30V 10A power supply keeps blowing 2N3055s

[anacierdem]

Where did you buy the 2N3055 you use ?

From a local supplier here in Turkey. It was always been hard to find good quality components here. This is not the first time Im having problems with transistors here. Farnell seem to ship here as well I will give it a try after choosing a bjt.

It is availability rather than price that prevents selecting another transistor. 2N3055 was rather cheap so I bought some to give it a try.

[Kleinstein]

How about the 2SC5200RTU? From a fast scanning through the parameters it looks like that would work as a replacement or did I oversee a important parameter? The SOA looks identical to the old guy. And it looks like it's not that expensive like the MJ14002G.

If you don't have a reliable source, I would stay clear of the higher end audio transistors like the 2SC5200. They are good, but here the chance to find fakes get high.

If availability is poor, it may be the other way around. Check what BJTs you can get and than check if they are suitable - especially FBSOA. Also check the price - if it is too low for the given part expect fakes. Under such conditions the 2N3055 might be still a good bet, though it would take a few more to get the full 10 A.  At that power level, I would consider a different circuit, like using two transformer taps to reduce the power dissipation.

[Twoflower]

@Kleinstein. Fair comment, especially the audio gear parts. I had just looked at Mouser and found they're listed for about 2.80€ a piece.

I liked it because the much easier to handle package. Modify a heat sink to handle more TO-3s is a bit cumbersome if you don't have a decent workshop.

[Kleinstein]

If it is just for the case, there is the TIP3055 - essentially the same die as an 2N3055, but in TO247. Another common one is the TIP35 - a little more robust of original.

[anacierdem]

I'll give Farnell a shot and select one of the available ones. Their Turkish distributor is near our office, I'll ask them first.

[IanMacdonald]

The 2N3771-2N3773 series are a more robust transistor than the 3055 and relatively inexpensive.  The die in these is much larger than in most 3055's.

Expecting 300W dissipation from 4x3055 is pushing things a bit. Even with a good h/s I doubt if they'd take that. About half that is OK. Years ago I built a 30v 5a PSU with 4x3055 on individual 3C/W heatsinks, direct mounted with the h/sinks insulated, and it worked well. The temperature cutout would operate after maybe 15min if the output was shorted at max current though. 

The other options are foldback current limiting, or a relay to select a lower transformer tap at low output voltages. Which is what most commercial linear PSUs do.

[m3vuv]

those are fake chineese crap,i have some identical,pass more than 1 amp and they fry,the die in a gen 3055 is about 8mm square not 3mm like these fakes.

[not1xor1]

Hi,

I'm working on building a current limited bench power supply at http://electronics-diy.com/30v-10a-variable-bench-power-supply.php. 

Schematic;

There are plenty of LM723 schematic better than that.
At a quick glance:
- at 30V the feedback network would dissipate 2W. Wire wound resistors are inductive and may cause oscillations.
- In the worst case 2N3055 have a current gain of 20. 10A/20 = .5A. At minimum output voltage the single BD139 would dissipate more than 12W.
- I guess C8 should be connected between the voltage reference and ground in order to reduce noise. A smaller value capacitor might be useful between the reference input of the 723 and ground to reduce the noise picked up by the pot wires (if any).

You are likely to find several more reliable LM723 based projects here in the old messages in this forum.

[rsjsouza]

Guys, this is a four year old thread. The issue was probably corrected a long time ago...

[jonpaul]

To OP:

My Advise: Dont buy junk knockoffs of 2N3055 from China.

Dont blindly copy a circuit without careful analysis and actual design.

SOA is only one of many specs of BJT that can kill the parts

Current sharing needs work.

Load inductive, capacitive or reistive? Inductive loads need more protection eg diodes.

Bon Chance


Jon

[strawberry]

Fast test for fakes:
short B & C
connect ~5A source
measure voltage drop across transistor
if voltage runs wild without case getting warm = fake, small die
silicon temp coefficient ~2mV/K