AEVA Forums

I'v been getting that.

I'll try to be more constructive. A while ago, I found a paper on exactly this:
A Four Quadrant Adjustable Speed Drive For Series Wound DC Motors

The author uses a reversing contactor, but has a good description of motor operation and control.

I see the options for no-reverse contactors as follows:

1) A really high current power supply that can reverse the field while the motor is being driven by the PWM. I would guess that the field is in the order of 5-10 volts, mostly the resitance of the copper straps in the pole windings. This is a tough one because it's exactly in the dread zone for semiconductors. It also needs an impressive down-converter for the high current (500-1000A), kind of like a DC spot welder, except that spot welders are all AC.

2) Sepex. The field windings need to be re-done so that the copper is appropriate to lowish currents that are easy to control with a small (50 amp?) H-bridge controller.

Sepex should be no less efficient than Series, as the power needed for the field is the same regardless. If the power in the windings is a constant, the voltage and current are determined by the copper geometry.
Thin wire = lower current, but more turns and higher voltage.

A field winding and current that is compatible with the power supply (160V?) is needed to make the driving electronics practical.

An analog is an automotive alternator. A few amps of field current can produce a lot of amps of output current. I think that typically, this is a factor of 30 or more.

In a series motor, the field windings are carefully chosen to provide
sufficient field when fed with the same current as the armature.

Sepex probably gets a bad name because the field current on the controllers is too little to fully realise the motor capabilities.

Sepex is like a shunt confiruration, except that the field winding current is controlled by a separate controller channel. The field current is proportional to the armature current, so the field PWM must vary according to the current in the armature, whereas the armature PWM is more related to the speed of the motor.

For braking, the field current can increased so the motor's back emf is greater than the battery voltage, causing current to flow through the transistor intrinsic diodes towards the battery. A Sepex DC motor will be considerably less effective at braking energy recovery than an indiction motor will be.

For reverse, a H-Bridge is needed to allow the field to be reversed and motor turn the other way. Because the field current is relatively low, this is likely to be only 4 big transistors with non isolated gate drivers like the IRL2113 ($2 last time I bought any) or its kin.

Overall, the Sepex can provide some operational improvements, but because it involves rewinding the field, it's benefit is marginal, if at all.

Speedily wrote: I still think a small motor and a battery is the way to go with your setup EV2GO

Just trying to get the value of option is perspective.
How much would a second small motor and mechanical gearing etc with an extra reverse drive suitable motor controller cost?

I did this exercise once before, I figure if your going to try do it with a starter you want a light weight high torque one like a Tilton which can set you back $350-$500 for starter alone. Of course you can buy cheaper but if I was going to do it I want to do it right first time.

bga wrote: I couldn't interest you in one of these reversing contactors?

Thanks for the Link.
The cost looks very good. Price: AU$185 inc GST (@2010/10/15)

"This particular assembly is basically two ZJWH400A contactors back to back"

Manufacturer: ZJWH400A-2T DC contactor

Schematic


The Manufactures spec on the 400A has:
7. Voltage fluctuating：0.7~1.1Ue

That suggests to me the per contact-pair voltage drop at rated current of 400A is 0.7V when cool and 1.1V when hot.

This would be for contacts in good condition

It also is very generic info, not that discriminating.
For the reversing pair it would be double that.

So at 400A at 1V+1V = 800W of power loss in the contactor.
This is just for forward motion.

So if the conacts act just as a resistor but they do get hot so resistance increases. But say for 2V/400A= 0.005Ohm.

If cruising current at 50Kmh is (eg Pe of 20kW/(32% of 300V) = 200A
Thats 200A x 5mOhm = 200W loss, not to bad compared Pe of 20KW

So what happens at 1000A.
I^2 x R = 1000A x 1000A x 0.005ohm = 5kW
5KW for 20Sec is plenty of heat.

But I'm inclined to think the voltage drop would be more than 1.1V.

The other issues is making sure they are never opened under current with significant applied voltage. Arcing obviously will burn them out very quick.

