Shunt mod worth while?

Totally worth it if you aren't stretching the distance. My commute is 3+ miles round trip and my son & I together weigh about 250. The shunt mod was totally worth it. I freely admit thought that I didn't do the mod myself and I just got a fresh set of batteries 2 days ago and am ready to ride again once my local temps get back under 90°. All the instructions are here: http://www.bergerweb.net/xb500/

I understand your trepidation, I paused while considering this mod myself. I ultimately chose to purchase an extra controller and modify it, and keep the original controller as a spare. The extra controller was about $65 at the time.

As for whether it is worth it, I think it is very worth it if you have any even minor grades to climb. Mine (unmodified) would climb a minor grade at about 10-15 mph, after shunt mod it would climb almost 20 mph. After the extra battery was installed, it now climbs the same grade over 20 mph, and is a lot more fun to ride. On the flat, it will go almost 30 mph.

I would say it is definitely worth it. Of course you take a risk in modifying your scooter, and that risk is all yours.

Good luck with your choice,

Tom

My weakest point now is the 12ah SLA batteries but I'll run these until they need replacing.

You can buy 12V, 14 AH batteries at Battery Mart. They are about $7.00 more than the 12AH batteries, and the difference is worth it, especially in battery life if you draw more current.

Did the 60v mod yesterday. All is good so far. Top speed is a lot higher (haven't clocked it yet) but struggles a bit to "max out" unless on very level ground. I tapped in my 5th battery just before the controller. This way the DC-DC convertor is still only getting 48v.
My weakest point now is the 12ah SLA batteries but I'll run these until they need replacing.

Can you post a picture or draw a diagram of where you tapped in? Dickey_b, what do you think of this?

Really simple hook-up. I didn't want to hack into my bike wiring yet and had already had to splice the proper connector onto my new controller so that's where I opened the circuit and added extensions. I used the positive wire (could be done either way) and connected my 5th battery between the plug and controller. Plug side went to the negative battery terminal and controller side to the positive terminal. DC/DC convertor gets it's power between the 4th and 5th battery which still measures 48v.

That's complete gibbrish to me. I'm a visual person. If I can see it, I can match it. :)

Now THAT makes sense to me! Thanks!!

Did the 60v mod yesterday. All is good so far. Top speed is a lot higher (haven't clocked it yet) but struggles a bit to "max out" unless on very level ground. I tapped in my 5th battery just before the controller. This way the DC-DC convertor is still only getting 48v.
My weakest point now is the 12ah SLA batteries but I'll run these until they need replacing.

Can you post a picture or draw a diagram of where you tapped in? Dickey_b, what do you think of this?

Well, AFB (aztecfembone), 68snewpy's method is great as a way of protecting the 12V converter from overvoltage, but isn't practical if you are retaining your old controller. Since he is using a different controller anyway, he had to re-wire his power switch in a different way, so splicing the battery in where he did made sense. But on an XB500 with a normal controller, the power for the 12V supply comes from inside the controller, so you wouldn't be able to furnish 60 volts to the controller without furnishing the same voltage to the 12V supply. The upside is that he can use the regular charger socket to charge his normal battery, and a 12V charger to charge his aux battery, without disconnecting anything. The 60V mod I describe on my website requires that I charge the regular battery by plugging my charger directly into the battery box, which means eventually I'll wear out the connectors and have to replace them. I'm wondering, though, where his battery is located? If it's back under the seat, then he had to run fairly long wires back from the front shroud area to the underseat compartment, and that would introduce voltage drop unless he used #10 wire.

Finally... many people on this forum have gone to 60 volts and had no trouble with the DC-DC 12V supply. I opened mine up, and all the components are rated at 64 Volts or greater. Bear in mind, though, that fully charged batteries could read as high as 66 volts.

