Sunday, February 27, 2011

Vacuum System Repair 2 and Tachometer Progress

Although the vacuum pump doesn't run all the time anymore, I've noticed that it does run more often when it's colder outside.  So I installed a gauge and valve near the pump so that I can isolate the pump from the rest of the system.  Like this:

The pump stayed on until the gauge read ~23 in Hg.  As soon as the pump kicked off the gauge steadily dropped to ~20 in Hg (took about 30 seconds), where the pump again kicked on to bring it up to 23.  Once it reached 23 and the pumped kicked off I shut the valve to isolate the pump from the rest of the system.  The gauge stayed rock steady at 23.  Opening the valve made the gauge dropped immediately.  So, there's something going on in the pump causing the vacuum to leak.  I think the answer is to put a check valve at the pump intake.

Also, some progress on getting the tachometer to work with the DMOC "speedometer" signal.  A search of the EVAlbum showed that there was one other person who used my motor/controller in a Saturn.  Oddly enough, it's another Greg, though he uses the extra g at the end of his name (Gregg Witmer).  Although he hasn't gotten his tach to work to his satisfaction, he sent a schematic of a circuit that he thought would.  Using that schematic as a starting point, a nice EE friend of mine and I succeeded in getting the tach needle on my "spare" instrument cluster to move.  Here it is at 1000 rpm equivalent:

The tachometer is designed to accept 6 high-to-low transitions in the square wave from the PCM for every engine revolution, so 1000 rpm on the tach means 6000 transitions per minute from the PCM, or 100 Hz.  Gregg's circuit actually uses a flip-flop chip to create a square wave from the pulse that the DMOC puts out, so the DMOC will need to put out twice as many pulses.  With our function generator at 1.58 kHz the needle moved up to nearly 8000 rpm, as expected.  Yes!

The tachometer draws more power than we expected, however.  The 1/2 watt resistor we used to pull the signal down to ground was getting mighty warm!  A quick calculation showed that we were actually pulling 250mA through it, which translates to 3 watts.  Oops.  A 5 watt MOSFET worked fine, though.  Next step is to try the circuit with the actual DMOC output to see if it'll drive the tach.  Then convert the breadboard circuit to something that'll be able to live in the car.

Monday, February 14, 2011

Bad vibrations

I have been worried for quite some time about vibrations in my drivetrain.  When I first got the car on the road last fall I was worried about the flywheel balance - even with the car standing still and in neutral, revving up the motor made the whole assembly wobble.  After replacing the clutch and having the flywheel balanced, this test produced very little shaking.  However, the car still shook like crazy while driving, especially at certain motor speeds (seemed to be between 3000 and 3500 rpms).  I mean, I thought the car was going to shake completely apart (not to mention reduce my fillings to dust!).

The other problem I've been having is that the clutch stops working from time to time, especially when it is very cold out.  Not good.  So, this weekend I replaced the clutch hydraulic assembly and reinstalled the original (rubber) transmission mounts.  No pictures, but here's how it went:

To access the top of the transmission I had to remove the accessory battery, vacuum pump, and the "high voltage" electrical box (the box that houses all my shunts for measuring current).  Then I removed the right-hand (driver's side) tray from the engine compartment.  Then I disconnected the slave cylinder from the transmission and the clutch pedal from the master cylinder pusher rod.  In order to pop the assembly off the firewall, I had to disconnect the brake master cylinder from the booster and nudge it away from the wheel well so there was room for the clutch master cylinder to be pulled straight back from the firewall.  Installation was opposite of removal, except for the part where I accidentally popped the top of the reservoir and poured brake fluid all over the splashguard.

Replacing the transmission mounts was pretty easy, though I ended up loosening the left side tray in case the motor came in contact with it when I raised the transmission.  Removed the driver's side front wheel, fender liner, and splashguards.  Removed the solid torque rod from the top of the transmission and chassis frame, removed the nut which attaches the lower transmission mount to the frame and the bolts which attach the mount to the side of the transmission.  Using a floor pump jack, I gently raised the transmission high enough to clear the stud which protrudes through the frame and removed the solid mount.  Installation was the reverse of removal.  Lowered the transmission assembly back onto the frame and installed the nut on the stud sticking down into the frame.  Next I installed the torque rod on the top of the transmission and into the chassis frame.  I had to push back slightly on the transmission to get the bolt to go into the frame (which means that the hard mount on the motor side is probably a tad off center).

After reinstalling the equipment trays and all the components I took my first ride with the rubber mounts.  All I can say is the difference couldn't be greater.  It was so quiet I felt like I was in a cushy Cadillac.  And the clutch was working like a dream.  Oh, the EV grin was back with a vengeance!  I couldn't believe I waited so long to put those rubber mounts in.  The car is a joy to drive now!  There are still a few motor speeds where you can feel vibrations, which I think is related to the fact that the motor is still hard-mounted, but I no longer feel as though the car is about to disintegrate.  I'm back in business!

