The GPSy EV Project

I'm beginning a new project that uses micro servos. Here's a little table that I'm using to compare various options.

The conversion for torque from kg/cm to oz/in is to multiply by 14.7. So a servo with a torque of 1.0 kg/cm has an imperial torque of 14.7 oz/in.

I just had a nice conversation with one of the staff at BatterySpec.com, which is where I bought my Tempest TR22-12 sealed-lead acid (SLA) battery. I had inquired about a warranty exchange. They denied it but they gave me a really good reason why -- which is why I'm giving them two thumbs up.

The reason my TR22-12 died on my Piaggio is because I had it in a 2 serial 2 parallel configuration with slightly lower capacity Rhino batteries. I had them configured as a bank of Rhinos and a bank of Tempests. The internal resistance of the Tempests is lower than the Rhinos, so the Tempests were discharging faster and this is why one of them failed.

If anything, I should have done a mixed bank of Rhino-Tempest paired with another Rhino-Tempest. This would have protected the Tempests. But in general, mixing different capacity batteries is a big no-no -- which is why my warranty claim was denied.

In any case, my TR35-12 should be good for its purpose so we'll see where to go from there!

All in all, I'm very happy with BatterySpec and glad that they took the trouble to call me to tell me why my warranty claim was denied and why I shouldn't be doing what I had been doing.

I stopped riding about a month ago because I had severe sciatica. It's getting better now and so I thought I'd get back on the horse. One thing I had noticed last month is that my Piaggio was barely making it to work where previously it had been able to go there with energy to spare. I thought it was simply the cold weather.

For a related project, I bought a relatively sophisticated balancing charger, the Turnigy Accucel-8 150W 7A Balancer/Charger. One of the reasons I got this particular unit was that it could charge up to 36 volts of SLA; 32.4 volts of NiMH; or 8 cells = 29.6 volts of LiPo. Most other intelligent chargers I've seen can't handle this high a voltage. I want to experiment with switching over to NiMH or LiPO but I didn't want to have to take the packs apart to charge them.

One of the interesting aspects of the charger/discharger/balancer is that it has an accurate ampere-hour gauge, so you can tell exactly how many Ah you are putting in -- or taking out of the battery. I used this to test the various batteries I have in my stable and found quite a few bad packs.

What surprised me, though, was that one of the SLAs that I had been using in the Piaggio had gone bad. My original setup was 2S2P or two parallel sets of 2 serial 12 volt batteries:

 Battery 1: Rhino SLA17-12  (rated 18Ah) = 10.508 Ah @ 2A discharge
 Battery 2: Rhino SLA17-12  (rated 18Ah) = 10.485 Ah @ 2A discharge
 Battery 3: Tempest TR22-12 (rated 22Ah) = 12.092 Ah @ 2A discharge
 Battery 4: Tempest TR22-12 (rated 22Ah) = 1.880 Ah @ 2A discharge

Yikes!!! Very bad. And I had used it for less than two months... But no wonder I wasn't getting any mileage. I was going to just replace the battery with another similar sized one when I decided to see how many Ah were in the larger batteries that were in my power scooter.

 BigBatt #1: Tempest TR35-12 (rated 35Ah) = 30.612 Ah @ 2A discharge
 BigBatt #2: Tempest TR35-12 (rated 35Ah) = 30.539 Ah @ 2A discharge

Wowza..... This blew my mind a little.

• Revelation 1: The big batteries were getting significantly closer to their nominal ratings than the little ones.
• Revelation 2: Two big batteries had more amp-hours than four little ones == 30 Ah @ 24 volts vs. 22 Ah @ 24v
• Revelation 3: The big batteries were 11.8 kg each x 2 = 23.6 kg
• Revelation 4: The little batteries were 6.8 kg each x 4 = 27.2 kg

This all added up to .... it was better to use two big SLAs rather than four little ones because: 1) I would gain 50% more amp-hours; 2) it was lighter; 3) it was smaller; 4) and in hindsight, it would have been much cheaper.

That was easy enough to accept but it meant that I would have to redesign the battery carrier system. This was something I had been planning on doing anyay. More posts to follow.

