420A 24S ESC by aliens

[Arlo1]

. . . .SOMEBODY should offer a Colossus!

Long story short my second colossus was for luke then he moved away. Now he has WAY more badder assed motors to work with.
My two colossus motors are mine and will always be mine I worked SO SO SO hard for getting where i am. I need them because I will be working on this controller stuff non stop and I need the colossus stock and rewound to test with everyday. SOON more info from me I promise.

I did offer to test this controller on one of or all 4 of my motors and my dyno. But Luke is better setup then me so I hope this all goes well.

 

[h0tr0d]

subscribed !!!

 

[brunotollot]

Here the pictures of the old and the new PCB. The new version will be bigger for better heat dissipation, with thickness multilayers copper foil for better current carrying and 36 fet (12 per phase). The old was 24.

 

[johnrobholmes]

For all the trouble to get heat out of the board, why not use legged FETS and install heatsinks on each parallel bank? Then your cooling fans could REALLY remove the heat!

 

[neptronix]

Wow, OK. I didn't know the previous controller was a 24FET. If you would have said that from the start of the thread, i wouldn't have expressed as much doubt. Usually these RC Controllers are using a lot less FETs than necessary and crappy heatsinks. They are rated at peak / burst amperage with no mention of the constant current, which is often super low in comparison to peak.

A 36FET sensorless controller that actually lives up to it's promises would be amazing.

 

[Byte]

Here the pictures of the old and the new PCB. The new version will be bigger for better heat dissipation, with thickness multilayers copper foil for better current carrying and 36 fet (12 per phase). The old was 24.

View attachment 1

So the one Luke has received is still the 24 FET, right?

 

[Byte]

Wow, OK. I didn't know the previous controller was a 24FET. If you would have said that from the start of the thread, i wouldn't have expressed as much doubt. Usually these RC Controllers are using a lot less FETs than necessary and crappy heatsinks. They are rated at peak / burst amperage with no mention of the constant current, which is often super low in comparison to peak.

A 36FET sensorless controller that actually lives up to it's promises would be amazing.

Do you know how much FETs a 160HV uses? Couldn't find it on the internet...

 

[brunotollot]

For all the trouble to get heat out of the board, why not use legged FETS and install heatsinks on each parallel bank? Then your cooling fans could REALLY remove the heat!

You mean to use the to 220 instead the 2d pak? If yes, the point is the size. Anyway is an RC controller. I should designed just one for E bike but it take time and test. Maybe when i finish the E bike i will consider to do that. You are right. Like an E bike controller where the fet are connected to the box enclosure. Much better heat dissipation.

So the one Luke has received is still the 24 FET, right?

Yes.

Do you know how much FETs a 160HV uses? Couldn't find it on the internet...

Honestly i don't know.

 

[Byte]

Any news Luke?

 

[liveforphysics]

Any news Luke?

Waiting on programming cable.

 

[Dauntless]

Just don't want this thread to die off before the tests can be run.

So why is this controller so big and heavy? The RC world normally likes them light, why is this one different?

 

[brunotollot]

Just don't want this thread to die off before the tests can be run.

So why is this controller so big and heavy? The RC world normally likes them light, why is this one different?

The USB program cable will be to Luke by friday. Don't worry.

The RC controller normally are for toys. But sometimes the toys ave very big and powerfull. This is why it is different...

 

[Dauntless]

I get it, an underdog controller that can provide the speed of lightning and roar of thunder for the overachiever. We got a few of those around here.

[youtube]9bdpyamRS4c[/youtube]

 

[CogHog]

Subscribed

 

[brunotollot]

Hi. Here 2 pictures of the sensored model.

 

[flathill]

Does the sensored model switch to sensorless at high rpm (does it have phase advance)?

Does it switch to sensorless if the halls fail?

Now that I think about it a physical switch to switch between sensored and sensorless mode would be awesome, as sometimes the controller won't be able to detect hall failure.

