Is there any inherent advantage to using TO-3 packages vs their flat pack counterparts?
The only thing I can think of is a possible thermal advantage, with better case to heatsink or case to ambient cooling. Maybe junction to case as well, if the die bonding is better in a metal herm sealed package. But I can't find any numbers in the spec sheets to know for sure.
Any thoughts on this?
The only thing I can think of is a possible thermal advantage, with better case to heatsink or case to ambient cooling. Maybe junction to case as well, if the die bonding is better in a metal herm sealed package. But I can't find any numbers in the spec sheets to know for sure.
Any thoughts on this?
Actually the manufacturers spec a better thermal figure
for the plastics, but I'm not sure that it's simply them
being less conservative.
In any case, the TO-3's are becoming more expensive and
less available, and I have only a few years worth left in stock.
Sonically, I don't hear a difference, the reliability and
measurements look the same, and the customers seem to care
less as time goes on.
for the plastics, but I'm not sure that it's simply them
being less conservative.
In any case, the TO-3's are becoming more expensive and
less available, and I have only a few years worth left in stock.
Sonically, I don't hear a difference, the reliability and
measurements look the same, and the customers seem to care
less as time goes on.
Huh?, the metal packages heat up slower?? In my experience they heat up the same.
The plastic packages save production costs, so that's what industry wants. They are just as reliable as long as they are clamped to the heatsink. I'm not convinced the single screw has as good contact (just being old?)
-Chris
The plastic packages save production costs, so that's what industry wants. They are just as reliable as long as they are clamped to the heatsink. I'm not convinced the single screw has as good contact (just being old?)
-Chris
HI,
Generally the plastic version has a lower Tjmax than TO3.
Calculating back to Rjc shows that most plastic have a lower thermal resistance than TO3. Does the manufacturer recognise that the olastic part is potentially more vulnerable at elevated temps?
This confirms earlier suspicion that a bigger heatsink is recommended for plastic to keep them cooler,
If you intend operating at elevated temp then 200 deg C versions have a better SOA (I believe that 200 degC is only available in TO3).
Generally the plastic version has a lower Tjmax than TO3.
Calculating back to Rjc shows that most plastic have a lower thermal resistance than TO3. Does the manufacturer recognise that the olastic part is potentially more vulnerable at elevated temps?
This confirms earlier suspicion that a bigger heatsink is recommended for plastic to keep them cooler,
If you intend operating at elevated temp then 200 deg C versions have a better SOA (I believe that 200 degC is only available in TO3).
TO3's have a much higher torque rating than plastic.
(i broke a couple of the square dude casings)
Because of the eliptic shape of the TO3 housing, and two holes for torqueing them, pressure is distributed more evenly, contact surface is much better.
Plastic cases have elastic behavior when torqueing, not really a point with the metal TO3, also a reason for bad connecting surface.
Resistance case to heatsink is 0.30 compared to 0.50 for a TO247, at the best.
With class A amplifiers that is a major disadvantage when using plasticos, my view.
The difference of 0.20 is for every output device.
That means the heatsink will have to surrender the 0.20, 0.20 is quite a difference in heatsink.
To make it worse, the 0.20 has to be divided by the number of devices mounted on it.
As Mr Pass said: fat chance the TO3's will be there for very much longer.
Not just for output devices, also for other active parts.
The mil. spec. often came in TO3 cases.
I am not keen on spending leasure time marching, but mil spec devices always had the same effect as a dirty mag.
I am about to raid NP's stash !
(i broke a couple of the square dude casings)
Because of the eliptic shape of the TO3 housing, and two holes for torqueing them, pressure is distributed more evenly, contact surface is much better.
Plastic cases have elastic behavior when torqueing, not really a point with the metal TO3, also a reason for bad connecting surface.
Resistance case to heatsink is 0.30 compared to 0.50 for a TO247, at the best.
With class A amplifiers that is a major disadvantage when using plasticos, my view.
The difference of 0.20 is for every output device.
That means the heatsink will have to surrender the 0.20, 0.20 is quite a difference in heatsink.
To make it worse, the 0.20 has to be divided by the number of devices mounted on it.
As Mr Pass said: fat chance the TO3's will be there for very much longer.
Not just for output devices, also for other active parts.
The mil. spec. often came in TO3 cases.
I am not keen on spending leasure time marching, but mil spec devices always had the same effect as a dirty mag.
I am about to raid NP's stash !
That was accidental, i can be clumsy, Per.
I noticed a few times that bolts loosened on plastic cases, a professional audio friend already adviced me to torque output devices well, especially on class A stages.
My thought was that plastic and metal do not mix well, with large temperature differences.
Thermal expansion for plastic is much different from the one for metal.
Even combining alloy and metal causes expansion problems.
A major problem with ships, btw : cruise ships have alloy superstructures and steel hulls.
I noticed a few times that bolts loosened on plastic cases, a professional audio friend already adviced me to torque output devices well, especially on class A stages.
My thought was that plastic and metal do not mix well, with large temperature differences.
Thermal expansion for plastic is much different from the one for metal.
