above the nut which holds the post in place. I found the nut on each post significantly looser than on other batteries. I'm glad I discovered this since I wouldn't want the post to allow air inside and/or work loose.

After filing the tops and putting a coat of NOALOX on each post. I then put NOALOX in each hole using a cotton swab. I followed this with the tap then a cotton swab again to get any filings out of the hole followed by NOALOX again. This may have been overkill but I did find that there was a significant amount of inconsistency in the threads. I don't think it would have made a difference in getting the bolts to work but I wanted good conduction to the bolts. I'm using a BMS from Black-Sheep Technology which mounts on top of a brass bolt with a tapped hole for a screw which holds the BMS module. I have the BMS version for the TS-LFP90AHA which has the same post spacing as the 100AH battery. The brass bolt makes sure I get good conduction to the BMS module and having the screw in the bolt makes sure I don't over stress the pcb.

Right now I have the batteries hooked up in two separate parallel strings getting their initial 4.2v charge as required by the TS documentation. I have one string hooked up to my bench top lab power supply and the other set hooked up to a 4.2V smart charger. I initially did a bulk charge up to around 4 vpc using my Zivan NG-1. So far I have pulled well over 14KWh from the wall. [edit: I do not recommend charging any LiFePO4 cell to 4.2V. Even TS has lowered their max voltage to 4.0V. It shortens their life and there is very little energy above 3.45v any way. I'm now only charging to 3.485vpc. See my January 2011 blogs.]

Now I'm working on a way to mount the batteries in the Gizmo battery box. The lead acid batteries were not bolted down, nor was the battery box. It just sat in the Gizmo frame. I don't want the Li batteries bouncing up and breaking a BMS board or something if I hit a big bump.

Thursday, December 24, 2009

Lithium Batteries Finally Arrived!

Thanks to the hard work of EV Components I finally have 40 TS-LFP100AHA cells for use in the Gizmo. I placed the order in September 2009 but due to several circumstances out of their control the shipment didn't arrive until now. Fortunately they were able to find a shipper who would ship to Seattle and customs didn't take a full month like it did on earlier shipments. Dave Kois, of EV Components was kind enough to let me drive up to pick them up today. I arrived while the truck was still unloading. They sure have their work cut out for them. Look at all those boxes of cells!

While unloading the forklift ran out of propane so they had to run fill the tank. They are anxious to get an electric forklift so this doesn't happen and also so they don't have to breath the fumes. Below is another picture. You can see the brand new Toyota RAV4 that Dave is converting.

Below are pictures of my cells in my 2000 Honda Insight. It was definitely over gross. I'm glad I'm not that heavy of a person. There wasn't much spring travel left.
The box in the back is the battery box out of the Gizmo. I took it out and had it sand blasted. The bottom and corners were badly eaten by battery acid. It will last much longer with Li batteries in it. I'm going to paint it black inside and out. I'm also planning on plugging the drain holes so that water doesn't get in from the bottom. I still need to put in a splash guard in front of the box so the batteries stay clean and dry. Last time I had the batteries out I put baking soda on the bottom of the box. When I pulled them out the entire bottom was wet with baking soda crystals on the sides of most of the batteries. I determined that most of the water entered from the two front drain holes which are right where the tires would splash water on the box.

Well, I'm off to get my batteries ready to install. I still have several things to do.
  • Primer and paint the battery box
  • drill mounting holes in the battery box to match up with some threaded holes in the frame of the Gizmo so the box doesn't bounce around. Even 500lbs of batteries would bounce on some of the bumps I've hit.
  • clean off the battery posts and coat with NOALOX. I'm going to use a very fine file for this with NOALOX on the file so the aluminum won't have time to oxidize.
  • Run an M8x1.25 bottoming tap down each hole to clean the threads. I'll use NOALOX here too. The BMS I'm planning on using will mount to the top of brass bolts which hold the connecting straps so I want good conductivity through the bolts.
  • connect the batteries up in parallel and give them their 4.2V initial charge. I have a lab power supply I can use. I also picked up a 4.2V Chinoz Smart Charger from EV Components to do this.
  • Rebundle the batteries in the arrangement I need to buddy pair them.
  • finalize my mounting design and mount the the batteries to the battery box.
  • It may not matter much but I'm going to put NOALOX on the ends of all of the connecting straps. They are four layers laminated with some shrink wrap. See the picture below.

