Mosfet Power dissipation question

[holmzy]

Hi guys,

I have a Nad C300 with the outputs blown. It uses original part STP55NE06L which is not readily available. The suggested replacement STP55NF06L is readily available. But I noticed that the amplifier now runs hotter even with the same bias etc.

The data sheet states the original is 130w power dissipation and the replacement is 95w. I dont understand the other parameters either

Can anyone please help me to understand if this is why it is running hotter than before?

[eccerr0r]

Re: Mosfet Power dissipation question

It appears the NF was supposed to be an "upgrade" to the NE, but these were meant to be run in switch mode. Running them in forward-active analog mode the improved parameters don't matter as much.

The 130W for the NE and the 95W for the NF are package limits and don't have anything to do with how much an application will dump heat. In all honesty dumping 130W from a TO-220 is tough to unlikely anyway.

I don't have the schematic handy but it's possible the bias point for the gate to position the transistor from cutoff into linear vs switching mode is different between the transistors and you may need to tweak the bias. The thresholds of the transistors are specced similarly but the range is fairly large which makes a difference when running in linear mode.

[holmzy]

Re: Mosfet Power dissipation question

Wowzer information overload!

I would love to fully understand it but as mentioned before the service manual states the idle at 160-180 and both times was set to 160 but was noticeably hotter with the NF.

I have attached link to service manual. If NF is an upgrade 95w is better? Lower numbers are better? Thanks

https://www.hifiengine.com/manual_li...nad/c300.shtml

[holmzy]

Re: Mosfet Power dissipation question

Also what do you mean the range is fairly large?

They were both operated in the same amplifier and bias set the same etc but I noticed the NF was hotter. I explained it again just so it was clear.

[R_J]

Re: Mosfet Power dissipation question

What is the offset voltage? is it near 0v.What are you using to measure the temperature? If the mosfets were indeed running hotter they would also be drawing more current.

[holmzy]

Re: Mosfet Power dissipation question

Hi, are you saying the DC offset can affect the temperature of the MOSFET? I just measured using my hands.

I dont know if it was drawing more current but like I mentioned I did adjust the bias according to the manual.

I did notice I had to turn the trimmer pot for the idle right down as it was very high with the NF (newer) MOSFET

[R_J]

Re: Mosfet Power dissipation question

if there is an offset one mosfet is likely on more than the other, it only takes a second to check the voltage. You may also have an issue with the driver circuit, How did the mosfets fail? if they shorted drain or source to gate, that usually causes other damage in the drive circuit.

[holmzy]

Re: Mosfet Power dissipation question

if there is an offset one mosfet is likely on more than the other, it only takes a second to check the voltage. You may also have an issue with the driver circuit, How did the mosfets fail? if they shorted drain or source to gate, that usually causes other damage in the drive circuit.

Well I got the amplifier and it was working with original transistors, I set all the bias and DC etc then I managed to short the cables.

So I ordered the NF replacement and noticed it was hotter than before and the bias was very high on previous setting, after adjusting it down I noticed it was hotter and as I was doing the DC I damaged to touch the leads and another short!

Both times the only damage was to the output MOSFETs

[holmzy]

Re: Mosfet Power dissipation question

But original questions still remains about the power dissipation factor. I still dont understand it and there is a big difference between the 95w and 130w. I can get the original part back in but at greater cost. I'm just trying to figure out if the extra heat was due to the newer NF MOSFET having 95w and old having 130w.

Would higher PD make it run hotter or cooler?

[R_J]

Re: Mosfet Power dissipation question

Normally It would suggest the 130Watt mosfet will handle a higher amount of power than the 95watt version. The 95 watt one should not operate at any higher temperature when the amp is operating at low power.

[eccerr0r]

Re: Mosfet Power dissipation question

The threshold voltage of the MOSFETs turning on is usually typical and actually is a range hence it's specified as such on the spec sheet. So you typically do have to tweak bias voltages when you swap a transistor out. When using the transistor as a saturated switch, the voltage is not as important as the previous stage will be switching well past threshold to turn these things on and off.

The 130W and 95W rating is how much power the transistors can theoretically dump before killing itself, assuming you have an infinite heat sink. I have to suspect the two are actually quite similar, despite the 25% reduction, because the derating for the NE is faster than the NF. In any case, you are NOT going to be typically, especially at idle, running at these levels, unless you are building an adjustable dummy load with the transistor.