My newest solution for more speed is to replace the controller with one that will drive the bike at 30 mph using a 48 volt battery pack. This winter, my project is to install a controller from a "Ride Green Capri" bike on my XB500. The Capri controller limits top speed to 30mph, using a 48V battery pack. My only problem is that I have to change the way my brake levers work if i'm going to use dynamic braking.
Dickey_b
Waste Not, Want Not

Of course it took many many many tries to find the correct wire sequence for the motor and hall sensors

Did you have to change the wiring sequence to both the motor power AND the hall sensors? I ask, because I have been trying to wire a different controller into my XB500. My first attempt was to use the same codes (Green -> Green, Yellow -> Yellow, Blue -> Blue) for the motor power, and also color to color for the hall sensors. That connection did not work... the motor would not move, and the current went pretty high. I re-arranged the power wiring to the motor, and got two conditions... Normal current, but the motor ran backwards, and another condition where the motor would sometimes not run, but when it did, it made an awful noise starting out, then would run very fast forward, but again, very high current. I guess my next step is to try re-arranging the hall sensor wiring. Any hints? What was your experience?

It took a lot of trial and error. Most combinations do as you descibe (bad noises, backwards, high current draw, no response, etc.). I thought at least the color combination would be the same between the motor leads and hall sensor but NO! I'd have to remove the seat to check the combo that worked for me but I could if you want.

The spreadsheet sheds a lot of lignt on things. I haven't tried it yet, but it does point out some basic facts... The hall sensor connector has 5 wires, two of which are power and ground, and the remaining three must be the signal wires, (yellow, green, blue). Since there are 6 possible combinations of connections between the motor and the controller drivers, and 6 possible connections to the hall sensors, there will be 36 total combinations. I'm about to embark on running through these to see which one works for me. I doubt if knowing your combination would do me any good, since you probably don't have the same controller as me, but thanks for the offer anyway. At least now I have hope for a solution, and a way of getting 30mph with only 48 volts.

Hello Richard.
Seems you are changing controllers and going back to 48V.

Are you staying with the original motor?

Let us know how it works out.

Regards.

Thom

Hello Richard.
Seems you are changing controllers and going back to 48V.

Are you staying with the original motor?

Yes, that's right, ttunes.
I need a top speed of 30 MPH for my bike to keep from being irritating to the SUV drivers in my town. I've gone from 48V to 60V, but that only upped my level road speed (accurately measured) from 18.7 mph to 23.4 mph. I began shopping around for bikes that will go 30 mph, and still qualify as a moped under Michigan state law. I found that the "Ride Green Capri" and the MotoFino MFE13 both fit that bill, and use a 48V battery pack. You can look over both their specs here:
http://travelectro.com/RideGreen_Capri.html
http://www.ecobikeusa.com/Motofino_Electric_Scooter.html
My wife won't let me just go out and buy another electric bike after having this one for less than 6 months. (and that seems reasonable to me), so I decided instead of buying a whole bike, I would buy just the controller for the Motofino, and wire it into my XB500.

I currently have my bike on a stand in the basement with the front shroud removed for easy access to the controller. I have substituted the Motofino controller into the bike as much as I can, and have found the correct arrangement of motor drive wires and hall sensor wires to make it run smoothly in the right direction. My next task is to rig up a method of providing the right polarity signal to the controller to activate the dynamic motor braking. I will probably use a photo-transistor for this purpose, since that seems to be the cheapest route. (The XB500 requires the brake switch to furnish a ground to the controller to shut down the motor and enable dynamic braking, while the Motofino controller requires a +5v signal to do the same thing.)

I have a small bicycle dynamometer which I will use to determine the difference in speed between the XB500 controller and the Motofino controller. If everything goes as expected, I will probably get a top speed of about 27mph with the Motofino controller. I already know that the XB500 maxes out at 18.7mph with a 48V battery pack.

