Hi Guys!
I hope some of you heard about IRS27951 or IRS27952 from IRF. These are 8pin half-bridge resonant controller ICs, they have a lot of improvements from the common-known IR2153 series.
- progammable dead-time
- increased FET-driving current
- easier variable frequency controll
- over-current sense with FET's Rdson
Due to its highly-integrated, with very few external components a stabilized-output high efficiency SMPS could be built from few hundred watts, up to 1kW.
The problem is that they are really new, and I haven't seen them anywhere (farnell, digikey, tme).
Anybody seen this IC somewhere?
I would really get my hands on few peaces.
I hope some of you heard about IRS27951 or IRS27952 from IRF. These are 8pin half-bridge resonant controller ICs, they have a lot of improvements from the common-known IR2153 series.
- progammable dead-time
- increased FET-driving current
- easier variable frequency controll
- over-current sense with FET's Rdson
Due to its highly-integrated, with very few external components a stabilized-output high efficiency SMPS could be built from few hundred watts, up to 1kW.
The problem is that they are really new, and I haven't seen them anywhere (farnell, digikey, tme).
Anybody seen this IC somewhere?
I would really get my hands on few peaces.
The bootstrap diode? that needs to withstand the full voltage+bootstrap voltage + margin. In hte appnote they use a 600V diode.
You have to be careful with the capacitor, due to high voltage present at high frequency. The voltage rating derates rather quickly with frequency. There is a risk of overheating so be very careful, if it blows all that resonance energy will go somewhere and problably kill everything, or as moderator expressed it "annhilating primary silicon".
You have to be careful with the capacitor, due to high voltage present at high frequency. The voltage rating derates rather quickly with frequency. There is a risk of overheating so be very careful, if it blows all that resonance energy will go somewhere and problably kill everything, or as moderator expressed it "annhilating primary silicon".
I have ordered a soic 8 to dip converter. Will try that and if that doesn't work will try removing a 27951 from its packing without touching it and try that.
The circuit is so simple that I must be doing something very silly.
I rigged up a 30 volt low voltage supply to my circuit so I can check it without killing myself on mains.
I found that the 27951 would only work at circa 14 volts VCC.
Above and below this voltage the signal just looked a mess.
I read through the datasheet again and this suggested using a star ground with the VB charge up ground separate from the RT and CT circuits and VCC.
I hacked my circuit to get as close to a star ground as possible and the circuit now works between 12 volts and 30 volts.
I then tried with a 90 volt supply and the circuit worked fine.
I tried again with mains voltage to the circuit but it seems to crash out and go into low power mode.
I am getting there slowly but have a little more work to do yet.
I will look at the auxiliary supply next to make sure the chip is getting enough power while working on mains.
I found that the 27951 would only work at circa 14 volts VCC.
Above and below this voltage the signal just looked a mess.
I read through the datasheet again and this suggested using a star ground with the VB charge up ground separate from the RT and CT circuits and VCC.
I hacked my circuit to get as close to a star ground as possible and the circuit now works between 12 volts and 30 volts.
I then tried with a 90 volt supply and the circuit worked fine.
I tried again with mains voltage to the circuit but it seems to crash out and go into low power mode.
I am getting there slowly but have a little more work to do yet.
I will look at the auxiliary supply next to make sure the chip is getting enough power while working on mains.
Last edited:
Nigel, I made multiple SMPSs with this IC. You can read about it
here on a Hungarian forum (translated to english with google for you)
I can help you, but I must ask some questions:
- could you draw the exact schematic you are using?
- what type of FETs are you using?
- you are balming things on satr ground etc... could you send a picture of your layout?
- What are you n and k values (turn ratio, ratio of magnetizing and leakage inductance)
- What type of transformer are you using? (core, gap, turns, wire)
I use a small aux baord for the controll, connecting in right next to the fets, this eliminates control common problems, however, this is a resonant converter and it is less sensitive to such problems then hard switching ones. So I think the problem is some other.
here on a Hungarian forum (translated to english with google for you)
I can help you, but I must ask some questions:
- could you draw the exact schematic you are using?
- what type of FETs are you using?
- you are balming things on satr ground etc... could you send a picture of your layout?
- What are you n and k values (turn ratio, ratio of magnetizing and leakage inductance)
- What type of transformer are you using? (core, gap, turns, wire)
I use a small aux baord for the controll, connecting in right next to the fets, this eliminates control common problems, however, this is a resonant converter and it is less sensitive to such problems then hard switching ones. So I think the problem is some other.
I have used the application note circuit.
I am using STP11NM60FD mosfets.
I am using an ET54 transformer with 36 primary turns and two 8 turn secondaries.
The aux transformer is 3 turns.
I am using STP11NM60FD mosfets.
I am using an ET54 transformer with 36 primary turns and two 8 turn secondaries.
The aux transformer is 3 turns.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
You are missing the MOST important things:
What is the magnetizing and leakage inductance of your transformer?
1.89 mH
I don't know how to calculate the leakage inductance.
I just wound on 36 turns on the primary and wound over the top two 8 turn secondaries.
Then Nigel I am sorry to say that you have did not understand the concept of an LLC converter at all.
In shorts: the LLC acronim stand for: series inductance (leakage inductance), magnetizing inductance and resoannt capacitor.
These three reactive elements dictate two resoanant frequencies, which the conveter uses to stabilize the secondary voltage.
You must design THESE VALUES to fit your converter. Reread the whole application note!
Leakage inductance cannto be exactly calculated, it MUST BE MEASURED: measure the priamry inductance while SHORTING the secondaries.
Your converter fails to start, because of the low leakaage inductance: the resoanant current goes to infinite due to the secondary cpaacitors, causes a high votlage drop on the FETs which triggers the IRS275x overcurrent protection. This is exactly the cause of error. You simply did not read the application note.
The inductance with secondaries shorted is 70uH.
I removed the secondary smoothing capacitors and the circuit runs so that confirms the 27951 is shutting down on over current.
I will read the app notes again.
I do have to admit I just took the app note circuit and scaled my transformer to fit my requirements.
70 uH could be a even good value depending on your projected frequency range. But your amgnetizing inductance is still large.
As you can see in the application note there is a value of "k" which is the ratio of the mangetizing inductance and the leakage inductance and is usually between 3 and 10 (while 1890 uh divided by 70 uH is 27).
Did you checked the hungarian link? You can find my schematic and my layout plan to, with pictures. You can see a picture of a completed converter too: you can see that the transformer uses a home-made split bobbin, and the use of the 3M yellow insulation tape.
I didn't use a split bobbin, I just wound the secondaries on top of the primary.
I tried a gapped core to bring the inductance down to 550uH with a leakage inductance of 170uh but the circuit still wont power up. I just get a brief pulse on the output then the 27951 goes into power down mode.
The circuit runs fine on 90 volts DC but wont run on 320 volts DC.
Last edited:
You must use a split-bobbin, otherwise you won't get the sufficient leakage inductance.
Did you calculate your max current? What is the soft-start freqncuy?
With 70 uH and a 100kHz switching the current is about 6 A, which with the 0,4 Ohm mosfet will trigger the IRS27951 (2V max voltage drop on the FET).
You must calcuate the three tags and the frequencyies (and everything...). In the link you can find an excel spreadsheet, it follows throught the application note 8so you will figure it out, even if its in hungarian.
Its really helpful so try using it. Or I encourage you to make a similar spreadsheet according to the application yourself, it will help your design.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.