Third Times the Charm, Class D Amp

[gnuarm]

I never did find a suitable class D amp to drive a ~1 watt speaker and detect if the speaker was open or shorted.  I know, everyone thinks I should just go class AB because it's "so little power".  But the entire board isn't using a half watt and I can't see wasting another watt or more on the inefficiencies of an analog amp. 

So I did manage to find a couple of motor driver chips to do the job.  One from MPS was a good candidate.  It actually will generate a 250 kHz PWM signal for setting the current into the load based on an incoming, slower PWM signal used like an audio signal.  That's perfect since the signal is coming from an FPGA.  It likely would be a good match to the job, but the data sheet doesn't describe the switching modes and when I asked the FAE he does everything but answer the question as if he thinks I am asking a wrong question. 

So I found another good choice from TI, DRV8876.  It has a current mirror and direct control over the PWM complete with documentation on the operating modes.  The simulation model has allowed me to properly design the output filter and I'm pretty confident it will work fine with very low power overhead.  But the current mirror doesn't seem to work in simulation.  I posted in the TI forum which is what they do for support these days and someone has replied, but not with anything useful as yet.  I found an earlier thread on the same issue where the poster is concluding the model doesn't work in that regard (even though the file itself says it is supposed to work).  The FAE in that thread simply replied they did "disbanded" the model writers and there will be no further work on it.  What the heck does "disbanded" mean???  Laid off?  Reassigned?  Shot? 

So other than not knowing for sure the current output works or exactly how it works, this seems like a good solution.  I'm hoping it is just the simulation that is broken and not the chip.  It won't be the first time a chip has a defective feature that wasn't fixed.