Series Pass Transistor

[excessor]

A few years back, there was a schematic for a power supply posted in the Answer section of the Tech Forum. It used a series pass transistor that was supposed to be MUCH better than the 2N3055's. Does anyone know the issue it was in? OR, what is the transistor number?

I need a 28 Volt DC 45 Amp Supply. I have all the big pieces except for a regulator circuit.

Thanks

excessor

[Bob Scott]

excessor,

I didn't see that particular article, but I do know that the 3055 is not a transistor with good looking curves. For example, the forward current rises with collector voltage.

The Motorola 2N3771, 2N3772, and 2N3773 family of transistors are superior. From top to bottom in the following list these transistors go from low to higher voltage and high to lower current depending on what you need.

2N3771: 40V 30A
2N3772: 60V 20A
2N3773: 140V 16A

I used these as universal replacement audio output transistors as a superior replacement for the 2N3055. They come in a couple of different packages including the TO-3 metal can. They are inexpensive and are available from DigiKey.

Maybe someone will pipe in with the exact transistor number from the article. If it is different, I'd like to compare the specs.

Bob

[Dean Huster]

I'm showing my ignorance here (won't be the first time -- I did author a magazine column, you know), but has anyone ever seen a design for a "Class D" equivalent of a series power supply regulator? I'm not talking about a conventional switcher here, but simply a PWM tag-on circuit to a conventional raw supply. It would seem to be the best of both worlds, eliminating the series-pass heat but not getting into an overly complex design issue as a typical SMPS does.

[Bob Scott]

Dean.

You should familiarize yourself with "Boost" and "Buck" regulators. I think you might find app notes at TI or National, definitely one of the two. That is what you are talking about.

Start with an unregulated supply, run the output through a DC choke, then chop up the output voltage from the choke with a PWM power transistor and sent that current into a filter capacitor. This "bucks" the current to a lower voltage. The PWM is feedback controlled of course.

Similar circuitry is used to boost PS voltage. Just switch the choke's DC output to ground momentarily and send the flyback effect's voltage boost through a rectifier into an electrolytic filter at a higher voltage.

You must already know this simple stuff! You are pulling our leg. Authors are the Gods of knowledge!

Thing is, you just need to change the choke in the boost regulator into a transformer and you then have a switching supply with isolation. It may not have sophisticated filtering yet, or power factor correction, current limiting, or overvoltage protection, but it is the basis of a switching supply.

[Robert Reed]

"I didn't see that particular article, but I do know that the 3055 is not a transistor with good looking curves. For example, the forward current rises with collector voltage."

Bob
I don't remember the 3055s exact specs, but 25 years ago I designed a regulated power supply for bench testing huge marine transmitters. I used (4) 3055s in parellel for the pass transistors which were driven by a 3054 transistor. The supply was rated 10-16 VDC at 30 amps max. output.
The regulation was superb all the way thru its range. It was in service for almost 20 years without any problems at all. What more could one ask from those pass transistors?

[Bob Scott]

The regulation was superb all the way thru its range. It was in service for almost 20 years without any problems at all. What more could one ask from those pass transistors?

I bought a dozen or so 2N3055's for 50 cents each from our wholesaler, way back in 1980. I did use them to replace all the outputs (4 per channel!) in a Grundig junk amp I picked up, and that lasted forever too. I left that amp in our tech shop playing background music for the techs downtown when I left in 1996.

I did try to replace the output transistors in a blown Dynaco 120 amp with 3055s. They were not good enough. High frequency (treble) rolloff. I suppose they are good in some uses like voltage regulator output transistors. But face the fact that they are old technology. Their technology is one step behind today's planar, mesa or whatever topology. They do have sloppy curves on a curve tracer that shows that they do leak current with increasing collector voltage. ie: The bias curves are supposed to flat-top, but they climb at a 20 degree angle on the tracer's scope, and the curves have big thermal hysteresis loops. 2N3055s do have a low gain bandwidth product (Ft). They are (were) inexpensive. I prefer the 2N377X series for power transistors in a TO-3 package. They are antiques vintage circa 1980 too, but they are a better transistor in all respects compared to the specs mentioned above.

Did anyone notice TO-3 packages are fading away? It looks like manufacturers are starting to charge a $5.00 surcharge for a TO-3 model vs, say, the TO-220AB package. I was at the TI or IRF website when I noticed that.

[Robert Reed]

Yea, I have noticed for several years now that any equivalent in a metal package (no matter what the outline) are much higher than their plastic counterparts. Thats why I mostly buy the TO-220s when available. The TO-3s do have an edge though when it comes to lower thermal resistance of junction to case.

[Dean Huster]

Authors are the Gods of knowledge!

No, authors are folks who are willing to chance making fools of themselves in print vs. containing their self-doubt to their own workbench.

Sometimes that apparent foolishness is well-deserved (that's my category) and sometimes it's the result of the publisher/editors. For instance, I never understood why the late Joe Carr kept publishing through Tab books, a publisher that is fraught with poor editing, poor illustrations, lousy binding and huge quantities of typos. I think had he used Howard W. Sams for his publisher, his practical, down-to-earth brilliance would have shone better.

Then there are the unnamed self-proclaimed "Gods of Knowledge" that I certainly don't enjoy reading because their writing style reflects their huge ego.

Now, as to the idea of the PWM regulator. I'm not sure if I was thinking on the same lines as the buck/boost ICs that are so common these days. No, I'm not pulling any legs here.