Generated object code loader can't find Stamp.
Phil Pilgrim (PhiPi)
Posts: 23,514
I've got a problem with a standalone object code loader generated under the Editor v.2.3.1 for a BS2pe v1.5. Although the editor can upload programs to this Stamp, the standalone executable generated by File->Generate Object Code->Single Executable cannot find the Stamp, even though it recognizes both the Loopback and Echo on the port. Has the standalone loader code been upgraded to handle the v.1.5 BS2pe?
Thanks,
Phil
Thanks,
Phil
Comments
I'm not sure about the BS2pe1.5 and editor v.2.3.1 specifically, but I think the problem may go deeper and further back than that. It is something I've been meaning to look into and take up with Jeff. I use the loader generator quite a bit for far away program updates, and people have been complaining that the loader doesn't work except as a workaround on one or two computers when we generate the loader in a particular way.
Here is an observation. Create loader on an old Fujitsu running WIN98 and having a real serial port. The loader works fine for programming a BS2pe of either version 1.3 or 1.1 or the Stache. What is more, I can take that loader to a new Toshiba running XP and that has a USB-keyspan pda adapter serial port. The loader created on the Fujitsu will program both chips and the Stache just fine. But now, if I take the source code and create the loader on the new Toshiba, that loader will program neither the two BS2pe revisions nor the Stache. Same source code, same USB-keyspan adapter. The Stamp IDE itself has no problem, as you observed. Now, I take that loader created on that new Toshiba back to the old Fujitsu, and, what do you know, it works fine to program both chips and the Stache. Go figure.
So the upshot is that I have been creating all my loaders on the old Fujitsu and asking my customers to keep around an old PC, because some of them are also having USB-adapter problems (that is a different issue!). The point is weirdness, that the loader created on the old Fujitsu behaves differently than the one created on the new Toshiba.
I don't have a BS2pe rev 1.5 to try, but my guess is that is not the issue. Do you have an older one you could try? (Note to readers: Phil and I have special numbered OEM versions of the BS2pe that have the brownout set at 2.6 instead of 4.2 volts.) I do have a rev 1.6 BS2p that I could try. Also, I might mention that this problem as I see it goes back to at least version 2.5.5 of the IDE. Not new in 2.3.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com
Thanks for the feedback, Tracy!
-Phil
Object code (or rather, a Stamp Loader.exe) generated from Stamp Editor Version 2.3.3 should correct the problem.
The code in the Stamp Loader is a derivative of that of the Stamp Editor, however, they have to be maintained separately.· As it turned out, someone <sheepish grin here> forgot to make one tiny change to the code in the Stamp Loader when that same person noted and fixed a problem in the Stamp Editor that dealt with serial ports.
----Ugly details here----
Specificially, some USB-to-Serial devices will process a Break condition and DTR sequence out-of-order.· For example, on a standard serial port, setting DTR high, then setting a break condition, pause, setting DTR low,·pause,·then end break will cause the port to do exactly that, in that order.· However, some USB-to-Serial devices will, with the same sequence of events, effectively generate the break first, then will wait until the break is completely over before setting and clearing the DTR pulse.· Ugh!
The fix (that I found quite by mistake while losing a pile of hair in front of the 'oscope) is to set DTR first, then break, pause, clear DTR, break, end break... this causes normal serial ports to faithfully reproduce what I instructed (which, luckily, works okay for the Stamp) and the USB-to-serial ports in question to start a break, then set DTR, pause, clear DTR, pause, then clear break (which is what I was trying to do in the first place).
----End of ugly details----
The editor v2.3.3 includes Stamp Loader v3.2, which swaps those two lines of code to match that of the Stamp Editor to support these devices.
Oddly enough, the code has been this way in the Stamp Loader for many, many months (years perhaps) and I hadn't heard people say that it wouldn't work.·
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
--Jeff Martin
· Sr. Software Engineer
· Parallax, Inc.
Post Edited (Jeff Martin (Parallax)) : 9/18/2007 10:38:49 PM GMT
-Phil
Tracy, when you get a chance, can you try it on the systems you've been struggling with?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
--Jeff Martin
· Sr. Software Engineer
· Parallax, Inc.
By the way, the version 2.3.3 IDE and loader continues to run perfectly fine on the ol' Fujitsu running Windows 98. The Parallax downloads page only mentions 2K/XP/Vista.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com
·· And the Editor may work under Windows 98 for quite some time.· We don't list it for two reasons...One is that we cannot support it on that O/S.· The second is related to the first, which is that we don't have a Windows 98 PC to test/verify compatibility on.· But, I am sure your feedback will help the forum members who continue to use Windows 98 and were wondering about compatibility.· As a side-note the newest version of the BASIC Stamp Editor is required for current versions of the BASIC Stamp Models listed in the firmware updates on our website.· Take care.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Tech Support
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
--Jeff Martin
· Sr. Software Engineer
· Parallax, Inc.
-Phil
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
--Jeff Martin
· Sr. Software Engineer
· Parallax, Inc.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com