Re: Re: ir rx tip

[ian.bird@xxxxxxx]

When I started playing with IR I remember reading if there were noise
problems both a resistor and cap were needed as close to the receiver as
possible (at least for the cap anyway). I have the bit from the PDF cut
out and pasted on my wall at home. Try having a look at the datasheet for
the TSOP 1838 for this info.

Ian






"Frank Mc Alinden" <fmcalind@xxxxxxx>
25/04/2003 09:41
Please respond to ukha_d


To:     <ukha_d@xxxxxxx>
cc:
Subject:        Re: [ukha_d] Re: ir rx  tip


Hi David
I  am using a .1 cap on the power rail adjacent the reg and the ic
(decoupling)...I have a 10 mF cap across the power rail adjacent the ir rx
and i need the 5v reg as the supply is 12v in......This has only happened
on a few units ......

Frank
----- Original Message -----
From: David Buckley
To: ukha_d@xxxxxxx
Sent: Friday, April 25, 2003 6:34 PM
Subject: [ukha_d] Re: ir rx tip


--- In ukha_d@xxxxxxx, "Frank Mc Alinden" <fmcalind@b...>
wrote:
> Hi All
>         Just a little tip for those who are making their own ir
receivers and may at some stage have the following problem.. I have
been building some FirM Rx Units and noticed while testing that with
some the Busy output was flashing briefly and quite often as  if ir
was being received...First covered the ir reciever to block any stray
ir but no good...replaced the ir receiver still no good replaced all
caps and ic ........and was ready to take all my ir receivers back to
my supplier to say they were faulty and thought i would try the
little 5v 100mA regulator and bingo the fault cleared replace the
regulator on another board and the fault also cleared...Im not sure
why the glitches were getting on the rx line unless maybe
noise..........so watch those pesky little regulators ..........

Not so suprising... those IR things have a high gain amplifier in
them, and it sounds like they also have poor power supply noise
immunity.  Thus any little bit of noise on the power rails (a very
common occurance) is shoving the aplifier over the edge.  You might
like to try a 0.1 bypass cap next to the thing instead of a
regulator, and see if that cleans the rails up sufficiently; in value
engineering terms a cap is cheaper than a reg and may achieve the
same result.

This flash feature is also present on the Xantech filtered recievers,
and is enogh of a problem that a slink-e will lock up if subjected to
it.  However in the Xantech I suspect the misbehavour mode is
slightly different; the've got a bandpass filter in them, and I
suspect its hunting (a bit like a PLL) when annoyed by a pulse of
IR...