aCeMadMod
1 kW
hi can show pic of what the pazza look like , i may like to take a bite ,, great work hope you can sell the broad soon,
thank you
thank you
Micro Lebowski Controller - DIY 3kw
- Thread starter Animalector
- Start date
Animalector
10 kW
Photos.
might order some parts and build on. should be easier than the last one to build
might order some parts and build on. should be easier than the last one to build
Lebowski
10 MW
Nice i'm especially happy about the way you buold the power rail, it keeps induvtance low which is good for the FETs 
Did you add provisions for temperature sensors ?
i'm especially happy about the way you buold the power rail, it keeps induvtance low which is good for the FETs Did you add provisions for temperature sensors ?
Animalector
10 kW
nope. I didn't re-do to CPU board. That's where it'll matter most. I can always loom the temp sensors and stick them onto the FETs later, but I really need to update the Brain Board to make that work.
I'll build one up. sometime soon...
I'll build one up. sometime soon...
Lebowski
10 MW
any updates ?
Animalector
10 kW
Not really no.. I haven't built the new driver \ output stage yet.. Got nothing to put it in. Plus it costs money. Currently building a house. So busy...
izeman
1 GW
i wish i could make it work with the mac. i'm now building a test stand so i can try different combinations on my work bench when it starts getting cold and wet outside.
@lebowski: have you had a look into the adaptto controller setup? they need a special setup for the mac as well. maybe something like that is needed here as well?! i can look it up later.
@lebowski: have you had a look into the adaptto controller setup? they need a special setup for the mac as well. maybe something like that is needed here as well?! i can look it up later.
Animalector
10 kW
interesting. maybe we should just email them to ask?
izeman
1 GW
i dropped them an email.
what i found so far is (from the manual):
BMC halls fix - option to help overcome the problem of starting motors in BMC / MAC by shifting the timing advance when you start back. At the start slightly reduced efficiency, but start going better.
and oleg mentioned that auto setup may not always work and there needs some manual tweaking to be done for the mac/bmc.
what i found so far is (from the manual):
BMC halls fix - option to help overcome the problem of starting motors in BMC / MAC by shifting the timing advance when you start back. At the start slightly reduced efficiency, but start going better.
and oleg mentioned that auto setup may not always work and there needs some manual tweaking to be done for the mac/bmc.
Lurkin
100 kW
Are these controllers ok for less power again? I.e. ~1kw? I am considering a 1kw Q11 direct drive for the next bike.
Does anyone have any boards left?
How much (in parts) do these cost to make?
Is it possible to use a CA v3 with these?
Does anyone have any boards left?
How much (in parts) do these cost to make?
Is it possible to use a CA v3 with these?
Animalector
10 kW
Hi Lurkin
No reason why they are not suited for less power. you could pick more suitable FETs for low power to make it lighter / cheaper, that might be the only thing worth considering.
I commented somewhere in email or on this thread about the cost for parts.. it's substantial in the order of $200 AUD plus board cost. then you have to build it. If you enjoy soldering that's great, but they're small components, and it takes time. The BOM is on the first post, you can price it in au.element14.com pretty easily.
I have PCB's you can buy if you like. but I think you should take on this project only if you want the satisfaction / challenge of building a controller yourself with good configurable control, because you can buy off the shelf small, FOC controllers pretty cheaply.. It's the "small with good power handling" that I was chasing. Plus the challenge of building one just to see if I could...
Andy
No reason why they are not suited for less power. you could pick more suitable FETs for low power to make it lighter / cheaper, that might be the only thing worth considering.
I commented somewhere in email or on this thread about the cost for parts.. it's substantial in the order of $200 AUD plus board cost. then you have to build it. If you enjoy soldering that's great, but they're small components, and it takes time. The BOM is on the first post, you can price it in au.element14.com pretty easily.
I have PCB's you can buy if you like. but I think you should take on this project only if you want the satisfaction / challenge of building a controller yourself with good configurable control, because you can buy off the shelf small, FOC controllers pretty cheaply.. It's the "small with good power handling" that I was chasing. Plus the challenge of building one just to see if I could...
Andy
Lurkin
100 kW
Fair call and about what I expected.
I think I would like to build one for the fun of it at some stage, will be picking up a 9c kit later this week which will suffice in the mean time.
It might be something I will buy to tinker with slowly rather than satisfy the current need for a better set up.
I think I would like to build one for the fun of it at some stage, will be picking up a 9c kit later this week which will suffice in the mean time.
It might be something I will buy to tinker with slowly rather than satisfy the current need for a better set up.
