SERIN noise immunity
Archiver
Posts: 46,084
Group:
I'm considering using the BS1 in a low power application involving wireless
comms. <See the background info below if interested in specifics> In short,
I'll have a digital signal feeding an input pin on the BS1. My serial data will
come in at 1200 or 2400 baud. Digital "1" should consistently be 5 volts.
Digital "0", however, will be 0 volts with occasional 5 volt spurs.
Fortunately, the duration of the spurs should be short relative to the bit
length of 2400 baud comms. My question -- how will the SERIN command be
affected by these spurs? I have two hypotheses:
1) As long as the spurs are sufficiently short (I need quantification here), the
data will be seen as digital 0, and everything will work fine.
OR,
2) As soon as the BS1 sees the spur, it will assume a digital 1 is coming in,
and the data and timing will get generally screwed up.
If # 2 is the case, I may be able to filter the digital stream to eliminate
pulses/spurs shorter than anything approaching a bit length.
Before I get too far down this path, I wanted to see if anyone out there has
dealt with this before. If so, please provide some insights. Thanks.
Clark Hughes
<background info>
I'm looking to use an on/off keyed wireless transmission scheme. During a mark
(or digital "1"), I should have a good signal, and I should successfully decode
a digital "1" on the receive end (the end with the BS1). During a space (or
digital "0"), however, I'll get random high frequency noise generated by
rotating machinery, etc. The RF receiver will decode this noise as mostly 0s
with occasional short duration 1s. The good news is that the noise should be
much higher frequency than the actual data, so I may be able to simply filter it
out with analog circuitry.
--
I'm considering using the BS1 in a low power application involving wireless
comms. <See the background info below if interested in specifics> In short,
I'll have a digital signal feeding an input pin on the BS1. My serial data will
come in at 1200 or 2400 baud. Digital "1" should consistently be 5 volts.
Digital "0", however, will be 0 volts with occasional 5 volt spurs.
Fortunately, the duration of the spurs should be short relative to the bit
length of 2400 baud comms. My question -- how will the SERIN command be
affected by these spurs? I have two hypotheses:
1) As long as the spurs are sufficiently short (I need quantification here), the
data will be seen as digital 0, and everything will work fine.
OR,
2) As soon as the BS1 sees the spur, it will assume a digital 1 is coming in,
and the data and timing will get generally screwed up.
If # 2 is the case, I may be able to filter the digital stream to eliminate
pulses/spurs shorter than anything approaching a bit length.
Before I get too far down this path, I wanted to see if anyone out there has
dealt with this before. If so, please provide some insights. Thanks.
Clark Hughes
<background info>
I'm looking to use an on/off keyed wireless transmission scheme. During a mark
(or digital "1"), I should have a good signal, and I should successfully decode
a digital "1" on the receive end (the end with the BS1). During a space (or
digital "0"), however, I'll get random high frequency noise generated by
rotating machinery, etc. The RF receiver will decode this noise as mostly 0s
with occasional short duration 1s. The good news is that the noise should be
much higher frequency than the actual data, so I may be able to simply filter it
out with analog circuitry.
--
Comments
check the state of the serial line at a point that is expected to be the
middle of the current bit -- no qualification time (debouncing) of the bit.
The key for you is to eliminate the noise on your serial signal.
-- Jon Williams
-- Parallax
In a message dated 3/6/02 8:48:58 AM Central Standard Time,
jchughes@a... writes:
> I'm considering using the BS1 in a low power application involving wireless
> comms. <See the background info below if interested in specifics> In
> short,
> I'll have a digital signal feeding an input pin on the BS1. My serial data
> will
> come in at 1200 or 2400 baud. Digital "1" should consistently be 5 volts.
> Digital "0", however, will be 0 volts with occasional 5 volt spurs.
> Fortunately, the duration of the spurs should be short relative to the bit
> length of 2400 baud comms. My question -- how will the SERIN command be
> affected by these spurs? I have two hypotheses:
> 1) As long as the spurs are sufficiently short (I need quantification
> here), the
> data will be seen as digital 0, and everything will work fine.
> OR,
> 2) As soon as the BS1 sees the spur, it will assume a digital 1 is coming
> in,
> and the data and timing will get generally screwed up.
> If # 2 is the case, I may be able to filter the digital stream to eliminate
> pulses/spurs shorter than anything approaching a bit length.
>
> Before I get too far down this path, I wanted to see if anyone out there has
> dealt with this before. If so, please provide some insights.
[noparse][[/noparse]Non-text portions of this message have been removed]
Clark
jonwms@a... wrote:
>
> Generally speaking ... most serial routines sync with the start bit, then
> check the state of the serial line at a point that is expected to be the
> middle of the current bit -- no qualification time (debouncing) of the bit.
> The key for you is to eliminate the noise on your serial signal.
>
> -- Jon Williams
> -- Parallax
>