Wednesday, December 31, 2014

Clutch Alignment Repairs

I've had another day to work on the car and made some good progress.  Last I showed you, the driveshaft was high compared to the motor shaft.  I took some closer measurements and it's hard to tell, but I've estimated that the shaft is about 0.045 inches off center.


The bellhousing and motor adapter are located to each other with alignment "sleeves" at two of the bolt holes.  These locations need to be moved by 0.045 inches, but that's easier said than done.  I decided to punch out the alignment sleeves and replace them with relocated pins.  You can see the two housings here.

Getting a pin pressed in the adapter and aligning it with a hole in the bellhousing in the perfect spot is a little tricky.  I finally decided I'd make a couple special plugs to do it.  You can see them in the picture below.  Each one has a diameter that tightly fits to the locating hole in the bellhousing.  Then the diameter of the other end of the plug is 0.090 inches smaller than the corresponding hole in the adapter plate.  That way I can put them together and shift the adapter plate within the clearance of the plug in the correct direction.  Since the radial clearance is half the diametral clearance, it should move 0.045 inches.  The plug on the left is in the alignment hole.  You can see the two diameters on the loose plug on the right.

After that, I drilled a 5/32 hole through the adapter plate and slightly into the bellhousing.  Now that the holes are lined up with each other between the two housings, I opened the hole in the adapter plate just enough to press a 3/16 pin into it, and opened the hole in the bellhousing so the pin just slips in.  Here you can see the pin pressed into the adapter plate.  In the above picture you can see the hole it mates with in the bellhousing.

Now it's time to test the fit.  I put the adapter plate back on the motor, put the bellhousing back in the car, and put the motor back in with it.  The bolt holes that hold everything together have enough clearance that the bolts still fit through.  And here's how the shafts line up after the new alignment.  Not perfect, but a huge improvement.  I was able to shift the joint between the bellhousing and the driveshaft tube just enough that it looks perfect to the eye.  Considering 0.045 looked pretty bad, I think it ought to be good enough.

So now it's time to put it all together again.  I pressed the pilot bearing into the coupler (shown below), lubed it up with some bearing grease, and remounted the coupler, flywheel, new clutch disc and pressure plate.  Then I put it all in the car along with a new throw out bearing.  

By this time I've put the motor in and out of the car about 4 times in one day (not counting the dozen times I did it the last time I was trying to fix this problem) and I'm really hoping this is it.  Two years ago when I spun up the motor, something in the driveshaft assembly (probably a bearing) made this nasty, loud, clacking noise.  This time I hooked the motor up to 12v and the spinning makes a noise more beautiful than a chorus of angels playing harps!  No laugh of a child, purr of a cat, or chink chink chink of coins dispensing from a slot machine could melt my heart that day like the soft whirring of the brushes in the motor did.  This problem was in my top 3 issues making the car a pain to drive.  Time to finish putting the car back together and tackle the other two!

Oh and happy new years to all of you!

Monday, December 29, 2014

Breakdown

It was a pretty good year for driving the electric car.  My goal was 5000 miles, and I made it to 4893.  I was on track to make it, but the Tuesday a week and a half before Christmas, something happened on the way home.  At first I noticed a very subtle looseness in the steering that would have been imperceptible if you hadn't been familiar with the car.  Then I noticed some brake squealing coming from the front driver's side wheel.  Finally I notice the speedometer isn't working!  I know the speedo cable attaches on that same wheel that the noise is coming from, so I figure the end of the cable broke off and was somehow scraping on the brake rotor.

I got home and pulled the wheel off.  To my surprise, the entire rotor assembly was loose!  The wheel mounts to the rotor, and rotor is supported by a pair of wheel bearings.  One of the bearings was completely pulverized, and the only thing keeping the wheel from falling off was the brake caliper!  Yikes, I dodged a bullet there.

Meanwhile, I've been planning some maintenance on the car for the Christmas break, so parking the car a few days early isn't a big deal.  When I put the car together a year and a half ago, I had to fit the motor in and out a number of times and I was using the old clutch to avoid banging up a new one.  Well, before I knew it, the car was all put together with the old clutch and I wasn't about to tear it all apart to replace the clutch.  Well, not too long after I started driving, I began to hear some clutch noise.  It started slow, and eventually got worse and began to vibrate.  I decided it was time to get a new one in there.  The biggest downside is it requires removing either the front or rear set of batteries to make the swap.  Here I've got 24 cells removed along with the BMS.  No turning back now.