Animalector
10 kW
some renders of the new layout. Upgrades are mounting bolts for driver board,no need for copper posts, more space for Caps,
At some stage I will re-do the CPU board, and use cheaper more common regulators and add temperature sensors,
At some stage I will re-do the CPU board, and use cheaper more common regulators and add temperature sensors,
izeman
1 GW
are the standoffs used as current transport? is steel good for that? i don't know if copper standoffs exist.
Animalector
10 kW
Nope they are just for mounting current path is through the board perpendicular to the FET board and driver. There are two.wires that run from the powerful to the driver for the Vbat but that's it the rest of the connections are 10 way plugs or inline headers from the driver to the FET board I'll show them in a photo soon.
izeman
1 GW
ok. that sounds interesting. biggest con of your v1 are the standoffs and the need to squeeze the twisted signal wires in between the boards. even though there should be no need to service and components it's good to be able to do so if needed.
i want to swap the 100B ACS758 for 200U, and can't, because the whole board is one big block.
if i don't use regen: will i still need B type current sensors? i guess yes, as the controller won't know if the ACS are installed that way or the other way, correct?
EDIT: doesn't matter. i ordered 150B now which will be enough anyway.
i want to swap the 100B ACS758 for 200U, and can't, because the whole board is one big block.
if i don't use regen: will i still need B type current sensors? i guess yes, as the controller won't know if the ACS are installed that way or the other way, correct?
EDIT: doesn't matter. i ordered 150B now which will be enough anyway.
Animalector
10 kW
Yeah when I did the stand-offs, I was actually thinking bullet connectors so one board slots onto the other, but it would have stood too high. Anyways it was one of those ideas that you think is awesome until you actually try to build it... Haha.
Don't know about the unidirectional current sensor. usually the voltage bias is at v_supply /2, and positive current is above, negative current is below, with a uni, 0amps is 0v (approx) up to 200amps at v_supply out. So if the lebowski chip can take some sort of offset, then yes you can use the 200u. But I think 150b will be fine that's still over 3kW yeah?
So is it.running the mac yet? You say in your email.that it's better but still not running properly.?
Andy
Don't know about the unidirectional current sensor. usually the voltage bias is at v_supply /2, and positive current is above, negative current is below, with a uni, 0amps is 0v (approx) up to 200amps at v_supply out. So if the lebowski chip can take some sort of offset, then yes you can use the 200u. But I think 150b will be fine that's still over 3kW yeah?
So is it.running the mac yet? You say in your email.that it's better but still not running properly.?
Andy
Lebowski
10 MW
Nope sensors must be bi-directional as all motor currents are AC...
izeman
1 GW
ok. thanks for confirmation. i ordered 150B. this will give a little headroom in case the MAC starts to work properly with the controller *ggg*Lebowski said:
Lebowski
10 MW
Animalector said:
A thought on the voltage regulators.
Izeman but also Arlo1 had some funny things going on where the chip reset itself during the FOC measurement.
I wonder whether this has to do with the dcdc converters. I can imagine that when you turn on a power-hungry function the controller
sees a dip of its input voltage due to the inductance of the wiring between battery and controller. A dcdc converter can be
a quite slow device, only 1 kHz bandwidth or so. It may not be able to maintain a proper 5V during a dip of the controller
input voltage. The controller IC reset when its supply dips below 4.7V (below this voltage ROM writes like when saving
the settings, are no longer reliable)
I think I was lucky with the design of my controller PCB, I've always derived the 5V not from the battery but from the 15V
gate drive. This provides an extra layer of isolation between the battery voltage and the 5V processor voltage. A dip in the
controller input voltage probably has an effect on the 15V (but hey, if this dips to 12 or so the FETS still switch on so no biggie)
but hardly any effect on the 5V.
So, my $0.02 being that deriving the 5V from the 15V might be better than deriving the 5V straight from the controller input voltage.....
p.s. v3.00 is coming, 64pins of goodness
izeman
1 GW
the dc/dc converter i installed is a 3A (!!) type. so it should be able to hold voltage quite nicely?!
of course i can't say how stable it is. what is the top most voltage i should set it to? maybe 5.1 or 5.2V? so if it ditched it may not go down to 4.7V.
of course i can't say how stable it is. what is the top most voltage i should set it to? maybe 5.1 or 5.2V? so if it ditched it may not go down to 4.7V.
Lebowski
10 MW
It is not only the amps it can deliver, it is also the speed with which it responds to input supply or load changes....izeman said:
izeman
1 GW
so for testing it may be a good idea to feed 5v from a battery? it shouldn't sag at at 200mA.Lebowski said:
shouldn't the added cap help as well?
Similar threads
