Harold in CR
100 kW
I forgot this was a brushless DC controller brain as I wanted an induction motor set up. Sorry.
Harold in CR
Harold in CR
The HI-Lebowski: a Lebowski SMD brain running a zombified Honda IMA Inverter: *a HOW-TO guide*
- Thread starter amberwolf
- Start date
kiwifiat
100 W
amberwolf said:arber333 said:
If you're referring to Kiwifiat's post up above:
I am presuming his is based on experience/usage with this same system (though I don't know with what kind of motor)...but I'm all for safer rather than sorrier.kiwifiat said:
Yes your presumption is correct, bear in mind that the IMA inverter is a baby compared to say a Leaf or Ampera/Volt inverter and has signifiacntly smaller IGBT's . I started at 3000ns and worked down to 1600ns where things deteriorate. Testing at 2000ns and 42,000 erpm revealed no issues and the great thing about the Lebowski controller is that you are free to set the deadtime to any value you choose.
Since I'm using large hubmotors, 23 pole pairs, at no more than around 300rpm, then i don't even need more than about 6900 erpm. Round that up to 10,000 for offground testing, maybe. Does that make any difference?
Also, I have edited the "SMDController.rtf" file that BobC created, originally posted over here
https://endless-sphere.com/forums/viewtopic.php?f=30&t=87104&start=25#p1331006
to correct the mistakes that led to the problems people had originally. I think I corrected them all, and updated the serial port graphic. Can someone verify that I did indeed fix it completely?
Anyone that wishes to help update the SMDcontroller updated 050620.zip (RTF) file linked below, to make it an easily-followed, greatly-informational document is welcome to download it, edit it, then reupload it in their own post so that I can incorporate usable material into the final version. (note that RTF is still disabled in the forum so you must put it in a zip file to upload it)
Updated version:
https://endless-sphere.com/forums/download/file.php?id=272897
Once it is finished, I'll see if i can get whereswally to replace the version in that post with the new one so no one makes that mistake again.
Also, once this thread has all the necessary documentation (what it is, how to use it, how to wire it up to something, etc) for the BobC SMDLebowski brain board as it exists today, I'll copy it all over to a post on that thread.
Today I'm wiring up the brain board in HIL1 to the pigtails, and hopefully the inverter board itself too. Am putting all my pinouts and wire colors for my own reference in one of the brain board wiring posts above. (link later)
The longest part of this is locating all the existing documentation, copying and linking (and fixing where needed) it in the reference posts at the begining of this thread, and also documenting what exactly I'm doing as I do it so I don't forget something.
Some thoughts about the Drive status LEDs:
I wish there was a connector to plug in external LEDs to for the bank of four drive-mode indicators, so I could put them somewhere I could see when it's mounted on the trike, for troubleshooting purposes later on if necessary.
For now I am planning to add a "window" to the IMA inverter casing just over their position, and put four clear plastic rods down to them.
Haven't decided exactly how to mechanically do it, but probably silicone the rods to the actual board/LEDs, then have them go up to *just* inside the casing surface, and then have a clear window of plastic glued to the casing that the rods can be seen thru without quite crawling under the trike.
Also considered leaving the rods long enough to extend out of the case, and pass thru four drilled holes in the plastic, and sand the ends of the rods into domes so they act as indicators that can be seen from the side if the ambient light is not too bright under the trike. Then I coudl see them a little easier without an inspection mirror on a long handle that I'd be sticking under the trike to look thru.
Also, I have edited the "SMDController.rtf" file that BobC created, originally posted over here
https://endless-sphere.com/forums/viewtopic.php?f=30&t=87104&start=25#p1331006
to correct the mistakes that led to the problems people had originally. I think I corrected them all, and updated the serial port graphic. Can someone verify that I did indeed fix it completely?
Anyone that wishes to help update the SMDcontroller updated 050620.zip (RTF) file linked below, to make it an easily-followed, greatly-informational document is welcome to download it, edit it, then reupload it in their own post so that I can incorporate usable material into the final version. (note that RTF is still disabled in the forum so you must put it in a zip file to upload it)
Updated version:
https://endless-sphere.com/forums/download/file.php?id=272897
Once it is finished, I'll see if i can get whereswally to replace the version in that post with the new one so no one makes that mistake again.
Also, once this thread has all the necessary documentation (what it is, how to use it, how to wire it up to something, etc) for the BobC SMDLebowski brain board as it exists today, I'll copy it all over to a post on that thread.
Today I'm wiring up the brain board in HIL1 to the pigtails, and hopefully the inverter board itself too. Am putting all my pinouts and wire colors for my own reference in one of the brain board wiring posts above. (link later)
The longest part of this is locating all the existing documentation, copying and linking (and fixing where needed) it in the reference posts at the begining of this thread, and also documenting what exactly I'm doing as I do it so I don't forget something.
Some thoughts about the Drive status LEDs:
I wish there was a connector to plug in external LEDs to for the bank of four drive-mode indicators, so I could put them somewhere I could see when it's mounted on the trike, for troubleshooting purposes later on if necessary.