2 hours of work later and the motor is pulled out enough to remove the clutch.  When I pulled the disk out I was expecting it to look pretty bad and I was a little disappointed that it looked okay.  Not bad enough to be causing me all the issues I've been seeing.

I dig a little deeper, and here's the problem.  The pilot bushing in the flywheel adapter is totally worn out!  Not a good sign if this needs replacing every 5-6k miles.  Something more significant must be wrong.  For that, we need to jump into the way back machine for the rest of the story.

About 2 years ago I was spinning the motor for the first time and excited for the day I get to finally drive the car.  When I first mounted the motor in the car and spun it up, it made a fairly significant clicking noise.  I don't recall if I blogged about that, but I had the motor in and out a dozen times in an attempt to fix it, mostly working on improving the pilot bushing.  Nothing I did seemed to fix the problem so I started to suspect there was an alignment problem between the motor and the bellhousing.  I loosened the mounting bolts a bit and used a jack to lift on the motor adapter to see if that made any difference.  What that did was change the angle between the motor shaft and the driveshaft.  I found that a very slight angle made a significant improvement in the noise, so I shimmed the joint and bolted it up.

The noise was all but gone, and I didn't look back because I was finally on the road.  Deep down I knew it was a bogus solution, but my desire to get the car rolling after 20 months of work was overwhelming so I chose to ignore it.  Well, apparently my days of looking the other way are over, and I find myself in the same spot I was in Feb of 2013.  What to do now?  Well, to start I need to fix the pilot bushing.  The original Porsche used a pilot bearing.  I didn't try this because I assumed it wouldn't fit with my flywheel coupler design.  Too bad I never checked the dimensions because it actually is smaller than the bushing I made!  With the increased load capacity of needle bearings vs. the phosphor bronze bushing, this ought to last longer.

My neighbor happened to be working in his garage the yesterday, and it also just happens that he is a professional mechanic and builds race cars, so I stopped by with the coupler to see what he thought.  He took one look and said the motor shaft and driveshaft centerlines aren't lined up.  He'd seen it a dozen times with poorly made motor adapters.  He sent me back home with a homework assignment of taking some careful measurements to figure out where the problem is.

I'm lucky enough to have an inspection panel on the clutch housing, so I have a 4x4 inch window to look at the parts.  Without the clutch and flywheel in place I can get a good view of the driveshaft and coupler.  I put the motor back in place and had a look.  Note that I've opened up the hole in the pilot bushing to accept the new bearing, but haven't pressed the bearing in yet.  Here you can see from the bottom that the shaft lines up fairly well left to right.

You can see a side view from the inspection hole, but I squeezed the camera in there and took a picture from the left side.  Now it's obvious there's an alignment problem in the vertical direction!  As unfortunate as it is to have this problem, I'm ecstatic to find a concrete problem that I can now solve!  Instead of being a poorly made adapter plate, I'll call this a poorly designed one since I took all the measurements.  Reworking the adapter is going to be interesting, but I've still got 7 days of vacation to figure out how to do it.  Better get working...

Wednesday, September 17, 2014

AC Update

I was thinking I hadn't blogged in a while, so I pulled up the site and realized I'd written most of this and never posted it!  I think I was waiting on taking a few pics before posting.  Whatever, enjoy and realize I probably wrote this sometime in June or July.


I've been driving with the AC working for a little while now so I thought it'd be interesting to share some info on how it's working out.  The biggest thing that's unique about how it works on my car compared to a normal car is that the AC compressor is driven by the main motor, so if the motor isn't spinning, the AC isn't actively cooling.  For example, when I'm stopped at a stoplight, it's like I turned the AC off for 30 seconds and just kept the fans running.  This wasn't a big deal when it was only 85 degrees out, but now that it's over 100, it starts to get a little warm if you have to wait through more than one light cycle.

