Re: A model proposal
Be careful with that formula. The orientation is not taken into account. A heatsink with vertical fins is more efficient than one with horizontal fins. The author does not derate his formula accordingly. E. g. he assumes a uniform disspation, when he calculates the thermal resistance of a case. Top and bottom however dissipate less than the sides, front and rear due to their orientation.
From http://www.heatsinkdesigner.com/ you can download a free trial version of a reliable heatsink calculator.
toolsresearch said:
Be careful with that formula. The orientation is not taken into account. A heatsink with vertical fins is more efficient than one with horizontal fins. The author does not derate his formula accordingly. E. g. he assumes a uniform disspation, when he calculates the thermal resistance of a case. Top and bottom however dissipate less than the sides, front and rear due to their orientation.
From http://www.heatsinkdesigner.com/ you can download a free trial version of a reliable heatsink calculator.
Re: Re: A model proposal
Wow, that program is really great! In trial mode length and width are fixed...? The 495 price sounds good; however, it is for an annual license!?
There is an Excel calculator at http://sound.westhost.com/download.htm
Gratis!
pacificblue said:
Wow, that program is really great! In trial mode length and width are fixed...? The 495 price sounds good; however, it is for an annual license!?
There is an Excel calculator at http://sound.westhost.com/download.htm
Gratis!
So how hot may a chipamp run?
The datasheets tell us for all models that 150 °C is the limit. They also tell us the maximum heat dissipation at 25 °C for certain models. And we know that we have to derate above 25 °C. But how does that work?
Not too difficult. The graphic solution is to make a graph, where the horizontal axis is the temperature and the vertical axis is the heat dissipation. Draw a line from the point (25 °C, maximum heat dissipation) to the point (150 °C, 0 W). Calculate Pd for your application. look in the graph, which temperature corresponds to that dissipation. You will find examples of such graphs in datasheets of power transistors.
The mathematical solution is Tic = (Tmax - (Pd * ((Tmax - T@Pdmax)/(Pdmax - Pd@Tmax)). Tmax is always 150 °C, T@Pdmax is always 25 °and Pd@Tmax is always 0 W, which leads us to the short practical form Tic = Tmax - (Pd * 125 / Pdmax)
E. g. LM3886T has a Pdmax of 125 W @ 25 °C. If this IC would have to be used with a power dissipation of 30 W (33,5 V rails with 8 Ohm, 24 V rails with 4 Ohm), the IC could operate safely up to a temperature of 150 °C - (30 W * 125 K / 125 W) = 120 °C.
The datasheets tell us the heat dissipation for the following ICs at 25 °C.
LM1875 derived from the graph on page 3 ~42 W
LM1876T 62,5 W
LM3875T 125 W
LM3886T 125 W
LM4780 125 W
Now many of us are interested in knowing about the performance of the isolated packages. Thanks to the Overture Design Guide, we can extract that information.
LM1876TF 47,5 W
LM3875TF 62,5 W
LM3886TF 62,5 W
Are those numbers shocking or what? That means the LM3886TF would only be safe up to 90 °C with 30 W Pd. Subtract its own heat resistance of 2 K/W and 0,2 K/W for thermal grease 2,2 K/W * 30 W = 66 K and the heatsink should not become warmer than 24 °C. Forget about the heatsink, buy an air condition instead.
To check the validity of those numbers I extracted the values for the ICs from the first list in the same way. The numbers coincided with the datasheet information with one exception. According to the Overture Design Guide the LM4780 has a a capability of 155 W. This seems logical, because the package is bigger.
But let us check. The specified 35 V rails for 8 Ohm lead to 65,73 W heat dissipation.
150°C - (65,73 * 125 K / 155 W) = ~97 °C
150°C - (65,73 * 125 K / 125 W) = ~84 °C
0,8 K/W for the IC and optimistic 0,2 K/W for isolation washer and thermal grease lead to 1 K/W * 65,73 w = 65,73 K. Heatsink temperatures could be ~31 °C with 155 W, but only ~17 °C with 125 W dissipation capability. Maintaining a heatsink below 31 °C with 65,73 W is already an effort that demands a heatsink with less than 0,091 K/W for ambient temperatures up to 25 °C. A fan will be necessary. 17 °C needs air conditioning or should only be operated outside on cold winter days. So it is probably safe to assume that the LM4780 is indeed capable of 155 W heat dissipation at 25 °C.
The datasheets tell us for all models that 150 °C is the limit. They also tell us the maximum heat dissipation at 25 °C for certain models. And we know that we have to derate above 25 °C. But how does that work?
Not too difficult. The graphic solution is to make a graph, where the horizontal axis is the temperature and the vertical axis is the heat dissipation. Draw a line from the point (25 °C, maximum heat dissipation) to the point (150 °C, 0 W). Calculate Pd for your application. look in the graph, which temperature corresponds to that dissipation. You will find examples of such graphs in datasheets of power transistors.
The mathematical solution is Tic = (Tmax - (Pd * ((Tmax - T@Pdmax)/(Pdmax - Pd@Tmax)). Tmax is always 150 °C, T@Pdmax is always 25 °and Pd@Tmax is always 0 W, which leads us to the short practical form Tic = Tmax - (Pd * 125 / Pdmax)
E. g. LM3886T has a Pdmax of 125 W @ 25 °C. If this IC would have to be used with a power dissipation of 30 W (33,5 V rails with 8 Ohm, 24 V rails with 4 Ohm), the IC could operate safely up to a temperature of 150 °C - (30 W * 125 K / 125 W) = 120 °C.
The datasheets tell us the heat dissipation for the following ICs at 25 °C.
LM1875 derived from the graph on page 3 ~42 W
LM1876T 62,5 W
LM3875T 125 W
LM3886T 125 W
LM4780 125 W
Now many of us are interested in knowing about the performance of the isolated packages. Thanks to the Overture Design Guide, we can extract that information.
LM1876TF 47,5 W
LM3875TF 62,5 W
LM3886TF 62,5 W
Are those numbers shocking or what? That means the LM3886TF would only be safe up to 90 °C with 30 W Pd. Subtract its own heat resistance of 2 K/W and 0,2 K/W for thermal grease 2,2 K/W * 30 W = 66 K and the heatsink should not become warmer than 24 °C. Forget about the heatsink, buy an air condition instead.
To check the validity of those numbers I extracted the values for the ICs from the first list in the same way. The numbers coincided with the datasheet information with one exception. According to the Overture Design Guide the LM4780 has a a capability of 155 W. This seems logical, because the package is bigger.
But let us check. The specified 35 V rails for 8 Ohm lead to 65,73 W heat dissipation.
150°C - (65,73 * 125 K / 155 W) = ~97 °C
150°C - (65,73 * 125 K / 125 W) = ~84 °C
0,8 K/W for the IC and optimistic 0,2 K/W for isolation washer and thermal grease lead to 1 K/W * 65,73 w = 65,73 K. Heatsink temperatures could be ~31 °C with 155 W, but only ~17 °C with 125 W dissipation capability. Maintaining a heatsink below 31 °C with 65,73 W is already an effort that demands a heatsink with less than 0,091 K/W for ambient temperatures up to 25 °C. A fan will be necessary. 17 °C needs air conditioning or should only be operated outside on cold winter days. So it is probably safe to assume that the LM4780 is indeed capable of 155 W heat dissipation at 25 °C.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.