Shop OBEX P1 Docs P2 Docs Learn Events
Servo signal attenuation on unpowered Prop — Parallax Forums

Servo signal attenuation on unpowered Prop

dwelvedwelve Posts: 21
edited 2010-12-20 23:23 in Propeller 1
Hello,

I have a servo line split to a receiver and to the Prop. When the Prop is unpowered, the servo signal is attenuated; for example, using a 10X probe, the signal goes from a 3.1 V to 2 V, with what appears to be capacitance charging on the signal from ~ 1 V to ~2 V [note the signal is a pulse width signal from ~1 ms to ~2 ms repeated at ~50 Hz].

When the Prop is powered, the signal is back to ~3.1 V, with no evidence of capacitance on the signal line.

Putting a series resistor in with the signal helps to alleviate the problem by keeping the signal at around ~3 V when the Prop is unpowered and getting rid of the capacitance.

Can someone explain to me the theoretical reasoning to the problem and the solution here? (The problem being that the servo signal is attenuated, which when connected to a powered R/C aircraft but an unpowered Prop will exhibit full control throws due to the signal attenuation. Having throttle go full blast is the real issue here).

I have some notions of what the CMOS inputs of the Prop might do when unpowered, but no definite proof in short research online.

I do not know if receiver is TTL or CMOS based, some operate at signals of around ~3 V, others at 5 V, others at 6 V. It doesn't matter which of the receivers I have tested so far, they all exhibit this problem.

=== updated information ===
Upon further testing, the servos may only attenuate when I have at least 7 servo lines connected to the unpowered Prop. The one line that has a series resistor is unaffected. I didn't have a scope on at this time... I can check it out more...

Is the combined switching of all seven servo lines turning on or affecting the Prop CMOS inputs?

Thanks!

Comments

  • BeanBean Posts: 8,129
    edited 2010-12-20 16:37
    The signal is going thru the IO pin protection diode and trying to power the propeller chip.
    This is a common problem.

    Bean
  • dwelvedwelve Posts: 21
    edited 2010-12-20 17:03
    Thanks for the quick reply.

    How does the signal going through the protection diode and trying to power the Prop chip lead to an attenuation of the input signal? Is the CMOS input of the Prop chip not high impedance when unpowered and with an active signal?

    Thanks.
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2010-12-20 17:15
    Think of it this way: when the Prop is unpowered, any input signal will try to power the whole chip. Not only that, it has to charge all the bypass and filter caps connected to Vdd, as well as to power any other loads (e.g. ICs) connected to Vdd. So an unpowered input is definitely not a high-impedance load. You can eliminate this problem by adding an open-collector gate between your receiver and the Prop input, pulled up to Vdd. I would recommend a device from the 74LVC logic family for this.

    -Phil
  • dwelvedwelve Posts: 21
    edited 2010-12-20 20:09
    Thanks for the comments, Bean and Phil,

    Is a resistor in series with the servo line an adequate alternative solution? Adding resistors in line to the cabling might be a quicker solution for me, but I suppose I'd have to cook up an inline board anyways so I might as well use something like this: http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=568-4501-1-ND ... on a separate board.

    Thanks for your help!

    Are there other useful tips for handling inputs to the Prop in a safe manner? I was considering optoisolating all digital inputs and outputs, but I only work with low-voltage and low-power signals.
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2010-12-20 23:23
    A resistor between the receiver-to-servo-line and the Prop input pin should help. Try something around 4.7K.

    -Phil
Sign In or Register to comment.