Now there are a few solutions to this problem.  I can put the car in neutral and spin the motor with the pedal, but that's easier said than done.  It's hard to keep the rpms just right, and before you know it you realize you were concentrating so hard on keeping even pressure on the throttle that you didn't realize traffic was moving again.  Another idea is to fix the min throttle setting by modifying the resistance at the throttle potentiometer or voltage signal going into the controller.  That's a little worrisome to me though because you'd have to be very careful that you don't introduce a failure mode that could take your car out of control.  Ideally I'll reprogram the controller to maintain a minimum rpm (idle), but that's going to take some development.  For now I'll deal with the heat or work on my throttle skillz.

Another common question is how does using the AC affect my range.  Well, it takes about 10% more energy to get somewhere with the AC running.  So instead of 250 Watts/mile it's closer to 275.  That should drop my range about 5 miles.

The other thing I didn't anticipate was all the noise it makes!  With a normal internal combustion engine, the noise must be drowned out, but it probably doubles or triples the amount of noise in my car.  Also, I'm used to cruising at higher rpms, but judging by the sounds it makes, the compressor likes it a little slower.  So I've started using 4 gear again and shift at lower speeds.  Then when I'm creeping to a stop I tend to pop the car out of gear before I'm getting too slow because there's a fair amount of vibration that happens if the compressor is spinning abnormally slow.

When I first got the AC hooked up, I did like I normally do and got it 98% complete and didn't bother finishing it off for a while because it was functional.  Well, the last 2% in this case involved one wire that turns everything on.  When I got the car, this mystery wire was just dangling behind the dash and had been cut off from somewhere (still unknown).  I later discovered that if you apply 12v to it, it turns on a relay that powers the blower fan that flows air through the AC ducts and powers the control unit.  The way I half-assed this one was to wirenut an extension wire to the end, then clip a jumper wire to a 12v supply.  Of course that supply is always active, so I have to unclip it every time I stop the car or it'll drain the battery.  I finally properly wired it into an ignition source so it turns on when you flip the key.



Well, I finally got to 99% complete and now the car looks like this!  I finally hard wired the power to the ignition and I found a replacement center bezel in with the two vents and controller on ebay.

I also wanted to hook up more than just those two center vents, so I ran some ducting to the left of the steering wheel and the crotch vent I've been dreaming of.  Oh yeah!


After that I decided to treat myself by replacing these weathered old bags

With these nice new visors.

Another problem I've been working on since about April is that my headlights stopped working.  I know I had headlight problems around October of last year, but back then the lights didn't want to pop up.  Now they don't want to turn on.  Over the last year or so I've really struggled trying to understand the "current flow" diagrams in the Haynes manual, but there are a few symbols and notations that never really made sense to me.  Well, after 2 hours of looking at it last night it all suddenly clicked and I instantly narrowed in on a problem with my ignition switch not sending power to the lights.  Luckily it's another VW part that goes for $10 so I'm back in business with working headlights!

Now that summer's cooled off, I feel a little more motivated to work on the car.  So maybe in the near future you'll get to read about my 240v outlet, fixing the clutch, fixing the transmission input shaft seal leak, or who knows, maybe I'll even finish the last 1% of the AC system!

Sunday, May 18, 2014

It's Payback Time

That's right, it's payback time!  I've broken 3000 miles and instead of doing my first oil change I'm going to add up all the payback I've made so far from this car project after bleeding into it for two years.

We'll start with mileage.  I've driven about 3200 miles so far, and when you compare the cost of electricity to the cost of gas, I save $0.205 for every mile I drive when compared to driving my Jeep Cherokee.  So that's $656 so far saved in gas!  Not too shabby.  There are a few people who've turned their nose up at the prospect of saving money with an electric car, but let's think about it.  The real thing to compare is what's being consumed that's different between the two cars.  In a gasoline powered car, you pay for gas and oil.  Gas currently costs about $3.50/gal, and at 16 mpg in my jeep, that's $0.219/mile.  If you spend $30 on an oil and air filter change every 5000 miles, that's another $0.006/mile.   In an electric car you pay for electrons and eventually your batteries will need to be replaced.  I pay $0.08 per kWhr and consume an average of 250 Whr per mile.  So that costs me $0.020/mile.  The battery pack cost about $5500, so the break even point is when I drive 26,830 miles.  According to the literature, if taken care of, these batteries should have 80% of their original capacity after 2000-3000 discharge cycles.  In my 3200 miles I've recharged the batteries 132 times, indicating I should get 50-75,000 miles of use out of the pack.  And even then my range should still be enough to get me to and from work.  If I assume the price of gas doesn't go up in the next 5 years, I should still at least break even if I compared my car to one that got 32 mpg.

