Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

donate to eham
   Home   Help Search  
Pages: [1]   Go Down
  Print  
Author Topic: transistor junction temperature  (Read 1360 times)
AJ4SN
Member

Posts: 70




Ignore
« on: November 14, 2017, 01:33:19 PM »

I am working on a 200 watt solid state amplifier, and I want to manage the transistor heat properly.  The VRF151 transistor that I plan to use has a maximum junction temperature of 200 degrees centigrade. I'd like to run it at about 130 degrees centigrade for longer life. It has a thermal resistance rating of 0.6 degrees centigrade per watt. My question is the relationship between case temperature and junction temperature.

If I am running two transistors in push pull for a total output of 200 watts, and the amplifier efficiency is 50%, then each transistor has to dissipate 100 watts. The thermal "drop" across the case is: 100 watts x 0.6 C/W = 60 degrees C.

In order to keep the transistor junction temperature at 130 degrees, the transistor case temperature must not exceed 70 degrees (130 - 60 = 70). Is that correct? Does the ambient air temperature enter into the calculation?

Thanks for your help!

Stan
aj4sn
Logged
AA4PB
Member

Posts: 14354




Ignore
« Reply #1 on: November 14, 2017, 02:06:40 PM »

Correct. The ambient temp comes into play when calculating the heat sink. The heat sink temperature rise above ambient will be rated in deg C per watt.

It's all like a series string of heat loss: junction - case - heat sink - ambient.
« Last Edit: November 14, 2017, 02:12:19 PM by AA4PB » Logged

Bob  AA4PB
Garrisonville, VA
N5EG
Member

Posts: 313


WWW

Ignore
« Reply #2 on: November 14, 2017, 02:13:01 PM »

Hi Stan - Yes, your calculations are in the right format. 
The thermal specification for the transistor is probably junction-to-case.
The computed temperature rise of 60C means the temperature rise of the
junction above the case of the transistor.

However there are some other sources of thermal resistance you also need to
include in your analysis:

1. Transistor case to heatsink thermal resistance.
2. Heatsink to ambient air thermal resistance. This thermal resistance will decrease
as you go from no forced air to forced air cooling.

Thermal resistance works like electrical resistance - you add the various
thermal resistances to compute the total thermal resistance.

As a hypothetical example:

Let's assume that the transistor-case-to-heatsink thermal resistance is 0.2 C/W, and that
the heatsink-to-forced-air thermal resistance is 0.5 C/W.    You'll need to find the actual
numbers for your exact hintsink, transistor mounting method, etc. These are just hypothetical.

The total thermal resistance from the transistor junction to ambient air would be:
0.6 C/W + 0.2 C/W + 0.5 C/W  or  1.3 C/W.

At 100 watts per transistor, then:

Tj = 100 * 1.3 = 130 C rise over the ambient air temperature.  If ambient air were
25C, then the heatsink temp would be  25C + 0.5 C/W * 100 W = 75 C
and the case of the transistor would be   75C + 0.2 C/W * 100 W = 95 C
while the junction would be  95C + 0.6 C/W * 100W = 155 C

-- Tom, N5EG





Logged
AJ4SN
Member

Posts: 70




Ignore
« Reply #3 on: November 15, 2017, 05:54:49 PM »

Bob, Tom,

Thanks for taking time to reply. I appreciate it. Tom, your explanation makes it very clear. I think this is a fascinating subject. I'm going to try to learn more about it as I continue with this project.

73
Stan
Logged
AJ4SN
Member

Posts: 70




Ignore
« Reply #4 on: November 21, 2017, 06:30:11 PM »

Thanks to your help, I think that I have a handle on the heat flow situation for this amp. I was able to find a very cool website that will calculate the thermal properties of various sizes of aluminum heat sinks including adjustments for the number and thickness of the fins:

https://www.myheatsinks.com/calculate/thermal-resistance-plate-fin/

I know that some of the high-powered solid state linear amps also utilize a copper heat spreader between the transistors and the aluminum heat sink. What is the effect of a heat spreader, and how do you calculate the thermal impact of adding one? Would a copper heat spreader benefit a relatively low powered (200 watt) amp like this one?

Stan
Logged
G3RZP
Member

Posts: 8140




Ignore
« Reply #5 on: November 23, 2017, 09:06:53 AM »

It gets more complicated because you have the thermal resistance between case and spreader and thermal  resistance spreader to heatsink - which includes the thermal resistance of the spreader itself. When I was working (32 years in semiconductors), we had one bipolar transistor array which we sold for $150 a pop. Guaranteed to work for 30 minutes at 270C - used in drilling bits for oil wells. They usually lasted 45 to 50 minutes...As you can imagine, doing the thermal calculations for that was fun.
Logged
AC7CW
Member

Posts: 1002




Ignore
« Reply #6 on: November 23, 2017, 10:09:45 AM »

Hints and Kinks: Black grackle paint increases heat flow from a bare aluminum heatsink to air. I don't know if it's better than anodizing. Grackle increases the surface area of the heatsink but introduces thermal resistance too.

Here's some info on power fets http://www.ixys.com/Documents/AppNotes/IXAN0061.pdf  It seems possible that a thermal reference attached to the case could be used to infer the junction temperature. I wonder if anybody incorporates that into their amps...
Logged

Novice 1958, 20WPM Extra now... (and get off my lawn)
G3RZP
Member

Posts: 8140




Ignore
« Reply #7 on: November 24, 2017, 05:27:12 AM »

Back in the early 1970s, RCA did a bipolar RF power transistor which had an on-chip diode for chip temperature measurement and bias correction. I've never understood why that didn't become common.
Logged
AJ4SN
Member

Posts: 70




Ignore
« Reply #8 on: November 24, 2017, 07:32:39 AM »

  ". . . It seems possible that a thermal reference attached to the case could be used to infer the junction temperature. I wonder if anybody incorporates that into their amps.." AC7CW.

Actually, that's what I plan to do. I am going to attach an LM35 temperature sensing IC to the case.  I've got an inexpensive digital voltmeter that will be mounted on the front panel. Since I know the thermal resistance of the case (0.5 degrees C per watt). Then I should be able to determine the junction temperature. The maximum temperature for the device is 200 degrees C, but I would like to keep it at about 130 degrees C for longer life. I am adding a fan control circuit with a front panel adjustment to enable "empirical finalizing." That should help me find the sweet spot for the fan. I'd prefer not to have it running when it's not needed. I do more listening than transmitting and I only want it to run when necessary.
Logged
G3RZP
Member

Posts: 8140




Ignore
« Reply #9 on: November 25, 2017, 02:37:55 AM »

The thermal resistance quoted for 'junction to case' is usually 'junction to mounting face of case', and not necessarily the 'junction to top of case' resistance.
Logged
W9IQ
Member

Posts: 1708




Ignore
« Reply #10 on: November 25, 2017, 03:24:24 AM »

The case temperature is generally a lagging indicator of problems. Not that it shouldn't be monitored but it is essential to monitor primary factors such as voltage, current, etc. to prevent device destruction.

- Glenn W9IQ
Logged

- Glenn W9IQ

I never make a mistake. I thought I did once but I was wrong.
Pages: [1]   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!