So the answer for your specific question is that the max power dissipation has no effect on actual power dissipation - only the bias determines heat generation.

Calculating actual heat generation is a bit tricky. The NF unit is "better" for switching supplies because it has a lower RdsON parameter. However if you're running in linear mode which I have to suspect because I don't have an account, this parameter has no meaning as you're not running it as a switch. However it will still have a resistance and will dissipate power, and depending on what the other parts of the circuit is doing, it can influence overall heat production.

[budm]

Re: Mosfet Power dissipation question

If you look at spec sheet ref to 25C ambient and both devices has 175C max Junction Temperature.
1) NE has Junction to Case thermal resistance of 1.15 C/Watt, so at 130W the junction will be 149.5C
2) NF has Junction to Case thermal resistance of 1.58 C/Watt which is higher thermal resistance than NE, so at 95W the junction will be 150.1C
3) You said the NF is hotter than the NF, so did you verify that the mounting, the thermal grease are applied properly for the heat sink to dissipate the heat from the device?

[holmzy]

Re: Mosfet Power dissipation question

Thanks for the replies.

There is no thermal grease just thermal pad and a screw. They were both mounted identically and exact same amp. It just seemed the NF was noticeably hotter at the same bias and volume.

Only difference was the NF bias was 100mv and the NE was 0mv bias. I was adjusting the 100mv before it blew (again)

If high DC offset has a serious affect on the MOSFET then I guess that answers that if not then I need original MOSFETS as replacements run a bit too hot!

[budm]

Re: Mosfet Power dissipation question

Thanks for the replies.

There is no thermal grease just thermal pad and a screw. They were both mounted identically and exact same amp. It just seemed the NF was noticeably hotter at the same bias and volume.

Only difference was the NF bias was 100mv and the NE was 0mv bias. I was adjusting the 100mv before it blew (again)

If high DC offset has a serious affect on the MOSFET then I guess that answers that if not then I need original MOSFETS as replacements run a bit too hot!

I cannot see why you 0mV bias, that means it is OFF. What is the bias setting procedure?
BTW, bias is not the same as offset.
Per service manua:
Offset Voltage is 0V +/- 30mV
Idiling current on D305/405 is 160 ! 180mV
So are you sure you follow the instruction in the service manual?

[holmzy]

Re: Mosfet Power dissipation question

The 0mv was the DC offset not the bias.

On the original transistors the bias was set to 160mv and the DC was set to 0mv (or as close to it).

The NF replacement transistors the bias was also set to 160mv but the DC was measuring 100mv. It was running like that for 15 minutes and I noticed it was very hot compared to the original only a few days back.

I thought it was running hot due to the high DC so when I went to adjust it unfortunately the DC wires touched between the channels and shorted the MOSFETs.

So now I need yet another set of replacement MOSFETs but I just wanted to know if I should go for the original which cost more or the cheaper replacement NF but the heat thing is an issue as it runs hot anyway. I just didnt want to put in the NF if it caused it to get more hot or if the high DC had something to do with it and it actually runs cooler?

[budm]

Re: Mosfet Power dissipation question

1) So you cannot adjust the offset Voltage to be within +/- 30mV?
2) What are the actual temperature readings?

[holmzy]

Re: Mosfet Power dissipation question

1) So you cannot adjust the offset Voltage to be within +/- 30mV?
2) What are the actual temperature readings?

Yes I can. I measured it after 15 minutes of the amplifier running realised the DC was high (speakers were not connected) and went to adjust it and it popped

I dont know the actual temps. I put my hand across the heatsink and NF was easily higher. The difference was very hot and very very hot if that makes sense

[budm]

Re: Mosfet Power dissipation question

Yes I can. I measured it after 15 minutes of the amplifier running realised the DC was high (speakers were not connected) and went to adjust it and it popped

I dont know the actual temps. I put my hand across the heatsink and NF was easily higher. The difference was very hot and very very hot if that makes sense

So why not adjust it to 0 then? If you adjust the off set pot and causes the MOSFET to blow then the pot may be bad and has open circuit or something when move the pot.

[holmzy]

Re: Mosfet Power dissipation question

So why not adjust it to 0 then? If you adjust the off set pot and causes the MOSFET to blow then the pot may be bad and has open circuit or something when move the pot.

I shorted the test leads by accident! The pot is fine

[budm]

Re: Mosfet Power dissipation question

So adjust it to be 0mV +/- 30mV offset then.