I have several other things to do but I'll post about them later.

Sunday, November 15, 2009

Battery failure ==> new pack

This summer my wife drove to work and back and then drove back to near work again to meet me at a store. Unfortunately there was no charging done in between trips. Normally this double trip would not have been a problem. I've gone 20 miles on a charge in the summer before with no problem. Apparently my range has decreased significantly. On my way home my PakTrakr gave me a message that battery #2 needed to be charged. I should have turned around and gone back to the Cowlitz PUD building where there are some EV charge plugs and plugged in for a while. Instead I tried to nurse the Gizmo home but I had one bridge to climb up and over and then 450 feet of elevation to gain on top of that. Well, one cell in battery #2 doesn't bubble on equalize anymore. I think I reversed the cell climbing the hill. Fortunately, my son and I picked up Gizmo #26 in Minneapolis, MN this summer. It had an aging pack of 6V flooded batteries which I've installed for now. I have a pack of TS-LFP100AHA on order from They are supposed to arrive the end of November 2009. I hope to have them installed durning my Christmas break.

While I had the batteries out I discovered a couple of things. First, the bottom of the aluminum battery boxes were badly etched from battery acid. I recommend that if you have metal battery boxes that you check them regularly for potential structural problems. Also, the battery box from the Gizmo I picked up in MN had cracked welds on all four corners. The Gizmo had been used as a pizza delivery vehicle so it has significantly more miles than Gizmo #31 does but it is another structural item to check.

Wednesday, April 29, 2009

I can SEE!

Wow! I just got back from a 10 mile run on rural winding roads and a short jaunt through town. The new headlight is much better than the old sealed beam. High beam is actually brighter than low beam now and I can light up the whole road if needed. I don't think I'll be getting many near misses with the elk which like to frequent my property. I can stand not having the adrenaline rush of nearly hitting one. With the old headlight It was difficult to see far enough ahead for comfort. Now, that is not a problem. This headlight mod is well worth the work it took. If any one wants to build one send me an email. I'll find out how much the guy who built the fiberglass cup wants to build you one too. I can give you measurements of my mounting plate if you want.

Friday, April 24, 2009

Splash Guards in place

I finally have the splash guards in place. Below is the right side. I split it into two pieces since this side has the new Gizmo Interface and I still need to get to it easily for programming as testing continues. It is a little difficult to get the splash guard out from around the front shock. The material is pushed up under the lip of the tub so no water will be able to get on the wiring from above. I'm hoping not too much enters from the front.
Here is the rear splash guard. The original one was not much wider than the fender. Carl Watkins of NEVCO said that when they had Gizmos come in for service they would widen this splash guard to protect the motor more. This one is bolted to the lip of the tub and zip-tied at the bottom. I also pushed the material up as high as I could since the hinge holes are exposed and water just runs down the back of the Gizmo and through these points.
Here is a close-up of the right hinge hole. You can see the plastic through the chip in the fiberglass.
The right side splash guard is one piece. I don't think I'll have to get access to this wiring as often as the other side. It wraps around the back corner a little ways. I'm trying to keep spray from getting on the terminal block mounted on the gold colored aluminum mount just visible through the plastic.In addition to this I bought some silicon spray from the Napa store and sprayed the controller and wiring. The spray insulates and helps keep water from sitting on things. I hope the controller will stay cleaner than when I had battery terminal spray on it. That stuff was a dirt magnet and the controller looked terrible. I feel much better about driving in the rain now.

Thursday, April 23, 2009

Gizmo Interface update.

As with any new item there is a testing and tweaking phase, then add a feature and do more testing and tweaking. The Gizmo Interface is coming along nicely. Variable regen is working on demand and the other basic functions work too.

Currently, while driving forward, I can press the activate regen button and vary the regen current with the throttle trigger. Unfortunately the controller only has a regen range of 50% of the max setting to the max setting. I wish it had a high and low setting so I could get a wider range of regen. I ran it with 250A max regen for a while but the low end was a little too agressive. I lowered the max to 200A which makes the low end more useable but I don't get the agressiveness I like at the high end.

I can also get regen if I pull the throttle while pulling on the brake. This is useful when doing a panic stop and you forget to hit the activate regen button.

