ChocoHolic, I didn't try 20954 yet, but IR introduced 20955 as stable replacement for this chip, and now released 20957 for the same reason but to replace 20955, so I guess 20954 was unstable also and probably even more, I mean problem of internal trigger latching. If you talking about OCP we had absolutely no problem around it, however we never used very heavy mosfets (irf540z[full range] & irf3205 [for the sub amp] only). BTW, about PCB layout, my car audio amp 4x150W passed CE-mark with some decent room from the limits. 
Hi Steven,
for me this information about the unchanged shut down and your results with proper PCB is quite interesting. So I see no reason for me to have a closer look again to the IR-chips. With MosFets like IRFB4227 or IRFB4321 I would need two additional totem pole drivers and if I then additionally have to implement the shut downs in seperate, then I do not see much adavantage of using the IR versus any other level shifter....
The nice thing about the IRS20954/20955/20957 (what happened with 20956???) would be low component count, if they would work like in the school book.
Hi IVX,
no trouble at lower power is fitting to my experiences even with the old 20954 and poor PCB.
You got the CE-Mark without EMI-struggle. This is sounding like you know about right PCB design and snubbering. My congrats!
...or do you know the right test institute?
for me this information about the unchanged shut down and your results with proper PCB is quite interesting. So I see no reason for me to have a closer look again to the IR-chips. With MosFets like IRFB4227 or IRFB4321 I would need two additional totem pole drivers and if I then additionally have to implement the shut downs in seperate, then I do not see much adavantage of using the IR versus any other level shifter....
The nice thing about the IRS20954/20955/20957 (what happened with 20956???) would be low component count, if they would work like in the school book.
Hi IVX,
no trouble at lower power is fitting to my experiences even with the old 20954 and poor PCB.
You got the CE-Mark without EMI-struggle. This is sounding like you know about right PCB design and snubbering. My congrats!
...or do you know the right test institute?
hi irs2092 with irf540st
hi..i have 2 irs2092 and many irf540st transistors..
http://www.oup.com/us/pdf/microcircuits/students/mos/IRF540-st.pdf
i was interested in class d in the past but now i like class a or ab..
is it possible to build an amp sounds better than krell ksa50 with my items?maybe they all have some advantages and disadvantages...
there is global crisis and global heating..i want to save energy but also want to have good sound..thanks
best regards
hi..i have 2 irs2092 and many irf540st transistors..
http://www.oup.com/us/pdf/microcircuits/students/mos/IRF540-st.pdf
i was interested in class d in the past but now i like class a or ab..
is it possible to build an amp sounds better than krell ksa50 with my items?maybe they all have some advantages and disadvantages...
there is global crisis and global heating..i want to save energy but also want to have good sound..thanks
best regards
ChocoHolic said:
CISPR20? No, EN55014 (Household appliances), but why you think that it was EN55020 (Broadcast receivers)?
I can check exactly after tomorrow, actually I didn't care about those CISPR/EN55*** numbers, I just use the power absorber (cables Disturbance Power Measurement) and HP analyzer in the my workshop to fit the disturbance under limit, and send to lab for testing.
The CSIPR20 and EN55020 are for " Sound and television broadcast receivers and associated equipment ".
Isn't HiFi-equipment inteneded to be used with radios or TVs?
That's why I was guessing the xxx20.
The title of the xx14 is sounding for me more like being intended for a fridge or a vaccum cleaner..
But I might be wrong. Defining which standard has to be applied is not always obvious and sometimes multiple choices are possible.
Isn't HiFi-equipment inteneded to be used with radios or TVs?
That's why I was guessing the xxx20.
The title of the xx14 is sounding for me more like being intended for a fridge or a vaccum cleaner..
But I might be wrong. Defining which standard has to be applied is not always obvious and sometimes multiple choices are possible.
hi chocolic
hi Mr chocolic i wanna ask some questions..we want to build 300watts/8 ohms class d amp with irs2092..but i dont know much about output filters..is ferroxcube rm14 3c90 material good for this? which cores are more suitable for our application?..would you recommend a quality transformer?we want to buy online.sorry for this questions.thank you very much..
hi Mr chocolic
i wanna ask some questions..we want to build 300watts/8 ohms class d amp with irs2092..but i dont know much about output filters..is ferroxcube rm14 3c90 material good for this? which cores are more suitable for our application?..would you recommend a quality transformer?we want to buy online.sorry for this questions.thank you very much..
Re: emc standards
Hello Anthony,
thanks for clarification, this is fitting to my guess in post #128.
And exactly fitting to coming steps for my smps design.
EN60065 is clear - what a nice fat safety standard!
...8mm creepage is already implemented. Quite a special fun in a three dimensional object like a smps transformer...
Hi Umut,
3C90 is a simple standard material, but should work OK for most filter designs.