So the suggested idea I'm thinking through is:
- Concept using 11" KOSTOV motor and Soliton Controller as basis.
- No changes to motor (ie rewind of field)
- No added losses in forward motion due to contact and extra wiring losses.
- Reverse of field current to allow motor torque reversal.
- Added benifit that field weaking can be achieved, this reduces the Kv allowing the RPM to increase at low torque. (Not efficient though)
- Added Benifit that the current can be varied to provide regen-braking at lower speeds.

bga wrote: I'v been getting that.

I'll try to be more constructive. A while ago, I found a paper on exactly this:
A Four Quadrant Adjustable Speed Drive For Series Wound DC Motors

The author uses a reversing contactor, but has a good description of motor operation and control.

Thanks for link it has some good info on the theory

bga wrote:

I see the options for no-reverse contactors as follows:

1) A really high current power supply that can reverse the field while the motor is being driven by the PWM. I would guess that the field is in the order of 5-10 volts, mostly the resitance of the copper straps in the pole windings. This is a tough one because it's exactly in the dread zone for semiconductors. It also needs an impressive down-converter for the high current (500-1000A), kind of like a DC spot welder, except that spot welders are all AC.

The reversein gof a car, trike or dragster doesn't need the full-ON power/torque that forward accelaeration needs/wants.
So currents don't need to go to 500-1000A. I suggest only about 100A for the 11" Kostov. This allows some trade of in effeciency in reverse to less than in forward as it isn't used very often or for long(Typically).

bga wrote:

2) Sepex. The field windings need to be re-done so that the copper is appropriate to lowish currents that are easy to control with a small (50 amp?) H-bridge controller.

Sepex should be no less efficient than Series, as the power needed for the field is the same regardless. If the power in the windings is a constant, the voltage and current are determined by the copper geometry.
Thin wire = lower current, but more turns and higher voltage.

A field winding and current that is compatible with the power supply (160V?) is needed to make the driving electronics practical.

An analog is an automotive alternator. A few amps of field current can produce a lot of amps of output current. I think that typically, this is a factor of 30 or more.

In a series motor, the field windings are carefully chosen to provide
sufficient field when fed with the same current as the armature.

Sepex probably gets a bad name because the field current on the controllers is too little to fully realise the motor capabilities.

Sepex is like a shunt confiruration, except that the field winding current is controlled by a separate controller channel. The field current is proportional to the armature current, so the field PWM must vary according to the current in the armature, whereas the armature PWM is more related to the speed of the motor.

For braking, the field current can increased so the motor's back emf is greater than the battery voltage, causing current to flow through the transistor intrinsic diodes towards the battery. A Sepex DC motor will be considerably less effective at braking energy recovery than an indiction motor will be.

I agree so far.. but it would be good to get the full capabilty of the Motor in forward and still have reverse.

bga wrote:For reverse, a H-Bridge is needed to allow the field to be reversed and motor turn the other way. Because the field current is relatively low, this is likely to be only 4 big transistors with non isolated gate drivers like the IRL2113 ($2 last time I bought any) or its kin.

Thats with a high turn field, so extra 4 IGBT's 50A/600V and driver circuit with 2 x IRL2113's on a PCB antishort logic and with just forward or reverse logic and PWM generator. Thats a not cheep to make your self. Compared to a $270 3V3 85A PSU with 250Vdc Supply.

bga wrote:Overall, the Sepex can provide some operational improvements, but because it involves rewinding the field, it's benefit is marginal, if at all.

Can you see any benifits in my approach?
Soliton is $4000
11 KOSTOV 250V 40kWCont $3500
So where do you get a SEPEX Motor and Controller With this kind of power?

The reverse Low Volt PSU on the field can be adapted to the Double Kostov or other motors and controllers.

The approach shown in the circuit in this post with the transformer buck inverter, would be better if the Capacitor was replaced with a seperate single 10A/600V IGBT, with PWM control.

