TIDA-01624: Options of selecting flexible substrate and ink

Hi, Brandon! 

Thank you for the feedback. May I know in your reference design, which substrate and ink type were considered? And also what was the trace thickness, substrate's dielectric constant and thickness? It would be really helpful if you could provide this information, if possible.

Best,

Muhtady

Hi Muhtady,

The total board thickness is 6.4mils. We used a polymide substrate which was 3.0 mils thick, with a dielectric constant of 3.3. The trace thickness for the RF portion was 6 mils, with all of the traces being copper. 

Best Regards,
Brandon Fisher

Hello, Brandon!

Thanks again. I am a bit concerned with the printed antenna performance. Do you have any separate document stating the dimension and test results for the antenna? How should I approach the antenna issues? 

Also, in the BOM file, C12 and C14 are zero in quantity. Is that alright? Because the footprints for the C12 and C14 are there in the layout. 

Best,

Muhtady

Hi Muhtady,

Muhtady Muhaisin said:

I am a bit concerned with the printed antenna performance. Do you have any separate document stating the dimension and test results for the antenna? How should I approach the antenna issues? 

We do not have an additional document discussing the RF performance of the antenna, but in practice we found it to be around 40ft without doing any additional matching on the board.

Its true that covering or folding the antenna can also affect the range, but this shouldn't occur in normal use as at least one side of the antenna should be left exposed when worn. If you want to minimize the possibility of folding, you could always add a stiffener behind the ground plane of the device. That would add some rigidity to the design, but should still be small enough that its not too uncomfortable to wear. 

Muhtady Muhaisin said:

Also, in the BOM file, C12 and C14 are zero in quantity. Is that alright? Because the footprints for the C12 and C14 are there in the layout. 

The BOM is correct, C12, C14, and R6 were included in the design for an optional Pi-matching network that was not necessary for our purposes, so the parallel portion went unpopulated. For optimum range, these footprints can be used to perform RF matching while the device is worn. That way you get the maximum possible range out of the antenna. 

Best Regards,
Brandon Fisher

Hi, Brandon! 

To avoid the variation in antenna performance, do you recommend the use of chip antennas? Do you have the experience of using any particular chip antenna with this circuit designed?

Best,

Muhtady

HI Muhtady,

We did not attempt to use a chip antenna for this design, as the intention was always to make it as flexible as possible. Also, since it was Bluetooth based design we were pretty satisfied with the range we were getting for this application.

That said, a chip antenna is an entirely viable solution, but you will have to account for its size when considering how a patch would be worn. You will also want to adjust the ground plane and elements of the layout depending on the specific chip antenna. 

Best Regards,
Brandon Fisher

Awesome!

Is there any compatible chip replacement for CC2640 which will provide me with on-chip RF antenna solution in the ISM band? I couldn't clearly find one but I remember coming across one design implementation using TI CC25XX SoC which provided on-chip antenna. Do you have any suggestions regarding that or do you recommend any compatible alternate for CC2640 with on-chip antenna feature?

Best,

Muhtady

Hi Muhtady,

I apologize for the delay. You should be able to use any 2.4 GHz antenna you can find. This TI design for the CC2541 uses one from Johanson. It would work with the CC2640 as well. 

Best Regards,
Brandon Fisher