RM14 is a good choice, the gap should be in the center only - for EMI reasons.
Also I would recommed HF-Litz, because of heat and distorsions - not solid wires.
Cheers
Markus
Anthony C Smith said:
Hello Anthony,
thanks for clarification, this is fitting to my guess in post #128.
And exactly fitting to coming steps for my smps design.
EN60065 is clear - what a nice fat safety standard!
...8mm creepage is already implemented. Quite a special fun in a three dimensional object like a smps transformer...
Hi Umut,
3C90 is a simple standard material, but should work OK for most filter designs.
RM14 is a good choice, the gap should be in the center only - for EMI reasons.
Also I would recommed HF-Litz, because of heat and distorsions - not solid wires.
Cheers
Markus
ok chocolic no problem..but i wanna say that we are interested in class d amps more than 300watts.and i trust you will make good circuits.when you finish your job and print extra pcbs,lets reduce the price of printing pcbs. tell me and i will pay you on paypal for some pcbs(i think quantity may be 4or more) and we are gonna built with my friend..thank you..
I prefer to sense current with a shunt in series with output inductor, and do the limiting in the modulator, so that the amplifier can operate and be stable in current limiting mode, without shutting down, "hicup", or chaotic mode due to peak current limiting with continuous inductor current and no slope compensation.
I think the Rds-on method is only ok as an ultimate safety measure to avoid output stage destruction when everything else fails.
These chips with integrated level shifting and low propagation delay are attractive, but I miss independent low and high side logic inputs in order to tailor dead time to my needs. Sometimes I need zero dead time, or even "negative" dead time (enlarging pulses rather than sortening them), to achieve proper timing at the gates, and those ICs don't allow to do that.
IR2110 is still my favourite, it's timing stability over temperature and part to part is hard to match. For my next design I'm going to try IR2011 with buffers, it's intended to produce over 3KW full range on 4 ohms. I have already achieved that power with full bridge and IR2110 without problems, but not full range because IR2110 can't drive so much gate charge directly at more than 150Khz without overheating.
I think the Rds-on method is only ok as an ultimate safety measure to avoid output stage destruction when everything else fails.
These chips with integrated level shifting and low propagation delay are attractive, but I miss independent low and high side logic inputs in order to tailor dead time to my needs. Sometimes I need zero dead time, or even "negative" dead time (enlarging pulses rather than sortening them), to achieve proper timing at the gates, and those ICs don't allow to do that.
IR2110 is still my favourite, it's timing stability over temperature and part to part is hard to match. For my next design I'm going to try IR2011 with buffers, it's intended to produce over 3KW full range on 4 ohms. I have already achieved that power with full bridge and IR2110 without problems, but not full range because IR2110 can't drive so much gate charge directly at more than 150Khz without overheating.
...hm, if things are coming to timing precisision I love the combination of the HCPLxxxx-isolators combined with the FAN3225...
Both not easy to handle - because of their speed already small layout imperfections cause what Agilent calls 'ambitious output' I love this wording ... but if you spend some love to the PCB and circuit design, then you get amazing results.
The FAN3225 is a double driver and has such a nice popagation delay matching, that you can parallel them with just a few Ohms.
But for very heavy MosFets they are of cause also not suitable. IMHO even in parallel they are not suited for gate charges above 100nC for full range applications.
Both not easy to handle - because of their speed already small layout imperfections cause what Agilent calls 'ambitious output'
I love this wording ... but if you spend some love to the PCB and circuit design, then you get amazing results.The FAN3225 is a double driver and has such a nice popagation delay matching, that you can parallel them with just a few Ohms.
But for very heavy MosFets they are of cause also not suitable. IMHO even in parallel they are not suited for gate charges above 100nC for full range applications.
Re: emc standards
Headache ahead for all SMPS designers
The EN60065 alone will soon not be sufficient anymore in case of SMPS powered amps. The EN60065 is explicitely naming that it is intended to be used for frequencies up to 30kHz. Above it is only applicable as long as there is no further directive. In fact the new directive/standard is more or less close to become reality.
So for new designs additional requirements should be considered. ...up to now I did not dig into this in detail, but I will have to - because my SMPS is running at 80kHz.
Ugly physics! ... that things change with frequency is quite uncomfortable, and now this topic is going to be implemented in the safety standards...
Anthony C Smith said:
Headache ahead for all SMPS designers The EN60065 alone will soon not be sufficient anymore in case of SMPS powered amps. The EN60065 is explicitely naming that it is intended to be used for frequencies up to 30kHz. Above it is only applicable as long as there is no further directive. In fact the new directive/standard is more or less close to become reality.
So for new designs additional requirements should be considered. ...up to now I did not dig into this in detail, but I will have to - because my SMPS is running at 80kHz.
Ugly physics! ... that things change with frequency is quite uncomfortable, and now this topic is going to be implemented in the safety standards...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.