I hope you prove the naysayers wrong and don't worry too much about advertised ratings. We need compact high power controllers and for most of us burst is more important than continuous power. The biggest hill in San Francisco only takes a couple minutes to climb. Even the all mighty sevcon gen4 controller can only sustain peak current for 2 minutes before it dials back the power to protect itself. The sevcon's continuous is approximately only 1/3 to 1/2 the advertised power. Imagine if you bought a new gsxr motorbike and after two minutes of hard riding the horsepower was cut in half. This is where liquid cooling has an advantage

 

[brunotollot]

Does the sensored model switch to sensorless at high rpm (does it have phase advance)?

Does it switch to sensorless if the halls fail?

Now that I think about it a physical switch to switch between sensored and sensorless mode would be awesome, as sometimes the controller won't be able to detect hall failure.

I hope you prove the naysayers wrong and don't worry too much about advertised ratings. We need compact high power controllers and for most of us burst is more important than continuous power. The biggest hill in San Francisco only takes a couple minutes to climb. Even the all mighty sevcon gen4 controller can only sustain peak current for 2 minutes before it dials back the power to protect itself. The sevcon's continuous is approximately only 1/3 to 1/2 the advertised power. Imagine if you bought a new gsxr motorbike and after two minutes of hard riding the horsepower was cut in half. This is where liquid cooling has an advantage

Hi. Sensored to sensorless at high rpm. Nice idea. Not now but i will work for that. Sensorless at high rpm suppose to be more fast. Now it work only in sensored mode. If someting wrong with the sensor it stop work. By program software you can switch it in sensorless. I'm still try a manual switch but i need a different firmware and a small logic PCB for the switch that work with the main one. It also have a temperature sensor. I'm trying to set two point. At x C° start the thermoelectric cooling system. At y C° start a safe mode and the controller stop work. I'm just waiting the new PCB for start.

 

[flathill]

Very cool! Hall sensors are the number one failure mode on hub motors so the sensored/sensorless switch would be crucial/awesome

Also this might be lost in translation but here thermoelectric cooling commonly means a peltier cooling. You might want to say fan cooling or active cooling rather than thermoelectric cooling. Super minor detail but just in case you are writing a manual or english software menu....

Also with the overtemp sensor it should limit power rather than just shut down. Think about needing to get across an intersection with cars coming....You dont want your ebike or escooter or eskatebaord to cut out in the middle of the intersection

Thanks!

 

[liveforphysics]

The package with the programming dongle arrived today.

Thank you Bruno! Now testing can begin!

 

[gwhy!1]

Very cool! Hall sensors are the number one failure mode on hub motors so the sensored/sensorless switch would be crucial/awesome

Thanks!

Im not interested in sensorless ( I never had a hall sensor fail ), but it would be nice to have the option

when are we going to be able to test/buy the sensored version, Im sure LFP will do some very good testing on the sensorless but I don't think any of the results will be truly relevant for the sensored version unless Luke fails to make the magic smoke come out of the one he is testing.

 

[brunotollot]

The package with the programming dongle arrived today.

Thank you Bruno! Now testing can begin!

Very good. I need info for the new version.

Thank you..

 

[Byte]

Luke, could you post some data of the 160HV ESC?

 

[brunotollot]

Wow, OK. I didn't know the previous controller was a 24FET. If you would have said that from the start of the thread, i wouldn't have expressed as much doubt. Usually these RC Controllers are using a lot less FETs than necessary and crappy heatsinks. They are rated at peak / burst amperage with no mention of the constant current, which is often super low in comparison to peak.A 36FET sensorless controller that actually lives up to it's promises would be amazing.

Hi. The High Amp models ESC are made by 2 PCB so actually the new version will be 36 + 36. The old one is 24 + 24. You can see in the pictures. Thanks

 

[Arlo1]

Luke any testing done yet?

 

[Dauntless]

Luke any testing done yet?

He might be saving it as a new years resolution.