Even combining alloy and metal causes expansion problems.
A major problem with ships, btw : cruise ships have alloy superstructures and steel hulls.
Upupa Epops said:Look at IR's Super TO 220 package- 200 - 300 W of Ptot is normal by this devices. Sure, they must be fitted by clips or by belt by two screws - all problems with plastic case is caused by one screw fitting
I like to use big steel washers on top, and lock nuts on the
other side. Others prefer bars which press more uniformly
on the surface, but I haven't seen any data as to their
superiority and the single screw is the manufacturer's
recommended aproach.
I haven't done any experiments specifically to measure it, but I have done quite a lot of thermal measurements involving various power devices and various mounting methods, and clamping with a bar always ends up showing significantly lower thermal resistance since more pressure can be applied and it can be applied directly over the die.Nelson Pass said:
Hi guys,
How comes that nobody talks about the cost. You guys are all rich I guess (or do you get them free? If you do, please tell me how to do so). I think the TO3 are much cheaper than their counterparts/alternatives. Tell me if I am wrong. (MJL21193/4, MJ15003/4). Hope my info is not too obselete.
How comes that nobody talks about the cost. You guys are all rich I guess (or do you get them free? If you do, please tell me how to do so). I think the TO3 are much cheaper than their counterparts/alternatives. Tell me if I am wrong. (MJL21193/4, MJ15003/4). Hope my info is not too obselete.
I just bought 50 original Toshiba's on the web this week, for a ridiculous price.
Erno Borbely in Germany lists them on his homepage for 12 Euro's each.
I have a box full of 2SK146's, Borbely wants $10 for each dual fet.
Over here a subsidiary of Ebay(since a month) has so much surplus for sale that sometimes i feel like a parts trader.
Regular parts from private persons, or company stock special parts surplus.
One time from a big aeroplane manufacturer gone bust, the other ocassion from an electronics repair guy retiring, or outdated components from active businesses.
Not only does that get me exotic parts that many would crave for, i find it fun to strike on a good deal. (probably heritage)
And if you like to buy new, nowadays parts can be bought over the internet for much less than local, and one need not cross the border like i did 15 ago.
Agreed, TO3's can be cheaper, i used to wrench old stuff apart for parts long ago.
If someone wants something different than a 3055 it can be an expensive hobby, how expensive is a choice or an attitude.
I read a lot here about ordering samples for free, i have not done that up till now.
Erno Borbely in Germany lists them on his homepage for 12 Euro's each.
I have a box full of 2SK146's, Borbely wants $10 for each dual fet.
Over here a subsidiary of Ebay(since a month) has so much surplus for sale that sometimes i feel like a parts trader.
Regular parts from private persons, or company stock special parts surplus.
One time from a big aeroplane manufacturer gone bust, the other ocassion from an electronics repair guy retiring, or outdated components from active businesses.
Not only does that get me exotic parts that many would crave for, i find it fun to strike on a good deal. (probably heritage)
And if you like to buy new, nowadays parts can be bought over the internet for much less than local, and one need not cross the border like i did 15 ago.
Agreed, TO3's can be cheaper, i used to wrench old stuff apart for parts long ago.
If someone wants something different than a 3055 it can be an expensive hobby, how expensive is a choice or an attitude.
I read a lot here about ordering samples for free, i have not done that up till now.
You can clearly see the difference between TO-247 and TO-3 packages thermal mode when they are mounted on heatsink thru insulation (BeO ceramic, or aluminium oxide, or mica, or silicon pad etc.) This is the common occurence for example in amplifiers with single heatsink, like car amps. At other same conditions TO-3 will run at lower case temperature than TO-247, simply because they have larger contact area with heatsink. With some types of insulation, it becomes the main limiting factor to maximum power dissipation and this will require bigger number of devices to connect in parallel. For example common thermal resistance of power BJT in real life is about 1,5K/W (die<->heatsink) and for TO-247 mounted thru insulation, it may rise to 3...5K/W (thus decreasing maximum dissipated power by corresponding number of times). TO-3 wins in such situations. For directly mounted on flat greased surface cases, this difference vanishes 
TO-264
Hey guys (and Gals),
All I hear is TO-3 -v- TO-220 -v- TO-247. How about the TO-264? This package has the same footprint as the TO-247, is physically bigger than the '247, and some chips using this package are rated at 300W dissipation. ON's MJL4203 and '4281 are NPN & PNP audio outputs, rated at 15A Ic, and 350Vce, with an F(t) of 30MHz. They're basically the high-voltage versions of the MJL1302 and '3281, respectively. Let's check these ones out.
Steve
Hey guys (and Gals),
All I hear is TO-3 -v- TO-220 -v- TO-247. How about the TO-264? This package has the same footprint as the TO-247, is physically bigger than the '247, and some chips using this package are rated at 300W dissipation. ON's MJL4203 and '4281 are NPN & PNP audio outputs, rated at 15A Ic, and 350Vce, with an F(t) of 30MHz. They're basically the high-voltage versions of the MJL1302 and '3281, respectively. Let's check these ones out.
Steve
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.