有关MOSFET的温升问题

我用3842做了一个单端反激24V/5A,Imax=2.7A,开关管选用IRFPF40,4.3A/900V/Rds=2.5ohm,满载很烫,现在是冬天,问题不大,怕到了夏天会挂掉.改用K2611,9A/900V/Rds=1.3ohm,还是很烫,用手摸感觉温度和以前差不多.对不起大家,这个是小日本的,只是做实验用,大家一定原谅啊!
我觉得有点奇怪,大家以为?还有,输出整流管用16A烫手,3倍裕量还不够吗?

全部回复(35)

只看楼主

正序查看

倒序查看

智慧大头

LV.8

2005-12-17 14:47

效率?

hulx

LV.5

2006-02-07 14:48

@智慧大头

效率?

限于条件,侧不了效率啊!
拜请各位前辈指点!

小林子

LV.2

2006-02-07 15:23

120W用反激可要下功夫啊.

superpowersupply

LV.6

2006-02-07 15:50

It must be overheat,but I have designed 135W power supply it is not overheat.Because of your mosfet ON-resistance too high---2.5OHM/1.3ohm.Did you calculate on-state heat?I think you can use the mosfet 20N60C3 instead your past used.But you should carefully design the snub circuit,the 20n60c3 VDS break down voltage is 600v ONLY.But it is possible.

hulx

LV.5

2006-02-08 10:07

@小林子

120W用反激可要下功夫啊.

说得对

superpowersupply

LV.6

2006-02-08 10:16

@小林子

120W用反激可要下功夫啊.

I don't think so,I designed 45V3A OUTPUT,universal input lead acid charger,it is works well.And it is not overheat.

hulx

LV.5

2006-02-08 10:32

@superpowersupply

Itmustbeoverheat,butIhavedesigned135Wpowersupplyitisnotoverheat.BecauseofyourmosfetON-resistancetoohigh---2.5OHM/1.3ohm.Didyoucalculateon-stateheat?Ithinkyoucanusethemosfet20N60C3insteadyourpastused.Butyoushouldcarefullydesignthesnubcircuit,the20n60c3VDSbreakdownvoltageis600vONLY.Butitispossible.

看一下英飞凌的手册,20N60C3的Rds是很小,还是TO220封装.只是使用20A的管子会不会很浪费?谁会提供参考价?

hulx

LV.5

2006-02-08 10:34

@superpowersupply

Idon'tthinkso,Idesigned45V3AOUTPUT,universalinputleadacidcharger,itisworkswell.Anditisnotoverheat.

说说你的产品的散热器尺寸可否?多谢你的回复.

superpowersupply

LV.6

2006-02-08 10:49

@hulx

看一下英飞凌的手册,20N60C3的Rds是很小,还是TO220封装.只是使用20A的管子会不会很浪费?谁会提供参考价?

I think rmb 5-8 Dollars?

superpowersupply

LV.6

2006-02-08 11:06

@hulx

说说你的产品的散热器尺寸可否?多谢你的回复.

about L100*H30*3mm

sunnytong

LV.4

2006-02-08 11:13

你可以算一下MOS上面的損耗就知道為何很燙了.

cf100

LV.7

2006-02-08 11:27

MOSFET发热原因很多,大多是电路结构上的问题.

比如,频率过高,电路自激,上升下降沿过缓.

sunnytong

LV.4

2006-02-08 13:57

@superpowersupply

Itmustbeoverheat,butIhavedesigned135Wpowersupplyitisnotoverheat.BecauseofyourmosfetON-resistancetoohigh---2.5OHM/1.3ohm.Didyoucalculateon-stateheat?Ithinkyoucanusethemosfet20N60C3insteadyourpastused.Butyoushouldcarefullydesignthesnubcircuit,the20n60c3VDSbreakdownvoltageis600vONLY.Butitispossible.

20n60c3的VDS break down voltage 不一定是600V 這要看尖峰維持時間,如果小於0.2uS電流小于3A是時 bread down voltage is 700V.請查證

cf100

LV.7

2006-02-08 14:24

用6N60做200W以下的反激,在现在的温度,温升不会超过40度.所以不是元件的问题.

hulx

LV.5

2006-02-08 14:39

@superpowersupply

Ithinkrmb5-8Dollars?

￥5-8还是$5-8?悄悄问一声:你用英文方便?

hulx

LV.5

2006-02-08 14:41

@sunnytong

20n60c3的VDSbreakdownvoltage不一定是600V這要看尖峰維持時間,如果小於0.2uS電流小于3A是時breaddownvoltageis700V.請查證

这个问题不是很大.
1139380842.pdf

hulx

LV.5

2006-02-08 14:46

@cf100

MOSFET发热原因很多,大多是电路结构上的问题.比如,频率过高,电路自激,上升下降沿过缓.

频率55K,edge还可以,自激就不好说,不知道什么样的波形算好,书上资料里面的都是理想状态波形.如果哪位方便上一个实际的波形看看?不知道可以不?

fish@mouse

LV.3

2006-02-08 15:43

你这个很烫,是多少度呀?

shine_sky

LV.2

2006-02-08 15:45

老兄,最好能把原理图发出来大家一起看一下,这样讨论起来方便些.
或者用IRFPF50,6.7A/900V/Rdson=1.6ohm.不过输入结电容比IRFPF40大一些可能要注意一下驱动.

superpowersupply

LV.6

2006-02-08 16:55

@hulx

￥5-8还是$5-8?悄悄问一声:你用英文方便?

RMB

lwp1975

LV.7

2006-02-08 22:09

我想应该是频率的问题,可适当地降低频率试试,减少器件的损耗,说不定有意外的收获,纯属个人意见,见笑了!哈哈...

hulx

LV.5

2006-02-09 23:28

@fish@mouse

你这个很烫,是多少度呀?

手可以受得了,能有70-80度的样子?

fish@mouse

LV.3

2006-02-10 09:55

@hulx

手可以受得了,能有70-80度的样子?

70,80度手还能受得了!你的散热器多大,怎么连接,冷却方式,还有磁芯型号?

superpowersupply

LV.6

2006-02-10 10:05

@sunnytong

20n60c3的VDSbreakdownvoltage不一定是600V這要看尖峰維持時間,如果小於0.2uS電流小于3A是時breaddownvoltageis700V.請查證

Thank you,but I failed to browse the 20N60C3 specification,But in my memory the VDS bread down voltage will be 600V MAX,would you please upload it?
Thank you evry much.