SigmaStudio v3.2 and defected USBi interface
For one of my customers I am developing currently a project - studio-grade near field active monitor, a kind of active loudspeaker. This product shall be introduced to the market in the middle of the 2011. Complete audio signal processing is done by ADAU1701 SigmaDSP (active crossover, tone/volume control, limiter, etc…). The first version of the hardware is already released and first samples should be produced soon. At the moment we develop SW for SigmaDSP. Yesterday I installed new version of SigmaStudio - v3.2 and unfortunately one problem has popped out. We develop the SigmaDSP SW using a standard ADAU1701 Mini Eval-Boards with the USBi interface. I do not know when exactly it happened, I worked already more than hour with it, but all at once I notice that I can not flash my eval-board with the USBi any more. I noticed also that SigmaStudio can not detect the USBi interface and on the USBi interface the yellow I2C LED is switched off. I would not think that the USBi interface could be electrically damaged because it was connected only to the original ADI eval-board and the board was supplied from the USBi interface – no external supply, no other external connections – aoutio out connected to headphones. I tried to disconnect and connect again the USBi board but it didn’t help. It looks the USBi interface somehow doesn’t work anymore.
I had in the office one more USBi interface. I connected it to the eval-board and flashed the board without any problem – it shows that the eval-board works OK. After flashing I tried to connect back the first USBi interface but it still didn't work Next I connected back the second (working) USBi interface and now the second USBi also doesn’t work! I have no clue what could happened??? Somehow the both USBi don't work and I can not develop my project anymore.
It looks like both USBi boards behave now similar – no USB communication / I2C LED off. I checked shortly voltages on the board and everything looks OK. I check also with the osciloscope the clock of the on-board microcontroller and it looks also OK – stable, nice 24MHz clock. For me it looks like the Cypress microcontroller lost his firmware in both cases – is it possible??? Everythink looks OK (electrically) but the Cypress doesn't work. There is also complettly nothing on the USB BUS after connecting (power-on) the USBi to USB.
I worked on the SW already for several days and I never had any problems with SigmaDSD<->USBi. I've been familiar with SigmaDSP and SigmaStudio for 2 years and I've also never seen a problem like that. The strange thing is, the second USBi was “defected” right after the first one and behaves now similarly. I know the USBi Firmware can be updated directly from SigmaStudio, therefore I suppose in this case, somehow by accident, SigmaStudio damaged the Cypress firmware trying flashing SigmaDSP eval-board.
Could someone from ADI support help me? Can I somehow reflash the USBi with the new firmware? Can I send my USBi somewhere to reflash it or can I get the boards replaced – the one I bought a week before, is practically brand new. Unfortunately I can not work on the project without the USBi interface.
Best Regards,
Tomasz
Audio Signal Processors
SigmaDSP Audio Processors
adau1701
Hello Thomasz,
It looks like you may have selected address 0xA2 in SigmaStudio and overwritten the USBi's self-boot EEPROM. You can read more details about this issue here: https://ez.analog.com/docs/DOC-1422
This information also appears in the application note for the USBi, but it's easy to overlook. The best solution would be to remove that option from SigmaStudio entirely. I have requested to the software team (via our internal bug-tracker) to remove address 0xA2 from the drop-down menu in SigmaStudio.
Because of the format of the compiled binary file that we download to the EEPROM, it is a little difficult to re-flash the EEPROM yourself unless you have a Cypress EZ-USB evaluation kit. So, you will have to send the boards back to us and we will reprogram them for you. I already contacted your local FAE. He will contact you directly about shipping information.
Hi BrettG,
Thanks a million for your replay. I didn't know about this case, probably I overlooked it. I used in previous project the AD1940 full development board therefore I didn't read the user's guide switching to mini ADAU1701 eval-board. However, I never used in my system any EEPROM with a different address than the one on the eval-board. In fact the core of SigmaDSP part in our HW is more or less copied from the reference design therefore the EEPROM has also this same address. However it could happened that after reinstalling SigmaStudio I set something accidentally wrong.
I am glad ADI can flash the boards for me. I already contacted local customer support about it.
I will send the boards directly to JeradL.
