电源次级肖特基二极管老是烧坏，求分析失效原因。

oped001

有个电源在客户那边老是烧坏，烧坏的元件全部是次级的肖特基二极管，过几天要到客户那边去现场确认失效原因，恳请各位朋友先帮忙一起分析一下！

这个产品不是我们家的，真正电路图不方便贴出来，但基本跟这张一个拓扑。烧坏的位置是D7，我用的是单管。目前的分析思路主要有三个方向：
1，先实测一下电源正常工作时D7两端电压是多少，有没有超标？
2，检查不良电源的稳压反馈回路有没有问题，如果稳压反馈回路有问题，D7承受的电压会不会超标？
3，检查电源负载瞬间变化极大时D7承受的反向电压会不会超标。

目前基本排除过电流打死二极管的可能性，因为换成200V的材料后没再发生这种问题，所以基本判断是过电压把二极管打死。如果各位朋友有这方面的经验，还请不吝赐教，谢谢！

gyzzg2030

15V的输出电压，起码得150V的整流管

oped001

gyzzg2030 发表于 2017-8-28 20:03
15V的输出电压，起码得150V的整流管

实际输出电压是12V，如果用实验室的AC source供电不一定能抓到反向电压超标，但要是到了市场上就两说了。所以想在实验室里面确认一下市场上供电最苛刻时加在D7两端的电压是多少，如果有这方面的经验还请兄弟指点一下，谢谢！

yunhuisong

100V太低了，风险很大。之前做的一个反激12V输出用的100V的总是偶尔有损坏的，后来换成200V就再也没出过问题了

rainbow

还有一个问题要注意，就是二极管的温度，肖特基的温度升高后反向漏电流增大，最大为50mA，大的漏电流又加剧了温度的上升，所以高温问题一定要注意。

funnynypd

Do you have the datasheet of D7. What brand is it? If it is a Chinese brand, you got to be very careful and try stay away from them.
We have learned a deep lesson on it.

oped001

rainbow 发表于 2017-8-28 23:41
还有一个问题要注意，就是二极管的温度，肖特基的温度升高后反向漏电流增大，最大为50mA，大的漏电流又加剧 ...

是这样，一般温度很苛刻的地方都用超快恢复二极管，不用肖特基。不过这次管子像是瞬间就被打死了，瞬间高压的可能性大一些，还是谢谢你！

oped001

funnynypd 发表于 2017-8-28 23:51
Do you have the datasheet of D7. What brand is it? If it is a Chinese brand, you got to be very care ...

I think you are '老外', honestly speak most of the diodes we used are made in China and Taiwan, even though we see lots of overseas brands. Can you imagine the quote if all diodes are manufactured in USA? it will be sky high price. So you have to audit your supplier by yourself，no matter Chinese brands or others.
We always say，别一棍子打死一船人..

funnynypd

> diodes we used are made in China and Taiwan,
Agree. Most of foreign brand of diodes can be made in China. As long as they stick with the same manufacturing standard as "made in USA", we have no problem.

The issue is: a few of those local Brand, they cannot match the same performance, but copy the data sheet from foreign brand. Then comes the headache.

funnynypd

For MBR20100CT, if you use "Diodes Incorporated" brand, or "Vishay", or "On Semi", or "Taiwan Semi", or even some reputable Chinese brand, you will have no trouble.
If you use a unknown brand, e.g. "jiangnan-江南", or many no-brand fab factory from ShiChuang province, I guarantee you will have the big headache as the first post mentioned.

MBR20100CT is a 2x5A diode, on our application for "Diodes Incorporated" brand it can working properly as the datasheet says.
For power semiconductors, it would be safe to stay with those reputable brands. Chinese brand or Foreign brand.

funnynypd

>瞬间高压的可能性大一些
Our experience shows: it is the working temperature matters.
"Diodes Incorporated" brand can survive at -40C to +155C, while many others may not, and get damaged due to excessive heat generated by the large current.
We learned the lesson recently just because the factory changed the Diode supplier (without notify us), and changed from "Diodes Incorporated" brand to a local unknown brand.
When a new batch of power supply was received couple of month ago, I was reported that during our incoming inspection, we got 100% failure rate just because of this supplier change.
There is corruption on Chinese government, there is also corruption on Factory Engineer and purchasing section (on components selection and purchasing). A few of those engineer is risking their life on this kind of corruption.
It looks like we are not alone. That same brand (or similiar brand) may hurt many other power supply manufacturers.

gyzzg2030

oped001 发表于 2017-8-28 21:25
实际输出电压是12V，如果用实验室的AC source供电不一定能抓到反向电压超标，但要是到了市场上就两说了。 ...

最大多少倍，这也要根据你的变压器变比来的，一般反激的电源，在市电最高输入情况下，会达到近10倍的输出电压的反峰

tang0571

学习了，marking

yan_hua

可以在后面加一个TVS管吸收反向的瞬态峰值电压，做电压钳位。需要样品可以找我[/捂脸]

dykwai1

试一下 D7 并联RC，看到不少电源这么做的，据说起抑制尖峰及保护作用

peter_zhang

大兄弟 先并联一个RC电路吸收尖峰，R≈几十到100Ω，C≈1nF 左右；
再者你算一下二极管要承受的反相电压时多少，(输入最大电压VDCmax/匝比N)+VOUT   为二极管承受的理论最大电压，
看你选的是否合适，并且留够余量，最大尖峰电压到肖特基的反相电压的80%左右差不多了。
最好是用示波器一看肖特基二端便知，根据具体波形调整RC并联值。

peter_zhang

yan_hua 发表于 2017-8-29 09:50
可以在后面加一个TVS管吸收反向的瞬态峰值电压，做电压钳位。需要样品可以找我[/捂脸] ...

这种反激电源，这里没有加TVS的做法，都是直接加RC.

yutianyiren

测试一下打火条件下的二极管应力，肖特基还是余量大点好，不耐cao。

mkliop

次级二极管没有尖峰吸收电路？

duanll

发一下你变压器参数和输入电压范围，没有这些说啥都没用。

oped001

电源一上电管子两端电压91V，满载时大约85V，实验室里面测的，设计余量是比较小。

zhiwei

要全范围测量。一般这里都要加RC或RCD吸收漏感尖峰。

fengyunyu

高手很多

qinxg

电源电路是计算出来的, 有没有把各个点的占空比, 电压,电流峰值计算出来?  D7的反射电压是多少?  把电源端调到270V,测试下D7两端电压.