See also:

• http://www.adv-transdata.com/RICE17A_comparison.htm
• http://www.tech-tools.com/mat-cc.htm Jerry Merrill [jerrym at tech-tools.com] of TechTools PO Box 462101 Garland, TX 75046-2101 FAX: (972) 494-5814 VOICE:(972) 272-9392 says

  1. The transdata chart implies that one must buy a new Mathias Family module with each Member module. This is incorrect. The whole idea of our modular architecture is to separate the functionally of the Family module (core processor) from the Member module (peripheral functions). This allows one to add support for devices within a given core (12bit or 14bit) for $50 - $200....period. This is why our chart separates the cost of the 'INITIAL DEVICE SUPPORT' from the 'ADDITIONAL DEVICE' support. The FAMILY module is a one-time expense included in the 'INITIAL DEVICE' cost.

  It also means that any features or enhancements to a family can be added to the FAMILY module and they become available to ALL member modules in that core. For example, when the newer 14bit core bondout became available from Microchip a couple of years ago, we offered a new 'ADVANCED FAMILY MODULE' that incorporated that chip. Because of the modularity, that one upgrade instantly upgraded ALL 14 bit devices. All of the member modules became operational at full speed and gained Data Break-pointing.

  Transdata is using this newer bondout (and offering data breakpoints) on its newest module only.

  2. The chart states our rated speed is 20MHz. In fact, Mathias operates at over 30MHz but is limited to 25MHz with current available bondouts. This is determined by the SLOWEST device in the system - the BONDOUTs themselves. BTW, 25MHz is not a limit because the bondouts operate at the full rated speed of the production parts.

  3. There seems to be some question as to what a 'no-skid' breakpoint is. By 'no-skid' we mean that our ICE breaks BEFORE executing a code breakpoint rather than AFTER. This is important. If an ICE breaks AFTER executing a breakpoint, and that breakpoint is set on a JMP/GOTO, SKIP or anything else that modifies the program counter, you can end up far away from the instruction that CAUSED the break. This makes debugging difficult. Also, if you are working in C, and you put a breakpoint on a line. You want to break BEFORE any _part_ of that line executes. Without 'no-skid' break-pointing, part of the line executes (the first opcode of the string of opcodes that make up that source line). Breaking BEFORE executing the line is more intuitive and makes debugging easier. Software debuggers break BEFORE executing the line. This is more difficult in an ICE, but doable.

  In all fairness, it is difficult to make a chart that accurately compares these two products because of architectural differences and the limited space available.

  1. The RICE17 supports Microchip's 17 series devices; Mathias does NOT. Because of this difference, the RICE17 MUST allow for memory expansion or include considerably more memory. Mathias includes the maximum amount of memory addressable by the 12 and 14bit devices. Therefore it would not be fair to say ours is fully populated and theirs requires expansion. Nor would it be fair to say theirs is expandable and ours is not. These are 17 series ONLY issues. - If you are doing 17 series parts, Mathias is not a contender. If you are doing 12 and/or 14bit parts, memory expansion is irrelevant; both ICEs include the maximum memory addressable by these devices.

  2. The RICE17 uses a single module system for device support. Mathias uses a two module (Family and Member) system. If you are doing a single device (or group of similar devices) then cost comparisons are simple; compare our 3 board solution (Mathias, Family & Member) to their 2 board solution (RICE, MODULE) ('INITIAL DEVICE SUPPORT'.) However, if you intend to do more than one group of devices, then it is a little more complicated; you must compare the INITIAL DEVICE support cost PLUS each ADDITIONAL DEVICE SUPPORT COSTS.

  3. The 12 bit and 14 bit core devices have different capabilities. It is difficult to note all exceptions in a chart. These differences are due to basic bondout differences. For example, we have DATA BREAKPOINTS, BREAK on STACK OVERFLOW, BREAK on STACK UNDERFLOW and STACK viewing on ALL 14bit devices. NONE of these features are available on 12bit devices(by any manufacturer).

  4. RICE17 has TRACE built-in, but TIMING is optional. Mathias does NOT include TRACE, but has two options; 1. Timing only option or 2. TRACE with TIMING option.

  Lastly, the relative importance of each of these points depends a whole lot on YOUR focus. Do you wish to emulate a single device or multiple devices? Do you do 12 AND 14 bit devices? Do you do 17 series? Do you see the value of Data break-pointing? Do you see the value of 'no-skid' breakpoints? Are the features YOU consider important available on ALL of the devices you wish to emulate?

  An important factor that can not be summarized in a chart is the intuitiveness of the software. Does it do WHAT you expect, WHEN you expect it? Does its work-flow follow (or adapt) to YOUR way of working? Only YOU can answer that!

  Download the software from both companies and evaluate that for yourself.

http://www.tech-tools.com/mat-cc.htm (TechTools) also have a comparison chart 9-1-98.

Questions:

• +

  .