GOLDENMOTOR CONTROLLER unlock without touching the shunt(Roughly translated from original web article by Jean Marc)
This has been a well-known problem for some time, the GoldenMotor controllers are restricted, and even if we program the maximum (30 Amps) you cannot get more than around 15 Amps. It seems that we actually get about half of the programmed value.
The most common solution to this problem is to modify the shunt. The shunt is a piece of resistive wire which produces a small voltage drop, which allows the current being drawn by the motor to be calculated (I = V / R) by the processor, which it then uses to regulate the power being consumed.
Here I have boosted my controller, before analysing its effect, by soldering a 10cm long piece of 0.7mm diameter wire in parallel with the original shunt:
I then managed to obtain 20 Amps instead of 15.
This modification worked perfectly, but it also has two risks:
A clear risk of decreasing the value of the shunt too much. I've seen pictures where the wire that had been added was a large copper wire or bracket component! Here the risk of exceeding the intensity bearable by the MOSFETs is great. But also the battery,
BMS, motor and fork dropouts will also suffer.
A hidden risk that only became apparent to me when I found the circuit diagram of the controller between the shunt and the microprocessor.
In fact, the voltage drop obtained from the shunt is handled by two different circuits:
Around U1-B we recognise a gain amplifier (R53 + R54) / R54 or 11. It provides a better resolution than if the signal was sent directly to the micro-processor.
But around U1-A we recognize a comparator whose output will go high when the current exceeds a threshold set by the resistors R1 and R2 in my circuit diagram. I did the maths; it's about 70 amps.
The chances are that this comparator can quickly cut the signal to the MOS in the case of a short circuit. (Based on electronic circuit breaker) From memory, it is matches precisely the 70A max current that the MOSFETs in the controller are rated for!
Here is the problem, if the shunt is modified - for example by reducing its resistance in half - we will also double the threshold protection! You wouldn't consider replacing a 16A household fuse with a 32A one! So this is why I decided to alter the gain of the amplifier, and by replacing R54 with a 220 Ohms (normalized value) we obtain a theoretical gain of 5.54, which is almost half the gain of origin. The current supplied would have to be doubled.
I measured the resistance of 220 Ohms SMT (over 221 is written as the right figure represents the number of zeros) on an old Gigaset. If you must buy it, its geometry is from 0603.
Here is the resistor that needs to be changed:
View of the modified resistance:
I know it's soldered on slightly crooked, but it not easy without an SMT clamp!
Anyway, here's a simple way to position the resistor ready for soldering:
First solder a short length of solder to one side of the resistor, then use it to align the resistor and then solder the other end to the PCB. Then remove the strip of solder and solder the free end to the PCB.
NOTE: you can see some other alternative methods for holding these tiny components in place while soldering in the attached document.
Before reassembling, I looked at the footprint of MOSFETs on the plastic insulation, it seemed that some transistors did not make contact with the plating insulation over their entire surface.
So I added thermal paste under the insulation, half a grain of rice next to each transistor, and half a grain of rice on each transistor.
Although this was not necessary, it cannot hurt.
I refitted the controller and it works perfectly, I initially programmed 25A/25A and the start is very smooth, the current increases with load and the maximum is obtained only if the load on the motor is sufficient.
With 25A/35A, the increasing current is faster, I think I'll stay stick with 25A/40A for now, because any more, and the throttle response is pretty brutal and it makes driving more difficult. Finally here it is directly in the "feeling" each rule as he prefers!
Your soldering irons! Oh yeah, talking about the soldering iron, it is best to use a soldering iron designed for SMT (surface mount components), if you don't have one, you can try to make one by filing down a solid copper tip to a suitable size.
Be careful also not to overheat.
Update June 8, 2012: After some time at 25 amps, as the controller seemed to cope, I wanted to increase it to 30 amps and having set the controller to the max, it was not enough, we had to replace the 220 ohms resistor with a 270 Ohm and I was able to get 29 amps. (And 1500 watts indicated on the Cycle Analyst) I have been using it at this level for 2 months without worries ... Jean Marc.
I would not recommend attempting this modification unless you fully accept and understand the following:- You will invalidate any warranty on the controller
- You will need to be pretty good at soldering and have an incredibly small soldering iron (preferably temperature controlled)
- You will need very good eyesight, and/or a decent sized magnifying glass just to see the component
- You will need a pair of very fine tweezers, along with a very steady hand (or two) and a lot of patience
- Your controller could easily become damaged, either during the soldering process, or while it's actually being used after the modification is successfully completed
- Your front wheel will struggle to remain in contact with the road surface at full throttle
Now, you can't say that you weren't warned!
Alan