I found a controller setting where the regen voltage can be changed. At first I thought this was an input voltage that the controller read to adjust the regen amount. In testing I discovered that this is the voltage in the armature (I think it is the difference in voltage but I'm not sure) at which point regen stops. I lowered this from 2.00V to 0.50V and now I get regen down to about 4mph. At these speeds the motor is spinning too slow to charge the batteries and it actually is a drain on the batteries. The batteries supply more current than generated at around 10mph but I'm trying to save the brake pads, not extend range. While I don't have a way to test it, regen probably comes out a wash as far as extending range goes. However, making the brake pads last an extra 500+ miles would be a great thing. As I expected, the motor doesn't get a chance to cool down during the coasting to a stop phase so it tends to warm up a bit. I'll probably add a blower to the motor to compensate for this. My 2 mile 400 foot elevation climb (most of this is in 1 mile) tends to heat things up quite a bit. I regularly see 250+A out of the battery. I'll have to check the armature current some time to see what it is.

The actual setting which controlls the regen amount is the Footbrake setting. I set this to come on at a voltage a little higher than the acceleration voltage and be at max just below the maximum voltage the throttle trigger sends out.

Next up: My new splash pannels. (Thanks goes to Fred in FL with Gizmo #30.)

Tuesday, April 21, 2009

Headlight is in!

I finally have a working adjustable headlight in my Gizmo! I did discover, however, that adjusting a Gizmo headlight is a bit different since the nose drops a fair bit when I sit in it compared to when it is empty. I finally just went out and drove it and then would stop and adjust the headlight and then drive some more. This headlight has a relatively sharp cutoff line so if it is aimed too low you can't see very far down the road.

This particular headlight has a 5W bulb in it too. I wired this small light to come on when the key is turned on and wired an ON-ON switch in the dash to turn on the headlight. The down position has the headlight off with the 5W light on. The up position turns off the 5W light and turns on the headlight. I guess you could say I have a running light mode.

As I talked about in my previous blog this mount has been a bit of a process. My goal was to build a unit which could be easily added to any Gizmo without having to do body work. Other than the less than careful extraction of the original fiberglass plug mold, I accomplished my goal.

Here are several pictures of the itterations I went through to get to the final product. I didn't take a picture of the original plastic sign material prototype. The next step was to see what it would be like to create a plastic mold of the headlight hole. I lined the hole with plastic wrap and filled it with wax. I supported a stick in the center of the wax to aid in pulling the plug out when it cooled. Here is the final carved result.

Next, I made a 1 inch thick model because I was thinking I would make the mount out of UHMW plastic.

Finally, the fiberglass cup idea was done. Below is an aluminum prototype plate made out of some damaged door kick-plate aluminum. I actually mounted the headlight and assiociated hardware to this plate. This showed that I would either have to carve the final aluminum plate thin around the mount holes and/or carve the plastic stops down on the plastic nuts. I ended up doing both. On this plate the top nut bent out the top of the hole and bent in the bottom.

Here is the fiberglas cup and final aluminum mounting plate. You can see the grinding marks above and below the top hole and to the sides of the side square hole. The eyebolts were mounted to the plate to hold them square and inplace while I used JB weld to glue them to the cup.

Here is the cup with the eyebolts cemented in place. If this doesn't work I will build another cup with the area where the screws go built up. I'll grind notches in the back side and install some blind nuts to bolt the front plate to this cup. You can see two screw holes beside the wire hole in the center of the cup. As it turned out, it was difficult to hold the nuts in place while the screws were tightened from inside the nose of the Gizmo and then install the headlight adjusting cup. If I have to take the unit out I'll be installing blind nuts in these holes.
Here is the finished unit with the aluminum plate painted. I didn't want to see it from the front.Here is the unit with the adjusting cup in place.
In this bottom view you can see that the adjusting cup is offset to the left side. This is to accomodate the adjusting screw on the right.

Here is the unit installed and ready for the headlight. You can see the damaged and partially repaiared portion of the nose.Here it is, ready for use!
If you have a Gizmo and want a similar unit send me an email. I can find out how much the guy who built the cup for me would charge to build you one too. I don't plan on making a kit out of this. I can give you a list of parts and dimensions I used to help you out, however.

