Formula Hybrid forums

Hello,

A race team member contacted me directly with a technical question. Out of fairness to the other teams, I will answer it in this forum instead.

> I would like to run [xxx] cells to run each motor separately

Why? Reliability?
Yea, like you're going to be able to still finish the race if one of the two system fails; you'll be back in the pit regardless. Simplicity and good engineering practices buy you more reliability than running separate systems. Arguably, two separate system cost you more in reliability because each set of components is not as high duty as when you have a single, more powerful system.

Why? Elegance?
Forget elegance: you want performance.

Your ticket to performance is:

A battery that uses a single series string (for maximum performance in the presence of a weaker cell, and to require only one fuse)
A single motor controller
A single electric motor

Now, if you must have 2 motors, one for each wheel, and they are DC shunt motors [edit], you may want to wire them in series, as it helps on curves (it acts as a mechanical differential). Still, even if wired in parallel, because of their internal resistance, they can handle curves just fine. (I'm not 100 % sure about this one; someone may want to chime in.)

Davide

_________________
Davide Andrea
Elithion.com

I think this is only true for DC motors (the majority of motors used at the competition). From what I've read, Voltage = RPM's, Current = Torque. So being in parallel will make them try to turn the same speed and may lead to skipping wheels, etc. In Series the motors will each have a different apparent resistance while experienceing the same current. This different resistance will mean a different voltage drop, and a different speed (allowing the car to turn easily).

I'm glad you are posting all this information for the teams. Sorry to get so annoyed at the hobby pack comment in the other thread.

Yes. Good point. Thanks. I edited my post accordingly.
Davide

_________________
Davide Andrea
Elithion.com

I think that if a team chooses to package their batteries external to the cockpit section of their vehicle, and if said vehicle is powered by two separate electric motors, then they would want separate packs due to the crazy isolation / HV protection rules. That way, they could run smaller wires (lower current) to motors in close proximity, rather than running a huge wire and conduit across the car. Arguably one option is safer than the other two...

As far as the reliability, it reminds me of a funny argument I heard at an FSAE comp about why a team chose to go with a four-cylinder over a single-cylinder engine. But in our case, why would battery death in one cell bank force the car into the pits if they are two laps from finishing endurance?

Overall interesting insight--thank you for posting it.

Since you cannot control the direction of the voltage drop, wouldn't this help in one turning direction and hurt you in another?

How exactly two motors would work--allowing slip like a differential or not--would depend on whether they are controlled by the same controller. On the same battery, but with separate controllers, is pretty much the same as having separate battery banks and separate controllers.

Youre' right, it won't, because you still have an engine. I stand corrected.

Let me add that, in my experience, cells do not suddenly go bad (the BMS makes sure that they are treated well). What happens, instead, is that a connection in the pack loosens or breaks. So, design your pack well, and build it well, using a torque wrench.

First, let me define "battery death in one cell bank" as: "one of the cells develops a failure that results in its resistance becoming very high, or in its capacity becoming noticeably lower". (The latter issue becomes the former issue once that cell is empty.)

Losing a connection in the pack has the same effect, but at least a reverse voltage on that connection is not a safety concern the way a reversed cell is.

Then, look at 4 topologies:
- two completely separate packs,
- two separate batteries that are then connected in parallel,
- a single pack with cells directly in parallel,
- a pack made up of half as many cells that are twice as big.

Under ideal conditions, all topologies are identical: all the packs have exactly the same voltage, total capacity and total current delivery capability.

But, if a single cell goes bad, the effects are different for the various scenarios.

Topology 1: two completely separate packs.
The battery with the bad cell cannot deliver any current (without a BMS, the voltage across the bad cell would be reversed, with worrisome effects; with a BMS, the voltage across the bad cell would drop below the limit the moment you tried to use that pack, and the BMS would shut down the motor controller). Regardless, you'd be running with only one motor (which will make the car pull to one side), and 1/2 the torque.

Topology 2: two separate batteries that are then connected in parallel.
As soon as you draw current, the pack voltage drops (say, 10 V). In the good battery, the drop is shared evenly. The battery with a bad cell cannot deliver current, so its cell voltages remain constant; the result is that the entire voltage drop appears on the bad cell, reversing it (-10 V in the example). The BMS will see that, and shut down the ENTIRE pack. So you're left with no electrical power. (Without a BMS, the effect on the reversed cell will be nasty.)

Topology 3: a single pack with cells directly in parallel.
The bad cell cannot deliver any current, but its buddy can. That pair of cells is the limiting factor to battery capacity and to battery current. Both motors continue to work, though the max torque is now 1/2 as much. But at least the torque will be balanced.

Topology 4: a pack made up of half as many cells that are twice as big.
The bad cell cannot deliver any current, so nor can the entire pack. So you're left with no electrical power.

So, the best topology is #3. Not ideal, but better than the alternatives.

But, Formula Hybrid requires you to use a fuse for every parallel connection, making a topology #3 pack cumbersome and expensive.

So, for you Formula Hybrid teams, that leaves topologies #1 and #4. The question is: "what's better? Losing electrical torque on just one side, or losing the entire electrical torque (evenly on both sides)? If the former, topology #1 is better; if the latter, topology #4 is better.

(Again, this is only considering the bad cell / bad connection issue. When looking at other issues, other factors may favor other topologies.)

Your thoughts?

Davide

_________________
Davide Andrea
Elithion.com

The speed of the motor which is being forced by the wheel gripping the ground is what determines the apparent resistance. Turning left the right wheel moves faster, develops higher apparent resistance, and takes more total power (but identical torque). Turning right, the opposite occurs. This is exactly the same behavior that happens in a differential and is desired to keep both wheels at the edge of their grip.

I'm not sure what you mean by controlling the direction of voltage drop. If you reverse the voltages, you will reverse the current and the car will start going backwards.