In addition to the fuel savings, I've had a few unexpected perks from work.  I won a contest for "commuter of the year" worth $50.  And I also have scored a free covered parking spot for driving an alternative fuel vehicle two years now.  Those spots go for about $400 each, and worth every penny during the summer when you can hop into a relatively cool 115 °F car rather than one in the sun at a blistering 150 °F!  (not a joke)

So in all, the car has paid me back $1506 in the first year of driving.  Not too shabby!

Speaking of blistering heat, last time I was talking about getting my AC running.  I had to mount a few fans to the condenser, install a new receiver/drier, mount the compressor, and wire it all up.  And I'm glad to say it all went pretty smoothly!  Except that the fans didn't want to fit, the receiver/drier had the wrong fittings because it's an aftermarket AC system, and the compressor...well, that actually was just fine.  Luckily the weather decided to stay in the 70s and 80s for a couple more weeks and I got all this done.

Here's the compressor and the third belt I bought that finally fit.

And here are the two condenser fans.



Last step is to charge it up and see what happens.  I've never done this, so I refer to my friend, Ed, for help.  To start, I'm not doing this the "correct" way.  As an engineer and a cheapskate, I like to do things just well enough, so I keep the whining to yourself if you want to tell me all the things I did "wrong" here.

Before you do anything, you need to get all the lines flushed out.  Well, I've already got a problem here because you're supposed to flush each component individually and you can't flush through the expansion valve.  The expansion valve on my car is buried so deep that I couldn't even get to it with the dashboard removed.  I decide I'm not gonna flush the evaporator or those lines, I'll just do the condenser.  Now, you're also supposed to use a special AC system solvent that will clean out any oils completely.  I'm sure those work great and all, but it's way easier to hook a hose to it and blast everything out that way!  A little compressed air and it all comes out.

Okay, the system needs a certain amount of oil in it.  I'm not sure what exactly the oil does, but I think the main function is keeping the compressor lubricated.  It spreads itself throughout the whole system, so you need a certain amount for the compressor, and another certain amount for the rest of the volume of the system.  Unfortunately, I can't figure out how much my specific, special, aftermarket system needs, so I guess it's about 6 oz.  There are also a number of different types of oil you can use.  The old R12 systems use mineral oil.  My plan is to switch to R134a, which typically uses an oil that is completely incompatible with mineral oil.  Since I didn't flush the whole system out, there's bound to be some left, so I use ester oil, which is supposed to be good for that.  There are all different opinions on where in the system you should put the oil at first, and there's a really convenient drain hole on the top of the compressor (to drain the shipping oil) that's just asking for me to fill it.  Of course, when I start pouring the hole is too small to pour fast so oil spills everywhere.  And now I really have no idea how much oil I've put in.  So it's a good thing I had no idea how much to put in in the first place because it really doesn't make a difference!

The receiver/drier's function is to filter the refrigerant and absorb any water in the system.  Because of the water part, you can't take the caps off until you're ready to charge or it'll quickly fill with moisture from the air.  I wait until the morning of, and of course I have problems getting it hooked up.  The current style fittings are all o-rings.  The old style fittings are flared fittings (what I have on the compressor).  And the really old style is barbed fittings.  The barbed fittings are supposed to have these clips that keep the hose clamps from blowing off under pressure.  The hose clamps have the clip, but the drier doesn't have the tang that it grabs onto.  Oh well, what's the worst that can happen?  They'll just blow off if the pressure gets too high right?

I'm not sure who's bright idea this was, but the drier is located in the front passenger wheel well.

Next step is the old R12 fittings should be replaced with R134a fittings.  Of course, the high pressure fitting doesn't fit on this compressor.  But luckily the low pressure fitting does.  Ed has all the connections you could want with his gauges, so no problems there.  After that you put a vacuum on the system.  This does two things.  First, you can see if it's leaking if you can't get the vacuum down low enough.  And second, it evaporates any water in the system, which I assume is bad for the compressor and could create vapor bubbles if it turns to a gas.  The vacuum drops quickly to 30 inches of mercury, so that's good, and we keep it there 30-40 minutes.

