DIY 6FET Controller help
- Thread starter SjwNz
- Start date
Lebowski
10 MW
PeakArlo1 said:
Futterama
1 kW
Corrected: Still 6FET with 190/200A phase and TO-247AC MOSFETs?zombiess said:
SjwNz
100 W
Back to 150amps now but it was good to see if it would handle 190 amps phase for a short time. Looks like I need to make a 12 fet controller now as I really want to run at 200 amp phase.
SjwNz
100 W
HighHopes said:
yes 12fets total (2 in parallel) with 200amp phase and the AvgBatt current will be around 70 - 80amps.
zombiess
10 MW
HighHopes said:
The only way I know of to verify their leg limits is to build a test setup and start monitoring the leg temperature and the case temperature until stability is reached at increasing current limits. This is on my todo list... still
Some of the above mentioned current limits could make the legs glow. I would like to know the true continuous power rating.
Luke and I have both tested the legs to 200amps on a single TO246 package. I can verify 150-200 amps continuous is possible on the legs with good cooling of the fet it self.zombiess said:HighHopes said:
The only way I know of to verify their leg limits is to build a test setup and start monitoring the leg temperature and the case temperature until stability is reached at increasing current limits. This is on my todo list... still
Some of the above mentioned current limits could make the legs glow. I would like to know the true continuous power rating.
zombiess
10 MW
Arlo1 said:
Are you referring to the videos where you apply these high currents and after 5 seconds they blow up? What was the most current you were able to pass through the MOSFET and what temperature did it reach? If you have done this and collected the data it will save me the trouble of doing it myself.
I want to place a MOSFET on a heat sink while supplying DC and then >95% PWM @ around 30kHz and let it run overnight while logging the temperature. Switching losses add a lot of heat.
Nozombiess said:
Jeremy I know you like to talk down to me but stop it.
Continuous means just that.
For my test.
(Luke has a more in depth setup I'm sure.)
I just took blown mosfets that had good drain legs and tested the power from the back out though the drain leg. All I did this for, was to see what kind of heat/limits they had.
200 amps is do able with good thermal paths and a 10-20 deg ambient temp.
All I wanted to know was if they can do the rated 160 amps and they can but you need to know what you are doing. People call it data sheet "Kool-Aid" but its not. As those people are not staying within ALL the parameters listed in the data sheet. When you listen to ALL the data you will be able to run them to the limits.
Its best you make a setup to do some of your own testing.
But remember turning it on to conduct though the silicon VS just conducting though the legs are 2 different things and the heat from the rdson or the switching alone will cause a temp rise in every part of the mosfet, then at that point you can get a thermal run away on a leg if you don't destroy the silicon first.
zombiess
10 MW
Arlo1 said:
There is nothing to stop, I'm not talking down to you, but I can see how it came across that way, sorry.
Did you try with the source leg in the current path?
zombiess said:
No I think I just did the drain as all the test was to prove is if the legs can do the 160 amps continuous and they can.
The test blowing them up after a few seconds of on time is through both legs but that was just to prove if they can do the rated amperage on the silicon and they can.
HighHopes
10 kW
- Joined
- Mar 28, 2013
- Messages
- 929
long time ago i once read on what lab conditions manufacturers use to define their silicon amp rating. it was a crazy unrealistic thing as i recall, something like emersing the mosfet in some sort of jelly and what not. no matter what the current ratings or how they are used (switching, continuous) in a realistic environment.. at the end of the day there are three limits for how much current/temperature you can use.
1. the junction temperature needs to be 25C less than datasheet limit or you risk thermal runnaway (here is where the 175C Tjmax can come in handy).
2. the rate of change of current has to be within the mechanical limits of the device (i don't think there is a datasheet parameter for this, but it speaks to the internal bond wires flexing and breaking off), this is one of the key factors that defines a "good" manufacturer. its not a thermal problem because the current comes & goes too quickly. if you have many devices in parallel, say more than 4, and do not manage the transient current sharing at ON/OFF this risk of mechanical failure can be a problem ("solved" by China controllers by switching very slow and putting a huge fan on it).
3. the delta temp between heatsink and junction temp has to be some max value or less. 40degC comes to mind, but that's just an old memory. this is another form of mechanical failure but not the wire bonds, some other part of the mosfet.
the above are just rules of thumb, you can push a device beyond any of the above but it starts to take away from the life of the component.
i suspect item #1 and more likely #3 to be something for investigation in your application but if you can get 200Amps while staying true to the rules above with only two devices in parallel then you're good to go.
1. the junction temperature needs to be 25C less than datasheet limit or you risk thermal runnaway (here is where the 175C Tjmax can come in handy).
