ADS1298ECG-FE device not recognized

Dean -

For programs that are compatible with Win7, you shouldn't have a problem running 32bit apps on a 64 bit machine.  ADS1298 is not compatible with Win7 yet, we are working to get this done.  We hope to have this done shortly.

However, regardless of the compatibility with Win7, all of our apps will remain 32bit with no plans to change them to 64bit.  We must keep compatibility with older systems, but the fact that it isn't 64bit shouldn't cause it to not work on your machine.  We have tested the upgraded version of ADCPro, which is a 32bit app, on both 32 and 64bit Win7 machines without issue.

Hi Greg,

Thanks for the reply.  The issue is not so much the application, which can be run in compatibility mode.  It is the driver, which must be 64 bit to match the O/S. It seems as though the driver is actually required for proper operation :-)

Actually, it's more like the INF file has to be rewritten to point to the 64 bit versions of the driver DLLs etc.  I was involved in a similar issue working at a different employer last year.

So for now, it won't run (properly) on a 64 bit system because it doesn't have the driver files installed (set up?) properly.  I've brought my personal notebook computer in from home, to load up the software and give it a try.

Regards,

Dean.

Dean -

Yes, it is the driver and hence the current version does not work on Win7.  We are updating the application also, since it needs to be ported forward with a newer version of LabVIEW to make it compatible with our other apps and resolve issues that the newer runtime engines create (if you happen to have another app that uses the newer files).  We are also adding a couple features and cleaning up the code a little while we're in there.

So in summary,  until we release the newer version (with updated drivers), this application will not run on Win7.  We will update the product page software requirements section and user guide software requirements section to indicate support for Win7 at that time.

Hi again.

I've now loaded the software onto my personal 32 bit Windows notebook computer, and I'm getting much further along.  Now I have the dreaded "DSP not ready" error coming up.

I looked at the schematics, checked and double-checked the settings jumpers, looked in the forums and pondered.  What I have are two stable configurations, see attached photos.  The only difference between the two is that one has ADS1x98ECG-FE JP18 & JP19 set for getting clock source from MMB0 and gives a "5" on the display with the top horizontal segment missing, the other for getting clock source from local oscillator and gives an "8" on the display with the lower right vertical segment missing.

I've also had JP2 & JP20 changed because I saw a note regarding this in one of the forum postings.  No difference at all.

Do you have any further suggestions?

Dean -

A couple things to check:

1.  the "8" with missing segment is actually an "e" as is error/invalid setup.  This generally is due to a hardware configuration that conflicts with what the software needs to operate.  While the EVM supports multiple configurations, the EVM software requires the setups allowed in the user guide only (this will be a subset of the overall settings).\

2.  the "b" missing a segment is the normal startup process...does a second driver (USBStyx) attempt to install once the broken "b" appears? Two drivers need to be installed for the software/hardware to connect.  The first driver should be installed, given that you get the broken "b".  This indicates that the firmware has downloaded to the DSP and the DSP is resetting using the new firmware.  However, at this point the new driver will activate (will need to be installed the first time).

You can reference the ADCPro document @ http://www.ti.com/lit/pdf/SLAU372  The drivers used are the same, so Section 4 and Figure 10,11, and 18 detail the installation and show the Device Manger screenshots.

Hi Greg,

Thanks for the note.  Aha, I knew that the patterns must mean something.

Both drivers are installed as per the directions, matching the procedure in document SLAU372.  When I first connect the EVM (MMB0 rev D+ADS1x98ECG-FE rev C), it enumerates as TMS320VC5509A (NI-VISA) as shown in the Device Manager.  Once the ADS1298ECG-FE software has run, it re-enumerates as LibUSB-Win32 / USBStyx. 

--> It seems as though the "e" error code indicates that the progress of loading & running is further along than the configuration with the "b" error code.  I would be inclined to continue from there, so it seems like I need to use oscillator on the ADS1x98ECG-FE board (not the clock from the MMB0).  Does that make sense?

For configuration of the board, I did not intentionally change anything, just pulled out the board and tried to use it. I just want to see it run, one time - then might mess around a bit.  Maybe :-)

--> I don't see a list of allowed jumper or firmware settings in the document SBAU171B to cross reference to check that the settings might be messed up (inadvertently of course).  Or am I just missing it?

--> What would be helpful is a one "known good" starting point setup (both hardware and software settings), so if things are messed up, I could set to the "known good" and expect the ADS1298ECG-FE to work.  Is there such a starting point?

Thanks,

Dean.

Further to this... sorry for the extra message, but it just occurred to me.

