Smallest RC Throttle Circuit

Discussion in 'Electric Bicycles' started by safe, Jan 27, 2010.

  1. safe

    safe Active Member

    Smallest RC Throttle Circuit

    There are a few RC throttle circuits already posted on the internet:

    http://endless-sphere.com/forums/viewtopic.php?f=28&t=8160

    http://endless-sphere.com/forums/viewtopic.php?f=9&t=11877&p=195857&hilit=+throttleizer#p195857

    ...and so I'm not the first to be playing around with the idea.

    What I want to do is to find the "easiest" way to get something to work. My idea goes like this...

    555 Timer Chips

    http://www.horrorseek.com/home/halloween/wolfstone/TechBase/com555_555TimerCalc.html

    The 555 Timer has a "Control" pin that most people seem not to be using. What the "Control" pin does is it adjusts the upper threshold so that an increase in it's value means a LONGER duty cycle and that's exactly what we want to have happen for our throttle. Most of the servo testers use a potentiometer along with a 555 timer, but the "Control" pin just sits there. So I was thinking you could just connect to the "Control" pin (or start from scratch with a 555 timer) and integrate the throttle signal (0 - 4 volts) so that the throttle goes directly inside the 555 timer to adjust the pulse.

    Op Amp Scaling

    http://www.stefanv.com/calculators/hp67_offset_gain.html

    So now you have a situation where the 555 Timer has a stable position internally of about 2 volts or so when the servo is closed and the motor power is off. If you simply can raise the internal voltage of the 555 Timer it will increase the duty cycle and that moves the servo to the open position. So all that needs to be done is for an Op Amp to scale a range from the throttle of from 0V to 4V so that it instead produces a range of about 1.5V to 4V. That's what this circuit would do.

    [​IMG]

    [​IMG]

    Okay... now I haven't tried this and am not sure what I might be missing... but it seems like this might simplify things a great deal.

    :thinking: You would just build this simple circuit and then attach it to a servo tester circuit through the "Control" pin.

    People that know circuits please give me your opinion...
     

    Attached Files:

    Last edited: Jan 27, 2010

  2. safe

    safe Active Member

    This might be even easier... all you need is an accurate voltage reference (can be done with an off the shelf chip) and two resistors.

    [​IMG]

    [​IMG]

    The way to go about this would be to accurately measure the low and high voltage readings for the "Control" pin on your chosen servo tester and then figure out what you need by reverse engineering. Alternatively you could build in some adjustability, but then that's getting complicated again. (maybe a couple of pots)
     

    Attached Files:

  3. safe

    safe Active Member

    This one uses a single 5 volt voltage supply for:

    1. Throttle supply

    2. Servo Tester supply (not shown in circuit, but assumed to be 555 timer based)

    [​IMG]

    [​IMG]

    [​IMG]
     

    Attached Files:

    Last edited: Jan 28, 2010
  4. safe

    safe Active Member

    Low Cost, Low Part Count

    This only requires six resistors, a capacitor, a 555 timer and an accurate 5V voltage regulator and this would allow you to buy an ESC without a BEC. If the ESC had a BEC then you could drop the 5V voltage regulator and just use the ESC supply instead.

    [​IMG]

    [​IMG]

    ..I'm not saying the idea is perfected yet, but it looks pretty good.

    You could then add another circuit (separate) that could be for current limiting that would be connected to the throttle before it goes through the resistor and pull the voltage down. By building it up as separate modules you can test each separately to more easily identify problems.

    --------------------------------------

    Note: I'm not actually planning on using an RC motor myself as I'm working on my own Halbach Axial Disc motor concept, but the ESC controllers are cheap and they are sensorless, so it's as good a way as any to go about adapting a brushless controller to a motor. My preference was to avoid doing any of this... (but it's looking to not be as bad as I had first thought) Also, I'm really cheap and so if it can be built for next to nothing that's something I like.
     

    Attached Files:

    Last edited: Jan 29, 2010
  5. safe

    safe Active Member

    SPICE

    This is the preferred SPICE file. There are a couple out there that really are hard to get not to crash in my SPICE program, but this one works pretty well:

    * ICM7555 MACROMODEL
    * ----------------------------
    * Revision 1.0 4/2006
    * ----------------------------
    * The ICM7555 is a general purpose RC timer capable of generating accurate
    * time delays or frequencies. This device feature and extremely low supply
    * current combined with virtually non-existant current spike during output
    * transitions.
    * ----------------------------
    * Connections
    * 1 = GND
    * 2 = TRIGGERB
    * 3 = OUTPUT
    * 4 = RESETB
    * 5 = CONTROL VOLTAGE
    * 6 = THRESHOLD
    * 7 = DISCHARGE
    * 8 = VCC
    *Parameters which are modeled.
    * 1) Full timer functionality
    * 2) Supply bias current and load current
    * 3) Output rise/fall times
    * 4) Reset threshold
    * 5) Output VOL/VOH
    * 6) Discharge voltage with current
    **************************
    .SUBCKT ICM7555 1 2 3 4 5 6 7 8
    XOPAMP 1 2 3 4 5 6 7 8 ICM7555_S

