Awww, Les , you're making me think

[Morgen 3Eman]

Hi Les,

So I've been thinking about your comments in another thread about longer phase wires vs longer supply wires.  I don't think it matters a whole bunch when the controller, battery and motor are all mounted on the back of a bike.  But in my case, where the battery is nearly 6 feet away from the motor, there is likely to be a difference in system performance if the controller is mounted near the battery vs near the motor.

I think we both agree that all the high current wiring should be as large and as short as possible.  Your suggestion that the phase wires length is less important than the supply wires got me to thinking about L dI/dt.  Since the bldc controller is a high current pulse modulating system with very fast voltage rise times, the inductive reactance increases with the length of the wires.  So the shape of the current pulse will clearly be affected by increasing the length of any wire carrying the full pulse current.  But since there are some filter caps across the supply lines at the controller cut board, the effects of those pulses will be reduced on the supply lines.   Are the effects going to be terribly noticeable to the driver?  Probably not.  Are they going to be noticeable to the controller?  I suspect the change in wave shape will have some effects on the back EMF signal that some controllers use to determine switching points. At any rate, it is common design practice to keep high current, fast rise time wiring as short as physically possible, just because of the wave shaping caused by the inductance.  (and radiated interference effects)

I haven't gotten my head around your regen comments yet......You gotta stop making me think about this stuff from my youth.

TTFN,
Dennis

[skylinenitro]

I agree with you Dennis, that the phase wires should be kept as short as possible.  Since we are limited to the size of the phase wires coming out of the motor hub, the phase wires should be kept as short as possible and should also be the same length.  Upgrading the phase wires to as big as can fit through the hub is a good idea, as well as mounting the controller as close to the motor as possible. 

Shorting the length of the phase wires might save you something like .n ohms, but when your pushing 1000+ watts, that point number ohms translates to wasted heat in the wires and the controller.

As for the power cables side, run as big as you can from the batteries to the controller, something like 10 gauge, 8 gauge, or in my case, 4 gauge    The voltage loss of 6+ ft in 4 gauge will be unnoticeable.  If you still feel uncomfortable about power loss, put a big capacitor at the end of your power cable right before it goes into the controller.  Too bad they don't make 1 farad caps in 48v flavors like the kind they make for 12v car audio systems 

A one farad cap would also do wonders for regen breaking 

[Bikemad]

With wind generators, it's usually recommended to locate the rectifier/regulator unit close to the batteries and run AC voltage over the three phase wires for the majority of the cable run, as this results in less voltage drop, making it more efficient. 
I would therefore assume that on a three phase motor, short battery leads and long phase wires would be preferable to long battery leads and short phase wires.

With the MPIII it should be relatively easy to install heavier gauge phase wires. By fabricating a suitable 3mm thick metal disc to replace the controller heat sink, you could drill separate exit holes to accomodate each of the phase wires if necessary.
If you're not using a brake disk, you could even discard the cooling fan assembly and run huge 20mm diameter cables if you were determined to go completely over the top.
Although, I suspect this would probably double the weight of the vehicle, and it would be almost impossible to solder those massive cables to the ends of the comparatively tiny phase windings and controller connections.

Too bad they don't make 1 farad caps in 48v flavors like the kind they make for 12v car audio systems 

A one farad cap would also do wonders for regen breaking 

If you were able to connect four of those in series could probably make your own 48V 1 Farad Capacitor, but I'm not sure how the poor old switch contacts would cope. You get a big enough spark by connecting the battery to a controller with just a couple of 470uF capacitors, so just imagine what the spark would be like if you simultaneously connected 2,128 of them (~1 Farad) to a High C rated LiPo pack with no BMS to protect it.

From what I've read, a 1 farad capacitor would fully charge from empty in around 5 seconds with just 10Amps of regen current at 60V. If it was already at battery voltage, the actual charging time would take less than a second!
Apart from a big kick for the first second of applying the brakes, I don't suppose it would make any discernible difference when braking continually down a long steep hill.

Alan
 

[Morgen 3Eman]

Hi Alan,

The reason AC is used for most long distance power transmission is that it is easy to run 1000 volts at 1000 amps through a transformer, send 1,000,000 volts at 1 amp to its destination, and then transform it back to 1,000 amps at something less than 1000 volts  the resistive losses are thus lower  (IxIxR)   

However, my brother was maintenance manager for a 6 or 7 hundred thousand volt DC power plant in North Dakota.. The diode stack was in a separate building, and I never got to see it.  The power was sent about 600 miles to someplace in
Minnesota.   I think he said it was a few 10's of amps on the transmission line.  So it can be done, but transformers are much cheaper. I would love to have seen the SCR's that were used at the load to chop the DC into a transformer for the voltage reduction, but I never got to see them. 