I didn't like the wooden saddlebag style racks on my Piaggio that I had originally made. The problem was that the Piaggio didn't have adequate mounting brackets for them, which meant they were putting strain on the rear fender. Also, they interfered with the pedal action.

I decided to change the mounting system to a rear rack style. I removed the elongated double seat and replaced it with a bicycle seat. I used my MIG welder to bodge up a rear carrier, seen below. I think it looks pretty good.

More postings to follow re: the decision to change batteries and the actual battery carrier itself.

Mike asks: I hope I'm not too much trouble. I was just wondering what diameter of pulley you ended up using with the HXT. What a pain, trying to find a 12mm one! According to my calipers, it's actually like 11.8mm, even better.

I used a 1/2" imperial pulley on my 12mm nominal motor axle.

1/2" = 12.7 mm. So you can use the imperial if you make a small shim out of brass or aluminum stock. I found the sides of a soda can worked great.

I also machined a notch on the axle so that it would grab the little allen bolt better.

Karen

Mike asks: I was looking at your bike and realized: Your CVT might not be working properly. Actually, I don't know if it is or not. Here's what's on my mind:My concern is that the CVT on your bike (which is different from the Motobecane) requires TWO movable pulleys to function. As the engine pulley closes, increasing the effective diameter, the tension on the belt pulls the cheeks of the driven pulley (on the wheel) apart, reducing the effective diameter at that end.With only a single pulley able to change ratio under load, would the electric motor not be subjected to varying belt tension as the bike accelerates and decelerates? I could be totally wrong about this - but as far as I've wrapped my brain around it, I thought I ought to ask.

My response:

Continuously Variable Transmissions (CVTs) on scooters have two movable pulleys -- the rear one is a simple spring and the front one is connected to a centrifugal arm. The faster the front pulley goes, the centrifugal arm forces it to become smaller. This forces the rear pulley to expand.

When the bike is at standstill, the front pulley to rear pulley ratio might be 2:1 but when it is at full speed it might be 1:2. This allows for low-speed torque and high-speed acceleration.

What I did with my EV conversion was to replace the front centrifugal pulley with a fixed pulley. This is because the old centrifugal pulley wouldn't fit on my electric motor. The end result is that I have a fixed gear ratio on my scooter (disabling the CVT) but one side benefit is that the belt is always tension properly since the rear pulley is always trying to become as large as possible.

A fixed gear on a little scooter isn't a problem as my ideal max speed is 30mph and electric motors have a lot of low-end torque. If I wanted the scooter to go 60 mph, I might try to machine the original back on, but at this point it's not worth it.

Mike A. also asked where I got the 12mm flanged bearing that supports the right side of my outrunner motor. I got it from Carr-McMaster, part #1434K15:

LineMcMasterPart NumberDescription/Customer ID#YouOrderedWeShippedBalanceDueUnitPriceExtendedAmount
1 1434K15 STEEL FLANGE-MOUNTED NEEDLE-ROLLER BEARING, FOR 12MM SHAFT DIAMETER, 60MM OVERALL LENGTH 2EA 2 0 10.26EA 20.52 

Pics and specs after jump.

Mike A. sent me a link to: http://www.1977mopeds.com/

They sell a limited number of parts for classic mopeds.

Me wanteeeee Brammo electric motorbikey........ me no haveee $15,000..... me very sad.

I'm now using my little moped to commute to work. It's about 5km each way so I have to recharge at work in order to get home. Luckily, there's a power outlet near the bike rack. Unfortunately with the little 1.5 amp charger I have, it takes about 6 hours to fully recharge so I'm stuck if I need to go back at lunch time.

I was curious how much juice it took to recharge my bike so I put my Kill-a-Watt on it while I charged it up overnight. The 36 volt SLA battery charger consumed 300 watt-hours (0.3 kWh) to bring me from pretty much zero to full charge.*

* The engine consumes 150 watt-hours for a full run but the batteries need 300 watt-hours to top up because of inefficiencies in charging / depleting.

At 12 cents a kilo-watt-hour in Connecticut, that's about 4 cents for a full charge, each way. Or about 8 cents a day.

Or $1.60 a month. Of which, I'm only paying 80 cents and mooching 80 cents off my workplace. Shhh.... don't tell anyone or they'll dock my salary!