AEVA Members: your Forum Login (here) is NOT the same as your new AEVA member login.
You do not need to change your existing Forum login.

Regen breaking on series wound dc motor

Open for any sort of non-technical discussion regarding EVs
Post Reply
User avatar
Alex_Brooy
Noobie
Posts: 25
Joined: Tue, 10 Nov 2009, 06:29
Real Name: Alex
Location: Ukraine
Contact:

Regen breaking on series wound dc motor

Post by Alex_Brooy »

Hi guys,
I did not know where to put this topic, so I thought its better to start it here.
I use my EV during 2 years. I think it is possible to get regenarative breaking on series wound dc motor (like Costov or AdvancedDC). We only need galvanicaly insulated source to power up the stator field.
Did somebody try to make it ? Or is there any factory made device? What do you think about this? Is there a strong demand in such a device?

Alex

User avatar
coulomb
Site Admin
Posts: 4040
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Regen breaking on series wound dc motor

Post by coulomb »

Alex_Brooy wrote: I think it is possible to get regenerative braking on series wound DC motors (like the Costov or AdvancedDC). We only need galvanically insulated source to power up the stator field. ... What do you think about this?
Oh, you mean some method of adding to the field current?
There is the problem that the armature current will go the wrong way when charging the battery, so the regeneration system would have to overcome the armature current (which would be the regen current, possibly a lot smaller than traction current) and reverse its direction. This would consume considerable power, since for say 200 A of regeneration, you need to push 400 A into the field. So during regeneration, the field losses would be tripled (the original field loss, another the same to overcome it, and another the same to reverse it the same amount). This tripled loss would eat considerably into the gains you could expect from regeneration. If the field was originally 10% of motor power, then you would be losing an extra 20% during regen. Regen would still likely be a net gain, just 20% less than you could have had if the field was co-operating with rather than hindering regen.

If you mean your isolated circuit to replace the field current only during regen, then there is the problem of switching the field current (you could possibly ensure that it only happens at near-zero motor current). If your plan is to replace the field current all the time with the isolated circuit, there is the problem of generating many hundreds of amps during acceleration, and tracking the armature current closely.

This is a frequently considered problem. Image
Is there a strong demand in such a device?
If you could get it to work with little extra cost, I'd say that something like half of the DC conversions would like to implement it. DC series wound conversions are 80-90% of all conversions (at a wild guess), so IF you could work out a cheap effective way to do it, then yes, my guess is that there would be strong demand.
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
1.4 kW solar with 1.2 kW Latronics inverter and FIT.
160 W solar, 2.5 kWh 24 V battery for lights.
Patching PIP-4048/5048 inverter-chargers.

User avatar
coulomb
Site Admin
Posts: 4040
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Regen breaking on series wound dc motor

Post by coulomb »

coulomb wrote: ... so the regeneration system would have to overcome the armature current (which would be the regen current, possibly a lot smaller than traction current) and reverse its direction. This would consume considerable power, since for say 200 A of regeneration, you need to push 400 A into the field.

Actually, this is woolly thinking. I was thinking that there would be 200 A of field current from the main controller to reverse. But of course, during regen, you would want to be sure that the main controller is off. Suppose you need 500 A of field current to get the armature to a high enough voltage to cause regen to happen. Suppose you get 50 A of regen to happen; that current is flowing through the field coils from the armature to the pack, and it is flowing in the wrong direction for keeping the back EMF high. So now you have 450 A flowing in the field, and the back EMF is too low to sustain regen, so the regeneration chokes itself. You will need 550 A of field to get 50 A of regen into the pack, or 700A to get 200 A of regen.

As the motor speed slows down, you will need more field current to get the same back EMF from the motor. Suppose 500 A net was enough for back EMF to slightly exceed pack voltage when the vehicle is doing 100 km/h; when it has slowed to 60 km/h, you will need 100/60 * 500 = 833 A net, or 933 A to get 100 A of regen. At that point, it might make sense to change down a gear or two, so that the motor is running faster, so less field current is needed for regen.

933 amps sounds like you will need a controller as large as the main one for regen. However, the field resistance (and possibly inductance) is a lot lower than the armature resistance, so the power needed is only about 10% of the power for the main controller. So a specially constructed smallish controller might be possible.

The other consideration is what the main controller will think of the strange voltages and currents at its output. Presumably if its PWM is guaranteed to be at zero, it won't care, but I'd have to wonder.

