Castle - Make A Special Edition Monster

[Pdelcast]

Originally Posted by GriffinRU

Try to take an infrared picture of your ESC with and without copper plate, I do not really care about arctic silver...as long as you can keep it as flat as possible, you can use graphite if you can maintain below 0.005". You know there is a step with sand paper right, not only for adhesion.

Sure, there are even better things you can do. If your R&D budget can afford 2 month clean-room run, I can share with you what can be better and we can talk patenting...

I have IR pictures of the ESC, and a copper plate doesn't make any significant difference on the 6oz board... the older 4oz board was a different story. And, a copper pin fin heatsink is actually an easier mod will have much more effect than a copper heat spreader.

BTW -- we are using the temp sensor on the processor -- so make sure any mods you do reduce the temp difference between the FETs and the processor, not the opposite.

What does that mean -- 2 month clean-room run? Are you thinking about getting On Semiconductor to give us some Fab time, or what?

[GriffinRU]

Originally Posted by Pdelcast

I have IR pictures of the ESC, and a copper plate doesn't make any significant difference on the 6oz board... the older 4oz board was a different story. And, a copper pin fin heatsink is actually an easier mod will have much more effect than a copper heat spreader.

BTW -- we are using the temp sensor on the processor -- so make sure any mods you do reduce the temp difference between the FETs and the processor, not the opposite.

What does that mean -- 2 month clean-room run? Are you thinking about getting On Semiconductor to give us some Fab time, or what?

Well, true for the last gen. pcb with only 6 fet's down there... still can see slight improvement. If you plan to add copper heatsink to MM, than I am sorry who needs MMM in 1/8 scale? You can do the same to MMM, of course. Just get plated copper, so it doesn't change its appeal :)

I know, on first versions I was adding copper spacer, but lately reusing your foam, works alright. :)

Something like that, but not at On Semiconductor, more like proto run at University and upon success, I do not know it is way too early.

[Pdelcast]

Originally Posted by GriffinRU

Well, true for the last gen. pcb with only 6 fet's down there... still can see slight improvement. If you plan to add copper heatsink to MM, than I am sorry who needs MMM in 1/8 scale? You can do the same to MMM, of course. Just get plated copper, so it doesn't change its appeal :)

I know, on first versions I was adding copper spacer, but lately reusing your foam, works alright. :)

Something like that, but not at On Semiconductor, more like proto run at University and upon success, I do not know it is way too early.

Well, the main problem we had with all of our Copper heatsink testing was that the heatsink is HEAVY -- and on crashes would sometimes rip the FETs off the board (or in half!)
The copper heat sinks did work very well, but were very expensive.

On Semiconductor has done some custom stuff for us in the past -- they might be willing to listen.

[GriffinRU]

Originally Posted by Pdelcast

Well, the main problem we had with all of our Copper heatsink testing was that the heatsink is HEAVY -- and on crashes would sometimes rip the FETs off the board (or in half!)
The copper heat sinks did work very well, but were very expensive.

On Semiconductor has done some custom stuff for us in the past -- they might be willing to listen.

There are ways to solve this, fet's case cannot support such load. Expensive way - hybrid heatsinks, all the way up to built-in heatpipes. Second one - pcb support for heatsink, you can solder heatsink to the board.

Well, now that is when we switch to PM.

By the way, I am pretty sure, you've seen that :)

you might as well except for Brian and a few others nobody has a clue as to what you and Patrick are talking about......:-)

[Pdelcast]

Originally Posted by GriffinRU

By the way, I am pretty sure, you've seen that :)

Sure is a lot easier to build a high power controller when you only have a single phase...

Probably just a couple Eupec IGBTs. Pretty simple. -- What isn't talked about is the RADIATOR that that controller requires -- and water pump, and plumbing... 2 watts of loss per amp just for forward drop, minimum.

IGBT drivers can be a challenge in three phase setups, but single phase is a breeze.

[JERRY2KONE]

Even in the best think tanks having a fresh perspective is always a good thing. We all know how secretive any company can be about protecting its investment on any project, so seeing you (Patrick) discuss specs with a fellow R/C junkie who obviously has some insight is great. We have all heard Griffen spew technical stuff in an effort to improve things with ESC MFG and he has modded some MM's with impresive results. So maybe if you two get to share some fresh ideas we will benefit in the long run. Nice to see you communicating on this. Good luck guys.

[Pdelcast]

Originally Posted by GriffinRU

By the way, I am pretty sure, you've seen that :)

You know, I've been thinking about this --

It wouldn't be too hard to do a 400V MOSFET based 3 phase controller for an EV. And it wouldn't be hard to do a PM motor for it either (expensive magnets though...) Both ST and Infineon have some nice 500V MOSFETs now (that don't have 1000nC charges.)

Need to sleep now, but I'm thinking about it.

[DRIFT_BUGGY]

Originally Posted by Pdelcast

About the wires- - I'm actually surprised you would want it changed -- it actually is more expensive for us to put wires on the controller than the plugs. . .

But, if you do want it changed, I'm sure we will exchange it, or update it with plugs.

Patrick

Thats great Patrick. Best customer service out there

[GriffinRU]

Originally Posted by Pdelcast

You know, I've been thinking about this --

It wouldn't be too hard to do a 400V MOSFET based 3 phase controller for an EV. And it wouldn't be hard to do a PM motor for it either (expensive magnets though...) Both ST and Infineon have some nice 500V MOSFETs now (that don't have 1000nC charges.)

