I am using this IR receiver in a circuit, and I have noticed that sometimes the unit generates an oscillating output when it is not being activated by a remote - I am guessing that this is due to the USB power supply.

I have seen other threads talking about 'filtering' the power supply with a cap and a resistor - however, could somebody explain how they come about calculating what size cap and resistor are required - and how they actually achieve the 'ripple filtering' effect?

For examples sake, my IR receiver is powered by 5v, directly from a netduino - and generates output pulses (when activated by infrared) in the 0.5-4.5ms range (see image: I dont know if that information is required, but there it is).

enter image description here

  • \$\begingroup\$ Just reporting back - I discovered the problem was NOT my power supply! Argh! I just recently bought a new Dell monitor (accessories.us.dell.com/sna/…) and it appears to pump out 6ms pulses of 38kHz IR from the buttons on the lower-right hand corner of the bezel!! If I cover it with my hand, my IR receiver stops producing the 'noise'!!! Go figure. \$\endgroup\$ – Adam Aug 10 '12 at 2:03

No, it's probably not the power supply.

It's not drawn in the block diagram in the datasheet, but the receiver will have AGC (Automatic Gain Control) to cope with big differences in signal amplitude. When no signal is being sent the AGC will just see the input noise and set the gain based on that. So you get digital noise out. I've seen it often with Vishay receivers: a continuous stream of random pulses, but when a code is sent it's received perfectly. The decoding software shouldn't have problems filtering the noise out, as it will have a much higher frequency than the signal.

A decoupling capacitor of about 4.7 µF with a 47 Ω series resistor on the power supply is recommended, but won't avoid the noise.

  • \$\begingroup\$ I totally agree and have seen the same thing from Vishay TSOPs and similar units by Sony. This is why IR protocols have a preamble and a checksum. The preamble allows the AGC to settle to the stronger signal, and the checksum and other protocol rules prevent random data from looking like a valid packet. \$\endgroup\$ – Olin Lathrop Aug 7 '12 at 11:25
  • \$\begingroup\$ @Olin - have you read the comments to this answer, about a 1 milli-degree precision requirement? I'd like to know what you think of it. \$\endgroup\$ – stevenvh Aug 7 '12 at 11:31
  • \$\begingroup\$ I hadn't noticed that before and just looked at it briefly. 1 mDeg is absurd, but I'm about to head out of town for 2 weeks so don't have time to think about it. I'll be at Microchip Masters, then exploring the hinterlands of Arizona, so probably won't be here much. \$\endgroup\$ – Olin Lathrop Aug 7 '12 at 18:35
  • \$\begingroup\$ @Olin - Enjoy your trip! \$\endgroup\$ – stevenvh Aug 7 '12 at 18:36
  • \$\begingroup\$ @stevenh - thanks. My decoding algorithm looks for the 'preamble' (as Olin has pointed out), which is 4.5ms, however my 'digital noise' is about 7ms. At any rate, my concern isnt that my algorithm will get confused (it doesnt) - my problem is that the digital noise is causing a microconrtroller interrupt every 7ms which is loading the CPU when it could be performing other useful work... Maybe I'm being a purist, but I hate the thought that 'sometimes' my CPU is processing this 'noise' looking for a signal, and sometimes it's not! Is this really how these types of circuits need to work? \$\endgroup\$ – Adam Aug 8 '12 at 4:06

If your load draws current through an resistor-capacitor (RC) filter, you would see some voltage drop (V=IR) across the resistor. You could:

  • choose the resistor small enough such that the voltage drop is acceptable.
  • use an inductor-capacitor (LC) filter instead of RC.

We don't know the actual spectrum of the noise on your USB bus. However, a common power supply filter on the USB slave side consists of a ferrite bead (inductor-like), 4.7uF or 10uF electrolytic, 10nF and 100nF ceramic capacitors. enter image description here Snippet taken from fig. 2.5 here.

More info on power supply filtering here (towards the bottom of the page).


I had a similar problem with a Vishay IR receiver. 1K resistor connected between +5V and the module's +V combined with a 10nF ceramic connected between module's +V and Com resolved the problem. Don't forget to install a 10K resistor between the module's output and +V. I monitored the output for hours with a digital storage oscilloscope and I got a perfectly clean output all the time.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.