Zener regulator current.

[davelectronic]

I've never got round to trying a zener regulator circuit,  the one i have found looks like it might be ok. Two questions really, first was, can this deliver up to 3 Amps as it claims. And am i right in thinking a 13 Volt 1.3 watt zener diode would be better to obtain 12 Volts, due to the transistor 0.7 voltage drop. Also possibly a silicone diode in series with the zener diode, on the 0 Volts side of the zener diode. But that's a maybe to gain another 0.7 Volts. Thanks for reading, thoughts on this appreciated.

[MK14]

As shown the circuit would supply up to about half an amp (as mentioned in the comments surrounding the apparent source of the circuit). Because the R1/680 resistor WON'T supply enough current for the relatively low Hfe (gain) power transistor.
Otherwise if the power transistor was sufficiently well cooled it could supply 3 amps.

There is no overload/overcurrent/overvoltage/overtemperature etc protection and the regulation would not be that good.

[4CX35000]

This is a popular simple circuit used in many textbooks which show a regulated power supply and can be used to study what the zener and transistor are doing. As mentioned not a ideal design to use. Often it is to show the temperature rise of the 2N3055 when under load and so on.

You can buy a dedicated voltage regulator such as the LM138 and comes with plenty of example circuits in the datasheet (see below).
http://www.ti.com/lit/ds/symlink/lm138.pdf

[MK14]

You can buy a dedicated voltage regulator such as the LM138 and comes with plenty of example circuits in the datasheet (see below).
http://www.ti.com/lit/ds/symlink/lm138.pdf

The LM338 plastic version is around £1.50
http://www.mouser.co.uk/Texas-Instruments/Linear-Voltage-Regulators/LM338-Series/_/N-1z0zls6Z5cg9gZ1yxyx8u

Whereas the LM138 metal can version, is an eye watering £42+
http://www.mouser.co.uk/Search/Refine.aspx?Keyword=LM138

[4CX35000]

I've got LM138 on my mind at the moment. I suppose it might be to with replacing one recently. Yes LM338 and others will do.

[madires]

TL431 plus driver transistor (BD135) for the 2N3055, creating a Darlington stage.

[MK14]

3Amp single 3 pin IC, at 77 pence one off price. (£0.77).
Why mess around with lots of discrete components, to make a much worse circuit, as regards protections ?

http://uk.farnell.com/fairchild-semiconductor/lm350t/v-reg-1-2v-to-33v-3a-to-220-3/dp/1417673

Or spend 70+ pence more, for the 5 amp version (LM338).

Unless it is a learning exercise, to learn how regulators work.
But if so, it is not really a proper regulator, just a transistor emitter follower. Since it lacks proper feedback.
But good enough for some applications, but risky at higher current/power levels, as it lacks even the most basic output protection.

tl;dr
The voltage in the original circuit, will tend to drift with temperature changes and not be very accurate.

[MK14]

TL431 plus driver transistor (BD135) for the 2N3055, creating a Darlington stage.

Those are good ways of improving the original circuit. The datasheet even has some application circuits, to make simple power supplies.

Such as this one:

[Zero999]

TL431 plus driver transistor (BD135) for the 2N3055, creating a Darlington stage.

Those are good ways of improving the original circuit. The datasheet even has some application circuits, to make simple power supplies.

Such as this one:

Yes, do that, except change the transistors for something bigger, obviously.

I'd recommend replacing both of the transistors with the TIP140.

[rdl]

Just for clarity,

The part numbers in the LM138 series which have a K suffix are packaged in a standard steel TO-CAN package, while those with a T suffix are packaged in a TO-220 plastic package.

The LM138 is rated for T_J = –55 C to 150 C, and the LM338 is rated for T_J = 0 C to 125 C.

The LM138 is only available in the metal can (TO-3 type) package. The LM338 is available in both the metal can (TO-3 type) and the TO-220 package (plastic with metal tab).

[davelectronic]

Thank you for all your replys, I have put together various circuits with 3 terminal linear regulators, using the LM317 and LM338, also 78XX regulators. I never tried the option of trying a zener diode as a regulator. Thought it would be good to try it, so the source is in the link below. If 680 ohms is to high, what value would I need to obtain the 3 Amps maximum current ? I have ordered some 3 watt 680 ohm resistors,  just going on the value of resistor from the source of the diagram. Thank you for the schematics, if the first attempt works OK  (my building it) sure I'd  love to try these improved circuits you've posted up. 

http://www.circuitstoday.com/12-v-high-current-regulator

[davelectronic]

I did find this zener calculator,  but as there's a series pass transistor in the circuit, not sure the same value, with out adjustment would apply.

http://www.reuk.co.uk/wordpress/electric-circuit/zener-diode-voltage-regulator/

[MK14]

If 680 ohms is to high, what value would I need to obtain the 3 Amps maximum current ?

If it was a 12V AC secondary transformer, so about 12 x root 2 - bridge diodes (about 2 volts ) = about 15V.

