RF noise emitted in receive mode 1

[Maxish]

Hello,

I'm attempting to meet FCC standards for my product. For the Transmit mode, its a go! However, while my transceiver(CC1010) is in recieve mode pending for data, I'm failing!

For some mind-boggling reason, noise enters the antenna which emmits 66dB(>45db FCC limit) at 928MHz(comm. frequency) while it is in receive mode!

The CC1010, from Chipcon, is single chip Transciever and 8051 MCU.

Can anyone please clue me in on how to fight this?

Kind Regards,
Oleg

 

[Comcokid]

FCC requires a check of emissions on a receiver to make sure the various LO used to convert a signal to the IF frequency do not leak back out through the antenna.

I'm not real familiar with the Chipcon CC1010, but a quick look at the data sheet indicated the input signal is converted down to a IF frequency. I could not quickly determine if the LO is Fo+IF or Fo-IF in frequency, or what the IF frequency was. I.e. - If you are receiving at 918, and the IF is 10.7 MHz, then the LO will be 928.7 MHz at the mixer (for a Fo+IF scheme). This LO (or VCO) may leak back out the receive input, or due to PCB layout shortcommings may be coupled directly from your VCO inductor to the RF input and radiate through the antenna.

Check your layout carefully for good ground - that is good RF grounds. Remember, a short trace is an inductor, a via is an inductor, the lead of a capacitor is an inductor. This is why at RF ground planes and wide ground traces with multiple vias are used, and smaller value SMT capacitors (27 to 100 pF) are used for bypasses. Inductors located in the same orientation nearby can couple like a transformer.

 

[VEBill]

According to the block diagram on the following link:

http://www.chipcon.com/files/CC1010_Brochure_1_0.pdf

...the one and only oscillator is permanently connected to the output power amplifier, even when the oscillator is being used as the receiver's local oscillator. In other words, broadcast the LO while receiving... I know it doesn't make sense, but that's what the block diagram shows and it matches the symptoms described.

I wonder if there is a way to disable the RF Power amplifier (output) - there must be...

 

[itsmoked]

Isn't this called hettorodynining? <=miss speeled. I thought it was generally ruled that using heterodyning in receivers instead of one of the other techniques is pointless these days because they *have* to leak back out the antenna?!?! I was told by several radio makers, "ah, that heterodyne receiver will never be licenseable"

 

[VEBill]

http://www.chipcon.com/files/CC1010_Data_Sheet_1_3.pdf

On this other link, Figure 19 shows a T/R switch where the VCO is redirected away from the PA in receive mode (makes more sense).

 

[VEBill]

"...they *have* to leak..."

No, they don't have to leak. No matter what, there is always a way to prevent any significant LO leakage to whatever degree you wish (at a price).

 

[IRstuff]

Aren't all AM radios heterodyned?

TTFN

 

[itsmoked]

Thanks VE1BLL for clarification.

IRstuff, I think one method is called "Direct Conversion".

 

[Maxish]

okay thanks a lot for the replies. So far I see a couple possible solutions

a)manually turning off PA (PA_POW=0x00) while recieving.

b)reconsider layout. I will attempt to match the reference design to the finest detail because the reference design board did marginally pass this FCC test with my bin file. This is a headache due to the 4 layer design over my existing 2 layer.

 

[itsmoked]

Maxish... I do not see you having a candle in a hurricane of a chance, making it happen without 4 layers, if the reference design barely makes it on a 4 layer.

 

[BrianR]

I have to agree with itsmoked, in you'll need 4 layers to get the grounding and PS decoupling you'll need especially for pins 2,3,6,7,8,9. It sounds like straight local oscillator leakage which could be coming from L1/L2 on the VCO or a grounding / bypassing problem. If you can, 4 layers with a complete ground plane will help. This is assuming you are receiving at 918 with a 10.7Mhz IF and 928 LO.