EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: Krotow on June 18, 2024, 09:14:48 pm
-
Stumbled upon a little problem - how to replace MOSFETs in IC driven H-bridge properly. I have H-Bridge motor driver based on Renesas HIP4081AIBZ driver. And weird Chinese MOSFETs that I would like replace to better ones like IRF3205SPbF. Particular driver I like because it take pretty small PCB space and use small surrounding discrete component count. But I believe that simply replacing one MOSFET to different one without changing at least bootstrap capacitor value and probably also gate current limiting resistors is not a good idea. Because MOSFET voltages and timings are different. Things that change however are not much and I started to think - maybe there are calculators that can help to spew out particular component values from driver IC and MOSFET datasheet numbers? To speed up process without recalculating whole shebang of high school physics book.
-
You're overthinking this.
I know the HIP4081 since back from the Harris days, and your concerns are baseless. The IRF3205 is not hard to drive,
Rule of thumb for this kind of circuit is bootstrap cap > 10 x gate capacitance.
In your case it will be 33 nF or 47 nF.
Done.
-
Rule of thumb for this kind of circuit is bootstrap cap > 10 x gate capacitance.
As an aside, where did you learn about this rule of thumb? Is it documented anywhere in an application note? Or is it something you just gained "from experience"?
-
Rule of thumb for this kind of circuit is bootstrap cap > 10 x gate capacitance.
As an aside, where did you learn about this rule of thumb? Is it documented anywhere in an application note? Or is it something you just gained "from experience"?
Both.
But this might help:
https://www.renesas.com/us/en/document/apn/an9405-hip4081a-80v-high-frequency-h-bridge-driver-hip4080a81aevalz (https://www.renesas.com/us/en/document/apn/an9405-hip4081a-80v-high-frequency-h-bridge-driver-hip4080a81aevalz)
-
Curious, in what respect is an ancient HEXFET like IRF3205 better than modern production (from any country)?
Assuming you get what you'd ordered -- but keep in mind, counterfeiting happens anywhere by anyone; the meaningful solution is buying through recognized/authorized distributors, and perhaps "receiving testing" if you like.
Without checking many parameters, general alternatives you might consider include AOB66616L, IPB026N06NATMA1, BUK964R8-60E, etc. Asian sourced parts will have comparable performance.
Tim
-
For the original question, you don't usually think too much about bootstrap capacitor sizes. Rarely you see any other value than 100nF tbf. Just make sure not to make it too small that it isn't enough to fully turn the FET on, or too large that it isn't charged fast enough at startup causing excessive dissipation in the FET (some fancy controller "pump" the lower FET before activating the highside FET driver to avoid this).
If there any things that may need an adjustment, it would be dead time values. Still it usually isn't necessary.
-
But this might help:
https://www.renesas.com/us/en/document/apn/an9405-hip4081a-80v-high-frequency-h-bridge-driver-hip4080a81aevalz (https://www.renesas.com/us/en/document/apn/an9405-hip4081a-80v-high-frequency-h-bridge-driver-hip4080a81aevalz)
Thanks. I already have it.
Curious, in what respect is an ancient HEXFET like IRF3205 better than modern production (from any country)?
Assuming you get what you'd ordered -- but keep in mind, counterfeiting happens anywhere by anyone; the meaningful solution is buying through recognized/authorized distributors, and perhaps "receiving testing" if you like.
IRF3205-s proved itself as being reliable even in not so standard situations. In my project I want to use IRFZ44NSTRLPBF - tested them before and they survived even being accidentally connected in wrong direction and baked alive from outside. Seems have spare magic smoke container inside :) I'm aware about popular component cloning. For project I'm sourced components from Mouser. I believe they usually have original components. I have no fear about limited stock size because it is very small project with unit count in few tens max. Switching to another MOSFET will be easy enough.
Without checking many parameters, general alternatives you might consider include AOB66616L, IPB026N06NATMA1, BUK964R8-60E, etc. Asian sourced parts will have comparable performance.
Didn't used these before. From datasheets they could be useful. Thanks about info.
The MOSFETs about I have doubt are Ruichips Semicon RU6188S (https://jlcpcb.com/partdetail/Shenzhen_ruichipsSemicon-RU6188SR/C2977204). Datasheet numbers look better than IRF44NZ, but we know about PMPO watts and old Asian tradition to skimp on everything as rule of existence. They are functional, but didn't tried to fry them yet. So have no proof are they can survive real world. Also boards where I "sourced" them was by words of DiodeGoneWrong - "super shody". Prove me wrong.
By the way, below is the reverse engineered motor controller circuit with "wrong" MOSFETs. Fine for me because I don't need current sensing. LDO is not required as well cause load is powered from 12V. Except I doubt about 10uF capacitance for bootstrap capacitor as too large. And 4.7 ohm gate resistors. Previous prototype with IRFZ44NSTRLPBF and IR2104 as bridge driver IC was rock stable with 10 ohm gate resistors.