Now's the moment of truth.  Is it going to work or not?  We prime the system and spin up the motor.  Once again, we don't know the size of the system so we don't know exactly how much refrigerant to put in.  After running it a while, the high pressure is good, but the low pressure side is reading low.  Apparently this indicates a stuck expansion valve.  I'm not exactly sure what the expansion valve is doing, but here's the quick story on how the system works.  The compressor spins and compresses the freon, and in the process, the freon becomes hot.  After this, you want the freon to be as cool as you can get it, so it passes through the condenser.  That's the thing in front of your radiator, and one or more fans blows over it to try and get it down to ambient temp.  Then this medium temp, high pressure freon travels to the expansion valve.  The valve is a big restriction in the system (like if you pinch off a garden hose) and the pressure just past the valve is low.  According to the laws of thermodynamics, the temperature has to drop significantly when the freon goes through this rapid pressure drop.  Now you have very cold freon that goes through your evaporator that looks like a mini radiator.  The fan inside your car blows over the cold evaporator, the air becomes cold, and if you're lucky, it's cold enough that you don't sweat your balls off on the drive home from work.  The freon is warmer now and returns to the compressor for another go around.

So this expansion valve is a variable orifice that (I think) changes depending on the flow to try and get something near 30 °F coming out.  The pressure on the low side on mine is lower than it's supposed to be, indicating the restriction is too high and the freon is "too cold" when it comes out.  The only real problem here is that you will be prone to having ice form on the evaporator.  Lucky for me, it's dry heat right now, so icing probably won't happen on my 20 minute commute.

After looking at the gauges for a while we hop in the car and realize it's freezing inside, though it was only about 75 degrees outside at the time so it's not a fair test.  But luckily it was still nice and cold a week later when it was 95 out, and hopefully it'll stay that way when it's 115.

Well, you've wasted another chunk of your life reading my blog that you'll never get back so I'll finish it up for now.  So long and good luck staying cool!

Monday, March 17, 2014

Air Conditioning Part 1 and BMS Repair Part 1 Too Many

I've had more trouble by far with one part of the car, and that's the battery monitoring system (BMS).  Two weeks or so ago over the weekend we had a big rainstorm here and I think my leaky hatch allowed a little water to drop on the two rear BMS slave boards.  Monday morning rolls around and the BMS checks the first 24 batteries and stops.  I'm not too concerned because I've had an intermittent issue with the power supply to the rear boards, so I just drive it in to work.  On the way home I have the hatch open to stow the sun roof so I wiggle the wires in hopes of bringing it back to life.  Well it came back to life, but now it's alerting me that something is wrong.  It does such a good job at getting my attention that I have to unplug the whole system to drive home cause it's so dang annoying!

I get home and take a closer look and it doesn't look good.  Lots of scorched areas (or carbon scoring if you're a Star Wars fan).





Yikes!  Most of that carbon scoring was superficial and must have been water boiling from current passing through it or something.  I get it all cleaned up and check all the parts that are easy to check and remarkably it seems like most everything is working.  I hook it back up, and alas, everything is not working.  8 of the cell voltages read as zero volts.  Long story short I figure out one of the traces was vaporized, so I replaced it with this jumper cable.


I hook it back up, and now the 8 cells read correctly, but there's another one reading too high.  I eventually deduce that a resistor somehow must be slightly low.  In fact, the resistance must have changed so slightly that it would be difficult to measure that change while on the board.  I was able to correct that by desoldering the resistor, cleaning it, and resoldering, and bam, it's all working again!  Amazing I was able to do that in just a few days and without an order to Mouser!

So now I need to make sure that never happens again so I have to figure out a way to ensure water can't get on there.  I scour the garage and come up with a plastic storage bin that, with a little whittling, fits perfectly over the top of those two boards.  And here there are!

Now a lot of people ask me if I have AC, and my answer is always no, but it will by summer.  Well if you live in Arizona like me, you'll realize that summer is nearly upon us!  We hit the 80's a month ago, the 90's could easily be just around the corner, and if we're unlucky it could break 100 in a month.  I have no idea how long this is gonna take, so I probably should have started this 6 months ago.  

