 Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me. |
Forum Index : Electronics : Ok what’[s the Ferrite Bead for on a FET?
| Author | Message | ||||
wiseguy Guru  Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
I'm glad you asked that because I now know & would like to share it with you! I have been chasing oscillations and ringing and other bad behaviour in my high speed FET power boards for a few days now. I was about to start a Warpverter in disgust but then I had a lightbulb moment - "I wonder if a ferrite bead might help". Here is a CRO shot of the crap that needed fixing. The top trace is the upper FET gate drive used for syncing the CRO. The bottom trace is the lower FET's Drain.  Here is an expanded view of the oscillations. Note the time cursors reveal a frequency of ~ 21MHz  OK so let me slip a Ferrite bead over the lower FETs Drain lead to see if it makes a difference.  Hmm slight change there, well that sorted out the LHS FETs. What does the RHS look like ?  Ok so lets slip another ferrite bead on the centre leg of the Lower FET.  OK, I'm convinced I just love Ferrite beads - and I finally see with my own eyes - very graphically - why they are used. FYI I tried combinations of 1 - 3 beads on each of the 4 FET's legs (12 in total) the best place was on the Drain of just the lower 2 FETs. Some of the FET's legs I tried had some really weird effects and distortions, so dont go crazy adding beads everywhere willy nilly, they dont all make it better. Using a 240V 300W Toroid with 2 x 6V windings, I powered up the inverter from 11.5V, the power stage idles at 0.5A and running a 29W light bulb (230VAC) it drew 3.5A. Efficiency ~ 80%. Incremental efficiency (minus the 0.5A idle) 94% at last I can now start on making it better ! Thats all for now folks! If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
mackoffgrid Guru Joined: 13/03/2017 Location: AustraliaPosts: 460 |
That's a very nice result. How many beads on the Drain Leg? thanks Andrew |
||||
| Solar Mike Guru Joined: 08/02/2015 Location: New ZealandPosts: 1138 |
Good one, normally I put them on the gate leads when parallel mosfets are used, or I use a high current smd version on the trace leading to the mosfet gate pcb connection. Cheers Mike |
||||
renewableMark Guru Joined: 09/12/2017 Location: AustraliaPosts: 1678 |
Funny you should bring that up, I ordered a big bag of TO247 compatible beads recently, not here yet. Was planning to only put them on the gate. Edit got 200 on the way yell out if anyone wants some. Cheers Caveman Mark Off grid eastern Melb |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
The beads I used are Electus (Jaycar?) LF1251 or FX1115. There was just the one bead used on the Drain. The way it reads above I used more than 1 bead - but that was always just one per leg. They are about 4.2mm diameter x 5mm long the hole in the centre is ~ 1.3mm. I just added a second choke to the other 6V winding and incremental efficience went up to 96.9% I guess I should add some heatsinks to the FETs now to avoid them frosting up  Mike - I already have smd inductors (1uH) in series with the gates but I just have 4 Fets in total (1 per leg) & no heatsinks yet. Mark, With regard to using them on the gates they do help with stopping oscillations when Mosfets are paralleled. My grief was due to the high speed drivers and Fast high speed layout. Putting them on the drain in my case helps, adding one to the gate lead made it worse - but I already have an smd inductor in the gate circuit. Maybe when i start paralleling FETs I'll have more information to add. If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
Ralph2k6 Senior Member  Joined: 24/09/2017 Location: AustraliaPosts: 129 |
This is bringing me some hope. I bought a 'kit' board and driver off Mad ages ago, maybe it's still able to be a stable long term power board.. Keep up the great progress fellas !  Ralph |
||||
| tinyt Guru Joined: 12/11/2017 Location: United StatesPosts: 438 |
Thanks wiseguy. How about just a bead on the gate lead and none on the drain. This is how I did mine. There is also some ringing on the gate, could be a chicken and egg thing. |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
