DIY Electric Car Forums banner

1 - 20 of 20 Posts

·
Registered
Joined
·
14 Posts
Discussion Starter #1
So I know some years back the guys at Rebirth Auto experimented with regeneration with series motors. I think they were WARP 9s, which are brush advanced. My questions is has anyone ever regenerated with their warp 9 and not have it blow up? How many amps? I built a homebrew controller that regens in full series mode with a Kostov 9, which has interpoles and it works very smoothly. I hate that Kostov is no longer in the game and would be interested in trying regen with other series motors. And yes, I know AC motors are the best yada yada yada....I'm a series motor man so..............yeah. Thanks guys
 

·
Registered
Joined
·
1,092 Posts
The good news is that Regen sucks and there's almost no point, so, ignore it and go build what you want.

A few days ago someone linked 3 different articles on Series DC regen but I couldn't quite grasp the particulars of how it happens.

Also, I imagine advanced timing makes regen even more difficult (not that you do it that often). You'd need the counter-advanced (retarded?) to not arc. But, 99.5% of the time you're driving forward not braking so, meh.
 

·
Registered
Joined
·
14 Posts
Discussion Starter #3
I think the statement regen sucks is a little over the top. I don't really grasp why it would suck when I use absolutely no mechanical brakes at all and can stop as fast as I want without wearing down brake pads, while charging the pack simultaneously. But yes, you are going forward most of the time, so I would understand why some folks wouldn't care.
 

·
Registered
Joined
·
966 Posts
There are 80v Curtis series regen controllers in French bread trucks

It really wasn’t super complex compared to an AC inverter but

1. Wears out the controller quickly (spikes and noise galore)
2. Wears brushes more rapidly
3. Leaf acid doesn’t regen well.

There were some external circuits posted here to do it but I think it’s more of a case that back in the dark ages it was hard
AC and sepex regen came along and is more efficient

So critical mass to make it mainstream and reliable never happened.

I have a strong feeling now that we can easily source 2000 volt 1000 amp rated parts the former arguments are invalid but mainstream suppliers will view it as it
we never did it so it doesn’t work because we never did it
And they had issues with it in the 1980’s when someone tried it in a specific use case on poorly designed hardware.
 

·
Registered
Joined
·
1,092 Posts
I think the statement regen sucks is a little over the top.
Naturally.

I don't really grasp why it would suck when I use absolutely no mechanical brakes at all and can stop as fast as I want without wearing down brake pads, while charging the pack simultaneously.
I mean regen sucks compared to most people's expectations of it. It's not a magical energy device. The energy lost in braking is miniscule and won't impact range.

In people's heads, when they hear and think of Regen they're thinking of like, 40% increase to their range. If you told them it was more like 4% no one would care. If you told someone that this amazing new system improved their gas mileage by 1 mile/gallon no one would care or think anything extra of it.

You can do magnetic braking on DC without doing regen.

On Forklifts this is called "plug" braking, where you use a double-throw-double-pole contactor to flip either the field or the armature (and, for some controllers, it matters which one gets reversed, protection diodes are the issue I think). This is functionally putting the car in reverse, so you're applying energy in order to slow the car down, making the motor burn double the energy it normally would IIRC.

There is also resistive breaking, where the field is toggled out with a massive resistor I think. That doesn't consume extra power, but it still dissipates it as heat.

Either of those options magnetically slows the vehicle without using the brakes, but neither of them actually adds energy back into the pack. Which is probably fine for most people, since it's a mild benefit compared to the ones you listed.

And they had issues with it in the 1980’s when someone tried it in a specific use case on poorly designed hardware.
I don't even really understand, at it's most basic, what is happening during DC regen. How do the wires change, how is the energy recovered, why was it considered impossible before but not now?

I probably couldn't understand the whole circuit, but, just a layman's explanation of the basics would help me.
 

·
Administrator
Joined
·
6,105 Posts
I probably couldn't understand the whole circuit, but, just a layman's explanation of the basics would help me.
Hi matt

The problem is control - it's easy to get a series motor to re-gen - but it goes into positive feendback
You ask for a little bit of torque and it feeds back and gives you too much
So you ask for a bit less and it drops to zero

When you regen you are adding some current to make it re-gen - BUT the re-gen then adds a lot more - positive feedback
 

·
Registered
Joined
·
1,092 Posts
When you regen you are adding some current to make it re-gen - BUT the re-gen then adds a lot more - positive feedback
That's what I first read like, 10 years ago. I still don't get how you hook it up. Like, what the actual wire arrangement change is.

With blind PWM control you'd have control be an issue, but, even my GE 1969 controller has current measuring and ramp up built into it. Surely it would be a trivial issue to measure the current and adjust the PWM on the fly to approximate smooth breaking, no?

Or is there a part of it that I'm missing?
 

·
Administrator
Joined
·
6,105 Posts
Hi Matt
It's the speed - the feedback loop is FAST - so when it builds up it build up FAST and you end up with 1000 amps of re-gen before the controller has a chance to get it's act together

The driver bites the steering wheel
Then it goes to zero - and you crash into the obstacle

It can be handled - but it's completely different to the way the controller works in power mode

The difference between balancing a ruler on the palm of your hand and holding the top and letting it dangle
 

·
Registered
Joined
·
197 Posts
Having a capacitor switched in off-throttle could help protect the battery and controller, but adding regen to a DC system takes it into the realms of complexity where you might as well just switch to ACIM.
 

