The price I want to pay for such device, is max. US$200 (also important to mention, I thought...).

That price level may and may not be reasonable

The issue is your production volume that will decice the cost.

If you make a million widgets, the cost of the widget is extremely important, the cost of tools does not really matter.

When I were involved with a huge volume project we spent more than $50.000 on tools and nobody blinked.

At another company we anticipated a manufacture of about ten widgets and the tool cost importance was so that "presidential approval" was required for all tools.

In my current environment (product a 2000/year, product b 10.000/year) the tool cost play a very small role.

So why do I use the SILabs chips - speed and no other reason. Would I use them just for the avialability of the built in debugger at the price they carry - no.

So, with the above said to emphasize that the SILabs were NOT chosen for the debugger some actual compares:

The SILabs F12x tools $149 chip++
The Philips LPC93x tools (Ceibo) $700 Chip--
The 89x51Rx2/89c66x series tools (Ceibo) $1500 chip "neutral"

The Ceibo tools are "clean" the SILabs tool has a lot of small "inconveniences" (in one release you could not change the value of a bit, in another, you could not scroll the xdata display) CtrlL for go to line is missing, but for all practical purposes does its job. This I would not use as a decision factor.

The way to decide is ((cost of processor) * (number of widgets to make)) + (cost of tools) - (cost of dev time saved by tools)


a personal opinion: One thing I have found is that integrating the debugger with the compiler makes more trouble than it is worth. Please note I say integrating not interfacing. As an example: when you integrate an accidential press of the spacebar while wanting to set a breakpoint brings up a message "recompile now". The so called IDEs are so wonderful in saving your sitting muscle that they save you when you so not want to be saved

Erik