For now I am planning to add a "window" to the IMA inverter casing just over their position, and put four clear plastic rods down to them.
Haven't decided exactly how to mechanically do it, but probably silicone the rods to the actual board/LEDs, then have them go up to *just* inside the casing surface, and then have a clear window of plastic glued to the casing that the rods can be seen thru without quite crawling under the trike.
Also considered leaving the rods long enough to extend out of the case, and pass thru four drilled holes in the plastic, and sand the ends of the rods into domes so they act as indicators that can be seen from the side if the ambient light is not too bright under the trike. Then I coudl see them a little easier without an inspection mirror on a long handle that I'd be sticking under the trike to look thru.
No actual wiring has yet been done, I spent the entire day*** looking up and verifying and documenting pinouts, etc., and drawing up (and redrawing) reference diagrams for how the actual wires I have will be hooked up (partly to make sure I"m doing it right, and partly so I have a reference later if I need to open stuff up and know where a particular wire goes).
So this is about how it will be for the first unit, HIL1, not counting the wires to the inverter itself from the brain (got too tired to see which color stripes are on which wires, so I could make sure they're marked right). It also doesn't include the reverse, because that's another connector I forgot. It will probably be another phono plug. If I had a mono-version of the 3.5mm phono jack I'd use that, cuz I have a few of the plugs on double-ended cables. (I could use a stereo version of the jack, I guess).
(EDIT: updated diagram with note explaining 5v source for throttle and ebrake)
*** not really the *entire* day, cuz the dogs variously insisted upon playtime, nomz, attentionalz, etc., that probably took a couple of hours in there a bit at a time, and there was lunch, and in the early morning before it got too melty a little bit of yard work while I still could....
So this is about how it will be for the first unit, HIL1, not counting the wires to the inverter itself from the brain (got too tired to see which color stripes are on which wires, so I could make sure they're marked right). It also doesn't include the reverse, because that's another connector I forgot. It will probably be another phono plug. If I had a mono-version of the 3.5mm phono jack I'd use that, cuz I have a few of the plugs on double-ended cables. (I could use a stereo version of the jack, I guess).
(EDIT: updated diagram with note explaining 5v source for throttle and ebrake)
*** not really the *entire* day, cuz the dogs variously insisted upon playtime, nomz, attentionalz, etc., that probably took a couple of hours in there a bit at a time, and there was lunch, and in the early morning before it got too melty a little bit of yard work while I still could....
kiwifiat
100 W
Your wiring diagram does not show any connection to 5v for the throttle. The controller will not run without the 5v going to the throttle.
Is there a particular reason the throttle *must* be powered by the brain board itself, and not an external 5v?
I've updated the diagram with a notation that the throttle signal comes from the Cycle Analyst v3, so it doesn't need the 5v from the controller (it has it's own), just a ground reference.
AFAIK, if I connect them, it will cause a conflict between the two slightly different 5v voltages, and potentially damage one or both 5v supplies?
This is why the CA v3 throttle signal to the present generic controllers doesn't use the 5v either, only ground and signal.
The ebrake (a hall-based cable-operated throttle pulled by a brake lever) is also powered by the CA's 5v, so it also doesn't use the 5v from the brain.
Additionally, since the two HIL's, once completed, will both share the same throttle and ebrake signals, then even if I were powering ebrake/throttle from the brain, I would only use *one* of the brains' 5v lines to power them, and not tie the 5v lines together? (just hypothetical, since I will still be powering both ebrake and throttle from the CAv3).
I've updated the diagram with a notation that the throttle signal comes from the Cycle Analyst v3, so it doesn't need the 5v from the controller (it has it's own), just a ground reference.
AFAIK, if I connect them, it will cause a conflict between the two slightly different 5v voltages, and potentially damage one or both 5v supplies?
This is why the CA v3 throttle signal to the present generic controllers doesn't use the 5v either, only ground and signal.
The ebrake (a hall-based cable-operated throttle pulled by a brake lever) is also powered by the CA's 5v, so it also doesn't use the 5v from the brain.
Additionally, since the two HIL's, once completed, will both share the same throttle and ebrake signals, then even if I were powering ebrake/throttle from the brain, I would only use *one* of the brains' 5v lines to power them, and not tie the 5v lines together? (just hypothetical, since I will still be powering both ebrake and throttle from the CAv3).
kiwifiat
100 W
amberwolf said:
That was not the point of the post. The throttle *must* be powered in order for the controller to function and none was shown. Now that you have updated your schematic there should be no room for errors.
amberwolf said:
Correct , you only need a common ground between the controller and the power source for the throttle and to limit the throttle output voltage to the range between 0-5V.
Ok. Just making sure I hadn't missed something.kiwifiat said:
This project has been a little more complex than anticipated, partly because all of the information I needed was not all in one place, in a clear, obvious format (which is what I'm trying to fix with this thread). While the Lebowski manual itself contains references to using various functions, it doesn't explicitly call out that some things must be wired this way or that, etc.
The same is true of the BobC buildup-and-test document; there's a number of things I think he simply didn't get to write before he died. This is another document I hope to perfect for others to use, but will probably need feedback on it as I go.
