has anyone used Abatron or Lauterbach?

[legacy]

• Abatron, Switzerland
• Lauterbach, Germany

they both do debuggers, supporting different architectures { ppc, mips32 , cpu32, arm, … }
licenses and tools are so expensive for hobby purposes that practically
putting hands on those seems possible only in professional experiences
job in avionics? job in automotive? are you a dev or tester-guy?
well, sooner or later, you might happen to use one of them

and in this case, I'd like to hear your experience

about licenses cost, here you can see the price list

[Sal Ammoniac]

I have used the Lauterbach debugger professionally, and it's a nice, high quality unit. We got excellent support from Lauterbach when we needed it.

I don't use one for hobby use--waaaaaay too expensive! For hobby use I stick to a $60 Segger J-Link EDU or a $21 ST-Link/V2.

[legacy]

2nd hand, on ebay, might save the hobby purpose

I happen to need one for a professional job task,
in this case target=powerpc4xx
while in my hobby the target is cpu32, or mips32

Abatron happens to support cpu32, bdi2k is deprecated
still supported, I might find a valid license for bdigdb

anyway, job has the priority, thk for your feedback 

[cfr]

At work we are still using Abatron BDI2000 for Coldfire processors.

For newer projects, which are mainly based on ARM Cortex-M microcontrollers, we got a Lauterbach uTrace, which is a little bit cheaper than their modular debugger platform, but only supports Cortex-M controllers.

I can confirm that Lauterbach tools are really powerfull and nice to use, once you get the hang of their software.

Regarding Abatron: You might be interested to hear the company decided to close down (see http://www.abatron.ch/news/eol-bdi-products/). Software support for BDI2000 and BDI3000 will end by the end of 2016, whereas hardware warranty and repair service will continue until mid 2018 (http://www.abatron.ch/fileadmin/user_upload/products/pdf/EOL_BDI_Products.pdf)

[stmdude]

I've used Lauterbach extensively, and I was pretty happy with them.  Didn't have to pay the bill for them though, that might be why I was happy.

It's a very powerful JTAG and associated software, once you get used to it. The software has a bit of a learning-curve, and its not very intuitive at all. Once you realize that the UI is all for show (and not useful), and just start typing the commands manually (or in a script), you've gotten half-way at least.

My background (since you asked):  Low-Level developer in the mobile-phone industry. Did bringups of new platforms, with associated troubleshooting.

[legacy]

thanks, exactly what I was wondering about the learning curve
I have less than 3 months of time to learn it

[stmdude]

3 months should be enough to get efficient in the basics, like flashing, breakpointing, and reading back code+variables. I.e, basic debugging.

You should check with the manufacturer of your hardware (or just the CPU if it's all custom) if they have Lauterbach (or other) scripts for working with their hardware. A lot of them do.
This would cut down on the time it would take for you to configure the Lauterbach to be able to even connect to your target.

The scripts I had for connecting to the Qualcomm chips were a few hundred lines of Lauterbach-script.  I think the minimum you could get that down to is about 10 lines. Setting types of interfaces, specifying the JTAG chain, speeds, cpu-cores and variants, resetting, etc.

[free_electron]

Lauterbach, Nohau , Ashling , American Arium , Arm Realview.  i've used all of em.
Any company still debugging with printf statements are neanderthals.

[ale500]

I've used Lauterbach for MPC56xx/57xx single and multicore processors. You need a extra license for multicore debugging. You open 3 instances for those 5775 . We even had that PowerTrace module with 1 GB RAM for tracing/profiling. That was a beast at 16 k€ !. If you spend enough time and you know your processor well and you have multiple monitors, there is almost no better way of debugging. Finding faulty code ? Debugging OSEK-aware code ? with threads and ASMP ? not a problem. The only drawback is that the scripting language is a bit... inconsistent. Some commands/expressions are unique to themselves but you can script everything, even time code from scripts... It goes so far beyond what you can do with a cheap debugger that if you don't find the bug... it is not the fault of the debugger, because you have almost every tool you can think of. For 5 k€ I'd expect less. Not like some other "professional" tools that only have a hefty pricetag but no professional functionality.
The price may seem hefty but, if they are paying you 100 k per year, the number of hours that you save with something so powerful pays it back in a month. It is for hobby expensive, because your time is the only not being paid, so to say.