BQ28Z560-R1 and External Thermistor Power-on

Hi Iderhn,

That's a rather strange behaviour. Is this repeatable accross different boards or just one board? Is the TEMPS bit set or cleared in the pack configuration register of your data flash?

thanks

Onyx

Hi Onyx,

We are seeing the exact same behavior with this part.  For about a week now we have been unable to communicate with this part but after seeing the above idea, we pulled our thermistor off the board and I2C communications started working.  Prior to removing the thermistor, it would never respond to its address.  We have another I2C device on the bus, so we have been able to verify that I2C communications are working appropriately with other parts.

As with the initial poster's comment, we are using the same layout as the reference design along with the same value for the thermistor (10K ohm).  Since we have not been able to communicate until just trying this idea, I do not know if the TEMPS bit is set or cleared.  It SHOULD be at it's default value, but I cannot say for certain.  I can say that it is 100% repeatable across our boards.

Also, we were first able to communicate with a board built up that only had this circuit on it.  It is the circuit we used to test with the EVAL module.

Any ideas on why the thermistor seems to be causing the issue?

Thanks,

Eric

Eric,

Can you hook your board to our EV2300 and then to a computer. Install the EVSW for the device and see if  you are able to communicate with the device this way without having to take out the thermistor?

If yes,ps send a screen shot of the DataRam screen and your gg files.

thanks

Onyx

Hi Onyx,

No.  Even with the EV2300 and the software, we were unable to communicate with the part UNTIL the thermistor was removed.  I'll will get a screen grab of the DataRam for you once I get it hooked back up.

Best,

Eric

Eric & Onyx,

It appears the TEMPS bit (enabling the external thermistor) is '1'/enabled out of the box.  Setting this to '0' and powering on with the thermistor connected results in successful communication.  I can then re-enable the TEMPS bit and the device continues to work, and did so across a few rapid power cycles (disconnect/reconnect to cell).  Not that this is a particularly common series of event because once the cell is attached, it should remain attached.  This is an issue with initial bring-up as each board will now have to go through an extra rework step to attach the thermistor after the cell is attached.

I unfortunately only have 1 board right now using a component I received as a sample from TI.  My actual order will not arrive for a few more weeks.

Hey guys,

I haven't seen this behavior on my EVMs here. Can you let me know what version of firmware is on your chips?

thanks

Onyx

Hi Iderhn,

Thanks for the update.  For the moment, we are going to use the internal temperature sensor as doing the process you described is not really an option for us.  It would be nice to understand why it's not working if the thermistor is attached.

Onyx - No screenshot yet but the TEMPS bit is definitely set by default.

Eric

The eval software reports 0.05 as the Version.

Eric

hey Guys,

Can you program one of your devices with this earlier version (0.04) so we see if this problem still persists.

http://e2e.ti.com/cfs-file.ashx/__key/communityserver-discussions-components-files/180/8547.bq28z560_5F00_v0_5F00_04_5F00_build_5F00_5.senc

thanks

Onyx

FYI: I programmed my EVM which had a  Version 0.04 with a version 0.05 and my board communicates fine without the need of removing the thermistor. Let me know the results of programming a board with the version 0.04

thanks

Onyx

Onyx,

Changing to version 0.04 did not improve the power-up behavior.  This is my sequence that I can reproduce with version 0.05.  The value of TEMPS does not matter.  Attached is a copy of the flash image.

8507.8547.bq28z560_v0_05 - ORIGINAL2.senc.zip

Steps:

0) Device disconnected/unpowered.  Thermistor is attached.

1) Power is applied (about 3.6V).  <--- Current draw is unmeasurable.

2) EV2300 is attached.  The I2C pullups are powered direct from 5V rail coming off EV2300.

3) bqEVSW for BQ28Z560 is started.  Correct version is selected from supported target list.  Click through NACK message.

4) Enable Load under autocycle menu (2 Pins -> Enable Load (I2C Vout = 5V)

5) Perform any type of I2C activity (e.g., read all dataRAM).  <--- Measured current jumps and holds at about 285uA

6) Disconnect one lead of thermistor.  <--- Current jumps around for a few seconds.

7) Device now accessible using the EV2300.  <--- Current follows pattern between 50 and 120uA limits.

Does this sequence work for you?

Iderhn,

Before starting the EVSW, you should wake your device up by either using the wake-up button(doesn't not work all the time) or momentarily short pack plus and cell plus..The EVSW is supposed to automatically detect the bq28z560 upon start up without you having to select the part from the target list.

My EVM always communicates. Do you have an EVM which you could test?

thanks

Onyx

Aha! That makes more sense now. I was wondering why your boards didn't work if they were a replica of our EVMs. good find Eric!