Hi TinyT, The ringing on the gate is simply the ringing present on the Drain being coupled back to the gate via Drain Gate capacitance. The gate signal has no ringing of its own. A bead placed just on the gate lead changed the shape of the ringing but in no way mitigated it. When I put it on the drain only it was like a signal from another planet compared to what I had been chasing, a bit like a whoa camel moment see here . FYI & anyone else interested here are some gate drive waveforms. Output rise and fall from TC4452 12A gate driver IC   Gate drive rise time against FET turn on (Drain falling)  Gate fall time against FET turn off (Drain rising)  So it can be seen that what appear to be relatively slow rise and fall times at the gates are still turning the FET on and off in a few tens of nano seconds. Lastly a picture of what happened to the supply current when I plugged in my Hakko desoldering tool as a test load. I probably should fit some heatsinks before driving it much harder.  The power supply for the control nano, 4 isolated gate supplies and FET drivers optos etc with unit idling or running is below 80mA from 12V. Note at 48V FET supply miller effect will increase this value but at 12V its almost irrelevant. If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
| BenandAmber Guru Joined: 16/02/2019 Location: United StatesPosts: 961 |
That's really cool with all you guys experimenting on here there's going to be a combined effort perfect in every way inverter one day be warned i am good parrot but Dumber than a box of rocks |
||||
| johnmc Senior Member Joined: 21/01/2011 Location: AustraliaPosts: 282 |
Good Day Wiseguy Thanks for the information on ferrite beads,really liked the camel carton.  johnmc |
||||
| mackoffgrid Guru Joined: 13/03/2017 Location: AustraliaPosts: 460 |
Just for reference I got some beads a while ago for just in case Ferrite Sleeve Core EMI 3.5x5x1.5mm These are a bit smaller than the Jaycar beads. Which ones did you get Mark? Cheers Andrew |
||||
| renewableMark Guru Joined: 09/12/2017 Location: AustraliaPosts: 1678 |
I got these that fit on 4008 fets. Cheers Caveman Mark Off grid eastern Melb |
||||
| renewableMark Guru Joined: 09/12/2017 Location: AustraliaPosts: 1678 |
 Got these in today, they fit nicely on the centre leg too. John, I'll stick some in the pack along with the other things that haven't turned up yet. Poida, you should have a play with them too since you have that nice DSO. Anyone else want some just yell out. Cheers Caveman Mark Off grid eastern Melb |
||||
LadyN Guru Joined: 26/01/2019 Location: United StatesPosts: 408 |
I usually see these beads on the middle pins of the MOSFETs |
||||
| Solar Mike Guru Joined: 08/02/2015 Location: New ZealandPosts: 1138 |
I have always only placed them in the gate circuit when paralleling multiple mosfets, but as wiseguy pointed out earlier, they worked best in that particular circuit on the drain. So perhaps best positioned after testing waveform with a scope. |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
Mark, that's not a bead its more like a small toroid............ Ok it seems there may be a slight downside to adding a bead to the drain leg. Yes it stops the 21MHz ringing I had - looking at the drain connection on the PCB it is well behaved now with no sign of bad behaviour. But are we really looking at the Drain - No we are not, we are looking at the other side of the inductor we just placed in series with the Drain - what is actually happening at the Drain is now on the other side of the ferrite bead directly at the Mosfet's Drain leg. Top trace is the Drain connection at the PCB side of the bead, bottom trace is on the FET's Drain on the FET side of the bead.  - where did that come from ? Well we all know that if we put current through an inductor (the ferrite bead) and then suddenly interrupt the current, we get a voltage spike, in this case way above the applied voltage. Yep we just generated about 30V along 5mm of solid FET Leg !! The only way I was able to tame it a bit was to put a resistor of 10 - 15 Ohms across the bead, the picture below shows the improvement. Bottom trace is the Drain on the FET side of the bead with a 15R across the bead.  My inverter that these signals are taken from is 12V in to - 230V~ out. This was an effect easily explained but was quite unexpected - it was just a teeny ferrite bead and it added nearly an extra 30V on top of the 12V to the Drains !! On a 48V system I guess it could present well over 100V to the FET. Maybe the FETs can avalanche this bit of energy without batting an eyelid - the main reason for highlighting it is the extra noise generated by the spike. Its also an inherent downside of fast switching - keeps the FETs cool but can generates a lot of harmonic noise. Last bit of semi useless information, the idling current from the 12V supply into the inverter whilst running is surprisingly (to me) low @ ~250mA, mostly magnetising current for the 250VA Toroid. There is a 1u capacitor on the 230V secondary. If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
| renewableMark Guru Joined: 09/12/2017 Location: AustraliaPosts: 1678 |
Sing out if you want any, got plenty left. Cheers Caveman Mark Off grid eastern Melb |
||||
 Print this page |
