In our device, we have used PAN1740 to transmit ECG data to Android phone. We found out that there are some data losses because of the BLE transmit failure within the stipulated time.
When we tried to reproduce the issue, we are seeing this very often when we cover the PAN1740 with bare hands keeping the phone and device(PAN1740) next to each other.
The same results were seen with the DA14580 Development board.
Our project is based on "throughput_eval_peripheral" project.
Is this behavior a known issue?
Has anyone seen this kind of behavior? Please help us understand the issue.
Thank you,
Bharath
Device:
Hi Bharath,
Well, its transmition over the air, so there will be loss and missing events, but not lost data since there will be retransmittions from the protocol. The throughput_eval demo is quite an old reference design based on an old SDK, have you tried to test this using the DSPS and do you still have experiencing data loss ? I mean, at some points there should be some missing connection events, but this shouldn't lead to missing data since the protocol will take care of that and retransmit in case there is no ack from the previous packet. Regarding the fact that you cover the board with your hand and you get that, are you observing any disconnections ? Do you have a sniffer log to check that the data are indeed missed and that isn't a problem from the application (perhaps overwrite the data that you receive) ? An idea would be to send an incremental value over the air and check with a sniffer log if all the values are send through the air.
Thanks MT_dialog
Hi MT_dialog,
Thanks for the reply.
The data loss we are observing is because our data buffers are overwritten before the data stored is sent over BLE to the Android phone. Also we understand that the unsent data will be sent again and hence no packet loss.
Our main concern is why does the transmission take such a long time when the MCU is covered with hands?
As we are using this MCU in a medical device, the occurrence of this case is very high. Is there a workaround for this problem?
Has any one else reported this kind of issues?
不幸的是,我们不能实现的throughput we require using DSPS application. Hence we are still using throughput_eval example based project.
Thank you,
Bharath
Hi Bharath,
This has to do with the antenna of the device and eventually the antenna used, and not the 580 module itself, the human body is conductive and changes the antenna characteristics, if the device is a wearable device, then you will have to include that parameter in the antenna RF design.
Thanks MT_dialog
Hi MT_dialog,
What do you mean by "include parameter in the antenna RF design" ?
How to we achieve that?
Thank you,
Bharath
Hi Bharath,
I mean that when designing the antenna, you will have to consider the human contact case in your RF design.
Thanks MT_dialog