Forums

Hi, 

I'm planning a next-gen power/clock board that will need to drive two rails @ 10A ea, with software (i2c bus) control of the voltage from 0.7V to 3.63V, though most of the time will be 0.8 to 1.2V.  This is to drive high-end analog circuits, where the ripple requirement is 1 to 5mV, with 1mV initial set-point accuracy across temperature!  Of course, I also want soft-start, over/under voltage lock-out, over-current protection (ideally, goes in to constant current mode), sequencing, load-step control, tight temperature-drift compensation, and voltage/current/temperature monitoring.  Is the UCD9224 controller paired with the PTD08D210W the right combination, or should I be looking at any different/newer TI products?  GE offers an RC series filter between the output and trim, to help suffer larger step-load with fewer caps ("Tunable Loop")--I know TI has the same thing under a different trademark, but is it offered in the PTD line?  For this brainstorming, cost is not an object ;)  Thanks for the tips!  Andrew.

Thanks for the reply!  I will not be relying on the on-module capacitance nor minimum output load; the output is based on a previous design with > 1mF of additional low-ESR bulk, followed by a CLC Pi filter.  Hoping the combination with the Powertrain modules will yield the desired < 5mV ripple target.  Cheers, Andrew.

Additional ripple reduction can be achieved by attaching additional bulk ceramic decoupling to the output of the PTD08D210 module.

But the voltage tolerance of 1mV, which would be a function of the controller not the driver, would not be possible.

The controller's voltage regulation tolerance is specified as 10mV on a 1V output at 25C and 20mV tolerance over temperature, this can be improved upon with calibration but I would not expect to be able to achieve the 1mV requirement.

Thanks for the reply -- are there any other TI solutions still in module form that could achieve this (prev. we used LTC2970 and digipots)?  

Actually,  would the ucd9224 accept bump up/down control via PMBus?  We do have true differential sense all the way at the final load; I could branch this to an external high-precision ADC and use my processor and my own software to request false voltage set-point to compensate over time/temperature.  We need to do run-time adjustment of the voltage, anyway--quick check that the ucd can be reconfigured at run time, vs. voltage is programmed once at MFG and no other usage model available?  Thanks, Andrew.

The output voltage can be modified on the fly with the PMBus using Vout_Command on all the UCD92xx controllers.

There is no PMBus bump-up/-down command.

The UCD9222/UCD9244 also have VID capability, either by VID pins or 8-bit serial VID PMBus command in addition to the Vout_Command.

Yes, there is differentiial sense to the load for the feedback voltage.