USBi Problem

Update #1

I installed everything on another Windows machine and everything installed fine and the device was recognized.

So it seems like there is some kind of problem with the driver. Is there a way to remove the driver so that it can be reinitialized?

Update #2

When I unplugged and replugged the USBi into the machine that it just installed correctly on, I got the same error. So there seems to be a problem using the USBi after the initial driver install, when it has been unplugged or the computer has been restarted/shut down.

I'm using Windows XP, by the way, any ideas as to what might be going on?

Hi mdot,

The USBi uses the Cypress EZ-USB FX2LP, which currently only has drivers for Windows. Neither it nor SigmaStudio has Linux driver support yet, so connecting the USBi to a Linux computer might have unpredictable results.

I have never seen a USBi work once on a machine and then cease to function directly after. Perhaps uninstalling the drivers and starting from scratch is the best option.

Does the device appear in the Device Manager when you have it plugged in to the machine?

Hi Brett,

The USBi uses the Cypress
EZ-USB FX2LP, which currently only has drivers for Windows. Neither it nor SigmaStudio has Linux driver support yet, so connecting the USBi to a Linux computer might have unpredictable results.

Yeah, the linux thing wasn't about thinking it would work, I was just trying anything to verify that the board wasn't fried.

I have never seen a USBi work once on a machine and then cease to function directly after. Perhaps uninstalling the drivers and starting from scratch is the best option.

Well I can guarantee that it was happened on two different Windows machines that I have tried. I uninstalled everything and reinstalled, didn't help. I even used Revo Uninstaller to clean up any thing that may have been left around, still causing problems, didn't help. When I logged out of my domain account, on my laptop, and just logged in as the mocal machine administrator, I plugged in the USBi and it went through its normal drive installing proceedure and again, was working fine.

I unplugged it, plugged it back it, "USB device not recognized".

Does the device appear in the Device Manager when you have it plugged in to the machine?

It does not. I get the little information bubble over the tray, when I click it, it just takes me to a "device" window basically saying that it doesn't know anything about the device. I don't know this this has anything to do with running 3.4 or not, but it is definitely weird.

I attached screenshots of the notification balloon and the window.

Message was edited by: mdot - Removed broken image link.

Brett,

I think I'm going to table this issue for the time being.

I'm seeing some irratic behavior, but I'm not ready to eliminate either the computers, or the methods that I am using...

Hi mdot,

Based on what you're saying, I wouldn't rule out our hardware either. The factory test for these boards basically involves plugging them into a machine that already has the driver installed, then running through a series of I2C and SPI read and write tests, with various addresses and data sequences. There is no unplugging/re-plugging of the boards, so if this failure is somehow related to the unplug/re-plug sequence, our factory test would not catch it. Let me know if you find anything out. We can always try sending you a new board that has undergone unplug/re-plug testing in order to see if this is a problem with the USBi or something else in your test system.

Let me know! It's really no problem to send you another board to try out.

Hi,

First thing, I want to know whether your giving power through USB or some external supply.

If you are giving the power through USB, please remove it and give supply through some external supply.

Steps to be followed:

Connect the one end of USB cable to Hardware and let the other end be free ie dont connect to PC.

Run the USB initialization code through VDSP tool.

Leave for (15-20) seconds

Connect the free end of USB cable to PC.

Make sure that you had installed the USB driver correctly and you have VDSP tool 5.0 update 8 version.

Thanks and Regards,

Prashant.

Prashant,

I'm out of town on business so I have not been able to actually try it on an external power supply. However, I was using it on a self powered USB hub, by itself, so I don't think it is a current drain thing.

Brett and Prashant,

I played with it a little more before I had to fly out and it seems to me that, be it my PC or the device itself, there is an non-trivial amount of time that must pass when the device is unplugged, before it can be recognized again. If this is not being observed by others, than there may be something funky going on with my board. You guys can do a quick experiement to see if you experience the same issue. Unplug you USBi, wait about a 5 count, then plug it back in. If your device is recognized, then i may have a quirky board. If your device is not recognized, I would be curious if Cypress would classify this as normal behavior.

Thanks for your follow ups guys!

Hi mdot,

I am able to unplug and re-plug the board repeatedly, with any amount of time in between, and I have experienced no problems. I still think that replacing the board might be the quickest way to see if the hardware is causing the problem.

Of course, if a replacement board doesn't solve the issue, then we're going to have to dig deeper into the possibility of driver issues, etc...

Brett,

I was trying to save both parties the effort of shipping, but I'm going to have to agree that there may be an issue with the board.

