John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
Jump to Latest
#38,201
john curl said:
JN, a VFET is not necessarily a VERTICAL FET.
Click to expand...

He didn't say that.

john curl said:
just like Demian just did by implying that a TO-92 part could easily replace a T0-220 part, even though the die inside are the same.
Click to expand...

He didn't say that.

Reading and responding to what's actually been written is much more useful than responding to the voices in your head.
 
#38,203
john curl said:
Scott, your attempted insertion of this schematic at this time was inappropriate. It would have answered nothing and would not have contributed to further understanding of the questions recently posed. Secondly, you didn't appear to 'understand' the schematic details to do anything but confuse the situation, just like Demian just did by implying that a TO-92 part could easily replace a T0-220 part, even though the die inside are the same.
That is something that JN might do, Demain! After all, with a 45 degree spread angle of heat flow, anything should be possible! I maintain that only liquid Freon would do the job properly, and maybe not well enough, given the plastic thermal resistance. '-)
Click to expand...

I knew exactly what the FET's were, I had no intention of refering to those very obscure parts. I find the circuit far more interesting than piles of paralleled unobtainable JFET's. I suspect I understand the circuit well enough, but if that were the case we might see the man behind the curtain.
 
#38,207
john curl said:
JN, a VFET is not necessarily a VERTICAL FET. It is an early power jfet that was made by NEC, Sony, and Yamaha, for example. It is a DEPLETION MODE device and it was discontinued, because it was too difficult to get a suitable yield. It is also sometimes called a Static Induction Transistor.
The power devices used in the JC-80 are called VMOS, because they are a vertical MOS structure, with ENHANCEMENT instead of DEPLETION as the fundamental property of the device. They cannot be easily interchanged.
Click to expand...
And the reason you felt it necessary to say that is????

As Sy stated, you seem to be reading what is not there.

john curl said:
Scott, your attempted insertion of this schematic at this time was inappropriate. It would have answered nothing and would not have contributed to further understanding of the questions recently posed. Secondly, you didn't appear to 'understand' the schematic details to do anything but confuse the situation, just like Demian just did by implying that a TO-92 part could easily replace a T0-220 part, even though the die inside are the same.
That is something that JN might do, Demain! After all, with a 45 degree spread angle of heat flow, anything should be possible!
Click to expand...
It all about the conduction path John. It's not about the package size. While you may be impressed with the amount of plastic or the color of the leads, neither contribute to thermal conduction. The world of electronics packaging has advanced in the last 30 years, perhaps you should come down off the pedestal and learn something.

john curl said:
I maintain that only liquid Freon would do the job properly,
Click to expand...

What is it with you and that liquid freon?? You've no idea what you are talking about here.


SY said:
Reading and responding to what's actually been written is much more useful than responding to the voices in your head.
Click to expand...

Where's the fun in that??

jn
 
Last edited:
#38,215
marce said:
Sot-89 on something like this:
http://www.ceramtec.com/files/ca_press_release_2009-11-09_1_en.pdf
or have an assembly made up with something like this:
Integrated Heat Pipe Assemblies - heat sinks
Ceramic heatsinks avoid EMC problems, or use the heat pipes to move the heat from the active devices to heatsinks moulded into the case.
Click to expand...

If I were in need of a good thermal interface, I'd go with a beryllia piece metalized as solderable on both sides. The stuff's got a thermal conductivity coefficient close to that of aluminum, and about 60% of copper.

http://materion.com/~/media/Files/P...s/CC-007BerylliumOxideCeramicsDesignGuide.pdf

For small quantity, I'd just buy some of the rf packages and use them, they already have mounting options built in. Just solder the entire package to a good piece of berquist, an alumina substrate, or directly to a good copper spreader if you need even better thermals than what the package offers as a simple screw down.

Ceramic Package|Beryllium Oxide Package|BeO package|Electronic Packaging|Materion Microelectronics - Materion

Edit: For most, just use a simple copper mounting pad and solder the device to it. While the heat flux density is high, use copper. As it spreads out, then you can use an insulator. Just don't expect an insulator like sil pad or mica to work well very close to the die, as the effective cross section of the heat path isn't large enough there. Even berquist board uses an epoxy directly under the foil, so if you solder the chip directly to the board, the thermal drop is high before it gets to the aluminum. Use a spreader over the higher resistivity materials.

jn
 
Last edited:
#38,218
scott wurcer said:
I thought I went through this exercise before (what's that definition of insane)? :headbash:
Click to expand...

Patient: Doctor, each night I have the same dream: I see mices playing football.
Doctor: I’ll prescribe you these pills. Start taking them tonight and in a few days you won’t have these dreams again.
Patient: Can I start taking them tomorrow because tonight are the finals?

George
 
#38,219
john curl said:
Hi again, everybody. I found that if you just 'Google' freon heat sink' or something similar, you can find out HOW to make your very own 'freon heat sink' and maybe its history. JN, please ignore this message. '-)
Click to expand...
Silly student.

Eventually, via continual random snarky erroneous statements, hopefully you will decide to learn something. But, maybe not.

However, there is a good aspect of your continual badgering, that is, others can learn and discuss.

The coolers you refer to are used to transport lots of heat from a large area device. Think about the area of a CPU chip. It's major. But the power density per square inch is low.

Now try to pull 280 watts out of an IRFP460 using that thing.What do you expect to happen?

First, you mount the case to the copper spreader used in quite a few of them.. you do so using what, grease? What's the thermal interface resistance? What's the power density?

Hint..it takes calculation.

The die is .250 by .360.

The spread at the bottom of the package is .330 by .440.

Now what's the thermal resistance of a greased contact to an electrically hot sink where the effective thermal channel is .145 sq inches? Or Mica with two interfaces?

The entire thermal mangement game is all about keeping the junction temperature as low as possible.

Tossing a "freon cooler" at a high power density semiconductor is not engineering.

No wonder you blow up silicon.

Try engineering a thermal solution instead.

jn
 
Status
Not open for further replies.
Top Bottom