Saturday, April 18, 2009

Progress on the Adjustable Headlight Mount

The headlight which came in my Gizmo was mounted in a specially shaped hole in the nose of the vehicle. The method was both inexpensive and made for a clean looking finish. There are three problems I had with the setup as it came. One was that water would collect under the headlight so I had to remember to swish it out before opening the hatch. The hole needed a slight down hill slope to it to drain water. The second issue was that the headlight was hard to aim. It was mounted with velcro at the back of the hole. I was able to get it close most of the time. Fortunately the beam was wide enough that when the aim was off it wasn't too bad to drive. The third thing had nothing to do with the design of the vehicle. It was the fact that the headlight was designed for a 4-headlight system. There is nothing wrong with this except that the wattage of the high beam element was lower than that of the low beam element. Turning on high beam meant that the road got harder to see even though the aim was higher. Initially I looked for a replacement sealedbeam headlight which was for a two headlight system and found that my '97 S-10 pickup had the right headlight. I was going to buy one and the plug that matched it (it has one angled pin like this |_\ whereas the existing headligh had a pin pattern like |_|) but I decided to see if I could come up with a way to have the standard aiming mechanism that a car has. I wanted to do as little body work on the Gizmo as possible. I gave up on making a standard size headlight mount and went with a smaller headlight. I found a Hella Headlamp upgrade in the 150mm rectangular size which included the mounting frame. I purchased it from Susquehanna MotorSports at They were most helpful and were willing to go measure the size of the headlight for me to see if it would fit. The only modification I had to make was file off the tip of the upper adjusting screw slot.

Below is a photo of the headlight hole. The velcro is visible in the hole. The dust is because the Gizmo has been sitting waiting for the replacement Gizmo Interface board.

Here is the back of the headlight. The piece of velcro on the bottom of the headlight came off because it was in water so much.
Once I had the replacement headlight I had to figure out how to mount it. I tried a wax mold of the hole as a pattern but that was a little difficult. I also carved a 1 inch thick piece of wax to try since I could get some UHMW plastic in that thickness but I was concerned about the heat generated by the headlight. At least the wax setup gave me a proof of concept item so I knew the headlight would fit. Finally, one of the members of our EV club, said that he did fiberglass work and could make a mold for a cup which would fit the hole. I could then mount the headlight mount to that and slip the whole unit into the hole. So, that is what we are doing. Below is the first step, building a plug the shape of the original hole.
Unfortunately there was a weak spot at the top of the hole in the Gizmo which cracked when removing the plug so I still will have to do some repair work on it. After the plug was made, a mold was made and then the cup which is going into the Gizmo. Below is the inside of the nearly finished cup and the aluminum plate I had available for the headlight mount.
Here is the back side of the cup and the aluminum plate. I'm planning on epoxying the four eyebolts into the cup. I mounted them to the aluminum plate to keep them square with the plate. I don't have any room to bolt from the outside of the cup so I hope the epoxy will hold the stainless steel eyebolts. I'm trying to decide between using some Plastic Steel Epoly and some Permatex Epoxy Metal Filler. I may go with the metal filler since it is supposed to handle a 2 inch gap without support. Only the tips of the eyebolts will be close to the fiberglass cup.

The two pieces as they will go together.Here are the two pieces as they will look when assembled.The two square holes hold the nylon "nuts" for the adjusting screws and the two holes in the lower left corner are for the tension spring. I went to a recking yard and got the screws and spring out of about a 1990 GM pickup. There is a whole list of vehicles from 1986 through 2002 which used the 150mm rectangular headlamp.

Sunday, April 12, 2009

Finally on the road! (Day light only.)

I finally got the Gizmo going enough to test drive it on the road beside my house. The programming of the new Gizmo interface is still being worked on. I tested things with the rear wheel jacked up and the front wheels blocked. Naturally, when the programmer is on the east side of the country and I'm on the other, I have the be the eyes, ears, and hands of the programmer. The strange thing was that the only way I could get the rear wheel to turn was to have regen on and enabled then pull the throttle trigger. A quick phone call and an email later and things started working like expected. We needed to get things to the point that it would be safe to drive while the other things, like regen settings and such, were figured out. I have to program the settings in the controller and give feedback to Ron on the tweaks needed in the interface.

I drove the Gizmo to the April 7, 2009 LCEVA meeting and then trailered it home since I don't have a working headlight yet. I'm still working on the new mount. Hopefully it will be a drop-in retrofit for other Gizmos.

