We use cookies to provide our visitors with an optimal site experience. View our privacy notice and cookie notice to learn more about how we use cookies and how to manage your settings. By proceeding on our website you consent to the use of cookies.
I have a Rumpelstiltskin problem (Thank you Adam Savage for that verbal shortcut) when searching for an SMPS IC. I’m trying to design a power supply which will output a positive voltage, but will source and sink current. What is the search term I should use to find such a chip?
I found some synchronous buck converters that spec a sink current, but that looks like it’s applicable to the gate drive pins.
Other details:
Nominally, this would be a buck or buck-boost supply, but I’m ok working around limitations.
The output voltage range should be at least 3.0-4.2V, but it would be great if it could do several times 4.2V. (Yes, I’m designing a battery simulator.)
Input would be 5V at a minimum, but up to 12V or 20V would be great.
Integrated FETs is a “would be nice”.
"Bidirectional DC-DC " would be the root term I’d start with. Thing is though, I can’t think of a lot of applications at single-or several-cell scale that call for bidirectional capability, so I don’t have a lot of hope that there’s going to be an IC or reference material out there targeting such a niche. There’s a fair bit of material and finished products targeting higher-power applications, but not a lot that works down to 3V.
That said, how important is it that you actually reclaim energy in load mode? At smaller scales a person can make a pretty respectable controlled load with an op amp, BJT, and handful of passives…
Yeah, this is a pretty niche application. I’m mainly trying to replicate what the big test equipment makers have probably done in their own ASICs. My reference point is the Keithley 2302.
Truly reclaiming (as in generating power to be stored or used elsewhere) would be an amazing feature, but I’ll consider that only once I get the basic design working. The main thing I’m looking for is regulating around a set voltage with currents that can be bidirectional. So when the device under test (DUT) sinks current, the SMPS will draw current from it’s source and maintain regulation. And when the DUT sources current, the SMPS (or as you’re suggesting, a secondary circuit) will absorb the current and either convert it to heat or do something more useful with it.
There is a part of me that considers making a constant current load (e.g. 500mA), then sourcing it with a standard SMPS capable of 2x 500mA. The combined circuit would effectively act the same except that at I=0mA, the supply is wasting 500mA continuously. Another alternative would be using a high-power op-amp, but that would just be wasting a lot of energy again.
I’d wager one frosty beverage that if you took a peek inside a '2302, you’d find that the power handling bits are mostly analog in nature. The power levels involved aren’t that large, switching is noisy and complex, and as a piece of test gear efficiency is at most secondary to getting good measurements.
It might be a different story were one talking about cycling batteries that cost $10 to charge, but at these scales one might be looking at $10 worth of power delivered over the life of the unit. It’s hard to pay for much extra complexity or effort on that sort of budget…
Yeah, it most likely is mostly analog in that instrument. And right now, cost isn’t my main driver. I’m just looking to see if it can be done.
I was referred some options which are just overkill for my application. I was recommended two part numbers: LT8714 and LT8708, which points me in the right direction. However, I’m hoping for a lower power and lower complexity option.
That LT8714 sound like good advice. There is a demo board that may get you started:
I’m curious to know how you will control the device. To simulate a battery you would likely need to monitor both current and voltage. You would then control the the switching / quadrant accordingly. With good test equipment, this could be done manually. However that’s a recipe for hours fiddling with the circuit to capture the data.
By the way, have you seen this video regarding the Keithley 2302.
