Showing posts with label austin. Show all posts
Showing posts with label austin. Show all posts

Tuesday, June 30, 2015

It's been a while... Small Improvements

Bad Luck

The Motenergy ME0913 motor started making a strange sound one day.  Kind of a brushing sound at low speed.  Careful listening indicated it was coming from the center of the motor.  Upon removing the fan cowl, a small bolt fell out.  It appeared to come from inside the motor.  Then with a flashlight, I saw a second bolt sitting on the interior coils.  This raised the mystery factor even further.  Then looking into the motor, I could see two empty holes in the center rotor where it attached to the hub that secured it to the main shaft.  There were still two bolts in place.  These are M6x12 flat head Philips bolts.  Well with a tweezers and a screw driver and some thread lock, two new screws went back in, the other two were tightened and the motor re-assembled.  Good advise from John at Motentery to not pull apart the motor housing/casing without first marking the alignment between the two halves since this assures the coils are aligned.  Luckily, I didn't have to separate the casing.

This reduced the brushing sound but there is still something not right.   One of the bolts that was fairly bent up and I suspect it dislodged one of the magnets in the rotor which is now rubbing as it spins.  Not too terrible sounding but not quite right.

Good news

Well with the motor back together, the timing advance circuit is working very well.  Currently, the circuit is advancing the timing from the motor as follows

<300 rpm - 0 degrees
300-600 rpm - 10 degrees
600-1200 rpm - 20 degrees
>1200 rpm - 40 degrees

This is in addition to some mechanical advance achieved by rotating the fan cowl.  The mechanical advance is limited to keep reverse functional since electric reverse is a nice feature for EVs.

The motor is running smooth but the 20-40 degree shift is barely detectable as a slight motor miss.  The next goal is to add a few more steps to lessen the amount of incremental shift per rpm band.

Regen

After some additional testing with regen, it has become evident that variable regen is a must and the best way to control this is tying into the braking system.  Found this stainless steel 500 psi pressure switch:
 stainless steel 500 psi Pressure transducer or sender for oil/fuel/air/water

The goal will be to tee this into a brake line and use it to signal the magnitude of regen for the Kelly KHB 72701 controller.  Using a switch on the brake pedal was too jarring since the controller was all or nothing with its regen efforts.  With this transducer, a 0.5V - 4.5V signal will be provided to the controller indicating the level of break pressure being applied.  By tweaking the controller, it should be able to supplement the mechanical brakes quite well and transparently.

Saturday, February 7, 2015

12V System Enhancements

12 Volts

After the challenges with the traction battery pack, I also noticed that the 12V-120V inverter that powers the cooling fan for the main motor was shutting off when the 12 Battery would start to sag below 12V.  Jane had a larger starter battery (the original from the the gas days) powering the 12V system along with 220V-12V charger to top it off at every charge.

However, under full loads, the 12V battery would run down and thus cascade into other problems.

Upgrades

First was to replace the 12V Starter battery with a DC-DC Converter that would step down the 84V traction battery to around 13V.  After checking the current draw of all of the 12V subsystems (headlights, blowers, radio, wipers, horn, driving lights, BMS) the maximum draw if everything was on at once would be approximately 30A or about 360W.  To provide plenty of headroom, I selected the Meanwell SD-500H-12


This is a 500W 40A DC DC converter that costs around $130.  It is not sealed, but since it will be in the Jane's truck, there was no concerns about dirt or moisture.  

To provide power for the radio when the ignition is off, a small 12V AGM 4.5AH battery is used.  The radio pulls about 3mA when off to maintain the settings, so this will run for months with issue.  Also, it is important that the hazard lights be operational when the key is off along with the horn.
After connecting the system, there was about a 50mA draw on the battery when everything was off which was traced to the DC-DC converter.  Apparently, it has a small load on the 12V battery side even when off.  This necessitated adding a 40A 12V relay between the DC-DC Converter and the rest of the 12V system.   This relay engages when ever the DC DC converter is active.  50mA equates to about 1.2AH per day so the 12V battery would have been drained in 3-4 days.  My goal is to be able to leave the car idle for at least a month without having to disconnect systems.

Also, to control the DC-DC convert, another contactor was added on the high-voltage side.  This contactor is activated anytime the BMS is on (running, charging or manual switch).


Testing

So once everything was connected back, it was time to test.  First just one system at a time.  The core electronics pull about 10A on the 12V system (Cooling blower, controllers, dashboard).  Then adding one system at a time:  Low beams (5A), wipers (4A), heater blower (4A), horn, hazard flashers, BMS (1A).  The current draw peaked at about 28A, well within the comfort zone of the DC-DC converter.

Benefits of the new setup:
1) Weight reduction - The old starter battery weighed about 30 lbs, which in a 1,000 car is 3%.  The new battery and DC-DC converter weigh less than 5 lbs so this should result in a small improvement to the running of the car.
2) Reliability - Since 12V system is now powered by the traction pack, it's run-time is directly linked to the overall run-time of the car.  
3) Proper Voltage on the 12V system.  A standard lead-acid starter battery has a resting voltage of around 12.6V but under load and given the losses in a car, at the fuse panel, the voltage was never much above 11.5V as shown on the dash volt-meter.  The DC-DC Converter is adjustable, so it is set at about 13.5V which is ideal for charging the AGM battery and yields about 12.9V at the fuse panel. This results in brighter headlights and the Blower Inverter having plenty of power.  

With this and the replaced LiFePO4 cells, I went for a test drive last night and everything looks to be operational.  Next steps - Finish up the timing advance system and rebalance the traction pack since it seems to be a bit off but with the relay boards providing full isolation, at least it should not drift out of balance any further.

Side Note: The DigiKey Scheme-It schematic design tool is really easy to use.  It is intuitive and flexible.  I am in the process of capturing an overall schematic for Jane and will post it here once done.