We can share info on the TPS62821 because we have used it with the DW1000. As you indicated, it is an inexpensive part that feels like it could do a good job. TI has been really good at producing low cost effective power supply parts lately and wins more of power designs these days.
There are, however, a couple of issues that you should be aware of with regards to the TPS62821 when used to drive a DW1000.
The TPS62821 is usually shown with a reference circuit using a 470 nH inductor. In the reference circuit, they are trying to get 1 A from the design, so they need a low value inductor to ramp current fast for the 2.2 MHz switching frequency.
In our testing, when using this low value inductor, it causes enough electrical noise to upset the DW1000 somewhat. The solution was to increase the inductor value to 1.0 uH or even higher. This will make the design somewhat less able to hit 1 A output, but you don’t need that, you need softer current spikes, and a larger inductor does that.
The best way to detect the issue is to put the DW1000 in continuous wave (CW) mode and then examine the spectral output. Ideally you get a perfect clean carrier, but some 38.4 MHz spurs are common and not too much of a concern if they are, say, 30 dB down from the carrier. A bad power supply will often show up as widening of the carrier or other artifacts. Anything that distorts the CW signal will distort the UWB signal and reduce effectiveness.
Due to the noise any SMPS generates, we recommend a ferrite on both input and output to keep out as much high frequency noise as possible. Layout is critical, particularly the loop from input cap ground, chip, inductor, output cap ground.
Another issue is that while the regulator is low Iq, when input voltage gets near output voltage (say 2.0 V in, 1.8 V out), the regulator can go into a 100% duty cycle mode. When this happens, Iq shoots up significantly. This is not documented in the TI datasheet, BTW. In the general use case of a 1.8 V regulator for a DW1000, this is usually not an issue as the DW1000 is also being given 3.0 V or 3.3 V from which the regulator runs, so you never operate in the low dropout region.
If you are using SLEEP or DEEPSLEEP mode, and need to wake up precisely at a given time to transmit or receive (say a time slotted network), there are some timing things to note with startup of the TPS62821 when it is controlled by EXTON. TPS62821 is generally up and stable in 1.5 ms, so not too shabby, but something to note if you need rapid entry into INIT state. TPS62821 turn on might be better handled by a microcontroller than EXTON to optimize this.
It is VERY easy to design bad power supplies for the DW1000, particularly on any design with an external clock input (say TCXO, or other clock source). The grounding of the DW1000 on your board is also critical. The DW1000, due to clock stability requirements, power supply requirements, external PLL filters, and operating in the higher GHz region, requires some care in board layout and design.
Hope this helps.
Mike Ciholas, President, Ciholas, Inc
3700 Bell Road, Newburgh, IN 47630 USA
+1 812 962 9408