Zener questions

Hi gearhead,

I just found another design similar to my friend's. What do you think about it ? Is it more efficient? Is it ok to let the 7812 which is rated at 1A to work at 866mA?

Allen

How does the circuit arrange things so that the six 2N3055's fairly evenly share the load current?

Jon

(snip)

Well, that explains what the .1 ohm resistor is there for (to help balance the current division between multiple transistors), but it still means that the wasted power is pretty scary at 170 watts per transistor.

Or lower. You want the input voltage as low as you can get it while still leaving a few volts to drop across the pass transistors at the bottom of the ripple voltage at low line. Any more than that is just room heater.

Do you really need 60 amps continuous capability?

I hate to use a pass transistor above the current where its gain falls to less than half the peak value. For a hot 2N3055:

Keep in mind that whatever you use for the pass transistors, they have to stay well below the safe operating area curve for DC. This transistor is red lined at about 10 volts drop at 10 amperes. The 17 volt drop in the design is outside this safe area, though the DC input may sag enough to just get in inside. Too close for me unless the input voltage is lowered. And if the output is ever shorted all bets are off, even if you add short current limit to the design, because there will be 24 volts extra dropped under that condition. It takes a moment for a fuse to blow, and in that moment, you could fry a half dozen transistors.

Much more capable as a 10 amp pass transistor. It has good gain out to 10 amps and a 15 volt 10 amp DC safe operating area. Pretty safe with a 5 to 10 volt drop at 10 amps.

I wouldn't use this as a linear pass transistor above 5 amps.

A cheap transistor that is good for linear operation out to about 7 amps is the TIP35. I would use about 8 of them in parallel for 60 amps. http://www.ortodoxism.ro/datasheets/PowerInnovations/mXqwzus.pdf

One of any of these would work for Q2, the next driver down the chain, but there should be a low value resistor from its emitter to the output, to speed up the response of those output transistors. Something like 2.7 ohms 2 watts.

A much smaller transistor could drive that. Perhaps Q3 could be a TIP31:

This chain could be used as the pass device for an LM723 based design, also, except that you would be able to add current limit and current limit foldback on short circuit functions.

TIP31 could also be used for Q4.

"John Fields" " Phil Allison "

" 2. Can I increase the wattage of a zener using a bipolar transistor arranged as below? "

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

A 2 terminal device, not 3.

There are not interchangeable and so one is NOT clumsy compared the other.

....... Phil

You can use the technique shown in that, but you really need to scrap the schematic and start from scratch. The one your link shows is totally inadequate for your 60A requirement.

Start at the input. Your transformer will need to be capable of providing over 100 amps at the secondary if you want a 60 amp supply. Your rectifier will need to be rated to handle that. That's two expensive items. Your filter caps will also be expensive. You do *not* want to use a 24 volt DC input to get

If you *really* need 60 amps at 12 volts, using a 12 volt deep cycle lead acid battery is likely to be cheaper and easier for you.

In the meantime, you can experiment with much smaller supplies - say 5 amps max - to learn, if learning interests you. At lower currents, the learning experience won't cost as much in blown up parts and burned fingers.

Ed

How can I tell if the transformer is having 100A capacity? May be my friend bluff me saying that his PSU is giving 60A.

The dimension of the transformer core is as below:

Calculate the heat dissipated in the 7812. Following the voltage drops conveniently labeled in the diagram, you have input to the 7812 of 24-4=20 volts. Output is 12, so you have 8 volts across the chip and .866 A, for about 7 watts. That's a lot of heat, and you would need a very good heatsink. Here's a link that shows a way to reduce the current through the chip:

Yes. I had forgotten that the OP eventually wrote, "The output was supposed to give 60A of regulated 24V supply." Sheesh!!

This alone should indicate, if linear approach is used, that a custom transformer is in order or else finding the right match. And that is going to take some very close examinations of the details to get done well. The transformer output should be closely crafted for the desired supply, at this level of current.

Average current is only one part of the consideration. And the OP writes, "The filtering Caps are 6 pieces of 10,000uF 50V electrolytic," almost as though one can just throw caps down without doing some thoughtful design. How much of the AC cycle will the caps be supplying current to the rest of the power supply? How long will the transformer have, then, to supply the necessary Coulombs to recharge them? What will that peak current look like and what implications are there for the transformer? How much ripple will get through the later supply and what is acceptable? What value of capacitor is then appropriate as a balance between all these things?

And so on.

Jon

Allen what purpose does this 1440 watt regulated power supply serve? Welding? If you need that kind of power and you try to do it with a linear regulator instead of getting practical about it and just putting a couple of car batteries together or using welding equipement, you may have headaches. Anyway, that transformer you put up the diagram of doesn't look to me like a piece of equipment in the 2 KiloVA category. If you really want to build a linear regulator, which I applaud and encourage you to do so, you should probably aim a little lower.

amps.http://www.ortodoxism.ro/datasheets/PowerInnovations/mXqwzus.pdf

TIP31:

Thanks very much John, I have studied all your suggestions very carefully and I will use it as a guide to improve the power supply unit.

But one thing I dont understand, which part of the datasheet of 2N3773 makes it a poor choice for my application? It seems to me it is better than 2N3772.

Allen

"John Fields"

A zener voltage clamp must have some impedance in series to allow it to work.

R load is just that.

The OP did not mention or ask for a voltage regulator.

His schem was for an active ( ie high powered) zener.

You are ONE F****NG IGNORANT ASSHOLE - Fields.

DROP F****NG DEAD.

........ Phil

Allen Bong wrote: (snip)

I reviewed the 2N3771 and 2N3773.

The 2N3772 is similar to the 2N3771 (both rated 20 amps and good for 10 amps), except that the 2N3772 has a higher voltage rating.

The 2N3773 is a 16 amp rated transistor that is good for about 5 to 7 amps.

If you buy a Motorola (or ON semi) 2N3773 - it will not be - it will be an MJ15003.

They have been selling their MJ15003 as one for 20 years by relabelling it as various other numbers, like MJ802.

Same goes for a great many "vintage" power transistors in TO3 pack.

....... Phil

Good info. I would have to have a very good reason to make anything with TO-3 transistors, when the TO-218 are easier to mount with fewer holes. That is why I added TIP35 to the choices. Have you seen substitutions for these?

Check out the sci.electronic.components thread "Counterfeit parts alert on Ebay"

Shame how single bolt mounting barely works.

Stick with the authorised agents and major parts suppliers and your chances of getting a fake device is absolutely minimal.

Buy from a anonymous, bargain priced, on-line " parts launderer" and you get a lucky dip.

....... Phil

"gearhead"

....... Phil