CC85xx RESET and programming problem

Mikelis Studers

Other Parts Discussed in Thread: CC8520, CC8530, CC8521

I have created a custom design for CC8520 for my product's application. There is a problem for CCDebugger to recognize the device. It seems that there is no answer from the device after CCDebugger tries to reset it. Therefore, CC8520 device is not found and cannot be programmed with any image. I have followed the reference design for both schematic and layout. The problem persists for some of the boards and not for the others (ration around 50/50) from the same manufacturing batch.

I have seen similar issue here: http://e2e.ti.com/support/wireless_connectivity/f/155/p/59518/223762.aspx#223762

Placing a pull-up resistor or RC filter on the RSTN pin didn't help to solve this issue.

Two interesting notes:

1) As some of the boards are recognized while others are not, makes me think there is no schematic or layout issues in the design.

2) If a target board is programmed once, it is always being recognized and is available for programming later (even it couldn't be recognized for couple of CCDebugger resets before). So it seems like the first initial programming of the device is somehow 'hard'.

PPW Configurator Firmware 1.4.2

CCDebugger Firmware version 43

Do you have any ideas? Similar issues encountered?

over 12 years ago

0 Fredrik K over 11 years ago

TI__Guru** 106705 points

Hi Mikelis,

Can you post your schematic?

Cheers,

Fredrik

0 Mikelis Studers over 11 years ago in reply to Fredrik K

Prodigy 250 points

Hi Fredrik,

What regards programming interface there are direct connections (no intermediate components in series, nor in parallel) from the custom programming pin header to CC8520's corresponding signals: CS_N, SCLK, MOSI, MISO, RSTN. +3.3V and GND also connected to their respective signals on the board.

Can I send the schematic as a DM or email?

Thanks!

0 Mark Davis over 11 years ago in reply to Mikelis Studers

Prodigy 205 points

I have a custom layout for the CC8530.

I am having the same issue without the level of success. Do you have any suggestions of how or what to monitor to verify the processor is even running?

I would be happy to send the schematics on a private channel.

Mark

0 DSPedro over 11 years ago

Prodigy 205 points

Same issue here. I built up a version of the CC85x1 USB board and now I cant get it to program. Would love a solution!

0 Pavel Ruzicka over 11 years ago in reply to DSPedro

Expert 1720 points

I had troubles with programming and I had wrong crystal. It is required crystal with 48MHz fundamental frequency. Many 48MHz crystals have 3rd order frequency which means that they work on 16MHz which is wrong. In many shops are not specified if crystal is running on the fundamental or 3rd order frequency. Please check it with manufacturer datasheet.

0 Mark Davis over 11 years ago in reply to Pavel Ruzicka

Prodigy 205 points

Pavel,

I am using X20-1010F16F48.000MHZ from HM International. This crystal was referenced in another post on this forum. From the datasheet, it is fundamental 48 Mhz. We removed a crystal from a TI Headset evaluation board and replaced it with a HM crystal. The board functioned properly and I was able to program it several times.

I am verifying with the manufacture as a double check.

Any other suggestions?

0 Pavel Ruzicka over 11 years ago in reply to Mark Davis

Expert 1720 points

Maybe there can be troubles with decoupling capacitors or PCB design. It is a good idea to keep PCB design recommendations in a datasheet.

0 DSPedro over 11 years ago in reply to DSPedro

Prodigy 205 points

I have solved the issue with my board. I was hand soldering all components and think some had a poor connection, particularly the crystal and cc8521 which have underside pads that cannot be reached by a hand solder. I used a heat gun to reflow the solder on the PCB and was then able to program all of my boards.

Hope this helps!

0 Mark Davis over 11 years ago in reply to DSPedro

Prodigy 205 points

We just solved our programming issue as well. The data sheets between the Epson crystal and the HM International crystal show the footprint differently. One from the top side, one from the bottom. Both show the crystal internals from the top.

We got it wrong.

By hay-wiring a crystal correctly, we were able to program the cc8530.

Time to print more boards. :(

0 Mikelis Studers over 11 years ago in reply to DSPedro

Prodigy 250 points

I also got most of the non-working boards fixed by reflowing the CC85xx chip. My guess is that the GND pad under the chip is sensitive to soldering quality. It would be nice to hear from TI how to avoid this problem, for example, suggestions for a proper reflow temperature profile or layout considerations, like number of vias under the CC85xx. Maybe there are some TI guidelines on this?

0 Fredrik K over 11 years ago in reply to Mikelis Studers

TI__Guru** 106705 points

Hi Guys,

I am happy to hear that you figured out the various issues with your boards.

Mikelis: Yes, we do have a document which should answer all your questions :-) (maybe except for the recommended number of vias in the EGP which is 16): http://www.ti.com/lit/an/slua271a/slua271a.pdf

Cheers,

Fredrik

0 Mikelis Studers over 11 years ago in reply to Fredrik K

Prodigy 250 points

Thanks Fredrik!

The resource you provided is very good for reference.

Mikelis

Audio

Audio forum

CC85xx RESET and programming problem