When I have finally made all the modifications I need and installed the Motofino controller, I will probably up my battery voltage again so that I get the full legal speed allowed in Michigan. That could mean adding only a 6 volt battery, which will be somewhat smaller than the 12V, 14amp-hour battery that I had in my under-seat compartment this past Fall. Upping the voltage by only 6 volts means that when my batteries are fully charged, I will be furnishing 61.2 volts to my controller, which does not exceed the 64Volt rating of the capacitors in the controller. By the way, the capacitors in the XB500 controller are also rated at 64 volts, so those of you who add a 12 volt battery to your pack are actually pushing past the limit when you have a fully charged battery. (each cell, when fully charged is 13.6 volts X 5 = 68 volts)

Expect to see my web-site updated when this winter project is complete.
http://www.bergerweb.net/xb500

You are a thinker that is for sure.
I was just going through my wish list of upgrades since my wife won't let me go out and toss down more dollars. Funny how they are .... LOL but I have to admit that this is a pretty pricey toy.
Lighting will come in under a $100 if I redo all 48V incandescents to LED with dropping resitors -- call that sometime when all the lights go kaput. And I should get a bonus of less current draw.
Installing metering for V and I will be about $22. A no brainer so I really have a better handle on the basic operations.
For uppin the speed I have had a "wild haired" idea. Since these scoots seem to operate on V/RPM I got to looking and thinking about starving the feedback from the motor hall effects. Basically I'm wonderign if I do a divide by 2 on the hall efects will the controller think it needs to increase the pulse width to the motor? Right now I do about 430 RPM for 22MPH. If I can increase that to about 586RPM I should be hitting close to 30MPH. I need a scope to verify and begin to qualify this idea. I can breadboard a circuit for what --less than $3-4 bucks? Guess I'll have to bug some friends from work to borrow a scope since the company took mine away when I went into management years ago....dang it.

so anyway -- I saw your post and was a bit more than curious. You do good work and are kind to docuement your findings on your web site.

take care.
Thom

Thom,

The Hall effect sensors don't really monitor RPM. They sense the magnets in the motor so that the controller knows when to switch on the FET's for that phase of the motor. I think that dividing by 2 would only give you a stuttering motor, with less speed and power. Now moving the sensors a few degrees in the motor might allow you some more speed, but at a cost of torque. Kind of like the timing advance in a ICE.

Hi James thanks for the thoughts. You may be right about the stuttering. I realize the hall effects trigger the phase wiring. Agreed - we are in sync there.

But what is limiting the speed? We all know these are limited to about 22-23MPH by design.

So my thought -- which may prove to be useless -- is that the controller is receving feedback to limit the pulse width to drive the motor. Again I may be wrong. Way wrong.

I know of only two things that are providing feedback or input to the controller-- the throttle and the hall effects inside the motor.
The throttle is an interesting theory becasue it may be driving an ADC that has been preset for a maximum. So no matter how much we twist the throtle we will only get so much of a pulse width to drive the motor. We have seen the blue wire that is disconnected in the states - the one that acts like a governor. so to me this is indeed very interesting. Thanks for making me think this through a little further.

The other feedback is from hall effects inside the motor. The idea of dividing by 2 implies reduced feedback which in turn could widen the pulse width. Concern for back EMF is present.

So anyway -- with winter here I've got a few months to try and figure this out. Plus I need a scope to see what is really going on.

Later.
Thom

Hi James thanks for the thoughts. You may be right about the stuttering. I realize the hall effects trigger the phase wiring. Agreed - we are in sync there.

But what is limiting the speed?....

I know of only two things that are providing feedback or input to the controller-- the throttle and the hall effects inside the motor.

Well, let's see now...
First of all, I have traced the wiring on my xb-500 controller just a little bit, and found two things.
1. The throttle input goes through a 10K resistor to an Analog-to-Digital pin on the microcontroller chip, while the three hall sensor inputs go to digital inputs on the microcontroller.
2. The throttle position determines the desired speed, regardless of how much power is required to get to that speed. In other words. If I set my throttle to the half way position, my bike will try to go about 10 mph. I have determined this by holding the throttle in one position, and driving over various terrain. Holding the throttle constant, if I'm on level ground, I will go about 10 mph with a current of about 4 amps. If I begin ascending a very slight grade, my speed will remain at about 10 mph, but my current will increase. If I begin decending a slight grade, my speed will remain at 10 mph, but my current will decrease. If the grade that I'm decending is steep enough, my speed may increase to more than 10 mph, but the motor current will decrease to zero.