EV2Go wrote:

Richo wrote: I think this setup would still need one main contactor.

The Soliton1 has a beefy built in contactor
....
The controller has 3 input and 3 outputs so perhaps there is something on the controller that might be of use.
Would image several hundred $300-$400 maybe sightly more.
Has an option for both 8KHz & 14KHz

The Soliton Contactor inside it is for the input from the Battery back.

Richo where do think the Big contactor needs to go?

Just wondering (because I don't know a fraction of what you guys know) if the controller is idling i.e no throttle percentage applied, and voltage in reverse set to 9v, would there still be voltage at the motor to be effectively pushed back with this rig?

Another way to do it would be to use one of the outputs on the controller to cut the 0-5v feed from the throttle, so the controller couldn't see a throttle input.

Or the switch that selects reverse cut the throttle voltage.

7circle wrote:

bga wrote: I couldn't interest you in one of these reversing contactors?

Thanks for the Link.
The cost looks very good. Price: AU$185 inc GST (@2010/10/15)

"This particular assembly is basically two ZJWH400A contactors back to back"

Manufacturer: ZJWH400A-2T DC contactor

Schematic


The Manufactures spec on the 400A has:
7. Voltage fluctuating：0.7~1.1Ue

That suggests to me the per contact-pair voltage drop at rated current of 400A is 0.7V when cool and 1.1V when hot.

This would be for contacts in good condition

It also is very generic info, not that discriminating.
For the reversing pair it would be double that.

So at 400A at 1V+1V = 800W of power loss in the contactor.
This is just for forward motion.

So if the conacts act just as a resistor but they do get hot so resistance increases. But say for 2V/400A= 0.005Ohm.

Had a look at the AllBright SW202 data sheet info
http://www.albrightinternational.com/fi ... EAFLET.pdf

They have 40mV/100A per Contact Typically.

So 4 x 40mV is 160mV at 100A
S0 400A would x4 so 0.64V

The idea of running 1000A through these is not going to be healthy.
Their fine with 250A continuous.
As long as you don't open them with over 100Vdc across them.

EV2GO wrote:Just wondering (because I don't know a fraction of what you guys know) if the controller is idling i.e no throttle percentage applied, and voltage in reverse set to 9v, would there still be voltage at the motor to be effectively pushed back with this rig?

The Soliton Controller allows for reverse setting so when it gets a REVERSE_SELECT Signal it limits its output voltage to say the 9V you mentioned. This 9V limit will limit the speed/torque in reverse.
A throttle limit of 10-100% may be set as well.

I'll assume Reverse Volts is set to 64V and Throttle limit is 60%

How the Soliton copes with a Reverse Select when it is Powering along at full throttle in forward motion is another thing.
I'll assume the driver can and will do it.

So with this "RIG" when the Reverse Button is selected (say on the dash) the Soliton gets this and it should calmly and it sets an Output
(to A REVERSE DASH LAMP/Buzzer).
Another Soliton Output is configured that PWM's (Base Freq 20Hz - 20Khz) the desired Field Current, this could just be an echo of the throttle value. This drives an opto that shorts the enable RC+ and RC- on the PSU (SP-480-3.3).
This enables the PSU only when the throttle is pressed.
If the PSU was enabled just by the Dash Button the PSU would inject 85A at upto 3.3V (280W) which would produce some torque in the motor as current can flow not just in the field but also through the armature/rotor.

The PWM to the RC (Remote Control) of the PSU would give sensitve reverse control.

For reverse, when the Throttle is pressed past the dead band the Soilton1 gives a little duty % to the Field PSU and a little duty % to the Main output.

For a full push on the throttle, the Soliton ignores any more than 60%, and sets 60% duty to the Field PSU (85x60%=51A) and a Drives the Main Output to a Max output Voltage of 64V (I think thats what I worked 200RPM would need with 100A Field Current).

