PFC
哪位大哥有用過L6561 transient mode controller, 現(xiàn)在我碰到振蕩問題無(wú)法解決,可以指點(diǎn)一下怎樣去調(diào)compensation for feedback ?Thank you very much .
PFC
全部回復(fù)(13)
正序查看
倒序查看
@cxzcjz
幾種方法幾種方法,僅供參考.1.調(diào)節(jié)反饋增益.2.過零檢測(cè)端對(duì)地加小電容.3.過零檢測(cè)端加偏置.4.如間歇振蕩,可提高動(dòng)態(tài)速度,如8-4加電阻.5.注意電壓取樣反饋線
6561
Thanks a lot, I tried to debug compensation, but no more help, I checkd the ZCD, no any abnormal. the problem occured at higher power output power, the ripple of buld cap occured little step follow the sinusoid waveform. did you meet this problem, yes, I will try according as your second idea, thanks a lot.
Thanks a lot, I tried to debug compensation, but no more help, I checkd the ZCD, no any abnormal. the problem occured at higher power output power, the ripple of buld cap occured little step follow the sinusoid waveform. did you meet this problem, yes, I will try according as your second idea, thanks a lot.
0
回復(fù)
@cxzcjz
幾種方法幾種方法,僅供參考.1.調(diào)節(jié)反饋增益.2.過零檢測(cè)端對(duì)地加小電容.3.過零檢測(cè)端加偏置.4.如間歇振蕩,可提高動(dòng)態(tài)速度,如8-4加電阻.5.注意電壓取樣反饋線
compensation
Could you explain for the 20Hz bandwidth for the feedback compensation, theoretic, it will improve power factor, so the cap of feedback compensation need to use bigger, but I found it will cause unsatable, do you have some idea?
Could you explain for the 20Hz bandwidth for the feedback compensation, theoretic, it will improve power factor, so the cap of feedback compensation need to use bigger, but I found it will cause unsatable, do you have some idea?
0
回復(fù)
@arthur.deng
compensationCouldyouexplainforthe20Hzbandwidthforthefeedbackcompensation,theoretic,itwillimprovepowerfactor,sothecapoffeedbackcompensationneedtousebigger,butIfounditwillcauseunsatable,doyouhavesomeidea?
About compensation
1、因?yàn)镻FC的輸出產(chǎn)生2培的工頻的紋波(100HZ),假如帶寬太高的話,反饋部分就會(huì)企圖去調(diào)節(jié)紋波,如果這樣的話
功率因數(shù)必然減低.這個(gè)帶寬的選擇有點(diǎn)類似開關(guān)電源的截止頻率通常設(shè)計(jì)在開關(guān)頻率的五分之一左右是一個(gè)道理.
2、如果反饋電容太大的話,也就是帶寬會(huì)更低,如果在穩(wěn)態(tài)情況下工作,功率因數(shù)是會(huì)提高,但是由于實(shí)際工作情況下很多參數(shù)都是在變化的,且反饋電容增大的話,會(huì)影響系統(tǒng)的響應(yīng)速度變慢,這樣的話必然會(huì)影起振蕩.
3、我想問你一個(gè)問題,不知你是否測(cè)過電感上的電流?我也曾經(jīng)用過L6561,我發(fā)現(xiàn)它并非理論上所說的臨界導(dǎo)電模式,電感上電流根本就沒有接近零點(diǎn),而有點(diǎn)象連續(xù)導(dǎo)電模式.實(shí)際可以和理論偏差,但是這個(gè)偏差也有點(diǎn)太大了,如果有哪位老大也用過L6561的話,麻煩你們測(cè)一下電感上的電流,小弟期待著你們的結(jié)果.
1、因?yàn)镻FC的輸出產(chǎn)生2培的工頻的紋波(100HZ),假如帶寬太高的話,反饋部分就會(huì)企圖去調(diào)節(jié)紋波,如果這樣的話
功率因數(shù)必然減低.這個(gè)帶寬的選擇有點(diǎn)類似開關(guān)電源的截止頻率通常設(shè)計(jì)在開關(guān)頻率的五分之一左右是一個(gè)道理.
2、如果反饋電容太大的話,也就是帶寬會(huì)更低,如果在穩(wěn)態(tài)情況下工作,功率因數(shù)是會(huì)提高,但是由于實(shí)際工作情況下很多參數(shù)都是在變化的,且反饋電容增大的話,會(huì)影響系統(tǒng)的響應(yīng)速度變慢,這樣的話必然會(huì)影起振蕩.