I still think shorter phase leads would be a better choice. The phase/phase crosstalk would be another problem made worse by longer phase leads, but the major thing IO don't like is the imaginary power loss as the V/I  phase angle increases with more inductance.   

You are right, 470 microfarads isn't much filter, but at a few tens of kilohertz  they can perform a pretty good smoothing of the supply into the FET H switches.  (not at 30 Amps!) 

Damn, I'd love  to the see the charging current pulse a 1 Farad cap could cause.  The power cables would have to be secured to keep them from flying about as the magnetic fields formed!   


TTFN,
Dennis

[skylinenitro]

Whenever I used my 1 farad cap in my car (12v) I had to put a 55w headlight bulb in between the cap and the battery to charge them.  And that took about 10 secs max, which is good because that's all my fingers could take before they burned from the bulb heat   

However 4x 1farad caps in series to make 48v would only give you a 1/4 farad total capacitance (1/((1/R)*(1/R)*(1/R)))

(In a Scotty voice) "I need more power captain!"

[Bikemad]

However 4x 1farad caps in series to make 48v would only give you a 1/4 farad total capacitance (1/((1/R)*(1/R)*(1/R)))

In that case, you'd need 16 x 12V 1 Farad capacitors wired in a 4S4P series/parallel configuration to make your own 48V 1 Farad capacitor, and then you'd probably need a trailer to put them in.

Now I'm even more confused, if a 48V 1/4 farad capacitor is able to store four times the amount of energy (compared to a single 12V 1 Farad capacitor) then presumably a 48V 1Farad capacitor is able to store sixteen times as much.

Anyway, would it really be worth all the hassle, additional weight, inconvenience, and expense, just for a single second of regen assistance, or did I get that wrong too? 

Alan
 

[Morgen 3Eman]

What in the world does one do with a 1 FARAD cap?  This must involve an interesting story....

TTFN,
Dennis

[truly_bent]

Back in the day, we used to use large caps like that to smooth out "build yer own" 24VDC industrial power supplies. You'd start with a 29VAC transformer secondary, rectify it, and add a large dollop of capacitance. Dead simple and worked like a charm. The more capacitance, the less ripple.

Don't see that anymore, except on older equipment. Everybody is using switching power supplies now, but back when switching supplies first started being taken seriously, they used to allow voltage spikes to pass through and on up the line, which of course had a negative effect on control equipment upstream.

Up until just a few years ago I still preferred to order power supplies which used a transformer - they had less noise on the waveform and guaranteed isolation. Now I just order switching supplies like the rest of the world.

Old school - we take our time comin' around.

[Bikemad]

What in the world does one do with a 1 FARAD cap?  This must involve an interesting story....

Dennis, these capacitors are commonly used with high wattage car audio systems to help subwoofers produce a more defined base response, as the battery alone is not always able to provide the necessary high current bursts without a considerable voltage drop. Without the capacitor, the high current pulses can sometimes cause the car headlamps to pulse in time with the beat of the bass notes.

1 Farad is the just the minimum recommended size, as installations typically require at least 1 Farad per 1000 Watts of output.
So I guess you'll need one of these tiny little 80 Farad capacitors when you install the 80kW music system in your Merc:


I assume these high capacity devices must have some form of integrated electronic pre-charge system which is automatically activated when they are first powered up.

Sixteen of those connected in a series/parallel configuration should allow your Morgen to accelerate away from a standstill at maximum power without any noticeable voltage sag whatsoever.

I would strongly advise against shorting out the output terminals of an 80 Farad capacitor with a screwdriver shaft to demonstrate whether it's charged up or not.

Alan
 

[skylinenitro]

lol, my old electronics teacher told us a story when he was in basic electrical school, that there was a classmate who couldnt wrap his head around how caps work, so the other kids used to charge up a 100v some mf cap and throw it at the metal garbage can and watch them explode.  Then they would charge up more caps and say to the kid "hey! catch!!" which is shortly (pun intended) followed by him saying "OWWW" and shaking his hands   

And ohh, a dead short across a cap will kill your cap and explode melted shards of metal... the bigger the farads the bigger the explosion.  don't try that at home

[Morgen 3Eman]

80 FARADS!!!!!!  IN ONE CAP!!!!Smaller than a room!!!!

I'm stunned how out of touch I have become.  Thanks for the photographic addition to my knowledge. Amazing.

TTFN,
Dennis