Zener questions

Not advised. One will have slightly lower zener voltage and will hog all the current, and get hot. Series is OK.

Yup. But zener shunt regulators are very inefficient and tend to regulate poorly.

Try this:

NPN

in--------------+----------------c e-----------------out | b 10-.6v | | | | +--------R---------+ | | | 10v Zener | | | | gnd

Or buy an LM317!

John

What's the purpose? Why do you want to do this???

That looks a lot more like a voltage reference, though I would probably expect to see it more like this:

That doesn't usually supply power, just a nice voltage if you don't tax it much.

In the back of my mind, I'm wondering if you are trying to supply regulated power using a zener and figuring on just increasing the current through it so that there is more available for the load. If that's your goal, you'd use a BJT this way:

You might consider this as both a temperature and Vbe compensator:

But I don't know what you are trying to do, either. So?

Jon

Umm, I think Q1 should be an NPN, flipped over.

John

Not reliably since the 2 zeners probably won't match very well so one will take more current.

That ought to sort of work but if you have a transistor it would make more sense to create a series regulator using it rather than use purely shunt regulation (it'll be more efficient).

Graham

I don't know what your requirements are how ever, to save components you could use a 78L10 100 ma voltage regulator which comes in various packages. I was just working with a 78L12 in a 92 Package today. they are 3 legged post regulators and very small.

For Question 1, I need to do this as I can only get 400mW and 1W zener locally. The spare part shop charges 4 to 5 times for a 5W zener, so I thought may be I can use five

As for Q #2, my friend showed me a circuit and wants me to duplicate it. The schematic is as follows:

Q1 10W 2N3055 0.1 ohm Vunreg (42V) ___ Vout (24V) o-----------------o---o----------- ------|___|----o---o------o | | \\ ^ | | | | --- | .-. | | | +----+ | | | o------- --+ | | |2k2 .-. | \\ ^ mje3055t | '-' | | | --- Q2 .-. | 4k7 | | | | | | | '-' +--- --+ | | 1k | | \\ ^ mje3055t '-' | | --- Q3 | | | | || | | +----------+--------o-----||--+ | | | \\| || | .-. | | mje3055t |-------o------->| | | | Q4

quoted text -

That's a good idea too. Never thought of that before. SO to get 13.6 I'll just use a 7812 and stick 3 1N4001 in series on the COM pin of the regulator, right?

Then where does my Vin comes from? The 24V regulated or 42V Vunreg ?

Allen

quoted text -

the 78Lxx and maybe the 78xx series are rated for something like 36 volts on the input. But I suppose you could tailor the components to work around that. something like a resistor divider;

I hope you realize that using the same part number for Q2,3 and 4 is a pretty big compromise for this circuit, since they have such different operating points.

With all the current gain in this output pass configuration (double darlington), there is no real need for high power zeners to make it work.

And, unless there is more stuff in the circuit that is not shown, there is no particular point in having 0.1 ohm, 10 watt resistor in series with the output current. It does, however, indicate that the highest possible output current is less than 10 amperes (or it would overheat). But at that current, the single 2N3055 is totally inadequate for the pass device. At 10 amperes, it would produce (42-24-1)*10=170 watts of heat. It is rated at 115 watts, and only if you keep the heat case at no more than 25 degrees C.

So what actual output current are you needing from this regulator?

I suggest you read the application notes for the LM723 regulator control IC as a replacement for this kludge (after you have learned lots from what it takes to make it work cheaply, if not well). It doesn't quite handle the 42 volt input, but fixing that is easier than making this circuit work well.

"John Fields" Allen Bong

An active zener is a voltage clamp, unlike a series regulator.

....... Phil

Yes. I was doing an ASCII flip and screwed up. Thanks for the catch!

Jon

"Jonathan Kirwan" John Farkin F****it

....... Phil

"John Fields" " Phil Allison "

An active zener IS a voltage clamp, unlike a series regulator.

Are you ASLEEP !!!!!!!!

....... Phil

Who designed that, the Geico caveman? Probably not, 'cause he would have done a better job. Check out the design in this next link and consider how in its simplicity and rationality it contrasts to your friend's circuit:

The following link explains some of the basics about linear voltage regulating. You should read a little bit before you jump into anything.

I seem to recall that the limits of the 78xx family was 35Vin wrt common, limited by power disssipation of the package. The TO-220 versions were good for 1W package dissipation I believe, depending on how much Cu was underneath. At 24Vin to 10V out, then that limits the OP to about 70mA for a hefty package. The TO92 variants are 100mW (I think

- don't have the datasheet in front of me) so that's about 7mA.

As you noted, we don't know the OP output requirements. In any event, there's going to be a fair amount of heat in any linear regulator in this case.

One thing - Rload seems to be upstream of the regulator. Is that really the case?

Cheers

PeteS

C.

Hi John,

You're quite right about something I didnt mention in the circuit. Q1 is actually 6 x 2N3055 in parallel mounted on 2 pieces of heatsink measuring 12" x 4" each. The heatsinks were cooled by 2 mini fans on each end. And I also know that using MJE3055Ts as drivers were overkill. Maybe the original designer had too many 3055 in hand or got them in bulk. Can you help me with what transistors are adequate for Q2-Q4 ? The output was supposed to give 60A of regulated 24V supply.

What is the best way to reduce the amount of heat wasted on Q1 ? Reduce the Vin to say 32V ? Change 2n3055 to something better? What about 2N3772 or 2N3773?

I have no intention to change the design to a SMPS but just to improve it. Any other suggestions are welcome.

I totlly agree with you and I'll study the A.N. of LM723 as suggested.

Allen

Thanks for your good suggestions and I will surely read all suggested. Boy, that's a lot to learn on a non-SMPS regulated power supply. You know, just the transformer alone weighs a ton!

Allen