Fechter's RC throttle interface ver. 1.8 technical reference

fechter

Administrator
Staff member
Joined
Dec 31, 2006
Messages
16,508
Location
California Bay Area, USA
For the discussion about this device and the history of development, see this thread:
http://endless-sphere.com/forums/viewtopic.php?f=28&t=8160


The throttle interface has the following functions:

Converts the signal from a standard bike throttle (hall or resistor) to a servo input for an RC motor controller. This unit was tuned specifically for the Castle Creations Phoenix 110HV, but should work any RC controller.
Has an on-the-fly adjustable current limiter that can adjust from near zero to around 95 amps (100A version). Current sensing is on the battery side.
Has a built-in BEC 5 volt regulator that powers the circuit, controller input and hall effect throttle. The BEC regulator output is limited to 100ma and short circuit protected. The input voltage can range from 10v to 60v. Higher voltage operation is possible, but may require changing R1.
Has an optional input for a brake switch cutoff and external battery low voltage cutoff signal.
Has a provision for adding overtemperature current limiting via external thermistor (beta).
Has a bank of capacitors to supplement the main capacitors in the RC controller.

The unit is installed between the battery and the RC controller. On the boxed units, the battery wires come out of the side with the adjustment pot.
Battery End.jpg
The controller wires come out the side with the servo connector.
Controller End.jpg
 
The colors on the servo connector don't match the colors on the Phoenix 110HV, but it should be fairly intuitive which way it goes. Black goes to brown.

The throttle connector wires are shown below. While the connector type matches a standard Crystalyte throttle, the pins are not in the same order, so will take some switching around to fit directly. Both sides of the throttle connector are supplied, so you can simply spice in the wires to any throttle cable.
Throttle connector wiring.jpg
 
To interface a brake lever switch or the LVC from an external battery management system, it will be necessary to solder a pair of wires on to the board. When the brake line is grounded, it will override the throttle and bring the output to zero. The connections are located up next to the main capacitors.
Ebrake connection points.jpg
 
The current limit adjusting pot controls the maximum battery current. The range goes from aproximately 5 amps to 95 amps. If you set the control midway, the limit will be about 50 amps. If the control is set above around 80 amps, the built in limiter in the controller might kick in.

The potentiometer shaft and body are both plastic, so should be protected against impact. Do not overtighten the nut or the threads can strip easily. The end of the shaft has a screwdriver slot that you can grab with your fingernail if the pot is hard to turn. You could also install a small knob if desired. The shaft diameter is 1 /8" (3.5mm).

Current limit adjusting pot.jpg
 
Phoenix controller programming:

The default settings on the Phoenix 110HV may need to be changed for best operation.

The default cutoff voltage is 'Auto LiPo', which measures the pack voltage when the controller is first turned on, then sets the LVC accordingly. This might be OK if you're using LiPo batteries. The other option is a fixed voltage cutoff which can be programmed to 12,18, 24, 30 or 36 volts.

The current limiting feature in the 110HV is more like a circuit breaker than a linear current limiter. If you hit the controller's current limit, the output either stops completely, or goes to sort of a half power mode. The default 'normal' limiting may be fine for most applications. If you hit the limiter too frequently, you could try the 'Insensitive' setting. I would not recommend using the 'Disabled' setting. Under nomal conditions, the built in limiter should never activate, as the limiting action of the interface circuit should keep the current below the built-in trip point.

The brake setting default is 'Soft Delayed Brake'. If your drive setup uses a freewheel at the motor, the brake setting won't matter much. If you are using direct drive, then it is advised to disable the braking, however I found it to be an interesting feature as it does regenerate back into the batteries when braking. The problem is there is no coast mode, so the brake comes on whenever you let off the throttle. You can adjust the delay, but the only way to 'coast' would be to keep the throttle on slightly.

Throttle type has option for 'Auto Calibrating Throttle' and 'Fixed Throttle'. Default is automatic. The automatic feature is nice and I recommend trying it first. The problem I had is my cheap throttle had a defect the caused the output to drop slightly as you advance the throttle position, then it came back up like normal. The Auto Calibrate circuit took the low spot as the zero throttle setting, which caused the throttle to be slightly on when I released the throttle, resulting in a bit of motor creep. By changing to 'Fixed Throttle', this problem was resolved.