I've been following the broad idea of regen on a series DC motor for a while, and as far as I know (I haven't read all that extensively), the idea of an isolated separate field controller with no contactors has not been considered. (No doubt others will correct me of I'm wrong there).

Actually, it may not need to be isolated, either. Controllers like to have the other end of the motor at B+, and conduct to B- when their MOSFET or IGBT is switched on. If you wire the motor such that the field has one end at B+ (so the main controller's output connects to the armature, not the field), then it may be possible to use a standard controller.

Maybe I'll post a circuit of this idea.
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
1.4 kW solar with 1.2 kW Latronics inverter and FIT.
160 W solar, 2.5 kWh 24 V battery for lights.
Patching PIP-4048/5048 inverter-chargers.

User avatar
coulomb
Site Admin
Posts: 4040
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Regen breaking on series wound dc motor

Post by coulomb »

coulomb wrote: Maybe I'll post a circuit of this idea.

I prepared a diagram of this, then realised that the field current would be in the wrong direction for regen. Oops. Image

So the field controller, whatever it is, would have to be rather tricky to achieve the required isolation (or really, reversed current flow; isolation isn't really the issue). A large transformer might do it, but that would be expensive and heavy, even though it would be operating at the PWM frequency. It would be difficult to prevent the transformer from saturating due to the large DC currents flowing.

I think that there's a reason that this is not easy; many people have looked at the problem, and concluded that it's not worth the cost and effort. Of course, it's possible to think of something that many others have overlooked, and you often learn a lot thinking about these sorts of problems.
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
1.4 kW solar with 1.2 kW Latronics inverter and FIT.
160 W solar, 2.5 kWh 24 V battery for lights.
Patching PIP-4048/5048 inverter-chargers.

User avatar
coulomb
Site Admin
Posts: 4040
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Regen breaking on series wound dc motor

Post by coulomb »

Maybe it could be done like this:

Image

This unfortunately requires a small auxilliary pack (about 10% of the main pack because the resistance of the field is about 10% of the resistance of the whole motor) for regen. This will be a nuisance to charge and keep balanced. For a 144 V main pack, the field pack could be say 24 V. Because of the low voltage, the IGBT as pictured would probably be replaced with a MOSFET, to reduce losses.

Also, the field controller is of a different topology than normal, since it has to push current the non-conventional way to the field. So you couldn't use an unmodified off-the-shelf 24 V 800 A controller, for example. You might be able to start with one and modify it. The gate driver would have to be isolated; I'm guessing this would not be common in a small controller.

As before, it is critical that the main controller is known to be off when the field controller is on, and vice versa. I have no idea now to make that fail-safe.

I'm guessing that the objection would be that the extra 10% of pack cells would be better used by incorporating them into the main pack, and forgetting about regen.

Perhaps the small pack could be replaced with a DC-DC converter from the main pack; 24 V power supplies that run on 110 V mains are common. However, the current requirements are huge, so it would not be any sort of ordinary power supply.

Perhaps there is some clever way of combining the DC-DC with the field controller, but I can't immediately think of a way, especially since the current needs to be reversed. There is a topology of switching converter with negative output; maybe that would work.

Good luck with the idea!
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
1.4 kW solar with 1.2 kW Latronics inverter and FIT.
160 W solar, 2.5 kWh 24 V battery for lights.
Patching PIP-4048/5048 inverter-chargers.

antiscab
Senior Member
Posts: 2599
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Regen breaking on series wound dc motor

Post by antiscab »

a series DC motor can be run as a sepex.

you do need a motor that has interpoles.
reversing the current on the armature (or the field) will put the ideal brush advance in the opposite direction.
if the difference between the ideal advance and the actual advance is too different, the result is arcing on the comm.

putting a boost converter between the motor and battery pack allows you to continue to run the motor in series mode.
In this case the motor acts as a current source rather than a voltage source.

Matt
Matt
2017 Renault zoe - 25'000km
2007 vectrix - 156'000km
1998 prius - needs Batt
1999 Prius - needs batt
2000 prius - has 200 x headway 38120 cells

User avatar
Alex_Brooy
Noobie
Posts: 25
Joined: Tue, 10 Nov 2009, 06:29
Real Name: Alex
Location: Ukraine
Contact:

Regen breaking on series wound dc motor

Post by Alex_Brooy »

well, actually I already have done such experimental device to switch the motor into genaration. I made HF tranformer, driven by IGBT, and diode rectifier after transformer.
Also there is a regulating resistor to regulate Field current from 0 to 230A. The total power I need to get 230A in the field wounding is about 5A at battery's 95V. The breaking power is about 2kW at 50km\h. I didn't make yet actual metering. I'll try to make a video with metering, soon.