Would be interesting to know what EVnetics thinks of this approach.
Their controller looks very nice so it would be sad to see a messy wiring and Ugly PSU (SP-480) near it. A smart looking metal box to house the extra gear in would be the go. A reversing contactor would also need a box.

I noticed I had 10A fuse on the 240Vdc to the PSU, it should be say 2A.

The Regen Boost and wouldn't be possible only field weakening in Forward motion at low max currents of 3.3V of Field voltage.

Explaination pending...

Another Update to Sketch. REVERSE SUPLY is the 240DC Contactor supply As the PSU has a PFC it won't have surge currents (I think as its rated for DC Input to 370VDC (SP-480-3.3))

The RC link needs OPTO ISOLATOR for PWM to pass to the RC+ and RC-

There is still an element of risk that this system won't work.
To get low voltage from the controller, the mark space of the PWM will be low. Consider, in normal operation, the field and armature are getting these pulses in sync.

In the proposed system, the 2.5 VDC source is supplying field current in the non-pulse-active part of the PWM cycle - when the armature is also not drawing current.
It will relay heavily on the field magnetic flux not responding quickly to the PWM pulses. The fluxes won't coincide (IMO). I think you will still go forwards.

Edit: Removed stutter

I'd started to consider an inductive filter a bit back in the posts.
Thanks Johny, for questioning the by-pass of the Firld inductance.


When the main controller pulses the motor with 240Vdc for 0.5% duty PWM at 8kHz the Field and Rotor inductances retard the current increase, so the Field will have large voltage across it.

I don't now what the Kostov Motor inductance is.
I'll just assume the Field L is bigger than the Rotor L.

So considering 300V across the Field the Low Voltage PSU has a diode that will bypass the Field and take all the current. (Bang!)
So if inductance is added between the diode (PSU O/P) and the Field winding, then the current increase can be retarded by the inductance.
The LV PSU has output capacitance and inductor could have significant resonance issue.

An inductor in the (+) Positive side would keep the (-) Negative side linked to the high side of the DC bus.

So with a switching PWM frequency of 8kHz, what inductance would give a significant impedance to attenuate current impulses.

So for a 30% duty of 300Vdc at 8Khz would be for
Tperiod x0.30=(1/f)xDuty% = (1/8000)x0.30 = 37.5 uSec

So for a current increase to 10A in 37.5 uSec the impedance needs to be: From dI/dt= V/L, L=Vx(delta(t)/delta(L))
So L=300Vx37.5uS/10A = 1.125 mH

I'm considering a toroidal core of 300VA size to cope with the Flux from (I^2 x R = 85Ax85Ax0.025Ohm = 180VA) plus 10A from the motor controller.

Looks do-able with 1.125 mH and 100A and 30% max duty at 8kHz.
Better still for 14kHz.

Any thoughts on the resonance issue?

My Sketch if the LV PSU was a dud so changed that.
Added the Inductor.
And moved Reverse relay to same side.
And Added a Diode to the O/P of the LV PSU.
Diode would need to cope with 600V/100A

More costs. But still feasible.

I got distracted on the ideas of 2 phase BLDC.

But as I'm not likely to use a brushed DC motor, you modtlikely won't see me modelling a prtotype of the above, just woarning if you were holding your breath.

I still like this concept without a seperate DC/DC:


There is no short cut path for current around the field as the transformer secondary windings can be made to have the inductance to block the high frequency through the diodes.

You still have the positive field current coming from the controller which is in the opposite direction to your injected reversing current. Won't this just cancel out and result in zero field current and hence zero motor current? Current is current no matter whether it is from a high frequency source or DC; the motors winding see it all as DC.

evric wrote: You still have the positive field current coming from the controller which is in the opposite direction to your injected reversing current. Won't this just cancel out and result in zero field current and hence zero motor current? Current is current no matter whether it is from a high frequency source or DC; the motors winding see it all as DC.