    .SUBCKT ICM7555_S 18 11 12 13 14 15 16 10
    *****************************************
    *COMPARATORS
    ES1 A11 18 10 18 1
    IABIAS A11 A12 10UA
    MA1 A13 A16 A12 A11 MOSP
    MA2 A14 A15 A12 A11 MOSP
    VAOS A17 A16 0.5M
    RAD1 A13 18 10K
    RAD2 A14 18 10K
    DAC3 A15 10 DA
    DAC4 18 A15 DA
    GCA1 18 A20 A13 A14 1000M
    RCA1 A20 18 100K
    VCA1 A21 18 1V
    DCA1 A20 A21 DY
    DCA2 18 A20 DY
    **************
    IBBIAS A11 B12 10UA
    MB1 B13 B16 B12 A11 MOSP
    MB2 B14 B15 B12 A11 MOSP
    VBOS B17 B16 0.5M
    RBD1 B13 18 10K
    RBD2 B14 18 10K
    DBC1 B16 10 DA
    DBC2 18 B16 DA
    GCB1 18 B20 B13 B14 100M
    RCB1 B20 18 100K
    VCB1 B21 18 1V
    DCB1 B20 B21 DY
    DCB2 18 B20 DY
    **************
    RB1 10 B22 60K
    RB2 B22 B23 60K
    RB3 B23 18 60K
    VR1 B23 A15 0
    VR2 B22 B17 0
    VIN1 A17 11 0
    VIN2 B15 15 0
    VIN3 B22 14 0
    *******************************************
    *SR LATCH
    EL1 A25 18 10 18 1
    RLA A25 A26 10K
    CLA A26 18 1P
    MLA1 A26 A27 18 18 MOSN
    MLA2 A26 A28 18 18 MOSN
    RLB A25 B26 10K
    CLB B26 18 1P
    MLB1 B26 B27 18 18 MOSN
    MLB2 B26 B28 18 18 MOSN
    MLB3 B26 B29 18 18 MOSN
    VFB1 B27 A26 0
    VFB2 A27 B26 0
    EFB1 A28 18 POLY(2) A20 18 10 18 0 0 0 0 1
    EFB2 B28 18 POLY(2) B20 18 10 18 0 0 0 0 1
    ****************************
    *PD
    EPD1 B33 18 B26 18 1
    RPD1 B33 B34 10K
    CPD1 B34 18 50P
    *OUTPUT
    MO1 B30 B34 A25 A25 MOSPA
    MO2 B30 B34 18 18 MOSNA
    MO3 B31 B30 A25 A25 MOSPA
    MO4 B31 B30 18 18 MOSNA
    CO1 B31 18 0.1P
    CO2 B30 18 0.1P
    VO1 B31 12 0
    *DISCHARGE
    MD1 16 B30 18 18 MOSNA
    *RESET
    RRST B29 A25 20K
    MRST B29 13 18 18 MOSN
    RRSTB 13 18 50G
    *SUPPLY CURRENT
    ISUP 10 18 12.3U
    FSUP 18 A36 VO1 1
    DSUP1 18 A36 DZ
    DSUP2 A36 A37 DZ
    RSUP A37 18 1
    GSUP 10 18 A37 18 1
    *******************************************
    .MODEL DA D(IS=100E-14 RS=0.5K)
    .MODEL MOSP PMOS(VTO=-0.7 KP=12.57E-4)
    .MODEL MOSN NMOS(VTO=0.7 KP=12.57E-3)
    .MODEL MOSPA PMOS(VTO=-2.0 KP=78.5E-4)
    .MODEL MOSNA NMOS(VTO=2.0 KP=78.5E-4)
    .MODEL DX D(IS=100E-14)
    .MODEL DZ D(N=10M)
    .MODEL DY D(IS=100E-14 N=0.1M)
    *******************************************
    .ENDS
    .ENDS
     
  6. safe

    safe Active Member

    50 Hz Desired

    Unfortunately it looks like analog servos do not like high frequencies. Servos are usually designed for a frequency of about 50 Hz and that means an average period of about 20 ms. The pulses are supposed to vary from 1.25 ms to 1.75 ms, so the actual duty cycle is down in the 10% range.

    So I need to add one more thing... a diode.

    This diode combined with adjustments in the resistor and capacitor values deliver the sparse pulse structure (low duty cycle) that a servo is going to want.

