How to run more current through fuse?

[ErikTheNorwegian]

Interesting, this tread reminded me off some experiences i had years ago. The fuse in a device blew,  Even the curent was under the limit. After some tests, i found out, running hot, the fuse blew faster under vibration.  So hanging the fuse holder into a shock absorbert case solved the problem and we could enjoy a couple of hundred milliamps more before it blew.
http://m.rs-online.com/h5/mobile/uk/catalog?url=%2Fweb%2Fp%2Frail-mount-fuse-holders%2F0354385A%2F

[Rerouter]

In the rules, the holder part number is specified,

http://www.digikey.com/product-detail/en/04980900ZXT/F3094-ND/553455

But that in its own looks like it should be very easy to add a thermocouple that is in contact via mounting pressure, equally having a low gauge wire coming away from the fuse block to the kill switch would increase the thermal mass of the fuse allowing you just a little more, without obviously doing everything you can to abuse the fuse, 

[ErikTheNorwegian]

Mount the fuse on a rubber damping plate...or add anti vibration ontoo the fuse holder.

http://www.antivibrationmagic.ca/index.php?page=info#Fuses

[Ian.M]

The fuse rules state: "We will be using the LittelFuse MIDI 498 series of fuses. The fuse holder they use is Digi-Key P/N F3094-ND".   It does NOT actually prescribe the use of a F3094-ND holder.  Even if the use of that holder is enforced,  if your fuse cooling scheme is sufficiently outrageous (but safe) the "7.2.3 | The “I Ain’t Even Mad” Clause" is likely to get you out of trouble.  Cryogenics or discharging volatile fluids would be likely to raise safety concerns.

I'd  have a couple of options that can easily be swapped in - A forced air cooled  heatsinked custom fuse holder, a F3094-ND holder with provision to immerse it in a bath of Fluorinert under partial vacuum (pump it down after each fuse inspection or change) to implement heatpipe cooling, and a F3094-ND holder with heavy cables and good airflow.

I suspect that the fuse temperature can be sensed to sufficient accuracy by measuring its resistance.  i.e. accurately measure the voltage drop across it and the current through it.  If the controller adaptively manages the fuse temperature, even if the judges force you to swap to an unmodified F3094-ND holder, you'll still be better off than any competitor without active fuse temperature management.

[nbritton]

Also I would implement a negative feedback current limiting circuit using an arduino that computes an exponentially weighted moving average of the current going through the fuse, that way you can always be at the maximum temperature rating of the fuse.

On second thought, a feedback current limiting circuit using a temperature probe on the fuse would be easier to implement. To set the device you would drive around a test track and using a potentiometer slowly inch up the value until you blow the fuse, then back off the potentiometer and replace the fuse. This has everything to do with heat, the fuse blows not because too much current is flowing through it, but because the current is causing the metal in the fuse to melt.

Furthermore, the rules do not state that you have to use the F3094-ND fuse holder, they simply recommend that part number for the people who don't know what fuse holder to get. I further recommend making a fuse holder from a block of copper or aluminum, so that the holder can act as a heat sink. Furthermore, the rules do not state that you can't use a fan to actively cool the fuse. In fact, you could even liquid cool the fuse by submersing it in water, plain old water has very high thermal capacity. A block of dry ice might work.

I'm assuming you're using DC motors, but if you used AC motors you could go with three phase power. The rules don't say that you can't use a three phase motor. I wonder if they would give you one fuse for each phase or just fuse one of the three phases? Ether-way you will have three time the power.

In the rule book it says motors must be 12-36 volts RMS maximum, so you could boost the 36V up to 144V through the fuse and then buck it back down 36V at the motor. 1440W / 36V = 40A veruse 1440W / 144V = 10A

[Zero999]

The rules clearly state "No vehicle may have a total peak power consumption of greater than 1440 watts." which implies that this figure can't even be exceeded during acceleration so cooling the fuse to stop it blowing won't be any good, if they disqualify you because you're exceeding the power limit.

Reducing weight and good aerodynamics would help but the rules state "The vehicle may not exceed 62” long by 36” wide." which will make that difficult to achieve, unless you get a dwarf driver.

[mikerj]

The rules clearly state "No vehicle may have a total peak power consumption of greater than 1440 watts." which implies that this figure can't even be exceeded during acceleration so cooling the fuse to stop it blowing won't be any good, if they disqualify you because you're exceeding the power limit.

Depends at what point this power consumption is measured.  If it's at the primary energy storage device, then you are free to draw 1440 watts from it all day long (or for as long as it lasts) but downstream you could have secondary energy storage that permits higher peak power to be delivered to the motor for short periods.  Capacitor banks are apparently out, but NiCd could be usable, or a kinetic storage device.

Basic liquid cooling with e.g. deionised water would probably be sufficient to raise the current handling of the fuse by a fair bit, i.e. immerse the entire fuse and connections in the coolant.

[T3sl4co1l]

On second thought, a feedback current limiting circuit using a temperature probe on the fuse would be easier to implement. To set the device you would drive around a test track and using a potentiometer slowly inch up the value until you blow the fuse, then back off the potentiometer and replace the fuse. This has everything to do with heat, the fuse blows not because too much current is flowing through it, but because the current is causing the metal in the fuse to melt.

