Project: Display of Speedometer/Odometer + Battery Info

[cgtech]

Might want to optoisolate (or otherwise isolate) that tach signal. But sounds like I'm probably preaching to the (plenty well informed) choir here..

[Wingsy]

Originally Posted by cgtech

Might want to optoisolate (or otherwise isolate) that tach signal. But sounds like I'm probably preaching to the (plenty well informed) choir here..

A good suggestion and one that I plan to do if the signal can withstand a couple mA extra drain. Certainly will need testing.

One thing I realized while trying to fall asleep a while ago is that my plan to switch to charging mode whenever I detect a charge current will not work. Because REGEN! Going downhill I'll sometimes see a charge current, so I'll have to qualify that charge mode switchover with the condition of the master switch being OFF. And I guess that means I should add a minus sign to the amperage display when in run mode and a charge current is measured, plus I'll need to subtract that from the amp-hour display. And then that brings up the question of how many charging amps would I see when going down a steep hill with a full load on the car. Anyone have any info on this? It matters some, because at present my scheme for the current measurement will be scaled so that it's full span will be 20 amps charge and 250 amps running. (I may just let it clip at 20A and lose a little data by not being able to measure over 20A charge - there shouldn't be much lost.)

Yet to do is to measure battery current while accelerating up my favorite hill with a full load. Think 250 amps will cover it 99% of the time?

[cgtech]

Considered wrapping a small segment of "speaker coil wire" around one charge port wire, and one main battery cable to create inductive ammeter indication. Just have to mathify it enough that your induced voltages appropriately utilize the range of your inputs.

[yawood]

Originally Posted by Wingsy

A good suggestion and one that I plan to do if the signal can withstand a couple mA extra drain. Certainly will need testing.

One thing I realized while trying to fall asleep a while ago is that my plan to switch to charging mode whenever I detect a charge current will not work. Because REGEN! Going downhill I'll sometimes see a charge current, so I'll have to qualify that charge mode switchover with the condition of the master switch being OFF.

You could use a magnet and sensor on a wheel to determine if the charge is occurring while the wheels are turning. If they are not (turning) your original detection of a charging current can be used.

[Sergio]

I think you have the right idea Yawood, but a separate sensor is not needed.

For code efficiency you should not even poll for the "charging started status" unless no pulses have come out of the sensor for a certain amount of time.

Once the program determines the cart has stopped (no pulses), it can then start polling for either movement or charging current to determine which routine is appropriate.

I would also recommend only resetting the AH counter if you detect charging current for over a certain amount of time.

I did not use an optocoupler when I did my interface with the speed sensor a few years back, did not see the need for galvanic isolation since I am using the same negative reference as the controller and the signal comes from a known digital source (hall-effect sensor).

I would suggest that on a Club Car attaching an optocoupler to an existing "input" as it is the case of the Green wire is not so easy, You have no access to the 5v supply where the pull-up resistor that keeps the Green wire at 5v when the hall-effect is OFF is connected.

Aside from the additional load on the hall-effect switch as mentioned, You still need a place to power the optcoupler that would not fry the 5v input when the hall-effect is OFF.

Edit:

Based on your numbers of 274 pulses/sec @ 20mph, it appears you have a 20" diameter tire, is this correct?

[Wingsy]

Battery Current Measurements
—————————————-

7.6mV across cable during charge at 15A; 0.5mV per amp

Voltage/current through shunt (cable) under different conditions:

Semi-Level gravel, low speed: 7-15mV, 14-30A
Sudden accel at bottom of 15° slope: 120-150mV, 240-300A
Climbing 15° slope, full throttle: 100-120mV, 200-240A
Level asphalt, full throttle: 60mV, 120A
Level asphalt, medium speed: 22mV, 43A

Let me start off by saying I don't completely trust these measurements, but it does give me an idea of what to expect. The reason I don't trust it is because I measured the voltage across the cable between battery 1 & 2 a while back and measured 10mv when charging at about 17A (ammeter on charger barely pegged), and today that measurement is 7.6mV at 15A. It is because of this that I'm going to take PingEye's suggestion and use a precision shunt. I haven't disturbed the cable and I measured at the same points, and still there was a noticeable change.

Sergio, yessir those are 20" tires. I measured the distance (circumference) it rolled in 1 rev to be 5' 3-3/8".

And this is (at present) the logic I'll use for switching between run mode and charge mode:
* Switch in OFF position and charge current detected (something like > 1A) to enter Charge Mode. Stays in this mode even if current drops to zero.
* Switch in RUN position to exit Charge Mode.
Entering charge mode also clears the odometer, timer and a-hr fields that will be displayed the next time it goes into RUN mode
It stays in charge mode until you turn on the switch so you can see the charging data before you begin your ride. Also, the charger can pause for a significant time near the end of a charge cycle and then resume charging (mine does anyway) so I need to stay in charge mode to be sure to capture this. Turning the switch on wipes the charging data just collected.

[garysmith39555]

Good luck....You lost me at "I'm!"

[yawood]

Originally Posted by Wingsy

And this is (at present) the logic I'll use for switching between run mode and charge mode:
* Switch in OFF position and charge current detected (something like > 1A) to enter Charge Mode. Stays in this mode even if current drops to zero.
* Switch in RUN position to exit Charge Mode.
Entering charge mode also clears the odometer, timer and a-hr fields that will be displayed the next time it goes into RUN mode
It stays in charge mode until you turn on the switch so you can see the charging data before you begin your ride. Also, the charger can pause for a significant time near the end of a charge cycle and then resume charging (mine does anyway) so I need to stay in charge mode to be sure to capture this. Turning the switch on wipes the charging data just collected.

The problem with that is that something manual has to happen (the key switched to OFF). It would be better to not rely on something like that because it might not always occur but it could still be in charge mode without regen.

[Wingsy]

Originally Posted by yawood

The problem with that is that something manual has to happen (the key switched to OFF). It would be better to not rely on something like that because it might not always occur but it could still be in charge mode without regen.

I was of the understanding that the key switch should be OFF when connected to the charger (although reading up on it a little more I find no clear cut instructions to do so). I turn mine off anyway whenever I get out of the car unless it's just for a minute or so. So the only down side I see by requiring the switch to be off is that if it's charged while the TOW/RUN switch is in the TOW position, the key switch would never be detected if on and the thing will think that's it's a regen charge. That will seldom if ever happen so no big deal. And I like the manual control especially since it's something I always do anyway.

[Wingsy]

I connected an oscilloscope across the battery cable to get a look at the charge & discharge currents over time. I knew it wasn't smooth DC but I was a bit surprised at what I did find with regard to discharge current. Here is what the charging current looks like:
BatCur2 Charge15A.png
0 amps is center horizontal grid line, 1st division below that represents around 30A, averaging around 15A. About what you would expect.

The surprise came when observing discharge current. I think it looks funky because of the way I had to measure it. My scope has to be plugged into the wall for power, so I obviously couldn't measure running current while actually moving. So I parked it with it's front end against a rock wall and gave it some throttle. This is what I got:
BatCur1 StalledLowThrottle.png

And expanding that:
BatCur2 StalledLowThrottle.png

I can't even begin to tell what is power switched from the controller and what is inductive ringing from the motor. I'm going to assume that the controller sees that there is no movement and is doing something to protect itself from overload. Anyway, I need a better test. I'll jack up the rear wheels and try again when it warms up outside. But without a load on the wheels I don't think I'm going to see what I'm looking for.