Talking about phase...

[z_p_e]

@ Bob Boyce.

Some time ago you posted a response to a question regarding a suitable MOSFET for fast switching:

Take a look at the MG400Q1US41, datasheet attached. A bit large, but they should have similar lower power devices in smaller packages. I have not looked.

Rise Time (tr) 0.3 uS typ, 0.6 uS max
Turn-on Time 0.4 uS typ, 0.8 uS max
Fall Time 0.2 uS typ, 0.5 uS max
Turn-off Time 0.8 uS typ, 1.5 uS max

Bob

This surprised me. The specs for the above device represent an eternity in terms of what many other devices out there can do....about one magnitude less.

Also, in your schematic available at oupower.com for the PWM3F board, you are using a Lumex OCP-PCP116 as a MOSFET driver, which exhibits a relatively long (3us best case) propagation delay as well. Rise and fall time specs are "ok". The output current of this device is only 50mA, and I am curious how this would drive a MOSFET gate sufficiently and to the degree that you have been emphasizing here?

I was under the impression from your posts here and at oupower.com, that it is important to switch the MOSFET's as quickly as possible, yet the above devices do not seem to fit the bill as you have prescribed.

I was wondering if you could comment on this, if you are still reading here.

Thanks,
Darren

[Bob Boyce]

@ Bob Boyce.

Some time ago you posted a response to a question regarding a suitable MOSFET for fast switching:

Take a look at the MG400Q1US41, datasheet attached. A bit large, but they should have similar lower power devices in smaller packages. I have not looked.

Rise Time (tr) 0.3 uS typ, 0.6 uS max
Turn-on Time 0.4 uS typ, 0.8 uS max
Fall Time 0.2 uS typ, 0.5 uS max
Turn-off Time 0.8 uS typ, 1.5 uS max

Bob

This surprised me. The specs for the above device represent an eternity in terms of what many other devices out there can do....about one magnitude less.

Also, in your schematic available at oupower.com for the PWM3F board, you are using a Lumex OCP-PCP116 as a MOSFET driver, which exhibits a relatively long (3us best case) propagation delay as well. Rise and fall time specs are "ok". The output current of this device is only 50mA, and I am curious how this would drive a MOSFET gate sufficiently and to the degree that you have been emphasizing here?

I was under the impression from your posts here and at oupower.com, that it is important to switch the MOSFET's as quickly as possible, yet the above devices do not seem to fit the bill as you have prescribed.

I was wondering if you could comment on this, if you are still reading here.

Thanks,
Darren

@Darren

I'm not sure that that response was a specific recomendation to use that exact part as the MOSFET, but just an example of a device. By the way, if you are not switching a boatload of current, switching times on even large devices can be much faster than published specs. Just take a look at the time vs voltage and current charts. On of my replicators does in fact use the MG400Q1US41 and drives them with TC4420 chips, but I have not used it myself yet.

As for the Lumex OCP-PCP116, do not sell it short so quickly. Try finding another opto-isolated chip that can accept a logic level input, amplify it, clean it up, power itself with internal regulator from a higher voltage (in case you need 24 volt operation) gate supply, and deliver a clean crisp output in excess of 100 Khz. Especially for the price!. It is a very nice chip for the application as long as it is driving small power FETs. I would not use that chip to directly drive a large FET with a lot of gate capacitance however. In the random phase pulsed mode operation that the PWM3 series runs in, propagation delay is not important. It does have very good rise and fall times, 3 times faster than similar chips, even from the same manufacturer. I used that chip because it was the best choice for the application in the PWM3 series.

Now for the rotational version, the PWM3 series board is not used. For that I suggest a more appropriate driver. Like the TC4420, or the TI chips UC3710, UCC37322, ect.

Bob

[z_p_e]

@Darren

I'm not sure that that response was a specific recomendation to use that exact part as the MOSFET, but just an example of a device. By the way, if you are not switching a boatload of current, switching times on even large devices can be much faster than published specs. Just take a look at the time vs voltage and current charts. On of my replicators does in fact use the MG400Q1US41 and drives them with TC4420 chips, but I have not used it myself yet.

Bob, I agree with the "lower ID / faster switching" phenomenon, thank you for pointing that out.

As for the Lumex OCP-PCP116, do not sell it short so quickly. Try finding another opto-isolated chip that can accept a logic level input, amplify it, clean it up, power itself with internal regulator from a higher voltage (in case you need 24 volt operation) gate supply, and deliver a clean crisp output in excess of 100 Khz. Especially for the price!. It is a very nice chip for the application as long as it is driving small power FETs. I would not use that chip to directly drive a large FET with a lot of gate capacitance however. In the random phase pulsed mode operation that the PWM3 series runs in, propagation delay is not important. It does have very good rise and fall times, 3 times faster than similar chips, even from the same manufacturer. I used that chip because it was the best choice for the application in the PWM3 series.

A question that arose for me was: Why do you need opto-coupling?

The MAX4420 has a TTL/CMOS compatible input, 4.5V - 18V operation, 25ns RT/FT (into 2500pF), 40ns Prop Delay, 6A peak current output, and should easily operate in excess of 1 MHz. Cost is about $2. This is the device I chose for my TP900 design, which can be found here: http://www.overunity.com/index.php?topic=2831.msg41629.

Now for the rotational version, the PWM3 series board is not used. For that I suggest a more appropriate driver. Like the TC4420, or the TI chips UC3710, UCC37322, ect.

I would recommend the MAX4420, as it is an upgrade to all older "4420-type" drivers.

Regards,
Darren

[gn0stik]

yes rich

with that nice video you made some time ago i thought you were far beond this

is

IS that video was of an experiment carried out in Jason's Lab. If you read the article you can see that. Also, it wasn't on this platform, it was on Otto's, to which updated info has not been forthcoming as of late. That camp has gone silent for some reason. Draw your own conclusions.

At any rate, that experiment did not seem to turn out to be overunity. I personally believe this is the approach to be following at the moment.

Rich

[innovation_station]

im glad to hear that

follow the path you choose i have chosen mine and now i will walk it

this thing operates on re as does most of the overunity devices built in the past but no one understands how it can work cuz there was no such explnation of this publicly  out there prior to the tpu

choose your path and remember this is open source so why hide in the shadows?


ist

remember sm wanted this public his gift to the world  and we will make it happin as he can not

@ rich  is there some reason YOU cannot be the CHAMP?