In all fairness to myself, I have actually done a fair amount of work on the AC over the last 18 months.  In fact, when I started when I took the dashboard out nearly 2 years ago.  The problem is that the AC is an aftermarket unit (i.e. not covered by the manuals), it broke some time ago, and a previous owner left a mess of wires with numerous cut ends in there to confuse me.  If you turn the system on, nothing happens.  No fans, no clicks, nothing.  I decide to start by trying to get the blower fan working.  Of course, that's buried in a (hopefully) watertight kick panel that holds the evaporator.  When I had the dash out I never was able to figure out how to get it out so I could have a look at where the wires go, so here I am trying to figure it out again.  

Luckily, with a fresh look, and a few years to mull it over in the back of my head, the wiring is actually pretty straight forward.  There's one mystery green wire that ends near the fuse box, which I correctly guess is the power supply for everything.  It actually turns on a relay, and I can make the relay click by applying 12v to green.  Now there's a 10 gauge dark blue wire coming from the relay over to a control box.  The controls are a slider that changes fan speed (0-4) and an ambiguous "cooler" slider that goes follows a tapered line.  I'm not sure if the coldest setting is setting "cooler" to the thin or thick end, but I can figure that out later.  The "cooler" slider mechanism has a temperature probe that goes into the evaporator to measure the refrigerant temp, and turns the compressor clutch on and off through a white wire leaving it.  I haven't checked yet, but hopefully that wire goes live when everything turns on.  The fan slider clicks in 1 of 5 positions, and there are a bundle of wires going off from it.  To control the fan speed, this switches in a series resistor to reduce the amount of current going through the fan motor.  The resistors are fairly low values and must run hot because the resistors are actually inside the AC ducts to keep them cool!  

So that's where my problems show up.  I get 12v going to the switch, but the fan doesn't spin.  I figure out that one of the wires has a bad connection somewhere, so I cut the connectors off and put new ones on.  That's fixed, but the fan still doesn't spin.  I trace all the wires and can verify that I get 12v all the way to the last wire that I assume goes into the fan itself.  12v going in and no spinning must mean there's either another connection that's bad or the ground isn't hooked up.  Like I said before, it's really hard to see anything in there so I scour the forums for over an hour to try and find a clue on what's going on.  I never found the answer there, but I did eventually get on my back with my face pressed into the firewall and found what I was looking for!  There's a grommet on the left (driver's) side of the box that houses the fan and evaporator, and there's a red and black wire coming out of it.  And guess what???  The black wire isn't connected to anything.  I attach that to a convenient screw and the fan is roaring back to life!  

Now that I know that's working, I only have one more major hurdle, and that's putting some electric fans on the condenser and shrouding it.  I've already built a mount for the compressor and have the belt and pulleys.  So a little wrench turning with a new compressor, a few wires, and I'll be ready to evacuate and charge the system and cross my fingers.  With a little luck I'll be cruising home with the windows UP before I know it!

Sunday, February 23, 2014

Vacuum Brake Improvement

A while back I posted about the inexpensive vacuum pump I cobbled together from a rotary vane pump and a 12v motor.  Today I'm gonna show you how I improved my efficiency by hooking it up to a vacuum switch.  One of my biggest issues with these commercial vacuum pumps made for EV cars is they're super expensive.  I finally found this reasonably priced ($25) vacuum switch at a website set up for applying veneers.

http://www.veneersupplies.com/products/Vacuum-Pressure-Controller-18-NPT.html

What's really nice is it's got built in hysteresis and an adjustable setpoint.  I teamed it up with a vacuum accumulator.  If you don't know what that is, it's basically a tank that stores vacuum.  Why do we want to do that?  The tank allows you to push the brake pedal  4-5 times before the pump has to run again for a few seconds.  The whole goal here is to have the vacuum pump run as little as possible because it draws up to 15 amps, and the less it runs, the more efficient my car will be.

So here's my accumulator.  It's got two 1/2 inch barbed fittings that I threaded in with a pipe thread that I tapped into the end cap.  Then that black thing in the middle is the vacuum switch.  To seal it all up I used some silicone.  At one time I figured out how much of a tank I needed, but that calculation is long gone.  So I'll just have to let you know how it works!


Here I've got it installed in the car.  Attached to that lower tube I've got a relay zip tied on.  The reason for that is the switch is only rated for a mere 10 amps, and I don't want it to burn up when the pump kicks on.  So the vacuum pump drives the relay, which in turn drives the pump.