We decided that it would be best and cheaper if I bought the programming interface to hook to my laptop and the Gizmo interface so that we wouldn't be sending boards back and forth across the country just for programming. The programming interface has now paid for itself in savings of shipping costs.

Here is a picture of the board:
This was the dryest location it would fit in. The old one was in the splash of the front right tire. The contactor is on the left and the DC-DC is on the right. I still need to bundle the wiring a little differently but this was good enough for some road testing.

In the next photo you see the cover over the Gizmo interface and that I have mounted the fuse block on the top cover. The fuse block used to be under the rear tail piece and would get water on and through it any time it rained, which is quite often here. This location should be much better since I'll install some splash guard plastic over the entire side when I'm done.

In addition to replacing the old Norm interface with the new Gizmo Interface, I replaced the old two wire speedometer pickup with a three wire version. The old one picked up too much interference when regen was activated and it also frequently gave spurious readings when it got wet. With this new one I actually can get a reading of 1mph. I rarely saw that with the old sensor. The sensor and mount can be purchased from Black Sheep Technology in the Gizmo Parts section.
The wiper relay used to be mounted under the tail peice beside the fuse block. Since I needed to add another relay I decided to move the wiper relay to the same location. The photo below is of the left side of the Gizmo. The spring and rod are for the mechanical emergency brake. I also wanted to get rid of the stack of ground wires going to the negative post of the battery so I installed a terminal block and connected the ground to each post. The gold sheet metal is actually an aluminum kick plate from the front door of one of my rental houses. The tenant's dog literally chewed one edge of it so they had to replace it. I kept it in case I needed some aluminum sheet. Well, here it is, at least part of it. I used two C-clamps and some pieces of wood to make a metal bender. A block of cedar and a hammer were used to make the bends clean. The lower right corner is bolted to the frame using an existing threaded hole. I zip-tied the other side and will probably just silicon the top edge so I don't have to drill more holes in the fiberglas tub. Even without anything holding the top edge, it doesn't rattle.

The relay on the right has a diode across the coil. It is inside the black heat-shrink tubing. I need to add a diode to the wiper relay and the high-beam relay. When the coil in the relay is de-energized the decay of the magnetic field causes a spike in the voltage. The diode shorts this out. I discovered that the headlight relay was wired wrong so that the headlight low-beam was on when the coil was energized and off when it wasn't. I'm wondering if the spike produced every time the headlight was turne on and any time the wiper was run along with any other transients may have been part of the cause for the old Norm circuit failing. It had no "surge protection" in it.

Sunday, March 1, 2009

Original Regen Circuit

My Gizmo came with a regen system in place. Various methods had been tried on this Gizmo including a pressure transducer on the brake line. What the end result was, was a single setting for regen which was activated by the brake light circuit. There is a toggle switch on the dash which was used to enable or disable regen. The difficulty in using this regen mode is anticipating when to lightly put the brake on so as not to stop too soon or too late. Right now I am in the process of installing a replacement "Norm circuit" on my Gizmo. This circuit read the hall effect sensor in the throttle handle and would change the state of the FS1 line (pin 4 on controller connector B) based on the state of the trigger. FS1 would be shorted to ground (FS1 closed) when the throttle trigger was pulled and then FS1 would be opened when the throttle was released. For regen to work FS1 has to be in an open state. When my original circuit became intermittent I decided to find a replacement. Ron Anderson of Black Sheep Technology was commissioned to build a replacement circuit. The new circuit will include variable regen, amoung other things. The purpose of this post is to document the original regen circuit. I will post about the new circuit when I get it installed and operational. Right now I'm rewiring. In a week or two I'll be testing out the programming of the new circuit.

Note that pin 11 on Controller Connector B is an analogue input used for regen with a voltage range of 3.5-0V. I don't understand why this input is from 3.5-0V when pin 10 (torque/speed command pin) has a range of 0-5V but that is what the controller manual says. Oh, the controller I'm talking about is a Sevcon SepEx PP745. Below is the circuit diagram for the original regen setup.

Sunday, February 22, 2009

Drive cutout issues

I have Gizmo #31. This Gizmo has a D&D SepEx motor and Sevcon PowerPak controller.