What is the next step in the process?

I'll send you a private message and we can take it from there.

For posterity, I received a new USBi two days after Brett's post, and it worked perfectly.

So it was a defective board...

My USBi is constantly giving me driver problems!

Hi Brett,

Please help me in solving the "USB Device not recognized" problem. My PC is not detecting the USB device. Please find the attachment of error message I am getting. I am running the VDSP USB module code given by ADI through JTAG and connected the USB to my pc, but my pc is not detecting the USB device. Please help me.

Thanks & Regards,

Prashant

Hi Brett,

Please help me in solving the "USB Device not recognized" problem. My PC is not detecting the USB device. Please find the attachment of error message I am getting. I am running the VDSP USB module code given by ADI through JTAG and connected the USB to my pc, but my pc is not detecting the USB device. Please help me.

Note: This is board is working fine ie USB is connecting earlier, but now its not connecting.

Thanks & Regards,

Prashant

Prashant, when you connect the USBi to the PC, do any of the USBi's LEDs light up?

Brett, I am not getting which LED you are mentioning.

Prashant, there are found LEDs on the EVAL-ADUSB2EBZ:

D1 (I2C monitor)

D2 (GPIO - unused)

D3 (SPI monitor)

D4 (Power)

When you plug the board into the USB port, D4 should light up, indicating power. The boot process then starts. After 250 ms or so, the I2C LED (D1) should light up. So, you should see D4 (red) and D1 (yellow) if the board is properly powered up.

Hi Brett,

Thanks for the response.

I am not using EVAL-ADUSB2EBZ, I am using Custom board.

Prashant, since this is a custom board I think I need to put you directly in touch with our software team. I'll have one of the software engineers contact you soon.

Brett,

Thanks for transfering my issue to software team, but please do it as early as possible.

I dont know how you are transfering but still now I didnt got any such mail or responses from ADI.

I am waiting for your response.

Thanks & Regards,

Prashant.

Prashant,

I spoke with the software team and they were not aware of any sort of collaboration with working on a custom USB interface for SigmaStudio. Actually, from your previous post, it looks like you're using Visual DSP, not SigmaStudio. Visual DSP is compatible with Sharc and Blackfin, but not with SigmaDSP. Are you using Sharc or Blackfin? If so, you should post this support request in the respective forum:

Sharc Forum:

https://ez.analog.com/community/dsp/sharc-processors

Blackfin Forum:

https://ez.analog.com/community/dsp/blackfin-processors

Or the Visual DSP Forum:

https://ez.analog.com/community/dsp/visualdsp

Hi BrettG!

My name is Chloe from China.

I have met the same difficulties as Mdot! It is about the USBI problem of ADAU1701.

I have tried every methods to solve the problem, just like what Mdot did, but the USBI still can not work.

My computer system is Windows XP, and I have installed SigmaStudio and the driver of Cypress.

When I plugged the USBI to my computer, only D1 and D4 (LED) could be lighted up.

(Here I have got two pictures for your check. Please find the attachment.)

But the compter reminded me that the "USB device not recognized" .

I bought this board about ADAU1701 and USBI from a Chinese ADI agency. And I received the board after 3 months. What a long time!

According to my experience, I am sure there is something wrong with the USBI.

Now I need another USBI board to try again.

Would you please send me another one for me now? It is very urgent for me!

You know, if I buy another one from the agency, it will still take me 3 months too!

I can not wait such a long time!

Would you please help me about this?

Your kindness help would be very appreciated.

I am looking forward to your earliest reply.

Thanks very much!

Chloe, I will contact you via private message.

Hello,

I have a problem with my USBi. Im using the EVAL-ADAU1442EBZ .When I plug it out and in my PC don't recognise the USBi anymore.

To get it working again, I need to start my PC in safe mode open the device manager and delete

the USB device. After that I restart in normal mode and install the USBi again and it's working again. Im using Windows XP.

At my Laptop (windows 7) the USBi stoped working 2 or 3 times while I was using it, without plug in/out.

Do you have any idea to solve the problem?

Hi

Your USBI is just to buy it? I suggest that you replace the USBI, I purchased the first piece of USBI not work, so I replace the USBI,than it is no problem.

Hi Christian,

Several users have reported problems with the USBi regarding drop-outs and "Device Not Recognized" errors. I wonder if this might have something to do with the Cypress drivers that work with the CY7C680. Can you verify that when you plug in the board, you always see the D4 and D1 LEDs in an "on" state?

Also, when you get a "Device Not Recognized" error, have you tried switching the USB port where you connect the USBi?