    [​IMG]

    [​IMG]
     

    Attached Files:

    Last edited: Jan 29, 2010
  7. SimpleSimon

    SimpleSimon Active Member

    Safe, in case you haven't noticed till now, I'll state it quite bluntly. From the lack of responses here I'd guess no one is terribly interested. Which isn't surprising, as I don't think anyone is very keen on a wireless servo-actuated throttle on their Mb, or even in throttle by wire on one.
     
  8. safe

    safe Active Member

    RC Motors are very popular in some circles.

    AussieJester used to post here and he has built a bike that uses one.

    Anyway... I was hoping to get feedback on this, but it looks like I'm on my own with it.
     
  9. safe

    safe Active Member

    Adjustable

    This could be it.

    If you add a 10K pot into the 555 Timer circuit it will shift the pulse width from narrow to wide across the whole spectrum.

    I'll have to post back onto this thread when I build it. (probably in the spring)

    [​IMG]

    [​IMG]

    5 Resistors
    1 Pot
    1 Diode
    1 555 Timer
    1 Capacitor
    1 Voltage Regulator

    ...total part count 10, cost would be maybe $5-$10.
     

    Attached Files:

    Last edited: Jan 30, 2010
  10. safe

    safe Active Member

    Efficiency

    Speaking of efficiency... I was a little worried about using a linear voltage regulator instead of a BEC because the rumor is that the linear voltage regulator is not very efficient.

    So I had to check it out.

    http://digikey.com/scripts/DkSearch/dksus.dll?Detail&name=KA278R51CTU-ND ($0.88)

    [​IMG]

    [​IMG]

    I found a nice little voltage regulator that manages to only drop 0.5 volts for a one amp current. My circuit only draws about 0.3 amps (see the graph) so that comes to:

    0.5 volts * 0.3 amps = 0.15 watts

    ...and if you compare that to a 1000 watt motor draining at full power you get:

    0.15 watts / 1000 watts * 100 = 0.015%

    So maybe it's true that the MOSFET based voltage regulators can save a little power (somehow) but it's not much. I'll design the wiring assembly so that when I'm not riding the voltage regulator is disconnected. I normally disconnect the power wires after every ride (and plug them into the charger) so if the battery is not connected it can't drain while the bike is parked. I remember a guy that spent $1000 for LiFePO4 batteries and he designed his circuit with a linear voltage regulator that drained the batteries when the bike was parked and that destroyed the batteries when he left the bike in storage for some time. So that's what needs to be avoided.
     

    Attached Files:

    Last edited: Feb 1, 2010
  11. Fabian

    Fabian Well-Known Member

    Hi safe

    Oh yeah - you are onto something great here

    There exists great potential to do something very interesting: an electronic speed shifter.

    I've raced motorbikes in the past and it would make for a very special and unique system of incorporating racing style speed shifter technology onto a motorised pushbike but taking it one step further and going for an electronic rapid fire gear change system.

    Wide open throttle gear changes would require precisely timed and integrated servo motors on the rear derailleur, throttle and clutch.
    Now we are talking about something really special and unique and also something very exclusive.

    I know there will always be people on this site saying (in a winging voice) "but for that expense, why not get a proper motorbike that has an auto clutch".
    I say hooey to them, they can shush it!!!

    If i had the money, i'd love to build a full carbon fibre and titanium pushbike loaded with the best technology straight out of the Ducati MotoGP race team.
    Yes, it might cost $50,000, but there's no-one else out there that would have one!

    Fabian
     
    Last edited: Feb 1, 2010
  12. safe

    safe Active Member

    Servos do have many applications.

    In this case it's a matter of necessity because if you want cheap RC parts like an ESC you need to adapt it to the ebike throttles.

    So rather than being "glorious" this circuit fits more into the "practical" and the "cheap".

    You can actually buy a circuit for this problem for about $100 or so from some of the other hobbyists... this was intended to help the hobbyist that wants to do it themselves.
     
    Last edited: Feb 2, 2010
  13. Fabian

    Fabian Well-Known Member

    I'm well aware of the R/C scene and the technology available as i have a reasonable collection of 90 and 50 sized (nitro fueled) radio controlled helicopters, with some tasty radio gear.

    What you're doing though is excellent as it's a custom designed concept.
     
    Last edited: Feb 2, 2010
  14. safe

    safe Active Member

    Well I want to be modest about it because the idea has been developed ahead of me. This was actually a big issue a year ago over on endless-sphere. Fechter (who is a regular there) created the first circuit using analog technology much like what I did. Later there was another guy that created a digital version. They are both still for sale to my knowledge so if you want to buy one they probably still have them for sale.

    The idea here was to build a smaller version of the analog circuit and pull out as much as possible so that you have a simple and streamlined design. At this point all I know is that the circuit appears to work perfectly in the SPICE simulation program, but it has not actually been built and tested yet. So I'll be updating the results this spring when I build it.
     
Loading...