My Advanced Realistic Humanoid Robots Project

[abeyer]

how shall we say? - a source of horrified fascination.

100% this... it's that slow motion train wreck you just can't look away from.

I've kind of gone back and forth on my thoughts on the thread... I think I reported this one when it first started, simply because it was poor form to restart the exact same topic immediately when the first one got locked. I'd kind of come around to being ok with it as long as it stayed in this thread, and wasn't leaking out to the rest of the forum, as I do think artbyrobot was at least trying to act in good faith for a while. However the last week seems to be just escalating his over the top refusal to have any meaningful engagement and just chirping about how great he/his design/his skills are. I'm no longer convinced the thread even has any entertainment value anymore.

[Analog Kid]

I really think a ban would be appropriate.
Treez was banned for less.

No. Why?

[Zero999]

I really think a ban would be appropriate.
Treez was banned for less.

Rubbish. Treez shat all over the forum, rather than keeping it to one thread.

[ebastler]

I really think a ban would be appropriate.
Treez was banned for less.

No. Why?

That would be folly! Wouldn't you agree?

[artbyrobot]

I have taken constructive feedback about 3-4 times in the past few days.  The most recent being that my highside mosfet needs 165v to drive its gate.  I thanked Andy for this.  So don't lie and say I'm not reading and considering the feedback.  I accept the feedback when it is solid.  I reject it when it is objectively false which is majority.

[Manul]

Lets think, how long would it take me to make a power supply... Perhaps two-three weeks, including waiting for PCB and components. I'm giving you 1 year. Go and build it. You will at least be busy with something more useful than displaying your empty arrogance. And no cheating.

https://www.tickcounter.com/countdown/6105768/world-class-elite-power-supply

[vk6zgo]

I really think a ban would be appropriate.
Treez was banned for less.

Rubbish. Treez shat all over the forum, rather than keeping it to one thread.

Yes, Treez was all over the place, but at least his posts made a small amount of sense.

[Simon]

I have taken constructive feedback about 3-4 times in the past few days.  The most recent being that my highside mosfet needs 165v to drive its gate.  I thanked Andy for this.  So don't lie and say I'm not reading and considering the feedback.  I accept the feedback when it is solid.  I reject it when it is objectively false which is majority.

165V to drive a MOSFET gate? that would be 10V less than the positive rail the source is connected to.

[artbyrobot]

Lets think, how long would it take me to make a power supply... Perhaps two-three weeks, including waiting for PCB and components. I'm giving you 1 year. Go and build it. You will at least be busy with something more useful than displaying your empty arrogance. And no cheating.

https://www.tickcounter.com/countdown/6105768/world-class-elite-power-supply

Well that timer has to start when making the power supply is the next on the to do list.  I was making one for my dust buster vacuum repair project which would serve as a first prototype for the robot's power supply.  That project has been put on hold since my mom bought me a new vacuum twice since then.  So I haven't felt the need to fix my first vacuum she bought me.  It's a perfect Christmas gift item it turns out so she keeps buying me them.  Their battery packs go bad after a few years for some reason.  But my mom likes buying me them. 

So when would I build the power supply for the robot?  Well, recall that the power supply is for an alternative power source than the batteries.  I think I can get away with just a off the shelf supply for charging the batteries in the early testing days.  So no need for a built in custom one that resides in the robot at first.  So then the build of it would happen after the arm and head are done, after the AI is massively progressed to the point of useful work being achieved, and most likely after the robot has built most of its own body.  So then the rest of that power supply the robot would build.  So I'm not going to be the one building the rest of it.  The robot can pick up where I left off.

[artbyrobot]

I have taken constructive feedback about 3-4 times in the past few days.  The most recent being that my highside mosfet needs 165v to drive its gate.  I thanked Andy for this.  So don't lie and say I'm not reading and considering the feedback.  I accept the feedback when it is solid.  I reject it when it is objectively false which is majority.

165V to drive a MOSFET gate? that would be 10V less than the positive rail the source is connected to.