Some users have reported that by closing SigmaStudio, swapping the USB port, and re-starting SigmaStudio, the problem goes away.

We can always try replacing the board as well.

-Can you verify that when you plug in the board, you always see the D4 and D1 LEDs in an "on" state?-

The LEDs D1 and d4 are always on.

-Also, when you get a "Device Not Recognized" error, have you tried switching the USB port where you connect the USBi?-

I tried connecting the USBi to different USB ports

-Some users have reported that by closing SigmaStudio, swapping the USB port, and re-starting SigmaStudio, the problem goes away.-

I tried it that way but it isnt working.

Just the way which I described in my first post is working.

I would like to replace the USBi board.

@BrettG can you contact me via mail, because when I send my problem to the support mail
I never got an answer.

regards,

christian

Haven't noticed this thread here. I also suffered from the same problem twice so far.

The first time I believed that that hrdware went over Jordan. Despite that suspicion I tried rebooting the EVAL board first then I tried to close and reopen the SigmaStudio software combined with unplugging and replugging of the USB - but to no avail. The next step was to de-install and re-install the driver - still not working. Even a de-installation and re-installtion of the SigmaStudio software didn't help. I then deinstalled SigmaStudio again and installed the latest updates of the OS. After the installtion of the SigmaStudio software it worked again.

Yesterday the same problem occurred again. But it worked again after a de- and reinstallation of SigmaStudio.

The annoying thing about this is that it appears suddenly when you do frequent link-compile-upload cycles. Only the first time when the problem apears you get a warning from SigmaStudio that there would be a communication problem. If yo don't notice that one you will try subsequent without being warned - you just wonder wha your chnages don't have any effect.

You may see that the colour of the USBi changes in the topology window. But even this may happen unnoticed since there is not abig difference between the yellow and orange colour indicating the USB status.

I suspect that it is a software issue rather than hardware - but who knows ?

Regards

Charles

charles.l wrote:

I suspect that it is a software issue rather than hardware - but who knows ?

I agree. In my experience, boards that suffer a hardware issue become completely nonfunctional, likely due to EEPROM data corruption. More here: Re: How to avoid corrupting the USBi's EEPROM

I expect that this might be a problem with the Cypress USB driver or our way of interfacing to it. However, I'm happy to report that we're working on a new USB interface that will not only provide SPI/I2C communication to the SigmaDSP, but will also stream audio to and from the PC. It will show up in Windows as a standard audio interface. I'm unable to provide a release date for that at the moment.

charles.l wrote:

You may see that the colour of the USBi changes in the topology window. But even this may happen unnoticed since there is not abig difference between the yellow and orange colour indicating the USB status.

True! I will petition the software team to change the colors to red and green, which would be much easier to distinguish.

Hi Brett,

I also suffered from the USBi Problem.

I work with the EVAL-ADAU1761Z. I link-compile-download several projects via this USBi to the eval-board and everything was ok.

But now: Communication failure.

I read the posts above and can say:

The LEDs D1 and D4 are always on

SigmaStudio recognizes the USBi because of the orange colour

If I start link-compile-download the statusbar first indicates Active:Downloaded but suddenly Compiled,Comms Failed and the software was not downloaded to the evalboard. After pushing the download button the D1 beginns to blink but never stops. I have to disconnect the USBi to stop that. If i don't disconnect before I restart SigmaStudio the colour of the USBi in the hardware config is red.

I did a software reset, described in AN-1006, but nothing happends.

I also swaped the USB-port and swaped PC, but nothing happends.

I need the USBi very urgend because of my thesis.

Would you please help me about this?

Do you have any idea to solve the problem?

Thank you very much.

Hello L.o.u.,

I have contacted you directly to arrange a replacement.

Brett,

I'm afraid I'm also having a USBi problem. I had mistakenly posted my question (actually a request for exchange of my ADUSB2EBZ, because I have determined by substitution that it's bad) in a new thread before discovering this one. I have deleted that thread and transferred my original text below:

My USBi (ADUSB2Z) has been intermittent since receipt a few weeks ago, and now stopped working altogether (gives "USB comm failure" errors). It shows up as "unknown device" in Device Manager. Re-installing the driver didn't help. A fellow engineer here has the identical set of eval boards so I swapped in his ADUSB2Z, which worked immediately with my ADAU1701. What should I do to ... get it replaced?

Please arrange for an exchange of this device for me.

Thanks, Bob

Hi KJBob,

I was pulled into several high-priority, time-critical projects at work over the past few weeks and unfortunately I had to de-prioritize my participation in the forum as a result. So, I apologize for making you wait so long for an answer. I am going to try to respond to everybody's questions within the next few days.

