NEW - ESC vs ESC

[Mercuriell]

Originally Posted by kimmik

Hey tony if you get a chance, i'd like to nominate pwm as your next point of discussion. I'm interested in a detailed breakdown and estimate of the pwm esc losses and pwm motor losses.

In particular the effect of fet drain source charge, voltage, current, switching frequency, afw on the switching losses. Would be nice to have a general idea of how much total loss it is.

That would be an interesting topic for discussion in the Electronics Forum but is Off Topic for this thread

[scotte]

I'd be interested seeing how the arcesc using Tony's logic board stacks up against this lot - or is AFW still a work in progress mate ?


Sent from my eyePhone using Tapatalk

[kimmik]

In his op he said he will discuss how to choose pwm and how it affects motors.

[ub3r]

Originally Posted by scotte

I'd be interested seeing how the arcesc using Tony's logic board stacks up against this lot - or is AFW still a work in progress mate ?

I might add it in the future.. But itll be a little unfair as might be bias towards my own ESC.

Originally Posted by kimmik

In his op he said he will discuss how to choose pwm and how it affects motors.

Ill get to that in the near future.

In the meantime, I've added a HW review of the Castle ICE2 ESCs..

[Mercuriell]

Interesting stuff on the CC - really a lot of very neat features especially that splitting of the PWM - really cool ! Certainly goes a long way to explaining the CC Ice HV problems and how CC have remedied them - inspires me to give them another go as I really like the control board - logs, firmware update, ease of programming etc

[galensk]

subscribed ! very interesting stuff !!!

[marcosp]

Simply fascinating!

[OnTheSnap]

Originally Posted by ub3r

I might add it in the future.. But itll be a little unfair as might be bias towards my own ESC.


Ill get to that in the near future.

In the meantime, I've added a HW review of the Castle ICE2 ESCs..

Great stuff. I recently switched back all my large models to Castle again and have been very happy. With 4.xx it works exceptionally well with external gov's now. I'm also running 4.02 alpha which includes a perfect soft start with ext gov as well. Good to see them come back strong after such a miserable year in 2011.

[Iceberg86300]

Originally Posted by kimmik

If k implements the kosmik mcu in the next jive I'll be all over it

What is the kosmik mcu.? I think there is zero chance they commissioned a new chip so does anyone know what it is? Freescale coldfire 32 bit, etc? Freescale actually has app notes for running a synchronous motor off of an 8 bit chip. Goto 16 or 32 and there would be tons of headroom for rpm control, governor, AFW, etc.

Steve
Sent from Tapatalk using my Inspire 4g through ATT (meaning it's a wonder you can see this!)

[mkovalcson]

FWIW I worked for a company that was a very small startup ( 25 employees) and we actually designed a chip that we put into production. The initial test run of 10 chips was about $300,000, but after that the pricing per chip was very good. This was in 1991. Things have gotten a lot cheaper since then.

I'm not saying that Kontroniks did or didn't have a custom chip made, but just that it isn't as big a deal as it used to be.

[neohp]

Ub3R, you've made one of the most interesting and well documented post on HF since a long time !

[fasse_sweden]

I agree, ME LIKE!
This is what it's all about, sharing knowledge and having a great time doing it. I hope to contribute with my knowledge s well.

Keep up the good work and keep them coming

Greetings from Sweden!

[Iceberg86300]

Originally Posted by mkovalcson

FWIW I worked for a company that was a very small startup ( 25 employees) and we actually designed a chip that we put into production. The initial test run of 10 chips was about $300,000, but after that the pricing per chip was very good. This was in 1991. Things have gotten a lot cheaper since then.

I'm not saying that Kontroniks did or didn't have a custom chip made, but just that it isn't as big a deal as it used to be.

I suppose your right. I just thought the economies of scale would be too small for a model ESC. Although I guess they could be using it for other motor control apps as well. Such as over designing it with encoder inputs or what not so they could use it in a servo application. Or even resell it, or they could even have an offshoot company.

Although I worked for a CNC manufacturer and we used multiple coldfire chips, but then again we needed a ton of processor power because we weren't just doing servo control.

Steve
Sent from Tapatalk using my Inspire 4g through ATT (meaning it's a wonder you can see this!)

[j2m]

This is just about the best thread I've seen for a long time.

The explanation of Passive and Active FW was very clear but it made me wonder - in AFW it's clear that there are lower losses in the controller under partial load but where does the the energy stored in the winding inductance go if it's not dissipated in the FET body diodes? Presumably it's dissipated in the winding and does this make the motor itself run hotter (perhaps only marginally)?

Would love to see a Hawk120 tested in this thread. I have three of them and they seem to run very cool under partial load - do they feature AFW does anyone know?

[CyprusFlyer]

It is my understanding that the energy is switched back to the supply, ie the battery by either diode or FET, of course the diode will drop approx 0.6v and generate heat. Also dont forget that three phases are overlapping here, so the next phase in the sequence will also be drawing current from the supply.

[j2m]

CyprusFlyer
"It is my understanding that the energy is switched back to the supply"

I don't believe that's the case. If you look at the circuit diagrams in Ub3r's explanation of FW, the FW current path appears to be back to ground via the body diode in Q2 or FET M2 in the case of AFW. As per his explanation, Q1/M1 will be switched off so there's no path for the back EMF to Udc.

[Scallybert]

Originally Posted by j2m

[...]where does the the energy stored in the winding inductance go if it's not dissipated in the FET body diodes?

The current flowing in the windinsg will be generating torque.

While it's still flowing in the 'right' direction (ie, not dropped past zero - impossible with a diode; presumably stopped by the AFW MOSFET switching off), that torque is powering the motor (reversed current would be braking).

So some of the saved power comes out of the shaft.

Some will be the in the windings (as said in the bit of the posting I cropped.)

[I guess...]

[ub3r]

Videos:

EPISODE 1: KOSMIK & 160HV Full Throttle Tests.
Helifreak HIgh Res Download: Episode 1

Youtube:
[ame]http://www.youtube.com/watch?v=z7jewAN1dTw&list=UU7lc_DFDwD3wDHd-ixH1gfA&index=2&feature=plcp[/ame]

EPISODE 2: KOSMIK & 160HV Partial Throttle Tests.
Helifreak HIgh Res Download: Episode 2

Youtube:
[ame]http://www.youtube.com/watch?v=AKN18ZTqoag&feature=channel&list=UL[/ame]

[ub3r]

I'm glad you're liking the thread guys.
More reviews & vids to come..

BTW the free-wheeling power which is saved in AFW, flows back into the motor and generate a little more torque. Remember this is not much.. Only roughly 0.5 volts for only the PWM-OFF time. The longer the OFF time (or lower the throttle), the power that is saved.

[dna001]

Awesome
Keep it up.