EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: merymery on May 02, 2021, 10:49:01 pm
-
Hello, I'm searching for a gate driver for my circuit with 10MHz switching frequency. I cannot use LMG1020 due to its case. I don't have a hot air station. I will do the soldering manually with a soldering iron. I found the EiceDRIVER ™ 1EDBx275F, but the operating frequency is given up to 10 MHz. I think it may not be healthy to run at 10MHz, the features of the circuit are below, I would be very happy if there are gate driver ICs that you can recommend, thank you :)
Vi = 3.7V Vo = 3 V Po = 3W SEPIC DC-DC Converter, no isolation needed
-
I don't think you will find many high speed gate drivers in a package with leads.
Does the LMG1025 work? https://www.ti.com/product/LMG1025-Q1 (https://www.ti.com/product/LMG1025-Q1)
Its the same die as the LMG1020 but in a DFN package instead of WLCSP. With some care (and a large via under the ground pad) it may be solderable without hot air.
However, given how cheap hot air stations are now days it seems like a really weird constraint that the package must be hand solderable with only an iron....
-
You are right, but unfortunately this is the stuation. Thanks for the suggestion, but I gave up LMG1025 as I had no experience with soldering.
The gate drivers I found of the IXYS brand are: IXRFD630, IXRFD615, DEIC421. Operating frequency up to 45MHz has been given. I think I can use one of these. Do you think it is appropriate?
-
Is there a particular reason you want to make a 10 MHz 3 W SEPIC converter? It will be more difficult, more expensive and much less efficient than a converter operating at a more reasonable frequency, something in the 200 kHz - 1 MHz range for example.
If you have to make a 10 MHz SEPIC, start with chosing the smallest MOSFET that will handle the power you need, this way gate charge is minimized and you have a realistic chance of driving the gate. If you pick a MOSFET with a threshold voltage below, say, 2.5 V, you might be able to use several 74LVC-series gates paralleled through resistors.
-
It's a project I'm going to do at school and my teacher wants me to work with these parameters. I will consider your mosfet selection and 74LVC-series gate driver suggestion, I have IRF7495PbF mosfet, can I drive this mosfet at 10 MHz with IXRFD630 gate driver?
-
There are a whole bunch of fast low side MOSFET drivers -
DIP IXDD604PI IXDI604PI Dual channel 4.5V ~ 35V 4A 9ns, 8ns
UCC27517 TTL Low-Side single 4.5V ~ 18V 4A 8ns,7ns
MAX17602 Dual Low Side 4V ~ 14V 4A 40ns,25ns
UCC27518DBVR CMOS Low Side single Inverting 4.5V ~ 18V 4A 8ns,7ns
UCC27519DBVR CMOS Low Side single non-Inv 4.5V ~ 18V 4A 8ns,7ns
NCP81074 Low Side Single - 4.5V ~ 20V 10A 4ns,4ns 8-SOIC
EL7104
LM5134BMF Low side Single 4v ~ 12.6V 4.5A 3ns,2ns
IX4340UE 5-AMP DUAL LOW-SIDE 5V ~ 20V 5A 7ns,7ns
EL7202CS Low-Side 4.5V ~ 15V 2A 7.5ns,10ns
VE3ELQ Nigel did some interesting experiments for Class E transmitter use with SiC and GAN FETs -
http://amfone.net/Amforum/index.php?topic=42262.0 (http://amfone.net/Amforum/index.php?topic=42262.0)
http://amfone.net/Amforum/index.php?topic=42504.0 (http://amfone.net/Amforum/index.php?topic=42504.0)
-
Using a 100 V 22 mohm MOSFET for a 3 W 3.7 V converter seems like a really bad choice. It will require 0.7 W just for the gate drive power at 10 V Vgs, which is a huge fraction of the target output power. The threshold voltage is also too high to practically drive it with 5 V logic.
The mentioned gate drivers like IXDD630 are not ideal in this application, as they are made to drive big MOSFETs at moderate speeds, even if they can operate at high frequencies. Looking at the IXDD630 for example, it will consume 10 W at 10 MHz, running from 12V without any load. This seems a bit ridiculous for a 3 W converter. You need a small driver to drive small MOSFETs fast and efficiently, unless you're aiming for 15 % efficiency or so.
If you can get a different MOSFET, this will be a lot easier. Something like a DMN2024UFDF-7 or a IRLML2030TRPBF could be a good starting point.
I still don't think 10 MHz is the right operating frequency for this application, it will be less efficient, less cheap and less compact than at 1 MHz most likely. Though if it's out of academic interest then it will at least be an interesting experiment.