Here's another view.  That bulge in the other tube is a check valve.  The reason for this is when the pump turns off, you don't want air to rush back into the tank.  The valve only allows flow from right to left, out towards the pump.


After I got it all hooked up, I tested it out, and naturally it wasn't wired right.  The switch has a great feature that allows you to hooks something up normally on or normally off to suit your needs.  Of course, this just means I get to swap wires and try again.

There's a knob on the top that changes the load on a spring inside the switch.  This causes the switch to flip at a different pressure.  The pump is supposed to be adjustable between 1.5 and 28 inches of mercury (FYI, a perfect vacuum is around 30 inches of mercury).  I set it in the middle and tested the brakes.  Still a little soft, so I crank it out a bit.  Ah, much better!  The pump goes back on after 4 pushes on the pedal, so that seems good, and it only runs about 3 seconds before it turns back off again.  That's really good!  It should reduce my overall power consumption by 3-4%, and give me a corresponding improvement in range.  Maybe someday I'll actually test that...

Tuesday, February 18, 2014

Danger: High Voltage

Lots of stuff going on for me lately.  I'm rebalancing the battery pack after 9 months or so, making some modifications to the controller parameters, and there's another car show at work to prepare for.

It wasn't long after getting the BMS up and running that the BMS started shutting down the charger because of cells reaching max voltage (rather than the whole pack reaching its desired voltage).  It's been doing that for two weeks now, so it's time to rebalance the batteries.  Any BMS typically draws some amount of current from the batteries, even when it's not turned on, so it's known that they can misbalance your cells.  I attempted to put some balance resistors on mine to minimize this, but I forgot to consider there are two banks of just 7 rather than 8 cells.  As a result, two cells are discharged slower than others.  Two times I've plugged in a 2 ohm resistor to bleed off some power to keep those cells from going overvoltage.  I think soon I'll try hooking in an extra resistor to each and see if that keeps them in balance longer.

I finally got around to balancing this weekend.  That involves doing a full charge like normal, then using a balance charger to top off every cell to 3.65 volts.  I was surprised at how long it took, but after 3 days I'm finally done!  The worst offender was nearly 4 amp-hours out of sync with the others, which was more than I thought there would be.  That effectively reduces my range by 7.5%!  Obviously my balancing resistors aren't quite configured right.  I'll either have to fix that or rebalance more often.  Here's a picture of the balancing process.


Last weekend, Joe came over and we had a chance to tinker with some of the controller parameters.  As I've boosted the max motor current, the pedal become more and more sensitive.  I'm to the point now that when I start up, it's very difficult to keep the car bouncing forward and backwards, making it look like I'm a kid learning to drive a stick shift.  We were able to back off on the current rise rate, and seems to have mostly corrected the problem.  If it gets worse with higher currents, we may have to make a more serious modification to the controller code to make it less sensitive at short pedal throw.

We also put a little circuit in the controller to measure the voltage spikes coming off it.  If you remember back to some of the posts on the controller design, you want to arrange some of the parts so that you minimize any stray inductance.  Stray inductance causes voltage spikes, and high voltage kills the electronic parts in your controller.  It's time to see how good my design really is.  At 600 amps, the voltage peaks out at a mere 265 volts.  At 600 amps, the pack voltage could be anywhere from 210-230 volts, so about 50v volt peaks.  This is good!  The components are rated to 600v, so I shouldn't have to worry about them wearing out early for voltage related issues.  For reference, a low voltage controller usually has 200v components and tells you to limit the pack to 150v, and these are typically 500 amp controllers.  So I feel I'm right on par, but with room to grow.

A couple weeks ago, we got an email at work saying we'd be having another employee car show.  This will be fun since last time I had to tow the car into the parking lot, and drive it the last half mile in with just 1 Hp.  I'll have lots of progress to show people.  The other good thing about a show is you want your car to look good, giving you some motivation to finish up a few of those cosmetic things you've been ignoring for a while.  I made a list of things to do.  So far I've put a new crank arm for the driver's side window, added rear seatbelts, and a few other things.  However, the thing I'm most proud of is the addition of a couple of stickers!  I wanted a couple "danger high voltage" stickers to put on anything where you could access the batteries and came across this beauty that's now displayed in my rear window.