Need to sleep now, but I'm thinking about it.

Too bad we are far from each other,
I am less concern about cooling, can fix that.

If you can get me some ESC's already pre-fabbed with caps and tvs, I would like to share with you one more aspect of this tuning. I am sure you aware of that, but still. Pushing 30V fet's towards 6S is not great and we have seen more than one catastrophic fet's failure. What I was trying to add to MM, but didn't find a good place, is tvs per phase, not across H-bridge. 3 TVS total, no TVS across power leads required if caps are good and leads are short.
Concern - V=L*di/dt
With higher Kv motors, motors inductance and fast high current switching can exceed fet rating pretty quick. It would be the best to place clamp that would be before fet's diodes.
If you have any catastrophic fet's failures (I had one unit from DickyT) than it might be related to that.
Neu 1515 1.5D has about 2.3uH @ 50Amps and 1us it can hit 115V, your are not clamping phase - but just a thought.

Need to go, see you next weekend,

[Pdelcast]

Originally Posted by GriffinRU

Too bad we are far from each other,
I am less concern about cooling, can fix that.

If you can get me some ESC's already pre-fabbed with caps and tvs, I would like to share with you one more aspect of this tuning. I am sure you aware of that, but still. Pushing 30V fet's towards 6S is not great and we have seen more than one catastrophic fet's failure. What I was trying to add to MM, but didn't find a good place, is tvs per phase, not across H-bridge. 3 TVS total, no TVS across power leads required if caps are good and leads are short.
Concern - V=L*di/dt
With higher Kv motors, motors inductance and fast high current switching can exceed fet rating pretty quick. It would be the best to place clamp that would be before fet's diodes.
If you have any catastrophic fet's failures (I had one unit from DickyT) than it might be related to that.
Neu 1515 1.5D has about 2.3uH @ 50Amps and 1us it can hit 115V, your are not clamping phase - but just a thought.

Need to go, see you next weekend,

You'll find there are much better (faster and more robust) solutions than TVS.

Notice you never see those high voltages (115V) in a MMM or MM, because the body diodes clamp WAY before avalanche occurs. But there are several improvements that can be made beyond what the MMM has. First -- the body diodes do clamp to rail +- 1.5V or so, but they are slow. The snubbers are faster, but can't handle enough power -- bigger snubbers make a huge difference.

The whole idea is giving that energy a place to go that is LOW resistance path, not a high resistance path like a TVS. That energy of the inductor doesn't have to be wasted in a resistive part -- it can be recaptured, re-routed, restored.

[GriffinRU]

Originally Posted by Pdelcast

You'll find there are much better (faster and more robust) solutions than TVS.

Notice you never see those high voltages (115V) in a MMM or MM, because the body diodes clamp WAY before avalanche occurs. But there are several improvements that can be made beyond what the MMM has. First -- the body diodes do clamp to rail +- 1.5V or so, but they are slow. The snubbers are faster, but can't handle enough power -- bigger snubbers make a huge difference.

The whole idea is giving that energy a place to go that is LOW resistance path, not a high resistance path like a TVS. That energy of the inductor doesn't have to be wasted in a resistive part -- it can be recaptured, re-routed, restored.

That is where software comes to play...
Hundreds of ps are not that slow (it is possible to get faster devices but then substrate size (to maintain current capacity) will increase capacitance, thus slow you, limit you) , and there is no need to clamp all the way to zero, just to the safe Fet's level, and fet's are not that fast. But overtime both of them degradate, TVS first (easy to replace) fet second. Temp also change their response. Diodes are also rated to max Fet’s voltage, they are by-product of Fet. So you will exceed diode ratings if you do not clamp it earlier with something rated at higher voltage. You got the idea.
There is no much energy there to recuperate, but plenty to damage. Just switching fronts I am talking, nothing else.

[lutach]

Originally Posted by Pdelcast

You know, I've been thinking about this --

It wouldn't be too hard to do a 400V MOSFET based 3 phase controller for an EV. And it wouldn't be hard to do a PM motor for it either (expensive magnets though...) Both ST and Infineon have some nice 500V MOSFETs now (that don't have 1000nC charges.)

Need to sleep now, but I'm thinking about it.

What about Ixys for those High Voltage/High Power MOSFETs? I don't know how they compare with ST and Infineon, but most of the products offered by Ixys are on the higher voltage side.

[Pdelcast]

Originally Posted by lutach

What about Ixys for those High Voltage/High Power MOSFETs? I don't know how they compare with ST and Infineon, but most of the products offered by Ixys are on the higher voltage side.

Well, they do have some interesting parts -- but in the last few years they have really fallen behind the "big boys."

Fairchild also has some interesting high voltage parts, but is still a generation behind ST and Infineon in RDSon / die size.

[lutach]

Originally Posted by Pdelcast

Well, they do have some interesting parts -- but in the last few years they have really fallen behind the "big boys."

Fairchild also has some interesting high voltage parts, but is still a generation behind ST and Infineon in RDSon / die size.

The last time I spoke with someone there, they told me they were the leader in High Volt/High Power MOSFETs. I'm pretty sure some of the MOSFETs manufacturers can make some sort of special MOSFET for the demands of us RC guys. Have you also thought about contacting the foundries that makes the silicon's that goes into a MOSFET to actually make one that better RDSon or what ever other special need you might have? A custom Castle MOSFET would be nice, kind of like the Futaba processor seen in their radios .