3、我想問你一個(gè)問題,不知你是否測(cè)過電感上的電流?我也曾經(jīng)用過L6561,我發(fā)現(xiàn)它并非理論上所說的臨界導(dǎo)電模式,電感上電流根本就沒有接近零點(diǎn),而有點(diǎn)象連續(xù)導(dǎo)電模式.實(shí)際可以和理論偏差,但是這個(gè)偏差也有點(diǎn)太大了,如果有哪位老大也用過L6561的話,麻煩你們測(cè)一下電感上的電流,小弟期待著你們的結(jié)果.
0
回復(fù)
@darkhorse96
Aboutcompensation1、因?yàn)镻FC的輸出產(chǎn)生2培的工頻的紋波(100HZ),假如帶寬太高的話,反饋部分就會(huì)企圖去調(diào)節(jié)紋波,如果這樣的話功率因數(shù)必然減低.這個(gè)帶寬的選擇有點(diǎn)類似開關(guān)電源的截止頻率通常設(shè)計(jì)在開關(guān)頻率的五分之一左右是一個(gè)道理.2、如果反饋電容太大的話,也就是帶寬會(huì)更低,如果在穩(wěn)態(tài)情況下工作,功率因數(shù)是會(huì)提高,但是由于實(shí)際工作情況下很多參數(shù)都是在變化的,且反饋電容增大的話,會(huì)影響系統(tǒng)的響應(yīng)速度變慢,這樣的話必然會(huì)影起振蕩.3、我想問你一個(gè)問題,不知你是否測(cè)過電感上的電流?我也曾經(jīng)用過L6561,我發(fā)現(xiàn)它并非理論上所說的臨界導(dǎo)電模式,電感上電流根本就沒有接近零點(diǎn),而有點(diǎn)象連續(xù)導(dǎo)電模式.實(shí)際可以和理論偏差,但是這個(gè)偏差也有點(diǎn)太大了,如果有哪位老大也用過L6561的話,麻煩你們測(cè)一下電感上的電流,小弟期待著你們的結(jié)果.
about compensation
Thanks a lot! You are right! Very glad to meet you! I solved the problem, it is not cause by the compensation, in fact, the compensation of L6561 is easy to design, no more problem, for my problem, because the output power is 240W, so I design in DCM is a challenge for us and D/D is design by flyback, this two challenges for us, maybe for all the power design, but I am successful. Just now, I found the root cause, it caused by the MOS driver circuit, because L6561 only can provide 700mA source and sink current, this is difficult to driver the general MOSFET for higher power output, even design a totem pole driver circuit, so I changed to CoolMOS of infineon, all are OK, by the way, the VCC is not used too high, it will cause this problem, because the turn off time will be very longer, more need to notice, the Vcc of PFC and Vcc of D/D is better to separated from the winding of main transformer by two diode, for your concern, the ZCD is not a good, of cource, we can\'t do a really idealist in the actual design because there have more parasitical element we can not consider....any way, this technique only is tended to improve efficiency, if the efficiency can be accepted, I thought this problem is not a issue, do you think so? Thank you again, I would like to recognize you. My e_mail address is arthur_deng@apitech. com. tw. you are welcome .
Thanks a lot! You are right! Very glad to meet you! I solved the problem, it is not cause by the compensation, in fact, the compensation of L6561 is easy to design, no more problem, for my problem, because the output power is 240W, so I design in DCM is a challenge for us and D/D is design by flyback, this two challenges for us, maybe for all the power design, but I am successful. Just now, I found the root cause, it caused by the MOS driver circuit, because L6561 only can provide 700mA source and sink current, this is difficult to driver the general MOSFET for higher power output, even design a totem pole driver circuit, so I changed to CoolMOS of infineon, all are OK, by the way, the VCC is not used too high, it will cause this problem, because the turn off time will be very longer, more need to notice, the Vcc of PFC and Vcc of D/D is better to separated from the winding of main transformer by two diode, for your concern, the ZCD is not a good, of cource, we can\'t do a really idealist in the actual design because there have more parasitical element we can not consider....any way, this technique only is tended to improve efficiency, if the efficiency can be accepted, I thought this problem is not a issue, do you think so? Thank you again, I would like to recognize you. My e_mail address is arthur_deng@apitech. com. tw. you are welcome .
0
回復(fù)