In the Castle Link software, there is a setting for throttle response. Use the fast setting.

The default setting for timing advance is 'Standard Advance' I have not tried changing this.

The defalult setting for LVC Type is 'Hard Cutoff'. I would recommend using the 'Soft Cutoff'

The default setting for PWM switching rate is 13khz. I have not tried changing this setting.
 
Ver1_8 schematic.jpg

Starting at the upper left, R1, D1 and Q1 form an amplified zener pre-regulator with an output of around 11v. Q1 is a Darlington and is rated for 100v. This feeds U1, which is a 78L05 regulator. The 7805 provides 5v for the rest of the circuit and has internal short circuit protection.

U2 is an ICM7556 dual cmos timer chip. This is a dual version of a 555 timer.
The left side is an oscillator running at 50 hz to provide the 20ms time base for the servo signal. The right side generates the signal pulse, which varies from 1ms at zero throttle to 2ms at full throttle. The pulse width is modulated by the control voltage on pin 11.

Resistors R5 and R7 scale the output of the throttle to the required control voltage for pin 11.
R6 is a pull down resistor that will prevent a full throttle output in the event the throttle connection is broken. C5 limits the rate of change of the control voltage, essentially providing a slower ramp-up.

The Allegro current sensor, U4, has an output of 2.5v at zero current, and increases to 4v at 100 amps. The current signal on pin 3 is filtered by R10 and C7 to remove high frequency PWM hash. R11, R12 and RV1 form a voltage divider for setting the current limit. When the signal from the current sensor exceeds the set point of the divider, the output of U3a will go low, pulling down on the throttle signal through D2. D2 prevents the op amp from ever pulling up on the throttle.

R8, R9 and C6 form the proportional - integral feedback loop for the op-amp. For sudden changes, the loop gain will be quite low, but over longer time periods, the gain will be more like 50.

U3b, R13-16 form a comparator circuit for an external thermistor. The thermistor resistance is around 10k ohms at 25C (room temp). When the temperature gets up to around 80C, U3b will start to pull down through D3 and decrease the maximum current. From 80C to 100C, the maximum current limit will steadily decrease to zero. This part of the circuit is untested at this time and the parts are not populated on the beta units.
 
When mounting the box, I suggest mounting it in such a way that the hole where the throttle wires go in is located on the lowest side. Also the screws that hold the cover on should be on the bottom or lower side. This will minimize the amount of water that could get into the box during rain, and allow any water that does get in to drain out the hole on the bottom.

As with any electronic device, water invasion could cause malfunction, including the possibility of uncontrolled full throttle. That said, I think the box is pretty rain resistant. The box could also be located inside another enclosure. I don't think there would be any heat issues (not true for the controller itself).

When testing for the first time and calibrating the throttle, it is strongly suggested to disconnect the motor drive train or lift the drive wheel. An unexpected motor run can ruin your day and cause injury.
 
If you want to work on the board, remove the four screws holding the cover and lift it off.
View attachment 1

Next, you can pop the wires out of the slots and lift the board out of the box as shown. Be especially careful about the wires going to the adjuster pot, as they are delicate and easily broken. If you lay the board on the side shown, there will be enough slack in the wires. If you are going to do any amount of work on the board (like installing the temp cutoff parts) I suggest unbolting the adjuster pot from the box.
Circuit Board Removal.jpg
 
I discovered a minor design goof in the regulator section. Resistor R1 was changed at some point in the design to prevent the regulator from dropping out until the voltage gets down to around 10v. This is good.
I failed to recalculate what happens at the high end however. With the 15k specified, the dissipation reaches the rating at around 60v input. I originally wanted it to take up to 100v. Since most ESC's can't go over 50v, this won't be an issue for most applications. If you desire operation above 60v, then R1 should be increased, or you could lift one side and put another resistor in series with it.
 
Back
Top