I will contact you with a private message so we can arrange a replacement.

Brett,

Thanks for your kind assistance! BTW -- Back in the day when analog was really analog, I had enjoyed designing in AD chips (multipliers, RMS-DC converters, etc.) because they just worked without a lot of coaxing. I'm happy that AD has transitioned this fine quality into the digital age.

Hi all,

I just want to mention that I also have problems connecting with the USBi to my EVAL-ADAU1701MINIZ.

I try to compile and download the program but I get a communication error some seconds later.

Is there meanwhile a solution for this issue?

Regards

Christian

Hi Christian,

One thing to make sure of is that you're connecting the USBi to a USB 2.0 port. The USBi doesn't support USB 1.0, and I'm not sure if it works with USB 3.0 either. My laptop has several 3.0 ports and only one 2.0 port, and that's the one I've successfully used.

Bob

Hi Bob,

first I tried the USBi at the USB port on my docking station. Here it never worked.

Then I connected it directly to my Laptop and now it workes "sometimes".

Every 5th time or so when I try to download a program it fails and the whole USB connection is corrupted.

I need to disconnect the USBi, restart SigmaDSP studio and connect USB again and in 90% after that it works again for the next 4-5 downloads.

Once the download was successful the connection is ok i.e. I can change parameters without an USB failure.

And yes, according to the user manuals and the driver version all USB ports should be USB2.0.

Regards

Christian

ChristianH,

I'm back from my vacation now. Sorry for making you wait.

I'm curious - what OS do you have installed?

Thanks,

Brett

Hi Brett,

it's a HP 8730w Notebook with Win XP 32Bit SP3 on it.

Well maybe the solution is mentioned here:

https://ez.analog.com/message/38547#38547

My USBi was delivered some days ago but seems to be without that modifications.

Regards

Christian

There is something funny going on with the programming board.

What is baffeling is that it works sometimes. Does not just go dead

Below are my personal, random thoughts on what could be going wrong.

1. Windows recognises the card every single time it is plugged or unplugged. (All: kindly confirm if this is not true in your case). So the Cypress chip which is the front end to the OS looks ok. The Cypress chip then communicates to the microcontroller as and when required. This is where there could be some in-consistancy

2. Plugging the USB is analogous to a switch and may have de-bounce. Is it possible that the cypress chip is communicating to the microcontroller, or vice versa, too soon? Is there a delay? If yes, is it sufficient?

3. Would it be good to have a reset switch for the microcontroller (U3) so that in a bad communication event, one can reset the uC without having to un/re plugg the cable? With R5 in place, can I try shorting C1 to see if the system resumes communication?

Wish ADI had revealed more design details (firmware)

On the the brigh side,

Working on the eval board is a lot of fun. Lot of low level blocks, lot of high level blocks. Sigma DSP/Studio has made huge progress since I worked on 1st gen AD1954 eval board for which I had to enter my design theough OrCAD capture, a tool used to enter hardware designs schematics and circuit simulation.

Finally I can work on multiple ideas without having touch the soldering gun!

Keep it up ADI!!

kksl

Note: All: This post is my way to convey my thoughts to the engineers @ ADI. Kindly do NOT act by yourself on this.

Hi KKSL,

The symptoms people have been reporting with intermittent failures (but not complete failures) of the USBi are hard to debug because they are difficult to duplicate. I like your idea of the reset button, and I will look into including that in the hardware update. I think your idea in #2 is interesting... there might not be long enough of a delay in the power sequencing between the chips. I'm also looking into physical aspects of the design, like a lack of proper isolation between the Cypress chip and the bus, that could be playing a part. I'm hoping to release a new board within a few months when the PCB design is done.

I'm glad you enjoy the SigmaDSP. Hopefully we can fix up these USBi issues to make it even more enjoyable.

Hello Brett,

Here is some thought for the new programmer design:

1. The programmer should be able to independently reset the eval board. However, the reset button on the eval board should reset both the programmer IC as well as the sigmadsp on the eval board. Should not be difficult since the eval board (which we do not want to change) has its reset line coming to the 10-pin header.

2. The earlier programmer had a very good feature: It was able to store several programs from signastudio and load any on of them [without the need of PC (sigmastudio)], by using a rotary, selector switch. It allowed me to test several variations with minimal intervention. You must re-introduce this feature in the new design, even if you have to un-freeze the design.

Now on the problems of USBi:

On power-up, the board loads the signal flow from the EEPROM, or looks to the EEPROM to see if there is a valid data in there.