That's where he uses the 30V to 3V toroidal transformer to "boost" the field current. The PWM pulse current from the diodes will be 10 times the current applied directly from the controller.
It requires a custom toroid transformer but easy enough to wind as there won't be that many turns involved. A 110 VAC to 5V might be OK - so a non-custom would be easier to try to prove it works.

Anyone have a 240 VAC to 5V 300VA transformer with a split primary so we get rig it for 120 VAC input? Even an Iron core transformer would do to try it.

evric wrote: ... and result in zero field current and hence zero motor current?

Zero field current means no torque and no back EMF, which means very large current (assuming you have some voltage across the armature), and possibly very high (destructively high) speed (assuming that there is some residual field).

Zero field is a Bad Thing (TM).

coulomb wrote:

evric wrote: ... and result in zero field current and hence zero motor current?

Zero field current means no torque and no back EMF, which means very large current (assuming you have some voltage across the armature), and possibly very high (destructively high) speed (assuming that there is some residual field).

Zero field is a Bad Thing (TM).

Good point - I just let it slide by. You have touched on a point in that series connected DC motors are a lot safer to use than Sepex or Series simply because there can not be an absence of field current.
That poses yet another "lookout" for the proposed system for reversal - but I still think it would be a general benefit for many DCers to get it proved.

evric wrote: You still have the positive field current coming from the controller which is in the opposite direction to your injected reversing current. Won't this just cancel out and result in zero field current and hence zero motor current? Current is current no matter whether it is from a high frequency source or DC; the motors winding see it all as DC.

The idea is the impedance seen into the transformer is low and pushes power in to the secondary.
This will not just cancel the current coming into the field from the positive rail but exceed it causing the NET current to be negative.

Yes the average current and direction is what is important.

I thought I could do it with out a capacitor and DC choke but no success yet.

See the simulator info at bottom of post.

Johny wrote:That's where he uses the 30V to 3V toroidal transformer to "boost" the field current. The PWM pulse current from the diodes will be 10 times the current applied directly from the controller.
It requires a custom toroid transformer but easy enough to wind as there won't be that many turns involved. A 110 VAC to 5V might be OK - so a non-custom would be easier to try to prove it works.

Anyone have a 240 VAC to 5V 300VA transformer with a split primary so we get rig it for 120 VAC input? Even an Iron core transformer would do to try it.

The inductance of the transformer and output windings is very important in responding faster than the Field winding.

I think your on the spot with "not many turns involved"

See the simulator info below for transformer inductance.

coulomb wrote:

evric wrote: ... and result in zero field current and hence zero motor current?

Zero field current means no torque and no back EMF, which means very large current (assuming you have some voltage across the armature), and possibly very high (destructively high) speed (assuming that there is some residual field).

Zero field is a Bad Thing (TM).

If the situation occurred where there was no or very little field current it would not be dramatic. As the driver throttle movement is expecting torque control to move the car backwards. With small field current there would be very little torque.

If you were testing the motor with the reversing circuit on the bench with no load, then it could be dangerous to see over spin.

The idea is to generate negative field current not just cancel it out.

Johny wrote:....Good point - I just let it slide by. You have touched on a point in that series connected DC motors are a lot safer to use than Sepex or Series simply because there can not be an absence of field current.
That poses yet another "lookout" for the proposed system for reversal - but I still think it would be a general benefit for many DCers to get it proved.

Well I hope the simulated circuit helps



Here is a zip of the TXT file that can be imported into the FLASTAD simulator it pops up if you have the free Java Plug-In with your browser:
Simulator_Motor_Field_and_Rotor_Self_Feild_Excite_Block.zip

I guessed (not even an estimate) the Field inductance of 5mH, Rotor inductance of 1mH. If you can suggest better estimate please do.

