On the 20k setting on the multimeter the ground trimpot reads 1.29, the blue 1.19 and the throttle brown ( wiper) and blue is 2.96 when throttle is closed and zero when throttle is full open, I dont understand the relationship between these.teklektik wrote:What are the final resistances of the trimpots after adjustment and the resistance of the Magura itself between the wiper and 5v supply lead (ignore the Gnd connection)?
Thanks. I was looking to get a feel for the actual trimpot values and the effect of clocking the gears. Here's what we have:Denisesewa wrote:On the 20k setting on the multimeter the ground trimpot reads 1.29, the blue 1.19 and the throttle brown ( wiper) and blue is 2.96 when throttle is closed and zero when throttle is full open, I dont understand the relationship between these.
So I did some testing with the magura and have determined that the controller needs 1.25 volts to make the motor start turning and above 4 volts the controller must be going into an overvoltage safety mode which shuts things down, of course I can get reach these numbers using resistors BUT the controller also needs to see about 1 volt at closed throttle to activate, this leaves .20 to .25 volts of dead space inherent in the controller. I think I'll hook up the second stock controller which does not have a modified shunt and see if there is any differance although I cant see how the shunt would have anything to do with this, kinda hoping I just have a bad controller.teklektik wrote:Thanks. I was looking to get a feel for the actual trimpot values and the effect of clocking the gears. Here's what we have:Denisesewa wrote:On the 20k setting on the multimeter the ground trimpot reads 1.29, the blue 1.19 and the throttle brown ( wiper) and blue is 2.96 when throttle is closed and zero when throttle is full open, I dont understand the relationship between these.
Rhi = 1.19K (trimpot-hi)
Rlo = 1.29K (trimpot-lo)
Rc-t = 2.96K (resistance from clocked ZERO to WOT)
So (just winging it and guessing at your Rb-t and Vcc) the average Magura has an Rb-t of about 5.2K so your Rb-c = 5.2 - 2.96 ~= 2.2K. This puts the total resistance from Gnd to C at about 3.5K - which is pretty large making the controller Vmin (wheel begins to creep) probably 2v or more. This value is normally around 1v (hall throttles typically have a 1-4v output range - more or less) and so explains the ZERO throttle dead zone issue you have been having. The GNG throttles may not be top of the line, but they don't appear entirely to blame - it looks like you probably have to twist through 1/3 of a hall throttle before the GNG controller gets enough voltage to notice...
skyungjae has found a freewheel adapter for the GNG jackshaft (It has an 11mm diameter tip with a key). The stock jackshaft freewheel is weak and hard to find, and this $43 adapter allows the use of an off-the-shelf robust 30mm X 1.0 BMX freewheel. It changes the tooth-count from 12T to 13T, so thats slightly worse, but the BMX freewheels can take a pounding while also being easy and cheap to replace.While I had the jackshaft disassembled to fit the new adjustable side sheets, I put 2 new good quality SS6201 2RS bearings on the jackshaft, and rebuilt the 12T freewheel with new (1/8” 3.17mm) ball bearings and grease. It’s only a 2-pawl freewheel but there didn’t seem to have excessive wear on the ratchet teeth/pawls. Despite the fact that there was very little lubrication in the freewheel. I have only used the GNG kit for a few hundred miles, but I did note that some of the ball bearings showed signs of wear, with a grey colour. I suspect they are not hardened chrome steel, like the ones I have replaced them with. It all seems to run quieter and smoother now
Works! More detailed pictures on this post: https://endless-sphere.com/forums/viewto ... 15#p696663spinningmagnets wrote:skyungjae has found a freewheel adapter for the GNG jackshaft (It has an 11mm diameter tip with a key). The stock jackshaft freewheel is weak and hard to find, and this $43 adapter allows the use of an off-the-shelf robust 30mm X 1.0 BMX freewheel. It changes the tooth-count from 12T to 13T, so thats slightly worse, but the BMX freewheels can take a pounding while also being easy and cheap to replace.
the 12T FW has shallow pawl engagement to allow the lowest possible tooth-count (an uses an odd thread for mounting), but a 12T is weak and sometimes hard to find when your 12T breaks. 13T FWs are common, robust and affordable.
http://www.ebay.com/itm/Freewheel-Adapt ... 1024772949
tri-lobe does the same, but with 44T/52TI run a similar configuration on my GNG-Kona Stinky. (48T/48T) Can pedal normally at 70kph (43-MPH) and still climb in slowmotion
I am in the planning research stage of a build and think I am going to go with the 450 watt if I can get theses brackets and a custom crank. Heard of a lot of people breaking there cranks with the GNG kits.Infinion Controllers, this guy has the manufacturer custom order mod them for performance and durability.