Min 2N3055 Hfe at around 3 Amps, is about 20 (4 amps in datasheet).

So 3V across that resistor (R1), needs to supply about (3000/20 =) 150 milliamps.

R = 3/0.15 = 20

BUT that would not work in practice (or more accurately, it WOULD work, but the zener/resistor would get really HOT!), because the peak current (at the top of the ripple voltage, and when no output current), would be about 10 volts (across R1, estimated NOT calculated, as I don't know what capacitors or AC voltage transformer you are going to be using).

Which would then be about 0.5 Amps.

So the Zener would have 0.5 x 12 = 6 Watts, which is rather high. (Because the ripple voltage goes up and down, the average dissipation would be less than 6 watts. But with no output load, it would be 6 watts (est), all the time).

So it is NOT a good circuit for using at higher currents, sorry.

But another option is to use a power darlington transistor (instead of the 2N3055, the darlingtons huge gain would mean it needs little current to turn it on) or use a better designed circuit. 

N.B. Exact values, depend on the specific components you use, such as the transformers secondary, AC output voltage etc.

EDIT: Figures updates. I used 4 amps, instead of 3 amps, so original figures wrong. Now corrected.

[davelectronic]

I was thinking about 4700uf, yes 30 ohms is hugely less than the 680 ohms in the diagram. I've got 13 Volt 1.3 Watt zener diodes, the glass passivated type. I don't need as much as 3 Amps continuous, my load would be a 4 watt cb radio. I have noted this draws about 250mA in RX and 1.6 Amps in TX mode. So I guess what I'm looking for is at least 2 Amps at least. My transformer would have an AC secondary of 15 Volts. So this would be higher after the filtering capacitor. I'm not sure what value resistor in ohms, and power rating I need. I always struggle with the maths. Usually once I've worked it out I can build it, but the maths is my weak point, I wish I was better at it. Thank you for the help.

[MK14]

I'd be temped as a first try, to stick with the original R1 value (680\$\Omega\$), as it runs cooler, so less voltage drift with temperature of the reference and R1. But use a power darlington, with a very high dissipation.
Note that a darlington will have a much higher base voltage drop, as it has two rather than one, base diode junction voltage drops.

Best to check with the datasheet, but it would be probably something like 1.5 volts or so. Almost any darlington, will probably have plenty of Hfe (gain), from the original circuit.

But if you are determined to use the original circuit, you could also dramatically reduce the value of R1 (to some value, such as to 50 or 100 ) AND use a higher Hfe (gain), (single NON-darlington, i.e. normal) power transistor, compared to the 2N3055. I.e. The 2N3055 is rather an old transistor, and does not have a very good Hfe, at higher currents. Try it and see how it works out, if you are just experimenting and trying things.

N.B. The datasheet is giving the worst case guaranteed Hfe (datasheet guarantees 20, but typically it is 70, according to datasheet http://www.onsemi.com/pub/Collateral/2N3055-D.PDF). In practice, it will probably be better than that figure. So if you try one on the bench, you will probably get better results. Just DON'T go into mass production, expecting the Hfe, to be as good as the one you try on the bench. It is best to stick with the datasheet recommendations. I just mentioned this, so that you realize why you may get, much better than expected results.

[Zero999]

As mentioned above: you need more current gain.

Another thing which will help is to increase the size of the filter capacitor to keep the ripple voltage down.

One disadvantage of a Darlington pair is it increases the voltage drop. Another thing you could try is adding another transistor so it short out the base of the power transistor when the base starts to conduct, see attached. This has the advantage of being able to use a 6.2V zener diode and potential divider, which will have a better temperature coefficient than a 13V diode.

[4CX35000]

Thank you for all your replys, I have put together various circuits with 3 terminal linear regulators, using the LM317 and LM338, also 78XX regulators. I never tried the option of trying a zener diode as a regulator. Thought it would be good to try it, so the source is in the link below. If 680 ohms is to high, what value would I need to obtain the 3 Amps maximum current ? I have ordered some 3 watt 680 ohm resistors,  just going on the value of resistor from the source of the diagram. Thank you for the schematics, if the first attempt works OK  (my building it) sure I'd  love to try these improved circuits you've posted up. 

http://www.circuitstoday.com/12-v-high-current-regulator

I think the first time I was shown that circuit was in a school textbook, which had the transistor as a BFY51. Using that transistor I made a small PSU to provide 9 volt and it worked fine for building small circuits while at college. I built the circuit with some veroboard and placed it inside a small black box with a minature transformer and a 1 metre length of wires with clips on.

