Castle - Make A Special Edition Monster

[Pdelcast]

Originally Posted by Dagger Thrasher

One question though...how come you're adding a sensor port? I'm just curious, as you've been saying that sensors are near enough useless next to good sensorless software. Unless that's going to be geared towards crawler use?

Two reasons -- First is for crawlers, for low Kv sensor motors to allow good holding torque on really steep rock faces
The second reason is for marketing reasons.

[suicideneil]

Originally Posted by Pdelcast

The second reason is for marketing reasons.

Trying to kill-off Novak? that's mean; they only have small weedy motors, how can they fight back against a true ~100amp esc that will run any kind of motor?...

I dont have any issue with fans like I said before- so long as they are used as a substitute for normal airflow, instead of as a forced-cooling unit, its fine be me. I'd like to see a better fan on the MMM, buts thats in hand I gather.

[johnrobholmes]

Trying to get more of the crawling market huh? It will never work. You know that is such a small niche with just a bunch of grown men BSing on the rocks all day.


Can't wait to see if you have any programming tricks up your sleeves (like neutral brake for motors with little drag ability) or what kind of motor you will sensor up. I can't imagine you would just use a normal old two pole inrunner with a sensor, but maybe I am wrong.

[GriffinRU]

Patrick, I am not sure that D2-Pak7 has smaller die than SO8-FL and its current ratings are not that shy. And I agree on SO8-FL vs. LFPAK's.

Personally, I am looking forward to use DirectFET package, you can heatsink them top&bottom equally. For 30V ESC.

I do not see why software can be so complicated, 10 years ago maybe it was an art to make brushless sensoreless, today with all bells and whistles inside micro's and all the features... Cannot believe in that... Neither I have seen chinese ESC with as smooth performance as schulze...

Nice plans for:
8S 200A ->6720W
12S 125A -> 6300W
8S 325A -> 10920W
12S 200A -> 10080W
Just keep wondering what kind of capacitor (size wise) you are planning to use to absorb punishment. Even 2% of 6kW would be 134.4W - that is pretty good size for capacitor...and 2% is nothing you know. Or you plan to have batteries implanted inside fet's :)

-
Did somebody said something about sensors ...Hmm - > Marketing or Hype -> must be crawlers

[Pdelcast]

Originally Posted by GriffinRU

Patrick, I am not sure that D2-Pak7 has smaller die than SO8-FL and its current ratings are not that shy. And I agree on SO8-FL vs. LFPAK's.

No, you are right, D2Pak (To-263) has a larger die than SO8-FL, but SO8-FL has a larger die than DPAK (TO-252)

However, SO8-FL has larger silicon-footprint ratio than D2Pak

The DirectFET is a great package -- but IRF silicon sucks. (Infineon is releasing some DirectFET mosfets soon...)

Originally Posted by GriffinRU

I do not see why software can be so complicated, 10 years ago maybe it was an art to make brushless sensoreless, today with all bells and whistles inside micro's and all the features... Cannot believe in that... Neither I have seen chinese ESC with as smooth performance as schulze...

TRY it! It's VERY complex! If it wasn't complex, there would be a lot more companies doing this ... the hardware is (relatively) easy.

Originally Posted by GriffinRU

Just keep wondering what kind of capacitor (size wise) you are planning to use to absorb punishment. Even 2% of 6kW would be 134.4W - that is pretty good size for capacitor...and 2% is nothing you know. Or you plan to have batteries implanted inside fet's :)

Big ones. :)

Originally Posted by GriffinRU

Did somebody said something about sensors ...Hmm - > Marketing or Hype -> must be crawlers

Now you know WHY I get pissy about sensors. All that work I've gone through in the last year to make something that works WORSE than what we already have. Bleah.

[GriffinRU]

Originally Posted by Pdelcast

The DirectFET is a great package -- but IRF silicon sucks. (Infineon is releasing some DirectFET mosfets soon...)

You will be ahead of the game if you supply IRF with right silicon and get your own fet's. While it would take some time for Infineon to release new package. You know that packaging is the major problem in microfabs. If IRF knows how to package it would be no problem for them to process "your" silicon.

Ok, I will try to come up with software, you convinced me :) Done that in the past, so not a big deal, as you said hardware is relatively simple :)

See, I was hoping you write software, I will do the hardware - the output stage...and then even open source it -> deep inside I think it is the hardware which is complex. Software only become complex, when has tons of bandages and solutions to overcome hardware flaws not limits, limits must be identified at the design stage.

[Pdelcast]

Originally Posted by GriffinRU

You will be ahead of the game if you supply IRF with right silicon and get your own fet's. While it would take some time for Infineon to release new package. You know that packaging is the major problem in microfabs. If IRF knows how to package it would be no problem for them to process "your" silicon.

Ok, I will try to come up with software, you convinced me :) Done that in the past, so not a big deal, as you said hardware is relatively simple :)

See, I was hoping you write software, I will do the hardware - the output stage...and then even open source it -> deep inside I think it is the hardware which is complex. Software only become complex, when has tons of bandages and solutions to overcome hardware flaws not limits, limits must be identified at the design stage.

Well, give it a try. I think you will find it is MUCH more complex than you think... But isn't everything like that? It looks easy from the outside, but when you actually do it, you find it is much more difficult than you first thought. And it's not bandaids either -- just the core software is very complex because of the number of possible states. There aren't any "bandaids" in our software to overcome hardware flaws or limitations -- it is usually easier to mod the hardware than the software....

Why do you think there aren't any "generic" hardware sensorless brushless controller chips? All the hardware chips out there require tuning to a particular motor, and they still don't work well --

And Infineon isn't releasing a new package -- it is the DirectFet package, licensed from IR. Mainly because IR is still behind in the MOSFET arena -- and IR doesn't really have enough fab capacity to handle a lot of new business anyway.

[Dagger Thrasher]

Originally Posted by Pdelcast

Two reasons -- First is for crawlers, for low Kv sensor motors to allow good holding torque on really steep rock faces
The second reason is for marketing reasons.

Lol, fair enough! The Mamba Max is already a great crawler ESC, so the Pro with a switching BEC and a sensor port should be even better...especially if any sensored outrunners turn up at any point. Is the BEC going to be the same system (minus component issues) as the Monster's?