Wednesday, February 9, 2011

PakTrakr issues fixed

I received the newly-repaired PakTrakr remote from Ken Hall late last week.  Thanks, Ken.  I was able to quickly re-install the remotes to monitor all 13 of my batteries.  In order to avoid the max voltage limitation of each remote, I am now monitoring 5 with the first remote, 5 with the second, and 3 with the third.  As you'll recall from my previous post (1/31/11), during the "high voltage" phase of the charge cycle I was bumping up against this max voltage limitation (86 V, or 2.4 V/cell), causing the remote to misinterpret the voltage of the 6th battery and giving me all kinds of erroneous error messages.  Here's an image of a typical charging cycle:
Notice the voltage for Batt6 and Batt12.  Also, the Pack Voltage is only measured as ~195V, whereas my multimeter showed the expected 218V.

With the new monitoring scheme, I am able to correctly monitor all 13 batteries throughout the entire charging cycle.  Yay!

Now I just have to figure out why the voltage sags so much when the motor controller pulls on them hard.  That'll be another story.

Tuesday, February 1, 2011

Commuting again

Yesterday was a good day.  Any day I can commute gas-free is a good day!  In the morning the outside temperature was 18°F but the batteries were a toasty 70.  Even with the cabin heater drawing power to keep the windshield clear the batteries were fine all the way to work.  Ditto on the way home.  I even drove my son to the high school later in the evening before plugging in.

Of course, between today and tomorrow we're slated to get almost 2 feet of snow.  So I wimped out and drove the dino burner this morning.  The CdV would probably be fine even with a 2 hour commute, but why risk it?

Sunday, January 30, 2011

PakTrakr issues

Another thing that's been nagging at me is the fact that my PakTrakr system fritzes out during part of the charging cycle.  The 6th battery's voltage in each remote drops to 4 volts at the peak charging voltage, so it's forever saying that "Battery 6 is failing."  Manually checking the voltage of that 6th battery showed that it was approximately the same as the other batteries, just over 16 volts.
Looking at the PakTrakr serial logger output, each remote maxxed out at around 86.5 volts.  Six batteries at 16.4ish volts each is just over 98 volts.  If the remote can't detect over 86-87 volts, that would cheat the last battery by around 10 volts.  The reduced readings make sense, then.

According to Ken Hall at KJH Motor Company (maker of the PakTrakr system), the remote is designed to be limited to 90 volts (although mine clearly were limited to 86-87 volts).  That means the remote can only measure the 36 cells (6 12V batteries that have 6 cells each) up to 2.5 V/cell.  AGM and Gel lead acid batteries are charged at up to 2.45 and 2.4 volts respectively, so the remote can (barely) monitor 6 of those types of batteries.  Flooded lead acid batteries are charged up to 2.7 V/cell, beyond the capability of the remote.  Fortunately he also told me that there's no problem monitoring 5 batteries per remote (it will not mess up the display or serial logger).  Since I have 13 batteries I already had 3 remotes, so I'll be able to use a 5-5-3 monitoring scheme and eliminate the out-of-range problem.

Unfortunately, I killed my third remote (the one that was monitoring the 13th battery).  Apparently, one of the unused leads contacted another battery terminal while I was reinstalling it after some battery maintenance that burned out something inside it.  I've sent it back - hopefully it can be repaired...  Until then I'll be limited to monitoring 10 of my batteries.

Vacuum system repair

Last week, having gotten the batteries warm and the receptacle fixed, I started working the batteries up to be able to start commuting again.  Driving around I noticed that the vacuum pump was running a bit more than normal.  By the time I returned the pump was running continuously.

I had wanted to change the vacuum reservoir anyway, because I had to cut away some of the splashguard where the tie rods came through to make room for the reservoir.  This unfortunately allows water to splash into the engine compartment, and I have been worried about the DMOC and some of my other electrical connections.

$5 at a local junkyard got me a new splashguard.  The original reservoir is a 10" length of 4" ABS pipe.  In it's place I constructed a "U" shaped construction of 2" PVC pipe.  This construction has a smaller volume than the original, but I think it'll do the job and the smaller pipe fits nicely between the tray holding the DMOC and water heater and the firewall.
With the new system, the pump takes about half the time it did for the initial draw down.  It comes on more frequently than before, but for less time.  Every indication is that the system is now relatively leak-free.  Once again I appear to be ready to commute again.  It starts tomorrow.

Tuesday, January 25, 2011

Improving the plug

I haven't been totally happy with the receptacle in the fuel filler area of the car.  Because the Saturn's body is plastic there wasn't really much to attach the receptacle to.  Just a single mounting hole where the fuel filler pipe was mounted (above and to the right of the receptacle in this picture).  Because it is only held on with a single bolt it is difficult to fully engage the turn-lock connector.  Sometimes I even had to reach behind the fender liner to hold the receptacle to make the connection.  Not good.
From the back you can see the exposed terminals.  I don't really like that, either.  Especially when I have to reach back there to steady it when making the connection.  It might get splashed, too, so corrosion may be a problem down the road.
So, instead of a receptacle I decided to make a pigtail that could be pulled out when I want to plug it in.
Once it's connected I can stuff the connection back inside the car so the fuel filler door can be almost shut.  Hopefully that'll keep the snow out.
I might consider notching the door so it can be closed completely.  I'll have to wait and see how much water and snow gets inside.