My understanding was that the 120v standard AC input after going through the diode bridge rectifier would become 155v and so then 165v at the gate would be 10v when you probe the gate (165v) against the source (155v) due to their relative voltages being 10v difference.  This would net a postive 10v as the Vgs.  For you to say it would be 10v less than the positive rail of the source, wouldn't that mean the source voltage is 175v?

[Simon]

I have not been following the thread but I suspect that the power supply jibe is a reference to a former member who had a thing for power supplies.

[Simon]

I have taken constructive feedback about 3-4 times in the past few days.  The most recent being that my highside mosfet needs 165v to drive its gate.  I thanked Andy for this.  So don't lie and say I'm not reading and considering the feedback.  I accept the feedback when it is solid.  I reject it when it is objectively false which is majority.

165V to drive a MOSFET gate? that would be 10V less than the positive rail the source is connected to.

My understanding was that the 120v standard AC input after going through the diode bridge rectifier would become 155v and so then 165v at the gate would be 10v when you probe the gate (165v) against the source (155v) due to their relative voltages being 10v difference.  This would net a postive 10v as the Vgs.  For you to say it would be 10v less than the positive rail of the source, wouldn't that mean the source voltage is 175v?

Oh dear, I've opened a can of worms. No the voltage does not become 155V having gone through the rectifier. 120V is the RMS voltage. It is a sine wave that peaks at 169.7V. You can take off a couple of volts for the diodes if you want to be pedantic and call it 167.7V but that is just the peak. If you place a capcitor across the output it will hold that peak voltage, but as soon as you put a load across it you will get some random voltage between the RMS voltage and the peak.

So the only correct way to drive a P channel MOSFET connected to such an unstable supply is with -10V relative to the positive rail, whatever it may be. If you actually produced a fixed positive voltage to drive the gate with your results would range from not working to a blown mosfet.

[Manul]

I have not been following the thread but I suspect that the power supply jibe is a reference to a former member who had a thing for power supplies.

Not exactly... This below (if you have strong nerves):

[Simon]

I think that guy has missed his calling drawing train maps.

[vk6zgo]

That's what I had been thinking,  that uC outputs aren't high power drivers.
   Probably you could find/select one type that has SOME pin outs with your good power.

Nice narrative though;  just pour the stuff over it,  and done...

well I learned you just solder a little baby logic level mosfet onto the big non logic level mosfet and this baby mosfet acts like the switch to the switch, opening up a sufficient voltage power source to feed the non-logic level mosfet for you.  So it's like a little buddy that helps the big mosfet turn on since the microcontroller is too weak to do it by itself.  The whole system is flawed though.  If I were in charge 80 years ago, all mosfets would be logic level because all microcontrollers would output enough voltage to turn all mosfets on -

80 years ago, there were no MOSFETS or microcontrollers, nor were logic levels  those that are now common.
Using vacuum tubes, logic levels could be anything from tens of volts to hundreds of volts.
Digital computers occupied whole rooms, & many of the functions they are now used for were done using dedicated analog circuitry.
Protection circuits in power supplies would use vacuum tubes, gas thyratrons, thermal switches, specially designed circuit breakers or a combination of these.

Even 40 years ago, MOSFETS were mostly low power devices used in their linear mode.
Microprocessors (microcontrollers were not yet mainstream) required peripheral devices to function, logic levels were commonly either 5v or 15 v.
Many of the functions power MOSFETS are now used for were carried out by power Bi-Junction Transistors, (BJTs) Silicon Controlled Rectifiers (SCRs), or a combination of either or both with Zener diodes.

this would mean microcontrollers change or mosfets change but leaving them both incompatible was a horrible decision made by incompetent men who came before me.

Logic levels have changed, so what level would you have had them use?

 And I scoff at those men.  They should be ashamed of themselves.  They have left a legacy of incompetence and folly.  Men of excellence like me would have never put future generations through this curse.  So now I have to clean up behind them like I'm changing their dirty diapers they left behind for us to deal with.  What a JOKE.

"Those men" you scoff at sent men to the Moon---what have you done, apart from playing around with junk?

[artbyrobot]

I have not been following the thread but I suspect that the power supply jibe is a reference to a former member who had a thing for power supplies.