I also got someone on Etsy to print off a couple custom stickers since nobody had the right voltages in DC.



Sunday, January 19, 2014

Battery Monitoring System Fully Functional

It's been a long road, but I'm very happy to say that my battery monitoring system is finally fully functional!  There have been quite a few things that happened recently, so here's the whole timeline.

April-May 2013 - Installed all the wires that lead to the + and - terminals of all the batteries.

June-July - I completed the schematics for the BMS, ordered all the parts, and built the circuit boards.

August - Completed the preliminary code for the system and tested it on the bench.

Sept-Oct - Installed the BMS in the car and routed all the necessary wires.  Continued to work on the code.  Got the code working, but there's too much noise for it to work.

November - I was able to filter enough of the noise to get the BMS to work while the charger is running, but still won't work if the throttle is pressed at all.  I decide to reroute the signal wire (master bus) inside the car to shield it from the high voltage cables.  Just before I reroute the wires, something happens and I have some fried parts on the circuit boards.

Thanksgiving break - I figure out that two of the microcontrollers on the slave boards have shorted out.  In the process of testing I burn up a couple other parts on the boards.  I order new parts.

Early December - I get the new parts and install them.  The BMS is working again, but rerouting the wires did nothing to improve the EMI (noise) problem.

Christmas Break - I relocate the master board to see if that helps with the noise and it doesn't.  Then I try hooking up just the slave board at the front of the car (so it doesn't need the long wires going to the back of the car).  In the process I accidentally hook up two wires that shouldn't have touched and poof! two microcontrollers blow up!  Interesting enough, it's the same two micros that stopped working just before Thanksgiving...hmm, I might be onto something.  I think about it a bit and realize that I have two wires in a bundle that have around 150v differential between them.  It's conceivable that the insulation on the wires is not rated for that kind of voltage, so that could have been the reason why those two micros fried.  Luckily I had the foresight to order a few spares so I have just enough parts to rebuild the boards.  I retest with just the front slave board (careful not to cross the wires I did last time) and it works!  I'm successfully able to drive the car around and monitor 24 batteries!  Joe has an idea for me on how to fix the problem for the other 46 batteries, so it's time for another parts order.

Early January - And here we are now (so you know that it must have worked).  Joe's brilliant idea is that the noise is coming from the wires that go the length of the car (acting like antennae) that are directly connected to the master board.  There is a +12v, ground, and signal wire that do this.  His theory is that if these are completely isolated from the master board that the noise won't propagate in.  So I build a simple circuit that provides isolated power for the slave boards (that's what the 12v and ground was for), and has an optocoupler that isolates the signal.  I hook that up and bam, works like a charm!  Here's a couple pictures of the prototype board:



After I could tell it was working I soldered all the parts to an actual PCB and mounted it in the car.


Here's that baby in action!  It shows the min, average, and max cell voltages, the temperature for the three battery clusters, the BMS allowable charge current (x10), and the cell number that it's currently checking.  I changed something on each of the circuit boards and the calibration is off a bit.  That's why it's reading a range of 0.15 volts.  They're actually quite a bit closer than that.


I also wired up the necessary wires to get the BMS to talk to the charger.  The purpose is that the BMS makes sure no individual battery goes over 3.65v.  If one gets up there, it tells the charger to slow down and/or shut off to protect it.  At first when I got this hooked up, I had some "nuisance faults."  This is a different data line than the master bus line, so I added a capacitor and that helped a lot, but not enough.  So I changed the code on the charger.  The way it works, the BMS sends a square wave to the charger.  The lower the frequency, the more current it's allowing, and vice versa.  If the charger waits too long, it thinks the BMS came disconnected or stopped working and shuts down.  If there's a really short time period (like from noise on the line) it will think one of the voltages is too high and could shut down.  To fix this, I did two things.  To deal any errant short pulse I make the charger average 8 signals before making a change.  To deal with a long pulse that sneaks in (I'm not sure how this happens), I make it require 4 bad signals like that to happen in a minute for it to shut down.  I don't have to worry about it taking too long because 8 signals is still a small fraction of a second.  I tested it out by disconnecting a wire and it instantly shuts off.

So after having the car out of commission for a while, I'm finally back in business!  I plan to drive in to work again tomorrow morning.  Woo hoo!