I would suggest putting a small jumper on the USBi board in the future. Placing a small jumper on write protect pin of the EEPROM would solve the problem and user will have all I2C addresses available to use in custom projects.
Thanks a lot one more time,
Tomasz
Hi BreetG,
Finally I found one more USBi available; in the last moment I hold the shipment and copied the firmware. It was really easy! I would say much more easily than shipping the USBi somewhere over the ocean. Now all my USBis are working back.
The USBi firmware is not protected and anyone with minimal embedded engineering knowledge can copy the Firmware from one USBi to another. The issue is just to have a working USBi available when your one lost the Firmware.
I would suggest implementing in a future version of USBi a simple jumper on the WP pin to protect the Firmware. Having WP jumper on the USBi the user will have all I2C addresses available and the USBi Firmware will be safe.
BTW, if any one does have similar problem like I had, and is looking for the USBi Firmware, he can anytime contact me. I have the firmware saved as a binary file on my PC. The good think I learn is how to build your own USBi interface (copy the original one) – also never thought before it is so easy
We have considered locking the USBi's firmware if we do a redesign of that board. While you can read the data from one and write to another, as ChUml describes, we would recommend that you contact Analog Devices with any issues you might have with your USBi board. This way, you can always guarantee that you're getting a board with the most current, verified firmware.
Hi JeradL
You absolutely right – the best way to go having lost firmware in USBi is to contact Analog Devices, but…
No one at Analog Devices GmbH in Germany is able to update the USBi firmware!?! – I contacted the local ADI Support… lucky you guys in US
Therefore I have to send it over the ocean to US. It means for people like me (having rather small engineering office) to pay half of the new USBi price for the registered express parcel service or wait weeks using flat post sevice for the repaired device being not sure if my USBi is not lost on the way…
I would really like being able reflash ma USBi at local ADI office (at least in the country I live) and be sure having most current and verified firmware but in my case it is unfortunately not possible. I am definitely not a person who prefers doing everything by myself. I even did not have any serial EEPROM programmer. I've briefly estimated what is the fastest and less-expensive way to go with my problem, I found someone having USBi, I bought online a basic I2S programmer and copied the firmware – it cost me 35€ and about 4 days delay from the idea to having back my working USBi and being able to continue my SW development on the already slightly delayed project. And I still have the I2S programmer for future while the EEPROM is not protected.
It’s a pity, I could get the most current firmware directly from Analog Devices even though the firmware is not protected and anyone can have access to it. Such a support would help me much more
Nevertheless – the Engineering Zone makes way better support than any local office! Thanks a milliion!
I am facing the same problem, EVAL-ADUSB2EBZ cannot be recognized by PC, I2C LED does not light , before it woks fine, but recently it's bad , I am not considering I flashed the EEprom in EVAL-ADUSB2EBZ, I use Sigma studio 3.2 . Because EVAL-ADUSB2EBZ DATASHEET says it's address is 0X51, and ADI says they update software disable 0xA2, but in V3.2 we still can see 0xA2. I think I didi not operate to rewirte 0X51,or 0xA2, For ESD I am not sure. FAE let me check 3.3V and 1.8V regulator , It's ok. I checked that Q1 stops. What can I do right now?
Thanks!
Meican.Wu
I have one more EVAL-ADUSB2EBZ which is good, I swap eeprom between good EVAL-ADUSB2EBZ and bad one, the good board still good , the bad one also still bad , Is it provide EEPROM no problem? For the bad PCB, Q1 Vb=5V, Q2 Vb=0V, For good PCB, Q1 Vb=3.8V, Q2 Vb=0.7V , Seems MCU doesn't control LDO.
If ADI team can help this , much thanks for them.
Regards!
Meican.wu
Right now I fix the USBi.
Meican.wu
If you are having problems with your USBi board, or any other SigmaDSP evaluation board, I would suggest that you email sigmadsp@analog.com to request assistance. Contacting the SigmaDSP support team through that address is generally a better way to get help on issues specific to your hardware or setup. The public EngineerZone forum is a great place to discuss issues that may be applicable to many other engineers, but direct contact through email is a better way to get support for issues specific to your use.
Please log in to reply.