Ashling Microsystems Evaluation

USB and JTAG interface of Ashling EVBA evaluation board

The Ashling Microsystem’s EBVA7 Evaluation Board (around 2003), had a fairly complex built-in USB/ JTAG interface as shown above.

The bulk of the board was the debug section.

In the late 1980s we ordered a C compiler for an Acorn Archimedes workstation owned by a friend. Two years before we were in the UK when ARM was becoming available as an embedded chip. Almost fifteen years later our first embedded ARM adventure was the Ashling Microsystems EVBA7 evaluation board with their PathFinder debugger for a 60 MHz NXP LPC2106 target. The kit was announced in April, 2003—not sure when we ordered it.

Ashling Microsystems manufactured various debuggers and trace recorders. The trace port was taken out on the Philips chip with a Mictor header on the board—absolutely amazing at the time for such a low-cost device. We even considered taking the Mictor header off the board as it was impossible to buy “one-off” and the minimum order quantities cost more than the EVBA7 board. We were hoping to get a low-cost trace analyzer or to build one using a large FPGA. The ARM device was a bit small for any decent real-time kernel research, but it would be a start if the toolchain supported external trace analysis.

 

Ashling Microsystems evaluation board with ETM

The serial ports were buffered and there were several buffers for I/O. The NXP 2104 to 2106 devices had 128kBytes of Flash, however, the Ashling tools were limited to a 32kByte download.


 

 

 

 

 

 

 

 

 

PathFinder

The debug screen shows the source level interface, the registers and a memory dump. The unrestricted toolchain and trace port analyzer were too expensive, so the evaluation was done “for fun” as the trace interface was the first we had noticed on a low-cost board. Paying work at the time was embedded Linux on 400 MHz devices. Serious ARM work would have to wait a while—the evaluation board was passed on to a student for a project as I was lecturing at the time and was unable to get a trace analyzer.

We used the NXP LPC2106 in a project in 2006; see Philips LPC

 

Ashling Microsystems PathFinder debugger