The only difference between the two configurations is the positions of JP18 & JP19 on the ADS1x98ECG-FE, changing the source of the clock (on-board oscillator / MMB0 provided).

This conversation made me want to try this: http://e2e.ti.com/support/data_converters/precision_data_converters/f/73/p/164743/605581.aspx#605581    It shows that I should be using JP18 shorted 2-3 and JP19 shorted 1-2, for External OSC Clock.

Upon re-read of that conversation, I think my assumption in the previous message was wrong - the "e" code is better, as it avoids the dreaded "DSP not ready... retrying" message.

Dean -

I would say from looking at your pictures that the one with the "b" is the more correct.  Additionally, just so you know, it ("b") isn't an error code and will actually still be displayed when everything connects properly, with the addition of the decimal point.

For configurations, use the doc @ http://www.ti.com/lit/ug/sbau180/sbau180.pdf.  It is for the ADS1198, but both parts share the same hardware and we are currently updating all this information for the software update.  It should show Table 4, which you can use to verify your default configuration.  This should be the default configuration for the hardware and software to sync.

A quick note...are you using the included power supply with the EVM?  It looks (in the photos) as though you may be using something else that has a right angle connector.

Hi Greg,

OK, that makes sense. 

• SBAU180 does show JP2 & JP20 for bipolar mode, but I'm using unipolar, so I adjusted for that.
• JP18 isn't shown in table 4, probably an oversight that you might want to have them correct in the next document revision.

So other than that, there are the clock internal/external/enable jumpers JP18 & JP19, and I've tried all those already. 

You make a good point about the power supply.  I inherited this PDK from earlier work 6 months ago (I've been here about 5 weeks), and the original power supply is missing.  I'm using a bench supply set to 5.6 Vdc.  The bench supply is all I have at the moment, and it's a homebrew, not sure of its output capacity - although I did look inside and its construction and filtering are good, not clear on the regulator ICs used.  I will chase down a different supply, to see if maybe somehow that is part of the problem.  It could be plausible that this supply doesn't have adequate output capacity, causing the PDK to hiccup. 

It would be appropriate to chase the power supply angle for a bit.

Do you have make & model of original power supply, can I buy one from Digi-Key maybe?

Hi,

I found an L.T.E. GFP151U-0525 5 Vdc output 2.5 Amp in the right connector & polarity.  Could this be the missing adapter?  It's only 5 Vdc where the screening on the MMB0 says 6 Vdc.  I tried it but see no difference.

Dean.

Well as happened in at least one other case that I see in the forums, I changed computers and it worked on the new one.  I'm not all that concerned as to why it is that way, just wanted to note here that this is what fixed it.  It might encourage someone else with similar issues to simply try another computer, just to see!

I had brought in a honkin' big bench supply from home, used it to provide 5.9 Vdc on the nose with big capacity.  Seems that the MMB0 + ECG-FE takes about 250 mA when finally running, but of course you don't know what it does in the transient case.  However, I then went back and tried the little dinky adapter and it worked too. 

The 32 bit MS-Windows XP notebook computer that I brought from home did not work, as noted.  However, I actually use that notebook primarily on LINUX with VMware virtual machines to run MS-Windows 2000 & MS-Windows XP alongside.  I was able to make the program work on both virtual machines.

I've taken my personal notebook home and set up a 32 bit MS-Windows XP VirtualBox virtual machine on my MS-Windows 7 desktop here at work.  I tried to use it in the same way.  Well, I got the first part working, up to where it switches over from NS-VISA over to TTStyx, but can't connect TTStyx to the virtual machine, sigh.  It seems to be an MS-Windows 7 issue that if the driver doesn't pick up properly in MS-Windows 7, it won't allow you to assign it to the virtual machine. 

I'm going to leave this now, and move on to my own development with the ECG-FE board.  I've seen the demonstration software work and captured its settings, so I'm on the road now.

What does this tell me? 

1. That the power supply wasn't the issue, at least not in my case.

2. The program & driver work in both 32 bit MS-Windows 2000 and 32 bit MS-Windows XP.

3. The program & driver sometimes works in a virtual machine on some virtual systems on some architectures.  Not going to investigate further unless I have to.

4. If it doesn't work on one computer, try another.  Who knows why.

Dean.

Dean -

At this time, I can offer a beta of the Win7 ADS1298 software if you would be willing to try it out and let us know any problems you find.  If you would like this, please let me know your email address and I'll send you the details.

hello

have you solve this issue yet?

I can not install the drivers on my Laptop Win7 64bits, I'm using the ads129xecg-fe-setup-1.1.1 software 

best regards