Regen braking with a "backup" snubber braking

The Great Electric Vehicle Project has started.

To review
Vehicle is a 56 Chev 1/2 ton pickup. I intend to as much as possible ensure truck is easily returned to stock condition ie put an engine back in etc etc.

Drive motor is a 3600 RPM 50 HP Leeson 208 / 230 / 460 motor.

Motor is in a prototype cradle to enable final measurements prior to making foundation bracket.

VFD is not selected yet. Thinking Hitachi SJ 300 or Yaskowa or ABB. Open to suggestions of course. Reliability and safety are top considerations, cost is further down the list.

Since we have a few hills here in Seattle downhill speed control is to say least right handy.

I know I can feed DC into the "normal" AC line connections but I lose braking and I gotta ovesize the VFD.
SO
Battery will be tied direct to DC bus. DC bus voltage is at this time nominal 325 VDC (1.4 x 208 VAC). Battery fused of course and disconnect means of course.

I am told by several that this will work fine as a load bank (snubber) when motor is in generating "mode" ie overhauled on downhill.
HOWEVER
they are unsure about battery voltage getting too high and feel that I cannot use added braking resistors controlled by the VFD. They think an additional control must be emplaced to sense battery voltage and parallel resistance as needed to keep DC bus voltage (and the battery) below max set by VFD mfr to protect DC bus.

Worst case scenario for this problem - fully charged battery and braking is needed to go downhill. Battery is "full" and voltage will rise of course.

I have not selected battery nor battery chemistry yet. Am thinking lead acid since I know and understand that best (former submarine electrician) and am willing to pay the weight penalty. It is also more forgiving of my future mistakes which I am sure I will make. After I get bugs and oversights dealt with I may go with a lithium battery.

SECOND QUESTION
I will definitely use a vector VFD. I believe V/Hz will have too hard a time starting on hills. What I am uncertain on is whether I need to sense both motor shaft RPM and shaft position for this application. I do not think I need to sense shaft position since I do not care about accurate positioning of vehicle.

Thoughts and opinions on this please. Dick DV you can surely chime in here.

Dan Bentler

You are on the right track. Protecting the battery with a DB resistor is valid concern. Pay attention to the turn-on point for the DB transistor, and ensure it is tollerable for the battery bank.

Most high-end drives have DC bus terminals brought out, so no derate should be required.

As for the motor control method, closed loop vector would be a waste of hardware and complexity. Trust the open loop vector and auto-tune of any good drive to give you more than 100% starting torque. The only real issue is motor temperature affecting the winding resistance, so test the auto-tune in cold and hot motor conditions.

Just to expand on what Gene has said: Most drives will have a built-in brake chopper that you connect a resistor to. The chopper will dump power to the resistor when the DC bus reaches a certain(usually fixed) voltage. Closed will only help you at zero or very slow motor speed, so may not be worth it.

My own preferences would be an ABB ACS800 or 850 with the optional snubber and a full capacity resistor bank. Remember that the braking hp might be more than the motoring hp so size the drive and motor accordingly. The motor will give you 220% shortterm overload torque so I would size the drive one or two sizes bigger than the motor.

I would be comfortable with simple sensorless vector like ABB's DTC. Also, it may well be better to operate the drive in torque control mode rather than speed control. I think the accelerator would be super sensitive using speed control and the result would be herky-jerky motion.

You will need some form of bipolar torque or speed input so the braking/recharging begins to occur by simply lifting your foot from the accelerator. The mechanical brake system will need to be retained with the drive interlocked with it so they don't fight each other. Here too I think torque control would be preferred to speed control.

I'm a bit surprised at the motor selection, Dan. I would plan to run the motor to 120hz at full vehicle speed. I don't know what the power train ratio is but it seems like a 4 pole or even 6 pole motor would be the better choice. Just use minimum gearing for efficiency and figure the motor at 120hz at top speed. Whatever power train ratio that is would be preferred, in my opinion.

DickDV said:

My own preferences would be an ABB ACS800 or 850 with the optional snubber and a full capacity resistor bank. Remember that the braking hp might be more than the motoring hp so size the drive and motor accordingly. The motor will give you 220% shortterm overload torque so I would size the drive one or two sizes bigger than the motor.

I would be comfortable with simple sensorless vector like ABB's DTC. Also, it may well be better to operate the drive in torque control mode rather than speed control. I think the accelerator would be super sensitive using speed control and the result would be herky-jerky motion.

You will need some form of bipolar torque or speed input so the braking/recharging begins to occur by simply lifting your foot from the accelerator. The mechanical brake system will need to be retained with the drive interlocked with it so they don't fight each other. Here too I think torque control would be preferred to speed control.

I'm a bit surprised at the motor selection, Dan. I would plan to run the motor to 120hz at full vehicle speed. I don't know what the power train ratio is but it seems like a 4 pole or even 6 pole motor would be the better choice. Just use minimum gearing for efficiency and figure the motor at 120hz at top speed. Whatever power train ratio that is would be preferred, in my opinion.

Click to expand...

Dick another person feels the same way re motor selection and recomends 50 HP either 1800 or 1200 RPM. My main concern is wind resistance at higher speeds when motor is over baseline which puts me in the position of chasing increasing torque demand with decreasing "supply". Rear axle ratio is 3.90:1 rear axle shaft torque is estimated at 65 so driveline torque shouild be 17 or so not counting wind resistance.

I think 3600 RPM motor may do it OK with 75 ft lb torque. If I have the numbers ran correct a 1800 RPM motor at 3600 will be at 25% torque so its 150 ft lb at baseline or less is now down to 37 or so.