Does the programmer reset the sigmadsp before loading the new flow? Is there any chance that after reset is performed, there is an inconsistency and every once in a while, the sigmadsp would have already started loading from the EEPROM even before the programmer WRITE operation begins?

Also, does the programmer erase the sigmadsp COMPLETELY before loading the new flow? If not, will the residual data from the previous flow be causing mis-behaviour?

Lastly, after the compile-link-download button is pressed, and the board mis-behaves, the yellow LED continues to blink at a fast rate? Why should this happen? Does the programming firmware get into some endless loop for some reason?

Do let me know your thoughts.

kksl

I have a small update:

I have noticed a better 'hit-rate' by following the below method:

1. If there is a download failure, save your project and shut down the sigmastudio and remove the USB cable.

2. Re-start the sigmastudio. Then remove the existing instance of USBi.

3. Plug the USB cable and then add a fresh instance of USBi and make the necessary connections. It should appear yellow/ orange, not red.

4. First link/compile/connect and then link/compile/download

I hope this improves the situation for all other sufferers too.

Thanks for the suggestion about the method for reconnecting the USBi, KKSL. I wonder if this somehow "resets" the interface from a software perspective. I will try this as well.

The USBi does not overwrite the entire memory of the SigmaDSP with zeros upon download. However, that's OK. For program RAM on earlier SigmaDSPs, SigmaStudio will generate zeros to pad the end of program memory beyond the end of the actual program. For newer SigmaDSPs, this is not necessary because the last program instruction is always "jump to start"... In other words, the rest of program RAM can be filled with random data and it will be no problem, because the program counter will never go into that territory.

We discussed the option of adding memory to the programmer so the user could load programs onto target hardware without a computer. However, there are several different SigmaDSP architectures and each one would require specialized code in order to get this to work properly. There would also need to be some way to notify the USBi, using hardware controls, of what SigmaDSP it is interfacing with. The firmware is currently not set up to deal with that. So, for the moment, the complexity of implementing that functionality outweighs its benefits, so it is on hold.

When the LED keeps blinking indefinitely, this means that the target hardware has not responded to a USBi request. There is currently no time-out code in the USBi that will cause it to break this indefinite loop. I will look into the firmware to see how easy it would be to add a time-out.

Hello Brett,

Thanks a lot for catching my blunder with the logic levels!!

At present I am putting off porting ADUSB1 to the AD1701 eval board, but for a different reason: I am exploring some existing pain areas and some audio theories and need the continuity of work. I am so thrilled by the

capabilities of this sigmadsp platform that I will want to put off any interruptions in my work & the thought process.

Imagine to be able to practicaly verify the Haas effect in minutes!! I wonder if the 'virtual' sound theory is based on this effect.

Also, there are a few things in the audio world that have built enough suspense in my mind. I am slowly but surely finding answers to those.

Things like:

1. In the mono era, 600 seat capacity movie theater had a 40Wrms or a 60Wrms vlave amplifier that was more than sufficient to entertain even the noisiest, full-house crowd. Why, then, do we, in the modern era, require

600Wrms to 3000Wrms per channel amplifiers and yet go home with a dizzy head?

2. The most in-famous problem is: Its too loud, but, I can't hear anything!!

3. Next in the line is dialogue intelligibility. Surround sound platform promises equal acuostic power. I wonder how much of it is really being fulfilled. I have observed that the surround decoding needs a minimum

reference level (Volume set on 7 on Do**y decoders in movie theaters) for the system to perform nicely. Below the reference, the soundfield collapses. So I am FORCED to use a minimum level.

4. I understand that calibration can improve sound/ tonal quality. But to be forced to do calibration to achieve even basic sound is something ridiculous.

The best part about audio is also the worst part: that there are no rules. What sounds good to me can invite a bad comment from the next person.

One theory supports giving all kinds of controls to the end user (who almost always is audio-illeterate) to SELL a product when most experts struggle trying to correctly set a three band tone control!

I am looking into a possibility of taking sound recorded with the modern (digital) techniques and passing it through a DUMB (no end user controls) box, which of-course uses a sigmadsp, to give a good quality sound output that will address the above problems and be good for most people.

Let's see where it leads me to.....

Kindly check your PM

Thanks in advance for your time and efforts!

For USBi replacement, will you allow me to surrender my board to an AD office near me instead of sending it to US? It will save me a lot of efforts and money. I can pick up my replacement from the same AD office or it can be sent to my address.

Hi KKSL,

Yes, that's fine. Let's coordinate via PM.