User avatar
Alex_Brooy
Noobie
Posts: 25
Joined: Tue, 10 Nov 2009, 06:29
Real Name: Alex
Location: Ukraine
Contact:

Regen breaking on series wound dc motor

Post by Alex_Brooy »

This image shows my current wiring.
Image

That's what I think to make to avoid the subtraction between field exciting and regenarative currents
Image

User avatar
coulomb
Site Admin
Posts: 4040
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Regen breaking on series wound dc motor

Post by coulomb »

Alex_Brooy wrote: This image shows my current wiring.
Image
I think you'll have problems with the large DC component in the transformer. The core will tend to magnetise in one direction, and you might end up with saturation.
This's what I think I need to make to avoid the subtraction between field exciting and regenarative currents
Image

Interesting. I now realise that with my circuit, and the above, regeneration current from the armature (rotor) gets into the battery via the protection diode across the main controller's switch. I guess that should be OK. Though you could add yet another diode to carry the regen current instead of the controller's protection diodes. It would increase regen efficiency slightly (one diode drop instead of two).

I also realise that my circuit requires that the armature current goes through the field. Yours is better in that you can prevent the regeneration current from flowing (the wrong way) into the field, at the cost of the required isolation.

So your circuit is interesting, and it looks like it can be made to work. There are three extra high current diodes there, one of which has to handle full traction current (the others only need regen field current, which could be less than full traction field current). The diode handling full traction current will of course lose performance even when not regenerating, but in a 96V+ system, the 1V lost across this diode (~ 1%) should easily be made up by the benefits of regeneration.

The 2 kW of regen that you have achieved is on the low side; Prius manage up to about 20 kW. But it's an excellent proof of concept; well done. It will certainly be interesting to see where this goes.

As a point of interest, does your motor that you're experimenting with have interpoles?

Edit: fixed tags; added comment re adding another diode
Last edited by coulomb on Sun, 12 Dec 2010, 04:36, edited 1 time in total.
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
1.4 kW solar with 1.2 kW Latronics inverter and FIT.
160 W solar, 2.5 kWh 24 V battery for lights.
Patching PIP-4048/5048 inverter-chargers.

User avatar
Alex_Brooy
Noobie
Posts: 25
Joined: Tue, 10 Nov 2009, 06:29
Real Name: Alex
Location: Ukraine
Contact:

Regen breaking on series wound dc motor

Post by Alex_Brooy »

The first schematic is not exact realization, it is only to show the principle of work. Everything is fine with saturation&etc. One man helped me to make the device, he is professional power electronics developer. My experimental "recuperator" had been working for about 45 minutes without any damage.

I'm not clear about the "interpoles". What does it mean?
My series dc motor has 4 poles in stator field. Poles where connected in series, but I reconnected in parallel-series to make rpm higher. This is bulgarian motor DS3.6\7.5\14 for old forklift car. Today Kostov mtrs. produces the analogue motors.

User avatar
Alex_Brooy
Noobie
Posts: 25
Joined: Tue, 10 Nov 2009, 06:29
Real Name: Alex
Location: Ukraine
Contact:

Regen breaking on series wound dc motor

Post by Alex_Brooy »

Once again, this is experimental isntallation:
Image

I've got 53A, 100V regen power. -6A is zero point of amp clamp meter, so -59-(-6) = 53A
Image

User avatar
coulomb
Site Admin
Posts: 4040
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Regen breaking on series wound dc motor

Post by coulomb »

Alex_Brooy wrote: I'm not clear about the "interpoles". What does it mean?
These are special windings that attempt to counteract a nasty thing called "armature reaction". I'm rather hazy on the concept; it's something like the current in the armature windings producing a field that twists the effective field orientation.

First consider a motor without interpoles. The net effect of armature reaaction is that the best place to put the brushes depends on motor current. Hence, most motors designed for above about 72 VDC have "advanced timing", meaning that the brushes are moved to a position closer to where the minimum arcing point is when nearing full rated current. The fixed position of the brushes is not ideal for low currents, but that doesn't matter so much since arcing is not much of a problem at low currents.