We do have several 20% grades here in Seattle but they are short and can be gotten around fairly easily. Motor will be tested on a mile long 10% grade.

I am taking his and your feedback very seriously. I am thinking if I test with wrong motor I have data to select the right motor. I would also feel lots better having a $229 motor fail or not succeed than a 2,000 motor. I bought the motor from Boeing surplus for $229 - appears brand new. Motor mount etc will be designed for quick (8hr ??) change out.

I fully agree with your thoughts re braking with motor and with hydraulic brakes. While I dream of an independent control for regen braking I do not see how it can be done and will rely on trained foot on accelerator to set the speed control signal for snubber braking. Right now and maybe for preliminary testing I think I will not interlock motor with hydraulic brake system. Later I may put in a "high pressure" switch to shut VFD and motor down in a crash stop situation.

Dan Bentler

Dick DV

You used term " bipolar torque or speed input" I am not sure what you are referring to.
My guess is a sensor that not only gives RPM signal but also foward or reverse rotation.

I have a clear 2" long space on 1 7/8 dia motor shaft. If you can recommend a sensor (make model etc) that from your field experience can survive wet conditions that would be highly appreciated.

Dan Bentler

You won't need any kind of speed feedback on the motor shaft. I was referring to the speed or torque INPUT. The accelerator if you will. It needs to be arranged so there is a point in the middle which represents 0 torque with further depression being positive torque (acceleration) and less depression being negative torque (braking).

Further, even that won't be ideal because the zero torque point should move up and down with speed. Hmmm! That could be difficult to accomplish.

Dick DV
I have never gotten a good understanding of torque control. I clearly need to learn more. If you can send me links to good study material that would be great.

Back to original question where I was told I can have either brake resistor or a battery across the DC bus but not both. After some thinking a few scribbled drawings I do not think that will be a problem. ASSUMING I am not using the same terminals as would be used for braking resistors - which does not make any sense to me. Here is why
1. Battery is connected in parallel across DC bus
2. Braking resistor(s) and the "control switch" are paralleled to DC bus
3. The brake resistor "control switch" senses DC bus voltage
4. The "control switch" shuts to parallel the braking resistors across the DC bus to keep DC bus voltage going over desired setpoint. It reopens when DC bus volt are less than setpoint.

If I am incorrect I will have a voltage sensor and alarm to manually activate braking resistors.

Dan Bentler

I was wondering for that fact.
You are building an electrical car. I suppose you run it on batteries, but not on cable. So, why do you want to heat the world with the braking resistors, as you have mechanical brake?
Think you can adjust the stop mode not to "ramp" but to coast, so yo will save much battery capacity

The shunt resistor *never* goes directly across the DC Bus. It is switched across the bus, as you said, by a switch (transistor) which is controlled by a voltage sensing circuit for self-protection (usually analog hardware).

In this case, as suggested, it would be used only as an over-charge protection device, wasting energy, along with mechanical brakes when the bus voltage reaches too high of a level (after fully charging the batteries).

There are some manually adjustable, external shunt controllers (such as Bonitron) which could be used, but typically, there should be a combination of batteries which will line up well with the drive's self-protection, thus saving considerable money and complexity.

I wouldn't get too hung up on speed regulation vs. torque regulation too much, until the rest of the control is proven. There are plenty of ways to 'soften-up' the speed regulation loop in the drive to make the throttle feel more natural than a cruise control. Most high performance drives today offer lots of extra functions to customize the speed/torque control loops (ABB ACS800 series, Emerson Control Techniques Unidrive SP, etc.)

Speed control regulates motor speed regardless of torque, within the set torque limits. Torque control uses the same analog input to regulate torque regardless of speed, again within set limits. It's not complicated but it can be hard to visualize.

As Gene Bond says, modern high performance drives can offer other combinations of the two regulation modes, or varying degrees of "stiffness" in speed regulation. I'm just saying that making the accelerator work like we are use to is going to be a challenge.

mdim said:

I was wondering for that fact.
You are building an electrical car. I suppose you run it on batteries, but not on cable. So, why do you want to heat the world with the braking resistors, as you have mechanical brake?
Think you can adjust the stop mode not to "ramp" but to coast, so yo will save much battery capacity

Click to expand...

Yes I could turn off VFD and either coast or use hydraulic brake to stop. Hydraulic brakes emit heat also - lots of it.

I agree using motor to brake will consume battery energy to keep motor energized. However the energy derived from braking action (motor acting as generator) and put back in battery will be much greater.

We have some fairly steep hills in Seattle and some are long. The "test hill" is 9% and approx one mile. Downhill speed control is important to me - got a $195.00 speeding ticket on that hill.

While brake fade may be a thing of the past I notice heavy trucks still downshifting for down hill speed control and braking on the engine compresson - another form of dynamic braking.

U of Washington fleet mechanics told me they could quantify the dollar savings in avoided brake jobs from dynamic braking in the Prius.

Dan Bentler

I did some work on an S10 truck that was converted to a forklift drive. It uses a 48 volt drive about 10 years old. We took the motor drive battery etc and transplanted it to the truck. With some gear ratio adjustments it works well and he drives it to work every day.
The original settings for dynamic braking were set up for a loaded fork truck. The first time out when he lifted his foot off the "gas" dynamic braking almost sent him thru the windshield. A friendly forklift mechanic tweaked the program but you still have to remember that lifting causes braking, not coasting as in a car.
It is a different way to drive!

Gas

I do not see much differance -
let off gas pedal and engine slows thus vehicle slows.
let off gas pedal and motor slows thus vehicle slows

I am still unsure what I need for speed sensing ie just a RPM sensor
OR
RPM and direction sensor
OR RPM direction and motor shaft position sensor.

Dan Bentler