From the above two facts, I have concluded that the speed of the bike is calculated by the microprocessor, based on the frequency of the inputs from the hall sensors. Then the processor determines what speed is desired, and as that speed is approached, the pulse width to the motor is decreased. Therefore, the motor assumes that for every given desired speed, there is a corresponding pulse width that will maintain that speed on level ground.

Several things can be done to fool the controller and make the bike go faster.
1. Increasing the supply voltage will increase the power sent to the motor at any given desired speed, but the controller doesn't realize that, since it assumes a particular battery pack size.
2. Increasing the circumference of the rear wheel will obviously result in higher speed, but less torque.
3. If somehow we could make the controller think the bike speed is less than it actually is, we would be able to drive the bike faster.

My next experiment is to try upping the voltage supplied to the throttle. The three wires that connect the throttle to the controller are Ground, +5 Volts, and Throttle Signal. The +5Volts is supplied by the controller, which has an internal regulator. The signal coming from the throttle to the controller varies from 0.7volts to 4.3 volts, depending on the position of the throttle.
If I supply the throttle with 6 volts, instead of 5 volts, will the controller give me a higher speed? I'm about to find out, because I'm going to put a 1.5volt battery in series with the red wire that goes to the controller, thus furnishing it with 6.5 volts, and see if my motor runs faster at maximum throttle. I'll report back when I have finished this experiment.

OK, I tried the extra voltage to the throttle. When I did that, the bike (on a dynamometer) reached the same max speed as without the extra voltage to the throttle, but id did it when the throttle was at a lower setting. Twisting the throttle past that setting did not increase the speed any more. Oh Well..... It was worth a try, don't you think?

Definetly worth the effort. Thank you.
I am wondering if you are seeing the result of the ADC being clamped to a predetirmined level. hmmmm something else to look at.
Also on the hmmm side is your observation that the microprocessor seems to calculate a given speed based on the inputs from hall effect sensors. Which continues to lead me to believe if we starve the feedback we may get the desired operation. Essentially reducing the input from the hall effect should be instruction for the processor to increase the pulse width? And again -- more to look at.
I'm shopping ebay for a cheap scope.

Thom

I don't believe that the voltage is clamped to a level, or we wouldn't see a speed increase with added voltage, going from 48 to 60 v. My guess, and it is only that, is the limiting factor is the frequency of the pulses from the controller. The motor running in Synchronous mode is in step with the pulses that are timed by the Hall sensors. Increase the frequency of the pulses and the motor has no choice but to run faster. Increasing the voltage allows more power per pulse, but I think the real answer to a speed increase is in allowing a higher frequency in the controller. As the speed increases the pulse width would have to be reduced due to RPM. resulting in a loss of torque but an increase of speed. That's assuming we don't try to get more current into the shorter pulse.

Now that should be clear as mud, right?

Definetly worth the effort. Thank you.
I am wondering if you are seeing the result of the ADC being clamped to a predetirmined level. hmmmm something else to look at.
Also on the hmmm side is your observation that the microprocessor seems to calculate a given speed based on the inputs from hall effect sensors. Which continues to lead me to believe if we starve the feedback we may get the desired operation. Essentially reducing the input from the hall effect should be instruction for the processor to increase the pulse width? And again -- more to look at.

Well, after tracing some more on the controller circuit board, I find that the supply voltage to the microcontroller chip is +5 volts. which leads me to believe that regardless of how much voltage is supplied to the throttle control, the upper level input to the microcontroller will be clamped to something less than 5 volts. (saturation) In other words, the micorprocessor's a/d converter can't measure a voltage higher than it's own supply voltage. There is a 10K resistor in series with the input pin, probably to avoid burning out the chip in case someone like me tries just this kind of experimental trick.

I don't believe that "starviing" the feedback would do anything for us. The hall sensors are on when they are adjascent to the north pole of the magnets, and off when they are adjascent to the south poles of the magnets. They furnish a logic signal, and making that signal smaller would only make it harder for the processor to determine which of the two values is there.

