About PCB design and audit

Wireless Connectivity Ultra-Wideband
1

HI:
We designed the PCB by ourselves, but there was a problem, the transmission distance was very close.Can the technical support staff help to review PCB and schematic diagram?
thanks a lot
tangyu

2

Can you post the schematic and images of the PCB layers here? Also details on your layer stackup would help.

3

hi! can you post radio settings and meaning of “very close” as well?

6

Hi

Thank you for your attention.
We use dw1000 respectively made a tag1 (nrf52832 MCU, PCB size 18 mm * 32 mm), and tag2 (stm32 MCU, PCB size 30 mm x 30 mm). Under the same software configuration is not the same as the transmission distance of the two. Tag1 can transmit 20-30 m, and tag2 can transfer 50 to 60 m, same program logic and dw1000 configuration, just drive of MCU peripherals (because of different MCU). We are not sure whether the problem of hardware,For example, whether the PCB wiring on tag1 will affect the transmission of dw1000 signal or the transmitted power (or standing wave) of dw1000, or whether the power processing will affect the communication distance.The design has been done according to the impedance matching requirements on datasheet.
RF set: channel5, 6.8Mbps, 64Mhz, 128Symbols, 48data.
Could you please help to analyze what is wrong with the communication distance of the two circuit boards, or why the communication distance of tag1 is very close?

thanks a lot!

7

which antenna you are using ? what is power and bandwidth settings?

8

Hi! as to my experience, the GND plane size may affect the link budget with all other matters kept same, so the smaller the tag, the lower it is. I faced the trouble with distance once for one of my attempts of custom PCB. But that fail was epic - I got only centimeters of link budget… I believe, it was impedance matching issue. If you see comparable link distance for your tags, of course, its a matter of loosing decibels, but the source may not be a messy PCB design. For ex, antenna pattern may change, as long as you change the size and layout of the board.
PS. please note, my nic (decawave) is just a nic, derived from my email address and I’m not an employee or part of Decawave team.

10

Hi,
We used an external antenna, connected to PCB via IPEX4 connector.send power 15 and 15.5

11

Hi，
Thank you for your reply.Sorry, I missed a message.
From dw1000 to antenna we used IPEX4 connector.The antenna is independent, using the same external antenna.So the antenna is fine.

12

Any differences in power supply between the boards?
Same voltage/current capability?
Any filtering to prevent CPU power supply noise getting into the DW1000?
Is the PCB layout around the DW1000 the same on both boards? The layout of the power supply bypass capacitors on the RF power amplifier inputs can have a big impact.

13

hi

Thank you for your reply.

The power supply of the two boards is completely different, but the power output capacity is enough.
CPU power noise does not enter dw1000 because two ldos are used separately.
The layout around dw1000 on both PCBS is completely different.
I’m more interested in what you have said “The layout of The power supply bypass capacitors on The RF power amplifier inputs can have a big impact.”.Could you give us some details so that we can find out the problem ourselves?

14

If you look at APH001 (the hardware design guide) pins 18 and 19 (VDDPA) should have independent connections to the power plane with separate capacitors on each pin.
The DWM1001 schematic also has ferrite beads in the power supply into these pins. I’ve not done that but other people posting here have found that made a difference for them.