Questions on LT3080

[chopnhack]

Hi all, doing a little research on LDO linear regulators and came across a five pin package that I was not familiar with.
http://cds.linear.com/docs/en/datasheet/3080fc.pdf




In the schematic above they show Vcontrol being tied to the IN pin. In the pinout diagram they show Vcontrol as a separate pin as well as a n/c pin. My first thought was why isn't Vc internally connected and the NC omitted and this whole production could be offered in a 3 pin package?!

From what I read of the datasheet, they state that the current across the reference resistor sets the Vout.
Does the mean that the pin is there to allow for a different current fed through Vcontrol than provided by the supply input?
Is there another reason for the extra pin and it not being internally tied together?

This could be a possible candidate instead of the LM317 for the bench psu due to its LDO of 0.35V at 1.1A. It also has a low thermal resistance of 3deg C/watt. Worse case scenario, 19 watts x 3 = 57deg C plus ambient, no heat sink would even be required!! Pretty nifty IC - also has current limit with foldback and temperature protection.


Does soft start work in this schematic because of the delay until C1 is charged? What purpose do D1 and C2 serve?


In this paralleled design, does the full current run through the second LDO? I see that the pot R4 is a meg Ω whereas the first pot is only 100k. Would both pots have to be rated for those loads?

Thanks in advance!

 

[hevans1944]

In the "soft start" circuit, D1 and C2 are the soft start part. C2 charges through the internal current source to provide a soft start. When input power is removed, C2 discharges through D1.

In the second circuit, the two devices are in cascade, not paralleled. The first device acts as a variable current source to the second device, connected as a variable voltage source. Output current does not flow through either adjustable resistor.

The device is available in a 3-lead SOT-223 package, but with somewhat poorer thermal resistance junction-to-case and case-to-ambient. Best package appears to be the 5-lead DD-PAK followed closely by the 5-lead TO-220. Those may not need a heat sink with their modest current capability and if the input voltage is only a few tenths of a volt greater than the regulated output voltage. This may be intended mainly for point-of-use voltage regulation where several are mounted on the circuit board.

 

[Arouse1973]

The V control pin is separate so you can reduce the power dissipated by the internal transistor by running it from a voltage very close to the output voltage. As Hop mentioned I believe the package is used even with an NC pin for it's thermal properties.
Adam

 

[chopnhack]

Thanks Hop, I should have realized by now that C1 was a filtering capacitor...

Would conventional flow look like this then?

Incoming voltage filtered by C1 until it was full then feeding to In and Vcontrol, stopping at the pass transistor until its gate is activated by the op amp. I am not sure what the ammeter symbol stands for in the circuit. Flow would then fill up C2 and then run past D1.
How would this accomplish soft start? Potential is still following the path it did in the first diagram and only getting to D1 and C2 afterwards.

 

[chopnhack]

The V control pin is separate so you can reduce the power dissipated by the internal transistor by running it from a voltage very close to the output voltage.

Hi Adam. I am not sure I follow you on this one. The Vcontrol pin is hooked up to the op amp and the ammeter symbol device (not sure what that is supposed to be). The pass transistor is directly connected to the input voltage.

 

[hevans1944]

Hi Adam. I am not sure I follow you on this one. The Vcontrol pin is hooked up to the op amp and the ammeter symbol device (not sure what that is supposed to be). The pass transistor is directly connected to the input voltage.

Remember, these are functional schematics, not meant to portray the acutal circuitry inside the IC. The Vcontrol is simply the power source for the internal electronics. The "ammeter symbol" is used in lieu of two intersecting circles to represent a current source (saving space on the page) since clearly there is no actual ammeter on the actual part. Since this is an LDO (low drop-out) regulator by design, making power to the pass transistor available on a separate pin allows the user to drop that supply voltage to a value less than the voltage required at Vcontrol and thereby create a low voltage drop from IN to OUT reducing power dissipation in the pass transistor. Pretty good idea IMO. Note the teeny tiny itsy bitsy 3-lead version doesn't provide this capability, so you pays your moneys and you makes your choices: small size with modest power dissipation, or larger size with more power dissipation and the ability to tailor the overall design. I like it! Good find, John!

 

[chopnhack]

Remember, these are functional schematics, not meant to portray the acutal circuitry inside the IC.

Indeed, it makes it hard to follow along when they obfuscate the internals!! But I guess that is ok, it may be a trade secret.

I like it! Good find, John!

Thanks, I was pretty enthused when I found it, but I am not so sure after reading this statement from the datasheet:
<<
The LT3080 can be operated in two modes. Three-terminal mode has the control pin connected to the power input pin which gives a limitation of 1.35V dropout.
Alternatively, the “control” pin can be tied to a higher voltage and the power IN pin to a lower voltage giving 350mV dropout on the IN pin and minimizing the power dissipation.
This allows for a 1.1A supply regulating from 2.5VIN to 1.8VOUT or 1.8VIN to 1.2VOUT with low dissipation
>>

I am not sure if this is just an example or?
The datasheet shows various designs which is nice of them! Helps to envision what one could do with these IC's. I will see if I can construct a model tonight using a single IC to see how it behaves, range of Vout, etc. Then try to parallel to see what kind of current out is feasible.

 

[hevans1944]

Paralleling the outputs of two or more devices requires a low-value "current sharing resistor" in series with each output, recommended to be a short (as in "not long") circuit board trace.

 

[chopnhack]

Thank you, I saw that later in the datasheet.

I modeled the basic variable supply in spice, but the output was not adjusting as expected. Regardless of what value I use for the set pin resistor, I get a consistent 10.97VDC out... Any clues as to what setting in spice I might be missing? Again, I know that these are at best models to work from, but I figure that this is their own product, the least their software can do is replicate the circuit that is covered in the datasheet!!! There is incoming ripple with the datasheet selected capacitance. I played with it and found that 470μF seems to work well with no load.