Using CC2420 with CC2591

The CC2420 is different at the output from the 2520, more akin to the 2430 output (the first two digits give the series and so two with the "24" would be more similar).  I am not positive that it would be identical in terms of matching, but for the RF section, these two would work better.

Chris, Thanks for the reply. You got me thinking that I should spend a little more time reviewing the CC2xxx documentation. I had looked at the TI CC2430/CC2591 reference design and it shows an almost direct RF signal connection between the two components; no matching network components required (except R21). I just wanted to get TI/Chipcon confirmation that it is really that 'simple' to connect the two (obviously we will match the traces as done on the reference design).

Comparing the CC2420 and CC2430 reference designs, the RF matching networks are very similar, with just a couple of nH or pF difference in component values. The data sheets show the exact opposite matching inductors (27nH and 12nH swapped between data sheets) for a differential antenna application, and the same configuration, but slightly different component values, for a single-ended antenna configuration (maybe the slight value difference was due to tuning?).

The CC2590 data sheet shows a direct connection (except for L21 across RF N & P) between CC2590 and CC24xx, and mentions the CC2420 VREGOUT signal.

Application Note AN040 clearly states that:

"The design described in this application note is based on the

CC2400, but it can be used for CC2420 and CC2430 as they have the same RF front end."

Can TI/Chipcon please verify the RF connection between the CC2420/CC2591 is the same as the CC2430/CC2591 (and the CC2590; what is value of L21)? Thanks.

There is a very minor difference in the RF front end with the 2430 vs. the 2420/2400 but the same matching between the 2591 should be used for all three. 

I do not know if there is a typo on the data sheet, with so much documentation there is always the chance that something gets through, or there could be some other difference going on in the two circuits that I don't see on first glance.

The 2591 was designed to match the CC24xx/25xx parts.

Hi Chris,

The postings here are useful as I am also implementing a CC2420+CC2591 RF design.  I was going to use the CC2520/2591 reference design, but after reading the posts I am going to follow the CC2430/2591 design instead.

Have you made any new discoveries recently in regards to this topic, since all the other posts were from October of 2008?

The optimal load spec for the CC2420 RF port is 95+j187, and for the CC2430 it is 60+j164, according to the datasheets.  Are these numbers valid or is there really a typo?  That might give us an idea of the difference between two RF front-end designs.

Lastly, is there an optimal load spec for the CC2520 RF port?  I have another project which will use this part.

Thanks and best regards,

Barney

hello,

      I had desiged the cc2430+cc2591 followed TI reference design ,but the result is not profect .enddevice can't join the net ,and I don't kown why,did any of you

have the same problem?

Hello Fengqin,

I did not have problems with end devices not able to join the net on our system.  But I did see an oscillation (positive feedback) problem at the CC2591 output during transmission.  This probably was due to an impedance mismatch between the CC2430 and CC2591.  It may also affect the receive sensitivity.  This could be the root cause of your problem with the end device.

To solve the oscillation problem, I put a 8n2 0402-size inductor across the RF_N and RF_P pins (R21 on the TI reference design).  I used an inductor from Johanson Technologies with high SRF.

Hope that helps,

Barney

hello Barney,

         thanks for your reply ,the trace between cc2430 and cc2591 I  also fellowed the TI's reference design ,is that still have the mismatch problem ?

look forward to your reply!thank you so much!

Hello Fengqin,

You should first check whether you have an oscillation problem or not by looking at the antenna port of the CC2591 with a spectrum analyzer.  When you put the CC2430 + CC2591 in continuous transmit mode, then set the spectrum analyzer to center frequency=2.4GHz and span=5GHz, you should see large signals on both sides of 2.4GHz if there is an oscillation problem.

If that is the case, then try installing a 8n2 inductor on the resistor pad between RF_N and RF_P pins.  This helped my design to eliminate the oscillation.

Good luck!

but  I don't have spectrum analyzer.

Hi,

The match between CC24xx/CC25xx and CC2591 could be challenging without Spectrum analyzer and RF expertise.

So for high power ZigBee I would recommend to look at RC2400HP module from Radiocrafts with CC2530 and CC2591 embedded.

100mW output, all matching done and no oscillation.

Contact sales@radiocrafts.com for details on availability.

-Ørjan

Hi Fengqin,

Your problem with the end device could be caused by other issues.  Impedance mismatch and oscillation are just some of the possibilities.  Check the CC2430 register settings and make sure the values are correct for receive and transmit modes.  We have been using the development kits and SmartRF Studio for all of our Chipcon radio work.

If you are really stuck and need some outside help, we would like to offer our services to you.  I work for a company called Pacific Design Engineering (www.pde.com) based in Vancouver, Canada.  We have successfully launched many products with embedded Chipcon (CCxxxx) radio ICs.  We have experience with hardware and firmware design and testing.

Barney

Hello,everybody

I have a serious problom ,i have designed CC2430 module ,and cc2430_cc2591 module,but the two modules can't  transmit .can anyone help me ???thank you very much!!!

Hello everybody,

sorry @ Fengqin, I don´t have an answer for you.

But to the others: are there any news regarding this topic?

I´m currently designing a prototype of a CC2520 with CC2591. Through our distributor I´m in contact with TI in Oslo, and I was adviced to add a 2.4GHz-lowpass-filter between CC2591 and the antenna. We want to comply with the european standards (especially DIN EN 300 328) and I was told, that the original evaluation-design of TI has some clearly spurious emissions that wouldn´t pass these regulations.

Did anyone of you tested a "final" design in an anechoic chamber and can confirm the spurious emissions?

My first prototype is already done and it works very good. We can communicate through the modules and there were no problems. But the transmit-power is still not in its optimum due to some mismatch, I assume. There´s still some work to do at the network analyzer. At the moment I´m waiting for the second prototype with the mentioned lowpass-filter between CC2591 and the antenna.