no text

[daywalkerdha]

no text

[Zero999]

How long does it take to warm up?

Why not simply use a linear regulator? If the warm up time is not that long, then the heat sink can be undersized.

Another option is a timer relay connected to a resistor in series with the heater, so the resistor is bypassed, when the delay has elapsed.

[xani]

You're driving a heater. If regulator is near it it will just act as additional heat source, you don't need efficiency if all you do is heat stuff

[georges80]

If you want a switcher and you want constant voltage/constant current, then a LED driver chip can be considered such as the LT3755 (external switch though).

Use Linear Tech's parametric search in their LED subsection to look for internal switch for something that meets your Vin/Vout and Iout needs. Then verify that it has a Vfb (for voltage control) input in addition to output current sensing.

cheers,
george.

[Zero999]

With a linear one, the warmup is about two minutes. Regardless of the complexity, with a switching one(with limited input current) i would also have more output-power which would decrease the warmup time. The voltage has to be quite low in the beginning to stay below 1A(about 5V) and i would like to avoid this extra heatsinking by having a switching regulator.

If the current is limited to 1A, then the warm up time will be exactly the same, irrespective of whether a linear or switched mode regulator is used.

[T3sl4co1l]

With a linear one, the warmup is about two minutes. Regardless of the complexity, with a switching one(with limited input current) i would also have more output-power which would decrease the warmup time. The voltage has to be quite low in the beginning to stay below 1A(about 5V) and i would like to avoid this extra heatsinking by having a switching regulator.

If the current is limited to 1A, then the warm up time will be exactly the same, irrespective of whether a linear or switched mode regulator is used.

If the regulator's heat went into the heater element, it would be the same.  However, a switching regulator can limit input current while providing extra output current at lower voltages (lower resistances, presumably the heater has a notable positive tempco here).

I'm working on a switching current limiter circuit, actually, but it's probably not going to be cost effective compared to other options (like using a resistor and sitting your ass down and waiting for it to come up ).

Tim

[ogden]

Due to your double post, my answer in the other thread became wasted [sic]. So I am reposting here:

Consider MC34063A/KA34063A. It's oldskool, <100KHz, but does what you want. http://www.onsemi.com/PowerSolutions/product.do?id=MC34063A

[Zero999]

Oh I see. You need to limit the input, rather than the output current. Normally it's the other way round, hence my confusion.

Does it matter whether the current sensing is on the positive or the negative side? If low side is fine, then a comparator could be used to turn off the transistor, connected to the load, via an inductor and diode (like a buck converter) when the supply current exceeds a certain level and switch it back on again, after a certain time has elapsed.

[chris_leyson]

A flyback converter will give constant output power if you fix the duty cycle, might be worth considering.

[ocset]

you need two error amplifiers external to the controller..................have them each able to pull down the reference voltage............kind of connected in open collector method............... make it so that the current one wins and controls it when your current goes too high.