This is potentially as good as the [open loop] model, but is still missing things: the ambient near the fuse (or even on the fuse body, or at the terminals) is not the fusible element's temperature.

Ideally, one would combine both for a very accurate system.  A thermal model of the element-to-thermometer path is what's missing here.

Perhaps, the thermometer-to-ambient path still remains in the model, but rather than being purely open loop (i.e., where current is used to calculate element temp, and body temp, down to ambient), that node's state is 'informed' by real measurements: its value is the sum of external (as measured, but perhaps noisy or systematically inaccurate) input, and of the estimated (modeled) value.

This is something like a Kalman filter, implemented in analog (or DSP just as well, but in such simple terms that either would do).  A full expression of that would require a few more stages (including accounting for the confidence level of the measured data; which could be done in analog terms with a 'fading' circuit controlled by some measure of noise, perhaps).  Even without the full functionality, it's an interesting method.

Regarding peak power: without a logging instrument thrown on each battery, it's a preposterous claim!  One supposes the fuse acts as their monitor; so in that case, "peak" means over a period of [thermal time constant], which is hardly the electrical meaning of 'peak'!  And, in that case, gaming the fuse is not only desired, but encouraged, so it would seem!

Tim

[Rerouter]

I'm still going to have to say that actively monitoring the temperature of the fuse to allow yourself a higher peak current would fit quite nicely under the "I Aint even mad" exception, he is not giving himself much of an advantage, just stretching it to go as far as he can,

obviously he exceededs that mark and he will blow a fuse, using the recommended fuse holder may win him some benefit of the doubt with the judges as he will not be trying to bypass the protection of the fuse, just run the knife edge,

[NotABagel]

I'd concentrate on getting as much power from that 1440W into the wheels by keeping the controller and motor in their optimum efficiency at racing speeds. Maybe look at aerodynamics as well to streamline the body.

That leads into a second question: What motor should I get to maximize efficiency?  I was looking at converting alternators to brushless motors, as several teams had success with that, but I don't really know how efficient that will be.

Aerodynamics isn't really important for the race.  The track is very twisty, so you won't get up to very high speeds, and since you have to use a body from a power wheels car, you can't really play with aerodynamics much.

[Psi]

If you make a thermal model of the fuse (the time curves are provided in the datasheet), you can estimate the internal temperature of it, and control current flow so as to stay just below the melting threshold.

Tim

Have two fuses of the same type in series with the load.
Fuse 1 is the official unmodified fuse for the rules.
Fuse 2 is the same part number but has been modified to have a temp probe mounted hard up against the fuse wire.

You can then monitor the temp of fuse2 in real time and keep under the blow temp.

[NotABagel]

No vehicle may have a total peak power consumption of greater than 1440 watts.

So yes, you want to cheat.
[/quote]

I talked to one of the race organizers, who told me that if you can run more current through, you have earned the right to run more current through.  Also, everyone is pretty open about the fuse trickery.  For these reasons. I consider messing with the fuse to be aboveboard.

[NotABagel]

The dithering stuff is another way you could abuse it a little, using PWM to get an average / RMS voltage of say 36V, but actually be supplying peaks of 48V (maximum unlicensed voltage), with trails of 24V to get higher torque,

How do I realistically do this?  I have no electronic engineering knowlege, but I do have a teacher with an electronic engineering degree.  How much work would this entail?

[NotABagel]

Stick a boost converter inside the battery module. Have it start at 36V and boost it as high as maybe 200V if the inverter calls for it. Basically a scaled down version of what's done in modern Toyota hybrids.

How could I convert this back to DC for powering the motor?

[Rerouter]

I've rethought it and you can avoid the 24V stage, just relying on the bulk 48V PWM'd to get your torque,

For your controller you would have a standard PWM h bridge controller controlling the motor, however you can add a boost converter to the input of this stage, the stored energy would be low so you could have it respond fairly quickly

At low speed or while accelerating, or even when someone presses a boost button, You ramp up the voltage, scale back the h-bridge ratio (To keep within there voltage rules) and gain more torque for the same input power, then when cruising switch off the boost converter and the freewheeling diode will allow the power to flow through to the motors at the normal battery voltage, (yes there will be a small efficiency loss)

Tie this in with a thermal fold back from measuring the fuse temperature and / or resistance and you should have a nice number of gains that abide by there rules from how you have so far described it, you can make it as simple as that an op amp proportionally scales back the pwm ratio as the temperature of the fuse exceeds X degrees, it prevents the vehicle from going full limp, and prevents you from ever blowing the fuse outside of a genuine fault condition.