This used: (EDIT: there are a few High impedance resistors to keep the simulator happy)
- A 250Vdc Source as the Battery (no internal resistance added yet)
- A 10mOhm resistor as SHUNT for Sensing the Motor Current
- 100uF P.F. type capacitor to remove the DC from the transformer primary.
- transformer with 100:30 ratio and Primary inductance of 10uH.
- added tiny secondary resistance (100uOhm)
- added inductance to the center tap of 1uH this keeps the current continuous in charging the Cap.
- The current produced in the secondary, drive the Capacitor (5,000uF) positive WRT to the Battery positive rail.
- added output inductor (10uH) to filter the PWM spikes from the Field.
- the 1K resistor had a voltage source across it to simulate BEMF.
    (Should the BEMF always be Positive, or is it a function of torque)
- the transistor timing was done with an 8kHz square wave with Duty% adjustment (Right click in simulator to edit parameters.)
- the White switch above the 8kHz Square Wave Voltage Source, allows you to turn off the DUTY pulse to the main transistor (Peddle of throttle) and see the current drop off.
- The white switch below the 100uF capacitor turns off the reverse BOOST to the Field. When OFF the secondary circuit if ON will bypass and weaken the FIELD current.
- The white switch at the bottom of the FIELD disconnects the BOOST circuit.

Driver wants forward torque.
- Start with all switched open,
- Set the duty at say 30% (Step in Throttle)
- close the DUTY-SWITCH
and the motor Field and Rotor cruuent will ramp up with the same current as per Normal operation.

Driver wants reverse torque to drive backwards:
- Turn off all switches,
- Set Rotor Voltage to 0V as per 0 RPM.
- Turn on both reverse Field Boost switched
- Set Duty to say 5% (a low level you would expect for throttle reference)
- close the DUTY-SWITCH
and the pulses from the main transotor will enrgise the transoformer and ramp up the negative current in the FIRLD winding.
The Rotor will have the positive current.
The ROTOR and FIELD Currents are opposite and different magnitudes.

The image shows graphs on the bottom. The left top is Rotor Current which was positive and large until I change the rotor BEMF from 0V to 45V resulting in a reduction in rotor current. Below that on the left is the Field Current which was steadily heading negative.

I hope some of you can try out the simulator by opening the txt file in the ZIP and copy and pasting the contents into the FILE:IMPORT feature in the FALSTAD simulator.

I would like to use a Simulator that others are comfortable using.
Perhaps I need to get MATHCAD/ SIMULINK going on a PC. Been a while since I used it.

Any way the concepts looks feasible. Simulators only go so far.
Usually save a few blown components though. 250V/100A controllers are a bit scary and slow to work on and stay safe.

The concept could be remodelled for a 48V DC Golf cart style system.
Any suggestions for such a Motor Controller Combo to simulate?

And thanks very much for the feedback.

(EDIT added above: there are a few High impedance resistors to keep the simulator happy)

For Reverse Switch Configuration (Added Shunt to monitor Battery Current)
With DUTY at 5% and Rotor Voltage at 0V
FIELD :-113
ROTOR :+129A
CONTROLLER: 190A to 488A 5% Duty: 190 + (480-190)x0.05= 190A+14.5A=205A
Battery Peak: (-ve:out) -485A @ 5% Duty = -24.45A ( @ 250V = 6kW)

Values bounce around and are not RMS or Averaged so can't work out efficiency for Battery Back. Had to consider Peak Pos and Neg values and Duty Cycle.

So with RPM ROTOR Voltage at 20V and Duty at 5%
FIELD :-125A
ROTOR :+ 14A
CONTROLLER: 0A - 84A 5% Duty: 0 + (84-0)x0.05= 0A+4.2A=4.2A
Battery Peak: (-ve:out) -76A @ 5% Duty = -3.8A ( @ 250V = 950W)
ROTOR BEMF: -14.5A at 20V = 290W

Eff : Pmech/Pbatt = 290/950 = 30.5%

Not too bas for Slow Reverse at high torque.

I'm hopping my model and Simulator are doing the right thing, will have to hit the paper and pen to validate.

Johny wrote: Good point - I just let it slide by. You have touched on a point in that series connected DC motors are a lot safer to use than Sepex or Series simply because there can not be an absence of field current.