·
Registered
Joined
·
14 Posts
Discussion Starter #10
That's very interesting about the French bread trucks! As far as wearing out the brushes though, I could only see that if the motors did not have interpoles. I'm doing it with the Kostov 9, and it is very smooth with no arcing. And yes, I doubt it improves gas mileage, but its a nice feature to have which is why I was wondering if anyone has ever done it.
 

·
Registered
Joined
·
1,092 Posts
It's the speed - the feedback loop is FAST - so when it builds up it build up FAST and you end up with 1000 amps of re-gen before the controller has a chance to get it's act together
I guess it matters how fast "fast" is. Computers are fast, faster than me, faster than I can react to pulse off the brakes, sure. Are they not fast enough?

The difference between balancing a ruler on the palm of your hand and holding the top and letting it dangle
For example, that's a solved problem.

 

·
Registered
Joined
·
131 Posts
nobody has really bothered to sort it out w/modern controller design, because AC came of age. And with the additional high current switching and complexity, might as well AC.

Interpoles is "better" than not, but AC is better yet in numerous ways. I don't know if trading motor brushes for brake pads is a win. If anything you want to use the brakes on a brushed motor, esp if it is hard to find a replacement.
 

·
Registered
Joined
·
131 Posts
I don't even really understand, at it's most basic, what is happening during DC regen.
You increase the field strength so that the back emf is above the rail for a given rpm. Series makes that tricky since the current wants to keep the same amplitude in the field and armature.

edit, worth looking into sepex motors for understanding the relationships/quadrants. It was sort-of the original "field oriented control", but like putting power recovery turbines on large piston planes, it didn't take them long to realize they didn't need the piston part.

If you had fine control of the gear ratio you could also regen that way. Just downshifting would be rather uncontrolled though.
 

·
Registered
Joined
·
966 Posts
I don't even really understand, at it's most basic, what is happening during DC regen. How do the wires change, how is the energy recovered, why was it considered impossible before but not now?

I probably couldn't understand the whole circuit, but, just a layman's explanation of the basics would help me.

Easy to understand, the field voltage needs to be controlled independently on series to regen as current jumps a lot and is like a feedback loop, more current, more field which makes even more current.

Also easy to understand the attitude,
Curtis took $20 of parts to make a boost buck circuit to convert an off the shelf sort of reliable 120v controller that had MOSFETs and other components barely rated high enough for a non regen controller, downrated the controller to 96 volts and used it for regen with obvious results.

This then set the mood in the market that it didn’t work reliably even though the requirements were well understood and even though series regen had always been used on trains, cargo “trucks” mining crapola etc. with success.

Lee Hart has a lot of comments on series regen and as stated there are circuits floating around on this subject.

There are also retail regen controllers like the 1221r, Kelly and Zapi units.

Just remember, even a 48 volt series motor can have HV spikes during regen, having hv rated components with oversized amp ratings is key
 

·
Registered
Joined
·
14 Posts
Discussion Starter #15
No you do not increase the back emf so that it overcomes the bus voltage.... that's not how it works. That would be an uncontrolled regen situation, with no way to limit the current other than through cabling impedance. The converter changes topology to accomplish this, and it is a boosting function, with the back emf as the source/input voltage.

The way I do it is the same as it was done on locomotives, and the control works essentially the same way in forward motoring as it does in regen. The more you press the pedal, the more it brakes.

Trust me, I've flown down the road at 50 mph and regened all the way down to a stop, controlled all the way down, just as smooth as ever...sustained regen currents up to 400A so far. And no, my brushes do not spark or wear. I am confident my regen currents can be anything with a neutrally timed motor. And as such, I will be doing this with a WARP11HV as well.

I will upload video if anyone cares. I kind of want to put some lifeblood back into DC with this
 

·
Registered
Joined
·
966 Posts
No you do not increase the back emf so that it overcomes the bus voltage.... that's not how it works. That would be an uncontrolled regen situation, with no way to limit the current other than through cabling impedance. The converter changes topology to accomplish this, and it is a boosting function, with the back emf as the source/input voltage.

I will upload video if anyone cares. I kind of want to put some lifeblood back into DC with this
Post a circuit, I’m sure someone on here would integrate it into a Paul and Sabrina open revolt controller.

Couple your design with the solid state controls (asic) proposed here
http://www.diyelectriccar.com/forums/showthread.php?t=196337&highlight=Controller

And you might have a new diy winner
 

·
Registered
Joined
·
1,092 Posts
I'm certainly interested. I still don't quite grasp what's happening, but I understand it a lot better than I did before this thread.

Idiot's Guide to DC Regen would be a nice addition :p
 

·
Registered
Joined
·
131 Posts
start by understanding what happens when you move a magnet near a wire, or when you run current through a wire. That is really all that is going on.
 

·
Registered
Joined
·
8 Posts
Have you looked into *modular* switched reluctance? Since the current in such a design does not go though the controller, you should be able to use GaN or SiC devices to accrue the current as it goes into the voltage/current bus. Such a design should require an ABS like interface so it does not lock the wheels. Note that a modular design means that the controller can activate anywhere from 2 to whatever the total amount of stator poles is.
 

·
Registered
Joined
·
727 Posts
Well, I don't want to repeat the discussion about regen in a DIY build, for instance about bad weather conditions when traction is an issue, the integration of ABS and conventional braking,

but if you really want regen and you want to choose the motor that is worst of all in generator mode, go for series DC.
 
1 - 20 of 20 Posts
Top