Through trouble shooting intermittent drive cutout issues over the past couple of years I believe I have an understand of how the NORM circuit works on this vehicle. This vehicle also has circuitry installed to enable constant regen when the brake light comes on. The NORM circuit affects this too.

My NORM circuit has 5 wires attached to it. The pink wire is attached to the switched +12V line of the DC-DC converter. The White wire is connected to battery pack negative. The blue wire is connected to pin 4 on the controller connector B and is called the FS1 line according to the controller manual. In conjunction with the NORM circuit is a hall effect device in the right hand control handle to provide feedback based on throttle trigger position. One wire is labeled "-" (black wire on my rig) and is connected to the white wire on the NORM circuit (battery pack negative). On the opposite side of the hall effect device is another wire labeled "+" (pink wire on my rig). This wire is connected to the yellow wire on the NORM circuit. The NORM circuit has a voltage regulator which puts out a +5v on this yellow wire. In the center of the hall effect device, between the + and - wires, is a third wire (orange wire on my rig) which connects to pin 10 on controller connector B. This line is an analogue input from 0-5V and is what the controller reads as torque request. In addition to connecting to pin 10 on the controller the center wire from the hall effect device also connects to the orange wire on the NORM circuit. This is necessary because the NORM circuit needs to know when throttle input is requested so it can control the state of the FS1 line. The schematic is below:On power-up the NORM circuit must have FS1 in an open state so that the Controller will boot up without giving a sequence fault. If FS1 is closed (shorted to battery pack negative) then the controller will sequence fault as a safety precaution. When the throttle trigger is pulled it sends a portion of the 5v coming from the NORM circuit out to pin 10 on the controller. On my rig this ranges from 0.82-4.24v as seen by the controller. When the controller sees a voltage greater than its lower threshold it knows that torque is requested but it won't do anything if FS1 is open. This is why the NORM Circuit must have a connection to the output of the hall effect device. The NORM Circuit monitors the voltage from the hall effect device and when a voltage greater than the lower threshold is seen the NORM circuit closes the FS1 line and drive will commence. Provided of course a direction is selected and no other fault conditions exist. When the throttle trigger is released the NORM circuit opens FS1. On my rig regen is enabled by the brake light. Regen, however, is disabled if FS1 is closed. This is why FS1 must be open when no throttle input is selected. If I didn't have regen FS1 could remain closed after the controller boots up and I wouldn't notice.

I don't know what kind of mechanism is used to close FS1 but it seems to be what has caused nearly all of my drive cutout problems. I have had the following drive related issues:

1) No drive on power-up because FS1 was closed, this caused a controller sequence fault.

2) No drive after power-up because FS1 remained open even when drive was requested. If I held the throttle in for 5 seconds-5 minutes sometimes drive would start to work. This usually happened if the sun had been shining on the vehicle for a while but eventually it didn't matter.

3) Similar to #2 but would happen while driving but only if the throttle was released.

4) Drive would cut out while moving and throttle still pressed because FS1 would open.

5) Regen would start to work then cut out or just wouldn't work at all.

6) Regen would pulse on and off, usually at speeds below 25 or so but sometimes at all speeds.

I believe that all of these issues in my case have to do with the NORM circuit not properly handling the state of FS1. To deal with these issues I installed a SPDT switch in the FS2 line as shown in the schematic. The common side of the switch is connected to the FS1 line (pin 4 on controller connector B). One side of the switch is connected to the blue wire on the NORM circuit. The other side of the switch is connected to battery pack negative. I chose an on-off-on style of switch so I have some options. For normal operation the switch is set to connect the blue wire to FS1. If on power-up the controller gives a sequence fault (2 blinks) I switch the switch to the off position momentarily to clear the sequence fault then back to normal operation. If the NORM circuit is just not closing FS1 I switch the switch to short FS1 to battery pack negative and I can then drive. This is how I deal with items 1-4 above.

To deal with items 5 & 6 regen I switch the switch to the off position and regen works just great. No pulsing and it never cuts out until the speed is too slow, of course.

The problem with this is that the switch has to be switched back to one of the on positions for drive to work. This is a bit of a hassle but it sure beats being stranded.

Right now Ron Anderson of Black Sheep Technology is working on a replacement unit. We hope to integrate regen into the unit too so that others can easily add regen if they have the same controller setup as I do.

Hopefully this helps. Feel free to contact me if you have any questions.