Johny, you want to clarify that? My understanding was that Series DC was the least safe as they can spin to destruction in a no load situation. I have been known to be wrong of course!!

Squiggles wrote:

Johny wrote: Good point - I just let it slide by. You have touched on a point in that series connected DC motors are a lot safer to use than Sepex or Series simply because there can not be an absence of field current.

Johny, you want to clarify that? My understanding was that Series DC was the least safe as they can spin to destruction in a no load situation. I have been known to be wrong of course!!

It might get a bit confusing as one type of circuit with the independent DC/DC would be a Separately excited FIELD, and the other circuit that generates the negative FIELD current from the Main PWM is a series FIELD, as the Field Current is dependent on the main controller PWM.

A motor with a SHUNT FIELD to the rotor is another circuit again that's also has been considered. When the MAIN control is not used to get reverse, only a DC Voltage on the FIELD that would have current also flowing through the Rotor and Main Controller FLY back Diode.

The main point is to find a way to get reverse with out adding switches to the forward torque circuit that can use 1000A at 250Vdc.

I just realised the Topic "Ways to reverse Series DC Motor" is a bit generalized, I think I need to add "100% forward 30% reverse"

Squiggles wrote:

Johny wrote: Good point - I just let it slide by. You have touched on a point in that series connected DC motors are a lot safer to use than Sepex or Series simply because there can not be an absence of field current.

Johny, you want to clarify that? My understanding was that Series DC was the least safe as they can spin to destruction in a no load situation. I have been known to be wrong of course!!

You are correct of course. I was thinking purely electrical. I agree, mechanically series can be a disaster with no - or little - load.

OK, I get it your, point was that if there is no field current you are likely to over heat the rotor windings.
The Kostov Sepex motors are probably a little less at risk because of the interpoles. I wonder if in the case of field excitation being absent the interpoles actually provide a weak field in a series motor potentially giving the no-load destruction problem.

These long posts are hard to zoom in on the bit of interest..

7circle wrote:

bga wrote: I couldn't interest you in one of these reversing contactors?

Thanks for the Link.

...

I agree so far.. but it would be good to get the full capabilty of the Motor in forward and still have reverse.

bga wrote:For reverse, a H-Bridge is needed to allow the field to be reversed and motor turn the other way. Because the field current is relatively low, this is likely to be only 4 big transistors with non isolated gate drivers like the IRL2113 ($2 last time I bought any) or its kin.

Thats with a high turn field, so extra 4 IGBT's 50A/600V and driver circuit with 2 x IRL2113's on a PCB antishort logic and with just forward or reverse logic and PWM generator. Thats a not cheep to make your self. Compared to a $270 3V3 85A PSU with 250Vdc Supply.

bga wrote:Overall, the Sepex can provide some operational improvements, but because it involves rewinding the field, it's benefit is marginal, if at all.

Can you see any benifits in my approach?
Soliton is $4000
11 KOSTOV 250V 40kWCont $3500
So where do you get a SEPEX Motor and Controller With this kind of power?

A: It's probably in an old elevator and weighs about a tonne.
A2: I think this is one of the problems with the SEPEX approach. Because of the high currents in DC Series motors, it's essential to rewind the field coils to get them to operate at a controllable current. Much more tricky than rewinding a 3Phase IM, I suspect.
A3:...

I assume that the level of field magetisation is the determinant of output torque.

The problem that I've got with an 85 Amp /3 Volt PSU is the amount of torque that this will translate to in a typical Series motor. I had a look at the performance curves for the ADC FB1 9 inch motor, 85 Amps is about 8 FtLb of torque. (100 FtLb needs 450 Amps on this motor)

This may be OK is a perfectly level parking lot, but probably not much use if there is any slope to climb. More of the problem is that actually only 1/2 of this torque is available because the 85 Amp field supply is being counteracted by the forward armature drive current. It may not be able to turn the wheels when on jack stands.

I feel that the assumption that little torque is needed in reverse is wrong. Backing slowly up a driveway kerb can require a lot of torque for a brief period. The power is low because the speeds are low.