The problem is that when the armature current reverses, as it does during regen, the advanced position of the brushes is in totally the wrong direction; you would want retarded timing. So with regen, you have to reduce the advancement of brush timing, so that during regen, it's not so badly out of position. That means more arcing at high traction currents, of course.

With a motor that has interpoles, the interpole windings counteract armature reaction. The strength of the counter-reaction needed is proportional to motor current, so by putting the interpole windings in series with the normal field windings and the armature, armature reaction is greatly reduced. The ideal brush position is now neutral - not advanced or retarded. With interpoles, arcing is reduced at high traction currents, but also at low traction currents, and even at high regeneration currents.

Hence, interpoles are good for regeneration. Some would say that they are essential; I don't know. At 96 V, you can probably get away without interpoles, unless traction and/or regeneration currents are particularly high. Kostov motors (made in Bulgaria, I believe) mostly (possibly all) have interpoles, but I suspect that even a Bulgarian motor designed for a forklift would be made without interpoles.

Edit: where the neutral point -> where the minimum arcing point
Last edited by coulomb on Sun, 12 Dec 2010, 16:52, edited 1 time in total.
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
1.4 kW solar with 1.2 kW Latronics inverter and FIT.
160 W solar, 2.5 kWh 24 V battery for lights.
Patching PIP-4048/5048 inverter-chargers.

User avatar
gmacd33
Groupie
Posts: 102
Joined: Mon, 11 Jan 2010, 23:51
Real Name: Greg MacDonald
Location: Brisbane
Contact:

Regen breaking on series wound dc motor

Post by gmacd33 »

That's an interesting point about interpoles; I have heard before the voltage limitation before advanced timing is required, which prohibits regen. Anyone know of any disadvantages of using interpoles - ie why don't Netgain and ADC use them?!?
Greg MacDonald
Chief Technical Officer
Lithium Battery Systems

User avatar
EV2Go
Senior Member
Posts: 2059
Joined: Wed, 16 Jul 2008, 00:21
Real Name: Paul
Location: Brisbane 1963

Regen breaking on series wound dc motor

Post by EV2Go »

Define voltage limitation... The Kostov I am using is rated for 250v.
The extra cost associated perhaps?

antiscab
Senior Member
Posts: 2599
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Regen breaking on series wound dc motor

Post by antiscab »

netgain use interpoles on their most powerful motor, the 11" HV version.

Both netgain and adc used to do forklift motors only, which run on low enough voltage that interpoles are redundant.

Matt
Matt
2017 Renault zoe - 25'000km
2007 vectrix - 156'000km
1998 prius - needs Batt
1999 Prius - needs batt
2000 prius - has 200 x headway 38120 cells

7circle
Senior Member
Posts: 540
Joined: Thu, 04 Feb 2010, 02:18
Real Name: Ken
Location: Melbourne

Regen breaking on series wound dc motor

Post by 7circle »

Can you eleaaborate as to why they are redundent.

Is it just that the Arcing of air requires too high potential voltage to scale up from 96V?

antiscab
Senior Member
Posts: 2599
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Regen breaking on series wound dc motor

Post by antiscab »

at 0 rpm, the ideal brush position is neutral, regardless of current.

above 0 rpm, the ideal brush position shifts relative to both rpm and current.

at low voltage, it is not possible to have both high current and high rpm (speaking relatively).
so interpoles don't help as much as just putting a bigger comm and brushes on (its a cost thing)

the actual limits depend on the specific motor.

to go high rpm, voltage and current, you need either very large brushes that can deal with significant mis-match, or you use interpoles to reduce the shift in ideal brush position.

Matt
Matt
2017 Renault zoe - 25'000km
2007 vectrix - 156'000km
1998 prius - needs Batt
1999 Prius - needs batt
2000 prius - has 200 x headway 38120 cells

User avatar
Alex_Brooy
Noobie
Posts: 25
Joined: Tue, 10 Nov 2009, 06:29
Real Name: Alex
Location: Ukraine
Contact:

Regen breaking on series wound dc motor

Post by Alex_Brooy »

Well, I checked about interpoles.
My colleagues and I discussed this a secret mechanism called interpoles. One of the most experienced electricians said that to counteract the armature reaction, there exist such methods:
1. Compensating winding
2. additional poles
3. increase the air gap between the armature and main poles
4. manufacturer of the main poles of the cold-rolled steel
5. increase in the size of the main poles

Methods 3,4 & 5 are used mostly on old bulgarian motors. Large dc machines use 1& 2nd methods.

Post Reply