2. the rate of change of current has to be within the mechanical limits of the device (i don't think there is a datasheet parameter for this, but it speaks to the internal bond wires flexing and breaking off), this is one of the key factors that defines a "good" manufacturer. its not a thermal problem because the current comes & goes too quickly. if you have many devices in parallel, say more than 4, and do not manage the transient current sharing at ON/OFF this risk of mechanical failure can be a problem ("solved" by China controllers by switching very slow and putting a huge fan on it).
3. the delta temp between heatsink and junction temp has to be some max value or less. 40degC comes to mind, but that's just an old memory. this is another form of mechanical failure but not the wire bonds, some other part of the mosfet.
the above are just rules of thumb, you can push a device beyond any of the above but it starts to take away from the life of the component.
i suspect item #1 and more likely #3 to be something for investigation in your application but if you can get 200Amps while staying true to the rules above with only two devices in parallel then you're good to go.
SjwNz
100 W
Just came across these, maybe useful for someone.
I mite use them in a work project.
http://www.digikey.co.nz/en/product-highlight/r/recom-power/dcdc-converters?WT.srch=1&mkwid=sgrXOlHls&pcrid=64697050346&pkw=_cat%3Adigikey.co.nz&pmt=b&pdv=c
I mite use them in a work project.
http://www.digikey.co.nz/en/product-highlight/r/recom-power/dcdc-converters?WT.srch=1&mkwid=sgrXOlHls&pcrid=64697050346&pkw=_cat%3Adigikey.co.nz&pmt=b&pdv=c
HighHopes
10 kW
- Joined
- Mar 28, 2013
- Messages
- 929
i like the similar ones by murrata. unregulated is what you want here and capable of light load. there are other features that you want but its impossible for these and similar to provide it, so, for that, they find a limitation for inverters of about 30kW. if you go higher power, you need a different solution. the power supply you link runs about $12USD ea, i have seen them as low as $8 (murrata?).
recently.. for my "cost effective" low power drive i'm trying something new for me.. bootstrap! ha. i'm curious to see if i can make it work for $4.
recently.. for my "cost effective" low power drive i'm trying something new for me.. bootstrap! ha. i'm curious to see if i can make it work for $4.
SjwNz
100 W
Haha, look what I found.
HighHopes
10 kW
- Joined
- Mar 28, 2013
- Messages
- 929
::busted:: lol
o
but to be fair, for the power you guys wanted i had to get you something that could work so you could spend your time more usefully instead of watching endless fires. so i had to be harsh on the bootstrap. but i bet you could design with bootstrap now successfully up to 10kW since you know now how exactly a gate driver needs to work.
my application is <1kW so i'm not too worried and i was asked to make it cheap. so its new for me, i don't make anything cheap - as you found out. lol
o
but to be fair, for the power you guys wanted i had to get you something that could work so you could spend your time more usefully instead of watching endless fires. so i had to be harsh on the bootstrap. but i bet you could design with bootstrap now successfully up to 10kW since you know now how exactly a gate driver needs to work.
my application is <1kW so i'm not too worried and i was asked to make it cheap. so its new for me, i don't make anything cheap - as you found out. lol
SjwNz
100 W
Hi All
Its been a while since I did anything on my bike and I haven't used it for almost a year as I spend most of last year making this silly mower.
https://www.youtube.com/watch?v=vai5NOIvKbQ&t=3s
But over the last week I have removed the optical sensors use for motor position and gone to hall effect sensors.
The optical setup worked well but trying to keep sunlight away from the sensors and worried about water and mud damaging the
Black white pattern was a pain.
I went with an external hall setup and I did not use the motors magnets but fitted an external disc and drilled 28 holes to glue in 28 small magnets.
I used this external disc method as I remember reading on this forum some years ago where people were having problems with external halls that were using
the magnets in the rotor of RC motors when power ratings were getting around 3kw.
I made a small cover to help keep water out and will start using the bike again to commute to work.
View attachment 1
Its been a while since I did anything on my bike and I haven't used it for almost a year as I spend most of last year making this silly mower.
https://www.youtube.com/watch?v=vai5NOIvKbQ&t=3s
But over the last week I have removed the optical sensors use for motor position and gone to hall effect sensors.
The optical setup worked well but trying to keep sunlight away from the sensors and worried about water and mud damaging the
Black white pattern was a pain.
I went with an external hall setup and I did not use the motors magnets but fitted an external disc and drilled 28 holes to glue in 28 small magnets.
I used this external disc method as I remember reading on this forum some years ago where people were having problems with external halls that were using
the magnets in the rotor of RC motors when power ratings were getting around 3kw.
I made a small cover to help keep water out and will start using the bike again to commute to work.
View attachment 1
SjwNz
100 W
Hi All.
Came across this scope.
https://www.youtube.com/watch?v=I7ppDNLlEL4
Came across this scope.
https://www.youtube.com/watch?v=I7ppDNLlEL4
Alan B
100 GW
SjwNz said:
$9600
Similar threads
- Solved