In defence of the reversing contactors. I know of several cars using them in 'direct to diff' motor applications, all successfully.

bga wrote: I couldn't interest you in one of these reversing contactors?

These contactors don't appear ever to give contact resistance. High current 3 phase industrial contactors always give contact resistance in uOhms or mOhms. I have a feeling that the contactor's contact resistance would be way less than the battery pack resistance (Impedance) but it's hard to get a figure to verify that the losses (of using a reversing contactor) are worth bothering about.

The Allbright Contactors that have a spec listing 40mV/100A per contact.

This Previous Post had a link to the pdf Spec of a SW200 series reversing contactor.

bga wrote:I assume that the level of field magetisation is the determinant of output torque.

The problem that I've got with an 85 Amp /3 Volt PSU is the amount of torque that this will translate to in a typical Series motor. I had a look at the performance curves for the ADC FB1 9 inch motor, 85 Amps is about 8 FtLb of torque. (100 FtLb needs 450 Amps on this motor)

This may be OK is a perfectly level parking lot, but probably not much use if there is any slope to climb. More of the problem is that actually only 1/2 of this torque is available because the 85 Amp field supply is being counteracted by the forward armature drive current. It may not be able to turn the wheels when on jack stands.

I feel that the assumption that little torque is needed in reverse is wrong. Backing slowly up a driveway kerb can require a lot of torque for a brief period. The power is low because the speeds are low.

I touched on this earlier ...

7circle wrote:Assuming you need 5Km/h and 1000kg up 30deg Hill
P= 1000 x 9.8 x Sin30 x 5/3.6 = 6.8kW

That's heaps, too steep, 10deg is 1.1kW @ 5km/h
That's still large so of the speed was crawl at 1km/h and 10deg
power is 5th at 220W.

The motor torque is proportional to the Current in the Field and the Current in the Rotor.

T = Kt.Ia . Kf.If
from the Thesis you (bga) posted earlier.

So a low field current of 85A may be acceptable if the Rotor current is higher.

When a separate voltage is applied to te FIELD the FIRLD and rotor currents are not the same, so the torque curves for the motor in series aren't correct. So Kt and Kf for the motor would be needed to predict the torque.

So with 85A in the 0.025ohm Field.
And if the Rotor current was 85A then the DC/DC Power supply would have 170A coming out of it.
With a 0.075ohm Rotor and 85A that is 6.375V and the Rotor is at stand still. The Motor is absorbing (85^2 x 0.025 + 85^2 x 0.075) = 722W

The KOSTOV 11" Torque Curves don't go down that low and you can't extend the current line.

So it looks like you have a point BGA in that DC/DC needs to be rated to cope with the SUM of maximum "reverse mode" current in FIELD plus ROTOR.

So a 3.3V with a 85A current limit won't be big enough.
It could only cope with 42A ROTOR and 42A FIELD.
For the Kostov in with both FIELD windings in parallel having 0.025ohm DC resistance and the ROTOR with 0.075ohm.

A 3.3V DC/DCwith current limit at 260A, could get 130A in field and let 130A flow in the ROTOR. The ROTOR Voltage would be 9.75V at stall.

So a 3.3V 260A DC/DC nearing 1000W is more appropriate.
The costs of the DC/DC and the inductor and reverse mode relays for 260A are getting expensive.

So back to the reversing Contactor would you recommend the ALLBRIGHT SW200 or "equivalent" NANFENG 400A rated reversing contactors for a SOLITON1 1000A 300VDC controller.

I'm still worried that the Contacts will burn out with over 500A running through them regularly during hill climbs or windy roads.
Let alone 1000A drags.

What about a Gigavac GX14 or GX16 for forward motion and some piddly little thing for reverse?

From eRace's Contactors - GIGAVAC post.

"The GX14 & 16 series at 75 deg C are designed to continuously draw 350A+ while the GX16 can continuously draw 600A."