Infrared RC oscillator

[Michael]

So many of the xmtr circuits I've seen for infrared remote control use a
crystal or ceramic resonator in a cmos oscillator.

Is there any reason why an R/C network oscillator using low tolerance parts
wouldn't be accurate enough at such a low frequency?
(38 or 40 khz).

Thanks.

 

[Charles Schuler]

So many of the xmtr circuits I've seen for infrared remote control use a
crystal or ceramic resonator in a cmos oscillator.

Is there any reason why an R/C network oscillator using low tolerance
parts wouldn't be accurate enough at such a low frequency?
(38 or 40 khz).

Economics. It is easier to use a resonator and a frequency divider to get
the required stability ... the only significant source of error is the
resonator itself and they are quite stable and inexpensive. With RC
control, one has several drift sources and inexpensive capacitors have
temperature problems. Also, don't forget about trimming ... the resonator
approach works every time with no adjustments.

 

[Frank Bemelman]

So many of the xmtr circuits I've seen for infrared remote control use a
crystal or ceramic resonator in a cmos oscillator.

Is there any reason why an R/C network oscillator using low tolerance parts
wouldn't be accurate enough at such a low frequency?
(38 or 40 khz).

With all the things that can and will go wrong, a ceramic resonator
turns out to be the cheapest solution. That is reason enough.

 

[Pooh Bear]

So many of the xmtr circuits I've seen for infrared remote control use a
crystal or ceramic resonator in a cmos oscillator.

Is there any reason why an R/C network oscillator using low tolerance parts
wouldn't be accurate enough at such a low frequency?
(38 or 40 khz).

Your RC oscillator should be as accurate as the tolerance of the parts used. In
other words probably fine for this application.

I'd use a wein bridge configuration with a fixed gain of say ~ 4-5 since a sine
output isn't needed.

Graham

 

[Michael]

Yeah, I've used a ceramic resonator before with great results.

I took a 400khz and divided by 10 and the xmtr worked great.

I had these 38khz receiver modules in my parts drawer and been having some
difficulty as I couldn't find anything in a ceramic resonator that was
either 38khz or divide down to it.
(?)

Looked everywhere....digi key, mouser, newark...nothing.

Plenty of 40 and 400 CR though.

So I wasted alot of time breadboarding a 4011 with a 38khz crystal....it
works but the start up time is ridiculous and a little change in Vcc changes
the frequency so much.

I thought an R/C might be alright at only 38khz but I'll go back to the
ceramic resonator and 40khz.

Why no 38 khz CR parts? There is a 420 and 455 khz that divided can get
close.

 

[qrk]

So many of the xmtr circuits I've seen for infrared remote control use a
crystal or ceramic resonator in a cmos oscillator.

Is there any reason why an R/C network oscillator using low tolerance parts
wouldn't be accurate enough at such a low frequency?
(38 or 40 khz).

Thanks.

Linear Technology has some cute RC oscillators. Use a resistor to set
frequency. They are pretty accurate and stable for an RC oscillator.
The LTC6906 draws under 20uA. $2.50 for the LTC6906 might be a bit
steep if your dealing with large quantities. Digi-Key carries this
line.
http://www.linear.com/pc/categoryProducts.do

 

[Mike Harrison]

So many of the xmtr circuits I've seen for infrared remote control use a
crystal or ceramic resonator in a cmos oscillator.

Is there any reason why an R/C network oscillator using low tolerance parts
wouldn't be accurate enough at such a low frequency?
(38 or 40 khz).

Just because the frequency is low doesn't mean the accuracy required is any less. In practice it
gets tricky getting RC repeatability better then about 5%, especially where supply voltage is likely
to be variable, partly due to the high pulse currents driving the IR LED. A 5% frequency error will
put the carrier far enough out of band for some IR sensors to reduce range noticeably.
Resonators are cheap and stable and provide the best solution for this application - the proof of
this is simply that it's what's almost universally used in an extremely price-sensitive market.

 

[John Fields]

So many of the xmtr circuits I've seen for infrared remote control use a
crystal or ceramic resonator in a cmos oscillator.

Is there any reason why an R/C network oscillator using low tolerance parts
wouldn't be accurate enough at such a low frequency?
(38 or 40 khz).

 

[Ben Bradley]

Just because the frequency is low doesn't mean the accuracy required is any less. In practice it
gets tricky getting RC repeatability better then about 5%, especially where supply voltage is likely
to be variable, partly due to the high pulse currents driving the IR LED.

Actually a circuit using a 555 timer will have pretty good immunity
to voltage variations, OTOH the 555 may cost as much as a ceramic
resonator, and the frequency is still dependent on high-tolerance and
temperature-varying resistors and capacitors.