Not exactly... This below (if you have strong nerves):

Why are you posting a old revision of my schematic Manul?  Proves you are arguing in bad faith and cannot be trusted.  Everyone knows I have greatly improved on that version with many revisions since then over a multi-year period as I learned more about electronics.  Here is a more recent one which I already posted yesterday and Manul conveniently ignores here as a bad faith actor: 

Here's my updated switch mode power supply design.  It has changed many many times as I learn more and find my mistakes.  It is perhaps significantly different than the one that was shared and examined here earlier today.

http://artbyrobot.com/switch%20mode%20power%20supply%20copy.jpg

[artbyrobot]

I have taken constructive feedback about 3-4 times in the past few days.  The most recent being that my highside mosfet needs 165v to drive its gate.  I thanked Andy for this.  So don't lie and say I'm not reading and considering the feedback.  I accept the feedback when it is solid.  I reject it when it is objectively false which is majority.

165V to drive a MOSFET gate? that would be 10V less than the positive rail the source is connected to.

My understanding was that the 120v standard AC input after going through the diode bridge rectifier would become 155v and so then 165v at the gate would be 10v when you probe the gate (165v) against the source (155v) due to their relative voltages being 10v difference.  This would net a postive 10v as the Vgs.  For you to say it would be 10v less than the positive rail of the source, wouldn't that mean the source voltage is 175v?

Oh dear, I've opened a can of worms. No the voltage does not become 155V having gone through the rectifier. 120V is the RMS voltage. It is a sine wave that peaks at 169.7V. You can take off a couple of volts for the diodes if you want to be pedantic and call it 167.7V but that is just the peak. If you place a capcitor across the output it will hold that peak voltage, but as soon as you put a load across it you will get some random voltage between the RMS voltage and the peak.

So the only correct way to drive a P channel MOSFET connected to such an unstable supply is with -10V relative to the positive rail, whatever it may be. If you actually produced a fixed positive voltage to drive the gate with your results would range from not working to a blown mosfet.

I was planning to use a n channel mosfet.  I don't like p channel mosfets.

I don't like them because I don't understand the notion of providing a mosfet with negative voltage and how that is done and so I view this as a pandoras box for a electronics beginner.  I like to use components that make sense to me intuitively.

[Manul]

I have not been following the thread but I suspect that the power supply jibe is a reference to a former member who had a thing for power supplies.

Not exactly... This below (if you have strong nerves):

Why are you posting a old revision of my schematic Manul?  Proves you are arguing in bad faith and cannot be trusted.  Everyone knows I have greatly improved on that version with many revisions since then over a multi-year period as I learned more about electronics.  Here is a more recent one which I already posted yesterday and Manul conveniently ignores here as a bad faith actor: 

Here's my updated switch mode power supply design.  It has changed many many times as I learn more and find my mistakes.  It is perhaps significantly different than the one that was shared and examined here earlier today.

http://artbyrobot.com/switch%20mode%20power%20supply%20copy.jpg

Honestly I was not paying attention, cause who cares. But thanks for reminding of the updated schematic. Everyone is free to view it and judge for themselves, how much of "improvement" there is.

[xrunner]

Why are you posting a old revision of my schematic Manul?  Proves you are arguing in bad faith and cannot be trusted.  Everyone knows I have greatly improved on that version with many revisions since then over a multi-year period as I learned more about electronics. 

That ridiculous childish schematic. In some places you put the schematic symbol in-line and put the component physical likeness beside it, and in other places you put the component physical likeness in-line and the schematic symbol beside it.

It's ridiculous to ask anyone to examine such a mess. I won't look at it further unless it's re-drawn in a more professional manner.

[Kim Christensen]

with 155V across drain-source, it will heat up and let the magic smoke out ASAP.

no, it's a IRF 840 rated to 500v across drain-source.  So 155V across drain-source it would yawn at.

Hmm... lets re-examine my entire sentence... OK?

Yes, with a gate voltage of 5V the MOSFET will be partially on, and with 155V across drain-source, it will heat up and let the magic smoke out ASAP.

