Which regulator to use

[DSGarcia]

What is the difference between the LD1086 voltage regulators and the L78xx regulators (other than pinout)? Is one better than the other or generates less heat? What criteria do I use for a decision other than cost alone? I only need a fixed voltage but in a hostile environment (outdoors in a sealed enclosure).
Thanks,
Dale

 

[dknguyen]

Did you try comparing the datasheets? Some criteria are quiscent current, dropout voltage, input voltage range, noise, line/load regulation, and the protection that it has.

They are both almost exactly equal efficiency (in almost all cases due to the quiscent current usually beign much smaller than the output current) and generate the same amount of heat (by definition that all they turn all the "unused" voltage (the difference between input and output voltage) into heat. Same number for both regulators = same amount of heat.

 

[evandude]

LD1086 is a low-dropout regulator, the 78xx series aren't. If you needed to operate with a low dropout voltage, (input voltage only slightly higher than the desired output voltage) then the choice would be obvious.

Both will generate the same amount of heat, because they're both just linear regulators - with the same input and output voltage and current load, either regulator dissipates the same amount of power, by P=VI. Well, not counting differences in quiescent current, but that should be negligible.

Therefore, you just need to look at the other factors that dknguyen mentioned.

 

[DSGarcia]

I cannot tell much difference from the specs as far as my application goes.

Both provide the 5 volts I need, the LD1086 has a dropout voltage of 1.5 vs 2.0 and a bit less quiescent current. It is also a newer technology (don't know if that means better or not).

My main question is heat on a hot summer day as well as vibration, etc, as one part being more suitable than another. If no major differences, then the next consideration is space (which is a major factor).

The application circuit for the L78xx shows a 0.33uF cap on the input and a 0.1uF cap on the output (both small) whereas the LD1086 shows a 10uF cap on both the input and output (which are quite a bit larger). Is this a fair comparison and the correct circuit to use? (I will have a 47uF cap and 0.1uF cap for each chip--which are 20 and 40 pin devices)

Thanks,
Dale

 

[dknguyen]

They both generate the same heat, and they are both completely solid-state devices (no vibrating MEMs or tiny mechanical parts) so are equally vibration proof. Therefore they are the same. Pick the cheaper one.

THe caps on the output don't really matter as long as they are there and exceed a minimum value. Usually the bigger the better (as long as the initial current surge during power up is not too much for the regulator). You can't just look at the circuits and take them as absolute, especially for regulators where you can sort of pick almost anything. It's just what the engineers liked when writing up the datasheet. YOu also gotta look at the noise specs and line/load regulation specs they define since the circuit they give probably has those specs (and therefore they cannot be compared directly).

 

[evandude]

You might also want to look at maximum temperature ratings, and thermal resistance between the die and the case. I would expect them to be practically the same, but if they're very different it could make a difference in a hot environment under high load.

 

[Hero999]

The dropout voltage is proportional to the current and varies with the temperature, at the maximum current of 1.5A between 25°C and 100°C it wil be about 2.25V while at 20mA it will be just 1.5V.

 

[audioguru]

Hi Hero,
Your graph shows that the regulator has already dropped out. Its voltage has dropped 100mV. It is not regulating anymore.

Add 1V or so to its input then it will regulate.

 

[evandude]

Hi Hero,
Your graph shows that the regulator has already dropped out. Its voltage has dropped 100mV. It is not regulating anymore.

I think that depends on what your definition of dropout is. From that image, it seems the manufacturer is considering 'dropout' to be where the output drops at least 100mV from its nominal value, whereas you seem to consider some more stringent (but unspecified) value.

For a 5v regulator, 100mV is only 2%, which in many applications isn't bad.

 

[Hero999]

Yes, providing the regulator is operating above its dropout it doesn't matter. audioguru does have a point thought, it is a good idea to leave a margin of error as I think the datasheet gives a typical rather than worst case value. I dod disagree with adding 1V though, I would say between 250mV to 500mV sounds more sensible.

 

[audioguru]

Some spec's for a 7805 regulator have a 10V input.
Some spec's have an 8V input.

A dropped-out regulator passes ripple from the unregulated supply right through it.

The amount of dropout voltage is TYPICAL, it is not guaranteed. Some are worse.

Give a voltage regulator enough input voltage for it to do its job properly.

 

[Hero999]

A dropped-out regulator passes ripple from the unregulated supply right through it.

However if the minum valley is above the droput then the ripple won't be too bad.

The amount of dropout voltage is TYPICAL, it is not guaranteed. Some are worse.

I agree; in my opinion datasheets should always provide the worst case as well as the typical. Another thing to be aware off is the output tollerence, the LM317 has a tollerence of 5% so don't be suprised if the output is 5.25V when the resistors are calculated for 5V; the 0.25V should be added onto the dropout voltage requirement, hence my minimum 250mV statement.