[MK14]

Thank you for all your replys, I have put together various circuits with 3 terminal linear regulators, using the LM317 and LM338, also 78XX regulators. I never tried the option of trying a zener diode as a regulator. Thought it would be good to try it, so the source is in the link below. If 680 ohms is to high, what value would I need to obtain the 3 Amps maximum current ? I have ordered some 3 watt 680 ohm resistors,  just going on the value of resistor from the source of the diagram. Thank you for the schematics, if the first attempt works OK  (my building it) sure I'd  love to try these improved circuits you've posted up. 

http://www.circuitstoday.com/12-v-high-current-regulator

I think the first time I was shown that circuit was in a school textbook, which had the transistor as a BFY51. Using that transistor I made a small PSU to provide 9 volt and it worked fine for building small circuits while at college. I built the circuit with some veroboard and placed it inside a small black box with a minature transformer and a 1 metre length of wires with clips on.

It is a nice circuit, for beginners, and low output currents. Especially up to about 100 milliamps.
But probably NOT a good choice for high current power supplies. It lacks too much ability to drive the (base of the) power transistor, as shown in the original circuit.

[davelectronic]

Thanks for the extra information and the schematics, I kind of skipped this circuit. When I started from scratch (absolute beginner) I got ahead of myself a bit and built a fare few LM317, LM338, 7812 etc. These ended up in cases as usable power supplys, even added meters to one of them. And up to 4 x PNP pass transistors on a couple of others. I'd been watching some circuits on zener diodes, and a tutorial of Dave Jones, then been kind of itching to try it out. For what I use 100mA is of little use, but 2 Amps would drive a standard 4 Watt radio. I've got a 10000uf 35V electrolytic capacitor left over from a MOT project (not finished yet) And have a couple of Darlington MJ11015 transistors, but there PNP,  could order the NPN version, it's MJ11016 it's got a hfe of 1000 @20A bit over kill maybe. Tried the PNP MJ11015, that worked well in another psu I rescued off ebay. I will have a play about on a breadboard first, something else I skipped doh on my part. I just stared soldering up circuits from scratch. I should have breadboarded a lot more for the values measured in the circuits, then built them. Thanks again for all your replys and help.

[MK14]

Here is an old UK article about these types of circuits:

http://baec.tripod.com/JULY92/regulators.htm

They make some interesting extra points. Including that the extra gain from a second (or even third) transistor, improves the regulation a fair bit. By adding an extra transistor, it can do overcurrent protection (shown below). They put in a capacitor to improve the voltage stability as well.

[davelectronic]

That's interesting, I've bookmarked it, thanks for the link. Some ideas to try beyond the first circuit. I have looked about at 5 watt zener diodes, do you think the 1N5350B diode would be of use, given its higher power rating ? I've got 1.3 Watt 13 Volt diodes, but wandered if the 5 watt diode would give more flexibility.

[MK14]

That's interesting, I've bookmarked it, thanks for the link. Some ideas to try beyond the first circuit. I have looked about at 5 watt zener diodes, do you think the 1N5350B diode would be of use, given its higher power rating ? I've got 1.3 Watt 13 Volt diodes, but wandered if the 5 watt diode would give more flexibility.

The 5W 13V zeners, seem like a good idea (if you want to try this circuit). Because at lower wattage's, they will run much cooler, and so be much more accurate (as much less temperature rise), so the voltage will drift a lot less. They will also let you try, higher zener currents, for single transistor circuits.
If you end up going for the darlington circuits, the lower wattage zeners, should be ok, but the 5W ones, let you try more experimental options.

You don't necessarily get the full zener rating, as the datasheet may assume a certain area of copper (PCB) cooling. Which may not apply on breadboards or verostrip. So it is best to not be too close to the limit, anyway.

[davelectronic]

I was kind of hoping it might give a bit more flexibility with a higher upper power rating. Yes I think I'll order a couple of the darlington transistors, the MJ11016 has a big gain, but it's TO3. I might look for a T0220 transistor. Thanks again for all the help everyone. 

[MK14]

These ST Darlington TIP142's are 99p each (£0.99), 125 Watts and Hfe=1000.

TIP142. - Bipolar (BJT) Single Transistor, NPN, 100 V, 125 W, 10 A, 1000 hFE
Transistor Case Style: TO-247

http://uk.farnell.com/stmicroelectronics/tip142/transistor-darl-npn-100v-10a-to/dp/1653637

[davelectronic]

These ST Darlington TIP142's are 99p each (£0.99), 125 Watts and Hfe=1000.

TIP142. - Bipolar (BJT) Single Transistor, NPN, 100 V, 125 W, 10 A, 1000 hFE
Transistor Case Style: TO-247

http://uk.farnell.com/stmicroelectronics/tip142/transistor-darl-npn-100v-10a-to/dp/1653637

I was just looking through a UK stores list of transistors, that's one I thought might be ok, the only thing I noticed is its 0.5 Amps Base current maximum, as opposed to 1 Amp Base current for the MJ11016, but I'm guessing a Base current limit on the devise of 500mA for this application should be ok. It's got the hfe of 1000@5 Amps. Thanks for pointing this out as a viable transistor. I'll order a couple of the 5 watt zener diodes, and a couple of the TIP142 transistors. Thanks again for all the help.