Go back and look at that graph again. 5V on the gate means the MOSFET will be passing 6 amps or more while there is 155V across it. That's more than 930 watts of power dissipated in the MOSFET. Maybe have another look at the datasheet too...

If your voltage divider gives the MOSFET's gate 2V (Below turn-on threshold)

the 2v turn on theshold doesn't mean its fully on though and it is not fully on till 10v.  so I imagine say between 2v-6v the amount on is so partial that the current flow is minimal and can be safely ignored.  So my crowbar would be okay to let waste power as a partial short circuit while the gate is only partially open like this.  That is my theory I have to look into more I guess.

Nope. It cannot "be safely ignored"... See my comment about power dissipation above.

For a crowbar circuit, a SCR, zener diode, capacitor, and a resistor are all you really need. Google "SCR crowbar circuit"

I prefer to invent my own crowbar circuit to pave the way to more options of ways to make a crowbar.  Another upside is if I design it, it means I understand it.  Understanding what is going on is important for me.  At least I prefer it over black box.  I'm ok with black box often but not for something safety critical I guess.  I prefer to understand it.  My assumption is a SCR is a unknown.  Never heard of that acronym.  So I might now understand that.  So I prefer to not go that route as it is foreign to me.  And I prefer not to look up that google search suggestion if my existing solution works already and from what I can tell by what you wrote, it seems it will work fine.

But you don't actually understand it and that's why you don't understand why a SCR & zener is the superior solution here. If you'd actually built and tested this circuit, you'd know that.
How did you ever learn about MOSFETs? Were you as resistant to learning about them as you are resistant to learning about SCRs and other ways of doing things?

[Manul]

"New" schematic file edited with Photoshop CS2, timestamp 2024 11 10, 22:37:32. You made that "improvement" yesterday after hearing what people said, don't you? You posted it just after someone explained about the voltage levels. And you moved the fuse, which was also suggested here yesterday. Now you present it like it was done by you, some "iteration" of your power supply design. I applaud you.

[vk6zgo]

I have not been following the thread but I suspect that the power supply jibe is a reference to a former member who had a thing for power supplies.

Not exactly... This below (if you have strong nerves):

AAAAARRRRRRGGGGGHHHHH!!!!!

[Manul]

I have not been following the thread but I suspect that the power supply jibe is a reference to a former member who had a thing for power supplies.

Not exactly... This below (if you have strong nerves):

AAAAARRRRRRGGGGGHHHHH!!!!!

[artbyrobot]

"New" schematic file edited with Photoshop CS2, timestamp 2024 11 10, 22:37:32. You made that "improvement" yesterday after hearing what people said, don't you? You posted it just after someone explained about the voltage levels. And you moved the fuse, which was also suggested here yesterday. Now you present it like it was done by you, some "iteration" of your power supply design. I applaud you.

Here is proof the fuse moving was done by me in febuary on my schematic.  http://artbyrobot.com/switch%20mode%20power%20supply.psd

So enough with your foolish false accusations manul

Also the voltage to drive it was discussed in this february one although I didn't raise it enough and now it turns out even the 165v andy suggested is too low even.  It needs like 180v according to chatgpt just now.  Since the rectified voltage is 168v-170v peak dpeending on 115v-120v input before the diode brdige rectifier.  In my last update I did in February, I knew the voltage had to be higher than 10v more like 55v but didn't realize it had to be 10v higher than the supply voltage at the time I thought it had to be 10v higher than the drain voltage which I now understand is inaccurate but I was getting there already before anyone here said anything.

[Manul]

"New" schematic file edited with Photoshop CS2, timestamp 2024 11 10, 22:37:32. You made that "improvement" yesterday after hearing what people said, don't you? You posted it just after someone explained about the voltage levels. And you moved the fuse, which was also suggested here yesterday. Now you present it like it was done by you, some "iteration" of your power supply design. I applaud you.

Here is proof the fuse moving was done by me in febuary on my schematic.  http://artbyrobot.com/switch%20mode%20power%20supply.psd

You need 165V to switch on your highside MOSFET (i.e. 10V over your 155V main voltage)

Coincidence? No. Anyway, I don't believe you, so who cares.