Successful TPU-ECD replication !

[z_p_e]

Thank you nong for that very good writeup

Ok, Roberto, I've done a little more studying of that circuit connection.

It is not an error per se, but it is somewhat unconventional to configure the driver that way. What IR has done is configured the MOSFETs in a Source-Follower topology.

This is good for a low impedance drive, but has other disadvantages. There is a minimum 3.5V loss in the drive voltage due to the Vgs threshold, which explains why there were problems driving the Power MOSFET fully ON and OFF with only a 6V supply. Also, there is a rather large "gap" in the switching, which amounts to a "break-before-make" switch sequence.

At any rate, it will work if sufficient voltage drive is implemented, but it still is unconventional in the sense that Driver chips using complementary MOSFETs on the output, do not do this. They use an inverter topology. I would encourage you to try a "purpose-designed" MOSFET driver to see if there are any differences/improvements.

I'm using the MAX4420 as a driver chip. There are many good ones to choose from. No sense in redesigning the wheel

Regards,
Darren

[giantkiller]

Thank you nong for that very good writeup

Ok, Roberto, I've done a little more studying of that circuit connection.

It is not an error per se, but it is somewhat unconventional to configure the driver that way. What IR has done is configured the MOSFETs in a Source-Follower topology.

This is good for a low impedance drive, but has other disadvantages. There is a minimum 3.5V loss in the drive voltage due to the Vgs threshold, which explains why there were problems driving the Power MOSFET fully ON and OFF with only a 6V supply. Also, there is a rather large "gap" in the switching, which amounts to a "break-before-make" switch sequence.

At any rate, it will work if sufficient voltage drive is implemented, but it still is unconventional in the sense that Driver chips using complementary MOSFETs on the output, do not do this. They use an inverter topology. I would encourage you to try a "purpose-designed" MOSFET driver to see if there are any differences/improvements.

I'm using the MAX4420 as a driver chip. There are many good ones to choose from. No sense in redesigning the wheel

Regards,
Darren

Review: http://datasheets.maxim-ic.com/en/ds/MAX4420-MXT429.pdf
on page 5. They comment about vil and vih.  The input must not be allowed to stay between vil and vih for more than .5us. One interesting note is that ringing may be problem.

Saw that and it piqued my alarms.

--giantkiller. Ya just never know the rules being outside the boundaries.

[ronotte]

YES Darren,

you are right! I did not noticed that in fact it is a sort of double source follower.....of course it does show the potential problem you correctly said.... but on the positive side there is the 'very low output impedance' which is mandatory if you want charge correctly and full the Power Mosfet Gate capacitance + Miller effect capacitance. Said fact is necessary to put the Mosfet in avalache state that is our target. Sorry  about my typo on +6V: THE IR7307 MUST BE SUPPLIED WITH +12V.
In the past I spent lot of time trying to drive the Power Mosfet in every possible way -> always experienced big problems readily shown by my high bandwidth scope! The IR solution was the first to meet the full IRFP460 specifications, but of course the Micrel MIC4426/4427 are also a very good and easier alternative as well as other providing that are able to show less than 3 Ohm driving impedance to power mosfet gate! (GK : sorry I did not try the MAX4420 chip..so I can't say anything).

Regards to all from

Roberto

[z_p_e]

Roberto/GK,

The MAX4427 is an upgrade to the TC4427.

The MAX4420 is the "single" version of the MAX4427, but has double the drive  capacity.

Darren

[giantkiller]

Thanks Darren,

Maxim-ic has a very liberal samples policy also.

--giantkiller.