This however would make electronic breaking harder to recharge from, but can still be done if neccesary,

[Kalvin]

After watching the promotional movie on the site's front page, I would also pay attention to driving skills. Although you had a powerful car but you couldn't drive it efficiently you can still lose to less powerful car with a better driver and with less total weight. Then, I would look at the mechanical parts too. Maybe two separate motors, one for each back wheels instead of one (Frankly, I do not know how the drivetrain is implemented on these small cars) so that the wheels can get optimal torque at the curves. If you can select the tyres, select something that has as little rolling friction as possible and use high tyre pressure. Check the bearings. Think about the whole instead of concentrating on the fuse and maximum current only. Nice and very interesting project! 

[Fungus]

For those who haven't read them yet (including the OP, apparently) here are the rules: http://www.powerracingseries.org/rules

(please read before continuing)

Section "2.7.1 Fuses" says:

No vehicle may have a total peak power consumption of greater than 1440 watts.

...

Here are the fuses that the race requires.

    36v = 40A Fuse (Digikey #F3089-ND)
    28.8v = 50A Fuse (Digikey #F3090-ND)
    24v = 60A Fuse (Digikey #F3091-ND)
    12v = 125A Fuse (Digikey #F1879-ND)

36V x 40A = 1440W
28.8V x 50A = 1440W
24V x 60A = 1440W
12V x 125A = 1500W

If the peak power consumption isn't allowed to be more than 1440W than how will any sort of fuse fiddling help? The supplied fuses can already handle that.

So mostly you're limited to:
a) Building a lightweight car and finding the smallest driver possible.  (VERY important)
b) Learning to drive well (even more important than 'a')

...last of all, the electronics:
c) You can make sure you're drawing 1440W ALL the time, ie. you need to be drawing 1440W even when you're braking and storing it somewhere ready for a power boost coming out of the curves. Rule 2.5 says "no capacitor banks" so this will be difficult.

You could maybe spin up a flywheel with a motor during the 'storage' moments then use the same motor as a generator during the 'boost' moments. There's a lot of electronics knowledge needed to do that properly though.

In short, electronics 'cheating' is going to be difficult.

And: You can earn just as many points by building funny cars and dressing up, so... 

rules from how you have so far described it

Hint: Try reading the actual rules. Messing with the fuse won't help because Watts are also limited.

(though I'm not sure how they can monitor the Watts when you're out on the track)

After watching the promotional movie on the site's front page, I would also pay attention to driving skills. Although you had a powerful car but you couldn't drive it efficiently you can still lose to less powerful car with a better driver and with less total weight.

This.

[Kalvin]

How about the different fuse types 125A vs 40A. I have a hunch that the 125A is more tolerant to overloading and abuse than the 40A fuse.

[Fungus]

How about the different fuse types 125A vs 40A. I have a hunch that the 125A is more tolerant to overloading and abuse than the 40A fuse.

Maybe, but then you have to work at 12V which makes everything else worse.

Me? I'd go for the higher voltages.

[Rerouter]

For the motors, I have done some reading, and alternator motors get most of there power at higher RPM's, so if your going slow through a weaving course, using hobby motors may get you an advantage,

Edit: Or much better, see if you can find brushless cordless drills that have dud batteries or cosmetic damage, 2 of those motors would give you the best of all worlds, or even just see if they sell the motors as a spare part,

[Kalvin]

Can you use some sort of fuel cell onboard? If you run the battery at the maximum current 100% of the time and use the unused power to convert water into hydrogen and feed the hydrogen into a fuel cell to give you some extra boost.

[T3sl4co1l]

Can you use some sort of fuel cell onboard? If you run the battery at the maximum current 100% of the time and use the unused power to convert water into hydrogen and feed the hydrogen into a fuel cell to give you some extra boost.

I don't even know where to begin telling you what's wrong with this.

Unless it was intentional.  "Sufficiently developed satire is indistinguishable from the thing it aims to criticize."

Tim

[NotABagel]

(2) Store power down stream of the fuse, ( supercaps ) so, you can draw a bit more, when you need it...  This woudl let you run the system at an average of 100%,   ( ie, when you throttle off, your caps charge up )...

Yes, that was exactly my first thought. 
Exactly WHERE in the circuit do they require the fuse?

Unfortunately, the rules expressly prohibit capacitor banks.  It's too dangerous for even Power Wheels Racing.

[DanielS]

I'm guessing the "no capacitor" rule means no major energy storage devices of any sort allowed downstream from the fuse, electrical, chemical, thermal or electromechanical. Most of these would likely get ruled out as either too dangerous or self-out themselves as simply too bulky to compete.

Since fuses and batteries are not exact devices, the competition should be using calibrated and sealed power metering devices to enforce the 1440W (5s rolling average) limit.

[NotABagel]

For the motors, I have done some reading, and alternator motors get most of there power at higher RPM's, so if your going slow through a weaving course, using hobby motors may get you an advantage,

Edit: Or much better, see if you can find brushless cordless drills that have dud batteries or cosmetic damage, 2 of those motors would give you the best of all worlds, or even just see if they sell the motors as a spare part,

Well, my initial plan was to have two drill motors that power the kart through chain drive.  Unfortunately, the Ryobi drill motors I used simply didn't have enough startup torque to get the kart moving.  If I were to actually use drill motors again, I'd have to use a big reduction gearbox, as the lowest gear for the drill wasn't low enough.