CC3200: [Urgent help] PCB and SMA external antenna weak signal

Hi,

Can you compare RSSI of your board with LaunchPad? How looks RSSI from CC3220 and AP side? Was impedance of antenna trace properly calculated and measured at VNA? Can you provide results from VNA or from spectrum analyser during conductive RF test?

Jan

Hi Robert,

Nice! Excellent explanation. Thank you for a spending your valuable time with helping other users.

In the conjunction I can add link to this TI video.

Jan

Hi Jan D
Thanks for your suggestions.

I will check and confirm but could you please show me way to check by your way ? by firmware and show result on terminal or another ways?

About checking with VNA, i dont have this tools for checking, so will be more difficult when tuning antenna.

But i dont think its more complex design with pcb antenna at first, i have just follow some reference design of TI. :)

Trace width for antena have calculated based on stackup in Hyperlynx at 50Ohm.

Also, PCB antenna footprint have checked demensions based on TI guideline, so i dont think this is an issue :)

Hi MKTech,

For antenna design testing we use VNA as similar way as was described by Robert. Without proper equipment like a VNA or spectrum analyser and RF generator  is hard to properly design antenna circuits. There are ways how to to test antenna design without measurement equipment stated above but it  requires high level of knowledge of RF circuits. I agree with Robert, if you are able, please ask some RF designer for help with your antenna. You will save lot of time.

Good start point can be do RSSI measurement from CC3220 and AP side. Comparison with LaunchPad can give you raw idea how bad/good is your antenna design. But it will not help you fix issue with your design. For this measurement you will need Access Point with capability report current RSSI for each client. For example you can use hardware with OpenWRT and use iw command. At CC3220 side you can read RSSI by API sl_WlanRxStatGet(). Be aware that RSSI results depends on external conditions and from this reason your comparison will be very raw.

Jan

Hi Chen
Many thanks for your details explain.
Really i have a basic of transmission line and i used it for calculating an 50ohm trace.
I used Hyperlynx with my fab stackup file, so i think its not an issue.

I saw a wrong with filter placement from your suugestions, i will update it in next design .

If you know what I am talking about then you should know that L5 & C25 should be placed as close to the feeding point as possible. Then, with VNA, you measure the signal at the left-hand side of L5. For this, you also have to cut the trace at the left-hand side of L5.

My trouble is still occur with external antenna, so could you please help to check in this case ?

SMA conn :

www.digikey.com/.../7908399

2.4GHZ Antenna:

www.taoglas.com/.../

Thanks

Thanks Jan D
I will try your way to make a basic checking first.
I will let you know when have update.
About VNA, could you share me a link of product that i can buy for testing with lowest cost :)
Thanks

Hi MKTech,

There is many low cost VNAs at the market at range below $200. But I am not sure how good are, because I never tested them.

If you want cheap "real" VNA+spectrum analyser up to 3.2GHz my recommendation is Siglent SVA1032X. Price is about €3000 and this is pretty reasonable price for such equipment.

Be aware that having equipment is one thing and knowledge how use them is complete different thing.

Jan

Thank you Jan for the video link. I did learn a lot of antenna knowledge from TI.

Hi MKTech,

For the external antenna, the following items can be checked:

1) The trace impedance might be too far away from 50 ohm. But I don't think you have proper equipment to check this....

2) Something wrong with the RF filter. You can try to remove the RF filter and short pin-1 to pin-3 directly. When the RF filter is removed, more noise from environment will affect the desired signal. So, you should turn off all unnecessary electrical devices around you.

3) Wrong capacitor type is used for C74. All capacitors and inductors used in RF path must be designed for high-frequency.

For capacitor, in theory z = 1/wc, the higher the frequency, the lower the impedance. But real capacitors are not perfect component. The impedance will become larger and larger after certain frequency. That means it becomes an inductor and blocks the signal. So, you have to pick the right capacitor. You should also check this for L5 and C25.

External antenna should be already 50 ohm at the feed-point. So, in theory, no matching circuit is required. I guess C74 is used to block DC bias from RF output. I am not sure about this... but it should be ok if you use a capacitor designed for high-frequency.

Hi Chen
Thanks for your suugestions.
I will try with external antenna and let you know soon.
Thanks.

Hi everyone.
Im not sure about why dont have any TI engineer support in this case ?
Maybe i worng when press solved the problem ?
Can I revert status on this topic ?

Thanks

Hi MKTech,

Thanks for you support thus far. We can take it from here.

In addition to what is stated I think there are a few other pieces of information needed.

1) Can you please confirm your measurement setup? Are you using RadioTool? How are you measuring the output power? Can you provide the data conducted against 1DSSS and 54OFDM?

2) Can you also measure this on the TI LP? What are the results that you get?

I would first look at the above all conducted to get a baseline of the output power that you are getting. One thing that I do notice is that for the path going to the U.FL connector you have a 0.5pF capacitor. Can you change this to a 0 ohm resistor and look at your results. This value may be too low and also causing attenuation of the RF signa. Once we can confirm the conducted data, then you can start to look at your antenna matching.

Thanks,

Riz

Hello Rizwan Murji

Thanks for your feedback.
Based on comments of Mr.Chen and Mr.Jan D, i detect some issue on current pcb board.

- wrong antenna trace width due to missed stackup 

- wrong on matching circuit (L&C) for pcb antenna.

- wrong ac capacitor for external pcb antenna ( will change to 10pF based on CC3220SF dev board). We try to change 0R and gain improve about 10db.

Really we dont have VNA to check and measure at antenna port, we only can test on CC3220SF dev board.
Im doing update a new board, so can i send you to review before put an order to fabrication ?

Thanks.

HI,

When you are ready you can submit a design review here: https://www.ti.com/tool/SIMPLELINK-WIFI-DESIGN-REVIEWS.

I will go ahead and close this thread for now.

Thanks,

Riz