SparkFun Forums

[OldCow]

eZ430-F2013 (the little USB gadget) does not work with Windows 98 because Windows 98 does not *fully* support USB. The target interface
of that little USB gadget consists of 4 wires (Spy-Bi-Wire). IAR debugger
supports that.

When you use Parallel Port FET, IAR only supports JTAG-4-Wire. Thus
you have to solder 8 wires. Use up 8 pins on a 14 pin uC is too much.

I think you should be able to use the little USB gadget (under Windows 2K
or XP) to do in circuit emulation. I think what you need to do is:

(a) Unplug the tip section (F2013+LED) from the USB section.
(b) Connect 3 wires to your board, GND, RST, and TEST.

[Crane]

Thanks Old Cow,

Looks like two of the wires on the little connector go to Pins 10, 11 so the other two must be power and ground. That's an eentsie beentsie connector. Don't think I have anything close to it.

I'll have to give it some thought. Probably the best way to start is to solder a header onto the removable board and use it as an emulator. When I get familiar with the process I can try programming from the connector.

Thanks again. I think I can move on now.

[leon_heller]

Thanks Old Cow,

Looks like two of the wires on the little connector go to Pins 10, 11 so the other two must be power and ground. That's an eentsie beentsie connector. Don't think I have anything close to it.

I'll have to give it some thought. Probably the best way to start is to solder a header onto the removable board and use it as an emulator. When I get familiar with the process I can try programming from the connector.

Thanks again. I think I can move on now.

The connectors are made by Mill-Max and have 0.05" spacing leads. They are virtually impossible to buy from a distributor but I got some samples out of them and have made a PCB for the MP430F2001, which connects to the eZ430 in place of the 'F2013 board.

Leon

[OldCow]

Yes, those connections are exactly the ones you listed -- Spy-Bi-Wire.

I think you can get that kind of connector from Mouser. Or, you may
be able to solder something on the other side of PCB. But either way
the wires to your board should not be more than a few inches long.

Please post your results/experience as others and I can learn from it.

[Crane]

Leon,

It would be a nice product for SparkFun. Looks like it's the only way to go. I looked at all the stuff I have and nothing came close. Either too big or too small.

Old Cow,

All of the literature on the Olimex device says it works with KickStart so I spent some more time with it. I have a bit more understanding and confidence having seen the system work. No dice. I cannot get the printer port to work at all.

I wired a 14 pin header onto the USB gadget. It did not like the extra load on pins 10 and 11. The wires are only 2 inches long. When I disconnected pins 10 and 11 it worked fine.

I programmed the LED blinking demo into it, removed it from the USB thing and plugged it into my applications board. Nervously plugged in the power supply and, Ta Da, the little green LED blinks. It works! It works with my 10Mhz external oscillator. Now, that is progress.

I'll play with the LED program for a while to get my I/O initialization errors corrected then get back to programming the application. I think the little green LED will come in handy for debug.

[OldCow]

All of the literature on the Olimex device says it works with KickStart so I spent some more time with it. I have a bit more understanding and confidence having seen the system work. No dice. I cannot get the printer port to work at all.

I do not have the Olimex device. But I have a device from TI called "MSP-FETP430F 1.4". I believe the Olimex device is almost a copy of this TI device. I wired the JTAG connector of this TI device to a MSP430F2013 and it works fine with IAR Debugger. But I need 8 wires. IAR Debugger does not work when I use only 4 wires.

I also have a TI device called "eZ430-F2013". It has two (detachable) parts. The 1/2-inch smaller part is a F2013 "target". The 2-inch bigger part is a "tool". The "target" and the "tool" are connected through a 4-pin connector.
With the "target" plugged into the "tool", I can program/debug/run code. I unplugged the "target" from the "tool" and provide 3V to the "target" from a battery. This standing alone (without the "tool" attached) "target" can still run the existing code (no crystal or oscillator needed). But I cannot program/debug. Next, I added 3 wires (each about 4 inches long) from the "tool" to the now standing alone "target" and I can again program/debug (in addition to run) code.

I wired a 14-pin header onto the USB gadget. It did not like the extra load on pins 10 and 11. The wires are only 2 inches long. When I disconnected pins 10 and 11 it worked fine.

Is the "target" still connected to the "tool"? If so you do not need to add any wires. What is at the other end of the wire (or the header)? You are interfering with the communication between the "target" and the "tool". Sounds like you are trying to use the “toolâ€

[Crane]

Old Cow,

I think it like the one you described. I Added a 14 pin header to the detachable board. I program the device by plugging it into the USB connected tool. Then I remove it and plug the header into my application and it works fine. It gets power and clock from the application, but no in circuit debug or programming.

The problem I had was that it would not carry the load of the two wires attached to 10 and 11. But I may have had a flux or solder bridge. I'll try it again later.

[OldCow]

Okay.
You soldered a 14-pin header, which includes pin 10 and 11.
You plugged this into your "application".

I think there are things in your "application" that is connected
to the F2013 pin 10 and 11 now. I think those things make it
impossible for the "tool" to communicate with the F2013.

It is not the 2-inch of wire, it is those extra load in your
"application" that is causing the problem.

Pin 10 and 11 of the 14-pin header should not go to your
"application". They should go to the "tool" only.

The 10 MHz oscilloator is not needed, but it does not hurt either.

[Crane]

Old Cow,

Sorry, I was not clear.

After I download the program I unplug the board from the tool and plug the header into my application board. Pins 10,11 are not used. The downloaded program then runs on my board. The Osc drives Xin.

As I said, I may have had a flux or solder problem. I'll connect the two leads again to see if the problem persists.

Crane