The hall sensors serve two purposes.
1. They tell the microprocessor the position of the magnets relative to the position of the motor windings. For any given position of the wheel, one feedback wire will be 5 volts, one will be 0.5 volts, and one will be about ready to change from one level to the other. The signals on the sensor wires are all square waves when the wheel is turning, and the edges of the signals are 120 degrees from each other. There will be 24 positive and 24 negative transitions on each sensor for each rotation of the hub motor. (on an XB500).
2. the frequency of the signal coming from the sensors tells the microprocessor the speed of the motor. We could trick the microprocessor into thinking the motor is running half as fast by dividing the hall sensor frequency in half, but then we would also be giving it a signal that does not correspond to the correct position of the magnetic poles.

What we really need, then, is a motor which will go faster for each phase reversal of the hall sensors, or to be able to re-program the microprocessor.

That's my theory, and I'm sticking with it!

Brushless motor Fans,

I do not have a brushless motor but I am very interested in them. I have been following this thread with interest. From what you have all said I and using my electronics background of 35 years I have some ideas which may help answer some of your queries. So far we know that applying more voltage to the throttle input does not increase the speed of the motor. We also know that increasing the voltage to the controller thereby increasing the voltage to the motor does increase speed. Richard also makes a good point that if you change the pulses from the hall sensors you will be giving the microprocessor incorrect information about the rotational position of the motor so that is a bad idea. Lets try something new. Tell me when you lift the drive wheel off the ground and apply full throttle does the motor turn faster? I'll bet it does, even a brushed motor does this. If you go down a hill does it turn faster? Again I'll bet it does. I submit the one factor you have all been over looking is that no matter what type of electric motor you are using brushed or brushless given a fixed voltage with a fixed load it can only turn a fixed speed. I think the only way you can make these motors turn faster is to add voltage. The controller is already design to output the max voltage at full throttle. This has already be demonstrated by the fact that when more voltage is applied to the controller the motor receives more voltage and spins faster. If the controller was limiting the voltage or rpm of the motor then adding voltage to the input would not increase the speed of the motor. The microprocessor would still keep the motor from spinning any faster. I have a bike which had a controller with a fixed top speed. When you reached 16 MPH the controller would removed power from the motor. When I added more voltage to this controller all it did was accelerate faster. When 16 MPH was reached the controller still cut power to the motor. I changed controllers and eliminated this limitation, now the same motor on the same bike with the same rider weight using the same voltage can reach 23 MPH. I hope you understand this point. You have proven the microprocessor does not limit the speed by increasing the voltage so the real answer is to get a controller which can be used with a higher voltage and go for it.

Grandpa Chas S.

Brushless motor Fans,

Tell me when you lift the drive wheel off the ground and apply full throttle does the motor turn faster? I'll bet it does, even a brushed motor does this.

My bike currently is sitting in my basement on a bicycle dynamometer. I can apply a load of from about 5 Watts (minimum) to a little over 140 Watts by adjusting the dynamometer load. The dyno has a MPH reading, which I don't know if it's accurate, but can be used for relative measurements.
With minimum load on the dynamometer (5 watts), and a fully charged battery (52 volts), the bike will go 19.6 mph, and the battery pack current is 2.4 amps.
With a 50 watt load, the speed drops down to 19 mph, and the battery pack current is 3.6 amps.
With a 100 watt load, the speed is 18.6 mph, and the the battery pack current is 5.1 amps
All these measurements were taken at full throttle. So the answer to chas_stevenson's question is "Yes!" but please note that the controller is trying to compensate for the added load, so it is obviously using some sort of feedback to furnish more power to the motor when there is a higher load. I have also measured my motor current going down a hill, and found that when the hill is steep enough to allow me to go over 21 mph (actual), the current to my controller drops to zero. (well, maybe not zero, but less than 0.1 Amp, which is the lowest I can read with my 100.0 Amp meter.)

Now here is one additional fact. The XB500 motor and controller have specifications of max speed = 20mph.
The Motofino bike, which also has a 500W motor has a max speed of 30mph. so you can probably expect it to go about 26-28mph.
I can swap between an XB500 controller, and the Motofino controller, (same motor, different controller) and the max speed on the dynamometer will remain virtually the same.