Wide Voltage Range: As of July 2011, the regulator design on these controller circuits has been upgraded so that they will operate over full range of voltage with no need to open up and change onboard resistors. 19-60V with the 20A and 35A models, and 26-88V with the 25A and 40A models.
Better Mosfets: We supply the manufacturer with locally sourced mosfets that have significantly better characteristics than the stock models, which results in a controller that runs cooler and is more robust.
Reinforced Bus Bars: The power traces on the ciruitboard are reinforced with solid copper wire for better conductivity and less chance of burnt traces.
100V Capacitors: We have all of our controllers made with large 100V 105oC rated bus capacitors. These capacitors are better able to handle high ripple currents and don't need to be replaced if the controller is modified for 72V or higher operation.
Long Leads: The phase and battery cables are 120cm long, able to reach right to the motor and battery for easy disconnection of those items. They are also terminated with real (not knock off) Anderson Powerpoles.
Smaller Package: Because of the better mosfets, we can offer higher current capabilities in a smaller enclosure size.
On/Off Switch: These controllers have a rocker ON/OFF switch that shuts the controller off so that it draws no current. It eliminates the need for a high current switch inline with the battery pack, and it has much greater reliability than the latching 'push' style of button switch used by Crystalyte.
Full Speed Fwd/Rev Control: The direction of the hub motor can be reversed with a simple jumper on the 2 pin connector. This makes it compatible with all different styles of motors regardless of which way is their natural direction of rotation
Cycle Analyst Connector: Each controller has the 6-pin CA-DP connector for easy integration of the Cycle Analyst. We also include a printed label that has the measured RShunt value for that particular controller, so that calibration of the Cycle Analyst is a cinch.
Regenerative Braking: The controllers are configured for a maximum amount of regenerative braking when an ebrake lever is connected to the 4-pin connector, saving wear on brake pads and returning mechanical energy back to the battery pack
Denisesewa wrote:This is how I did my primary chain conversion using these parts >>the parts are from TNCscooters.com , # spk-103240 for the 65 tooth , #spk-106012 for the drive sprocket , and 2 ft of # chn-105250 chain , you will have to find another supplier for a #25 halflink , I used the original large belt pulley and trimmed the outer edge off , I then made a bushing to help center the sprocket on the pulley and kept adjust it untill I had virtualy no runout,
once I was satisfied I hotmelt glued the sprocket to the pully to keep it in place and then drilled and bolted the sprocket on,
I remover the outer " washer" from the motor drive pulley simply by bending it off with pliers and used a new file with the motor turning at a relativly slow speed to turn down the pulley to 10mm, by going slow and checking my work often I ended up with a nice snug fit for the new sprocket and again zero runout, as the shaft gets smaller material is removed very fast so check often. this took less than two hours of careful work.
CLICK ON THE IMAGES FOR VIDEO
The drive sprocket had to be tapped for set screws and two flats were filed on the shaft for the setscrews to bite against once I had correct chain alignment. To have minimal slack in the chain I used a half link instead of a master link.
The result is a smooth running primary drive.
the tensioner was made from an old dirt bike chain roller cut down ,the spring from the original secondary drive tensioner which isnt needed there and scraps of aluminum.
So far under hard use this has held up perfectly and is almost as quiet as the belt with the added bonus of eliminating the huge power consumption of the belt system, with " Lighningrod's adjustable sheets I hoping to do away with the tensioner and save even more power from being wasted.
The primary side only changes motor rpm, to change cadence ratio you need to change the chainring ratio ( differance between the tooth count on inside and outside chainrings), there are several people who have done this although I dont know where those threads are discussed other than the main GNG topic.Whiplash wrote:So you pedal slower? That seems perfect! Low speed climbing is what I'm after as well. Lots of goat trails and this is for my brothers to keep up with my mid drive MAC build. Thanks!