I have an IC with a wide Vin that’s enabled by an input over 2.5V on the PWM input pin.
I want to power it with 5ishV but implement a low voltage lockout of 10V by only powering it with 5V when Vin is greater than or equal to 10V.
I was/am considering regulating Vin with a 5.6V zener in parallel + a series resistor and a 10V zener in series in front but I saw this 5V regulator with Vin rated for -0.3 to 100V with the note that for Vout regulated 10V < Vin < 100 (those are actually less than or equals but I don’t know how to type that).
I thought it might make a nice solution if it turns on sharply at 10 volts. How should I expect this regulator to behave when Vin is less than 10V?

 

Welcome to AAC.

Your best bet is to use a comparator to check the input voltage and only switch the signal to the PWM when the signal is ≤10V.

 

Welcome to AAC.

Your best bet is to use a comparator to check the input voltage and only switch the signal to the PWM when the signal is ≤10V.

Thank you for your reply and your welcome,

To be clear, I don’t actually have a PWM signal. The IC I’m powering supports one but I’m not using one. I just need to supply it with 2.5V to enable the chip and since the max V the pin can handle is 7V I need to supply it with something between 2.5V and 7V. So as I see it I need to regulate my voltage source to 5V to create the correct supply for my enable signal.

The question is: if I use that regulator to create my 5V enable signal how does it behave below 10V? If it turns on relatively sharply at 10V. I don’t need to do anything else.

 

What you are suggesting is a kludge. You should provide the enable signal as a result of using a comparator to check the input using a comparator. The comparator’s output would enable the device, not the regulator’s. The behavior you are looking for is not something a regulator is designed to do.

The comparator can be made of just a few components and will be predictable, reliable, and accurate. If you use an incidental behavior of the regulator there is no guarantee it will perform consistently and there may be unexpected behavior in edge cases you don’t know to test for.

If you want to pursue that, you should get the part and test it with the circuit you have in mind. It is the only way you will really know if it behaves as expected.

 

Sometimes I think it is best to let people discover the shortcomings of an approach on their own. No amount of experience, logic or reason is likely to dissuade them from their quest. This might be one of those cases. Props to Ya’akov for trying to give good advice.

 

How should I expect this regulator to behave when Vin is less than 10V?


View attachment 290107

It behaves as though the zener is absent - the transistor will be turned on by the resistor between base and collector, so it will appear as a forward biassed diode (provided that it has enough base current).

 

Sometimes I think it is best to let people discover the shortcomings of an approach on their own. No amount of experience, logic or reason is likely to dissuade them from their quest. This might be one of those cases. Props to Ya’akov for trying to give good advice.

Eh, it might be a little early to throw that out there Papabravo. I didn’t ask my question to try to talk myself into doing something that doesn’t work. I asked what I asked because I’m looking to develop an understanding around the subject at hand so that I can make informed decisions. When I understand the drawbacks to one potential solution I’m better prepared to see the virtues of another. There hasn’t been any actual discussion of either yet.

 

So the problem as I see it, from the point of view of experience, is that shortcut solutions almost always end in tears. One workaround prompts another because the first doesn’t do quite what was intended and there is no headroom in the design.

To avoid redesigning the “right” way, some other kludge is bolted on—and it is a spiral down form there until it is either abandoned or replaced with the right thing which should have been done in the first place.

A comparator is the right thing for the circuit you want. Your proposal might work, but since it isn’t the intended function of the part, it might start out working and then stop or become unreliable, probably in mysterious ways.

The only other suggestion I have, which I know will cause consternation in some quarters, is to use a little 8-pin MCU like the ATTiny13A. They cost a couple of bucks, and need only a bypass cap to use. They can be programmed in a simple development board, and then put in circuit. They come as THT or SMT and use very little power.

The advantage is that you can change the behavior of the device with software, no hardware revisions required. So if you aren’t planning to make thousands of these things, it might offer advantages down the road since it can be programmed to do much smarter things than a simple comparator and doing that control the IC you are switching on in some unforeseen ways.

Anyway, the comparator is first choice based on best practices, the ATTiny13A is a choice based on flexibility and possibility. I wouldn’t use it in a production situation but if this is a 1 or 5 off thing, it could be handy.

If you want to explore the possibilities of using the regulator, I strongly advise empirical testing with the actual part that you are planing to use as the load. Also, make sure you definitely follow the datasheet concerning capacitors.

 

If you supply the 5V via MAX5033, the MAX5033 has a UVLO input which is programmable via two resistors. It can easily be set to disable the output if the input is <10V.

 

The only other suggestion I have, which I know will cause consternation in some quarters, is to use a little 8-pin MCU like the ATTiny13A. They cost a couple of bucks, and need only a bypass cap to use. They can be programmed in a simple development board, and then put in circuit. They come as THT or SMT and use very little power.

The advantage is that you can change the behavior of the device with software, no hardware revisions required. So if you aren’t planning to make thousands of these things, it might offer advantages down the road since it can be programmed to do much smarter things than a simple comparator and doing that control the IC you are switching on in some unforeseen ways.

This is an led drive so it would actually be nice to be able to program in some warning flashes as approaching UVLO.
I happen to be very short on board space but things are very miniaturized these days so this just might be feasible.