PWM controllers... shopping

looking for the ideal PWM controller.

handle just the PWM function, and I provide my own error amp, gate driver, and reference externally. The ideal part would look like this: oscillator set by resistor, comparator non-inverting input, and pulse out. Plus Vdd and ground equals 5 pins only! Does anything like this exist?

ver used them. Does anyone here have horror stories from these parts that I can't infer from the datasheets?

looking for. Vdd=6V max for example, kind of sucks. I do like how they use a 50uA current source as the reference pin, so you set Vref with a res istor or drive from low-impedance source, nice.

I can't get to a 5 pin device, but I bet I can do that in a 48 pin QFN usin g an FPGA. ICE5LP1K-SG48ITR50, under $3.00 qty 100.

There's not much in the control world that can't be done primarily digitall y these days.

ATTiny13A, 1k program memory, two PWM channels, 10 bit ADC, 20MHz clock. can do all sorts of stuff with that for 40 cent in 100s

You can design your own if you want using something like the Silego/Dialog GreenPAK OTP mixed-signal PLC:

They have a GUI system-designer it's pretty nice. Their prices are p. good they'll program the chips with your design probably like 30 cent per in quantities of thousands last I looked.

Last I heard Apple is doing that for their switching controllers they want to roll their own vs. off-the-shelf integrated types from the major manufacturers.

Yeah but it seems like overkill to have C code that just performs a comparator function... plus I wonder how much jitter you would get from using that 10 bit ADC

You would need an FPGA with an ADC input. If you already had one on the board, then you could get this PWM block for free.

I have used Greenpaks, I really like them. Great for glue logic. Not sure if their analog blocks are up to this task, though. I'll have a browse at their current devices, it's been a few years.

I don't recall learning about the engineering metric "overkill". A solution should be judged on the merits, not biases.

My concern with a small MCU is the processing limitation, but if it does the job, fine.

Your characterization of this as a comparator seems a bit short of the mark. Isn't this the bulk of the control algorithm? If it were just a comparator why not just use a comparator?

Oh, how would that work?

You can construct an ADC in any FPGA with differential inputs which is pretty much all of them.

You can make a decent analog-to-PWM converter with an RRIO opamp or comparator and a cap and three resistors.

I have a client whose final design is going to be about the size of a postage stamp thanks to those, he was sighing and saying "well I guess if we can get it to perhaps 3" by 3" with today's tech that's just what I'll have to work with"...Heh

IDK if the ATTiny13 has it but a number of devices from that line have hardware dead-time generators on the PWM outputs so the manufacturer clearly expected it could be used in a switch-controller role

the line has some other occasionally useful features in certain applications like the ones with differential ADCs can sense high-side up to several volts over the uP's own supply rail without additional hardware

Why do you need an external error amp, or all the other things? Internal ones are so nice.

Here are some starters, plus bonus memes:

TL494 is fine, but it's a big chip for most purposes, and isn't current mode as such. Very hungry, something like ~10mA operating. It can be used for average current mode pretty easily (use one err amp for low side current sense feedback, disable the other), but you do need to add an external voltage error amp. (In an isolated circuit, this is the secondary side TL431 or whatever, which works out nicely.)

MC34063 is a hysteretic controller, and not even a good one at that. The CS pin reduces oscillator frequency in an odd way.

At least the NCP3064 has proper cutoff under current limiting conditions; I don't feel bad using it, but I would still prefer something with lower output ripple.

TPS54xxx and others are good for local supplies. Hundreds to select from; picking any single one would be futile. Parametric search on ti.com, Digi-Key and others is the way to go.

Tim

-- Seven Transistor Labs, LLC Electrical Engineering Consultation and Design Website:

At the end of the day I'd really like to have the simplest IC possible to handle just the PWM function, and I provide my own error amp, gate driver, and reference externally. The ideal part would look like this: oscillator set by resistor, comparator non-inverting input, and pulse out. Plus Vdd and ground equals 5 pins only! Does anything like this exist?

I've also been checking out the famous TL494, and MC34063 since I have never used them. Does anyone here have horror stories from these parts that I can't infer from the datasheets?

Another one I discovered is MCP1632, pretty simple but not quite what I'm looking for. Vdd=6V max for example, kind of sucks. I do like how they use a 50uA current source as the reference pin, so you set Vref with a resistor or drive from low-impedance source, nice.

I guess the big question is, what are you wanting to do with the controller ?

Sometimes a micro is good enough. Sometimes one of thsoe Greenpaks are good. I use them.

I also use the UC3843 family of current mode controllers. They are similar to the Microchip device you linked to but are slightly different.

Depends on the application. And is it for a product where you would buy a lot of them ? Or a one off project ?

boB

There have been a dozen answers to the OP already, all promoting the predictable favourite solutions. For any question in electronics, Rick will inevitably claim an FPGA is the simplest, cheapest and most flexible solution. John Larkin will want to do it all in analogue electronics. Bitrex will immediately propose a small microcontroller. Then they will argue about the pros and cons of these.

Thank you for that breath of fresh air, asking what the thing is supposed to do. Only once that is figured out, is it possible to look at ideas for a solution.

These are....scurrilous accusations!!

That's the ideal, sure, but often "what do you want to do" is the hardest question to get a straight answer on, and by the time you get one and start pondering potential solutions it starts to feel like work one should be getting paid for. I dunno about other people but I still optimistically have to do that for another ~25 years before I have a shot at retirement...

Is retirement a good thing?

a lot of people used to working and the mental exercise that comes with it, wither and die surprisingly fast after they retire