Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

2n3055 Heatsink?

Status
Not open for further replies.

tom91

New Member
Hi,

I have a 2n3055 transistor which I am going to use in a darlington pair arrangement with a BC107. The 2n3055 is in a to3 case which I am not familiar. The transistor only has two legs, E and B and the case is C. Do I need to heatsink this device? What would be the best way of doing this? Also what would be the best way to connect the collector to the PCB?

Thanks,
Tom
 
tom91 said:
Hi,

I have a 2n3055 transistor which I am going to use in a darlington pair arrangement with a BC107. The 2n3055 is in a to3 case which I am not familiar. The transistor only has two legs, E and B and the case is C. Do I need to heatsink this device? What would be the best way of doing this? Also what would be the best way to connect the collector to the PCB?

TO3 is an EXTREMELY common case type, usually used for high power devices. Depending on how much power you intend to dissipate, it will almost certainly require a heatsink, quite possible a sizeable one (but with no idea of your use, we can't really comment?). It's VERY rare to use a 2N3055 without a heatsink!.

You need to mount the transistor on the heatsink with an insulating kit, these contain a mica or silicon washer, and two plastic insulators for the bolts. If you use a mica washer, you should smear each side with heatsink compound, which is freely available (as are the mounting kits). You connect to the collector with a soldering tag fastened to one of the bolts through the plastic insulators.
 
I think that a little BC107 will burn out or melt if it tries to drive a 2N3055.
At 10A, the 2N3055 has a min current gain of only 5, so the little BC107 will smoke trying to drive 2A.
 
BJT's are terrible for high current...

For a lot of applications, a transistor in the TO-92 package will drive a '3055 just fine (lower current / higher voltage applications).

As for the heat sink, you need to know how much power that transistor is absorbing. There are "thermal resistance" values on the transistor's data sheet, as well as the datasheets for the heat sinks and mica pads. If you add up all of those, you'll get a total thermal res. from junction to air--multiply this by the power going into the transistor and add it to the expected ambient air temp. to get the junction temperature. BE SURE that the power dissipation you calculated is below the "derating" value given on the data sheet for the junction temp you calculated!
 
i_build_stuff said:
BJT's are terrible for high current...

Err no, They are fantastic!!!! The problem is they are current-controlled devices and with increased load-current the beta comes down so you need quite a large base-drive
 
You can find 2N3055 whith up to ganin of 15.

Also MOSFETs come to mind.They have chanel resisstance insted of forward voltage like BJTs do.Also they are voltage controled and make much less heat but they are slow.
 
Someone Electro said:
You can find 2N3055 whith up to gain of 15.
Maybe, if you have a bucket full of 2N3055 transistors that were made at different times or made by different manufacturers and you test many to find one with a gain of 15 at 10A.
But what if you ordered some from your supplier and they were all made at the same time late on a Friday? Then their gain would be only 5, wouldn't it? Would they have any gain if they were made early on a Monday?
A 2N3055 transistor is rated for 15A. The datasheet doesn't list currents higher than 10A. Does it need 7A of base current (its max) to get 15A of collector current?

If a manufacturer shows on the datasheet that a transistor is guaranteed to have a gain of 5 or more, then I design the circuit so it will work with any transistor that meets that spec.
 
Styx said:
i_build_stuff said:
BJT's are terrible for high current...

Err no, They are fantastic!!!! The problem is they are current-controlled devices and with increased load-current the beta comes down so you need quite a large base-drive

Which, in most cases, is a horrible problem that turns the rest of the design into a nightmare...

For the "high current" values people are likely to bring up here, MOSFETs are easier to deal with (even though a BJT can have a higher max rated current).

For truly high-current power electronics (pretty much anything over 100A), you'd want to use SCRs or IGBTs.

Also they are voltage controled and make much less heat but they are slow.

MOSFETs are a lot better than they used to be--actually, some of the fastest switching devices are using them now.
 
MOSFETs can switch more curent than BJTs.This is becose they can go down to miliohms of full on resistance.BJTs are above 1 volt so at high curent an BJT will act more as a load and so making heat and too much heat sends semiconductors in silicium heaven.

The problem whith MOSFETs is there big gate capacitance.So you end up driving an voltage controled MOSFET whith somthings that can output amps of curent.Becouse if you charge the gate slow the MOSFET will be more of the time in that half conducting state in this state a lot of heat is made.So an better drive cirucit will make the MOSFET run cooler in high freq.
 
Hi,

I have a 2n3055 transistor which I am going to use in a darlington pair arrangement with a BC107. The 2n3055 is in a to3 case which I am not familiar. The transistor only has two legs, E and B and the case is C. Do I need to heatsink this device? What would be the best way of doing this? Also what would be the best way to connect the collector to the PCB?

Thanks,
nasib rahim+92555939555213
 
2N3055 Θjc = 1.5°C/W, max junction temp = 200°C, so with a perfect heatsink you can dissipate (200-25)/1.5 = ~120 W.

Figuring out Θca for power transistor packages is a little tricky, but somebody might have measured or calculated it and posted it on the Web.

Here's a link with data on a TO-220 package
**broken link removed**
with Θja = 75°C/W and with a max Tj of 125°C.

By scaling up from the surface area of a 220 package to that of a TO-3 and neglecting Θjc for the 220 you could guesstimate all the thermal resistances you need for a TO-3 package.

If the TO-3 package has four times the surface area of the TO-220 package, Θja = 1/4 *75°C/W = ~20°C/W. For Tj max = 200°C you could dissipate (200-25)/20 = ~9 W in the TO-3.
 
Last edited:
Hi Nasibrahim,

Whether a transistor requires a heatsink or not will depend
on the total power dissipation of that transistor in that
particular application and this applies to all transistors. :eek:

If you want to mount a transistor in a TO-3 housing directly
on a pcb you need to drill two 3 or 4 mm holes in the pcb
and bolt it (together with it's heatsink) to the pcb. Also make
the copper tracks wide enough to solder the nuts directly to
the pcb.

on1aag.
 
They make board level heatsinks specifically for TO-3 devices if you are dissipating low power.

An easier device to heatsink/mount is the TIP3055. It comes in a TO-18 package which is like a large TO-220. Easy to mount to a heatsink.
 
Interesting - a look at the datasheet table suggests a gain of 5 at 10 amps - the graph suggests a gain closer to 10. I would not claim to be skilled at reading the datasheets - just pointing out what I see in hopes that a comment or two might improve my ability to interpret the datasheets.
 
The table is worst case, while the graph shows typical. Although it's highly probable that you'll get a typical part, there's always the risk that some devices could approach worst case.
 
That's not true, the gain is still above 10 at 10A.
If it has 4V or more VCE then its minimum gain is only 5.
if you want it to saturate (and not even well) then its minimum gain is only 3.

You don't know the gain of a transistor until you use it. Design the circuit so that any of them will work.
 
Status
Not open for further replies.

New Articles From Microcontroller Tips

Back
Top