And the same is true of the various threads for various uses of it. Lots of detail on certain things, and nothing at all on others, etc.
Good. The CAv3 doesn't output anything more than that range, so it should be safe.
At present it is still set for the standard hall-throttle output range for the generic controllers still running the trike. I'll make a new setup for it that uses the full 0-5v output range once it's running the HILs...or maybe just use a second CA entirely since I have a spare, and just swap it out, keeping the other one with me to swap back if I run into problems.
Would be nice if the CA's preset function included the throttle ranges and proportional-regen-throttle and stuff like that, but it doesn't, so I can't use presets to swap back and forth, since the setups for those will be completely different between the HILs and the generics.
If I didn't need the CA to convert the PAS information into a throttle, and to do my speed limiting and such (with different limits for parking lot use vs streets vs bike paths), I'd just wire the throttle directly to the brains and not worry about it.
I guess this might be the first PAS-controlled large-EV controller setup.
kiwifiat
100 W
amberwolf said:
No doubt Bob would have sorted everything out had he not passed on unexpectedly. You could update the current docs but as I have said elsewhere there is a new BobC schematic and board layout. I have had some boards made at OSH Park and am testing an assembled one now. Once the testing phase is complete Whereswally606 will post the updated schematics, gerbers and build instructions then.
amberwolf said:
That would be entirely unnecessary as the Lebowski brain is far too smart to need that and will run perfectly on a HALL effect throttle. In fact if you had a throttle that put out say from 0.5 to 3V you would be able to calibrate it to function just as well as any other throttle. The only limitation is that you do not exceed the 5V maximum.
kiwifiat said:
Ah. Then the only reason to update the old docs would be to fix them for anyone wishing to build the old boards for whatever reason, which seems unlikely. I guess it would still be a good idea to replace the old file in Whereswally's post in the BobC SMD brainboard thread with the mistake in it with the corrected one I posted in the docs post early in the thread though, just so it doesn't have the incorrect references in it re: BAT vs BPOW.
Ok. My only thought was that control resolution would be better using the full voltage range, but I don't know that it matters.
So I could just setup the brains for the range the CA is presently setup to output, and leave it at that, unless I want to test if there is a difference.
Then setup the response for each brain's output to match the motor each one is using, since they are different winds (4504 and 4503) on each side, so that I get the same torque/response from each wheel for smooth acceleration and braking (without pulling to one side like it does now with the generics), and load-sharing.
Some thinking out loud, to be sure I've covered everything, before summarizing it as instructions to go in one of the reserved posts above, and then actually wiring it up.
Prepping the IMA boards for wiring:
The conformal coat (clear gunk) over the boards precludes just soldering wires to them. First the CC has to be removed. Scraping is tedious and presents the possibility of physical damage, so using that degreaser/solvent previously pictured is the best bet, to remove the CC from the areas around the points wires must go, such as gate drive and current sensor pads.
Using a q-tip or cotton ball dipped in the solvent to wipe the CC off only takes a few seconds to dissolve it, then wipe it away. Leaves a gooey ring around the area for a few minutes until that re-dries. I don't recommend using paper towels or the like, as there will be fibers everywhere stuck to things. I do recommend wearing gloves to keep it off your skin, and doing it outside or in a well-ventilated area.
Power supply to the brain board:
BPOW is the "battery positive" input pin, to power the brain board. This is pin 24 of the double-row header on the brain board;
the two pins direclty next to it above and to the side are not used for anything else to improve isolation. See schematic section at end of post. The schematic shows that a resistor, R1, needs to go between actual battery positive output and this pin, but that is not necessary if powering the board from 12v, with the board's own 15v DC-DC removed (like mine are, as supplied from Kiwifiat, to prevent potential problems since I am not going to use it anyway), then this resistor is not needed.
(if you are doing something different, you'll have to figure out what this resistor is supposed to be, as I haven't yet found any documentation telling me how to determine it, and the schematic says "TBD"--if anyone knows, post it, and I'll include that here).
Since I will be using a 12V DC-DC to provide gate drive power, I will also use that 12V directly connected to the BPOW pin instead.
Since I will be running two brains, each in it's own IMA inverter, they will each use one of the DC-DC's separate isolated 12v outputs.
So for this type of setup, on the brain board itself, pin 24 BPOW goes to the 12v pad of the not-installed capacitor at "C5" on the IMA inverter board, which at the same edge of the board as the large gray connector (that you won't be using), across that "blank space" at the edge, past the screwhead, and just next to a pair of large can-style capacitors (C195, C238), and one small SMD capacitor (C2). The pad you need is the one closest to the edge of the board.
The other pad the same size but inboard of it is the ground pad, and you'll wire that to the group of grounds from gates and current sensors (see later in this post, down at that section of wiring instructions) that then goes to pin 9 of the 24pin connector on the edge of the brainboard.
That is also where I will be soldering the wires from the 12vDCDC pigtail, since it is easy to stress-relieve that by ziptying that cable to the PCB edge rigth there, in the unused hole next to the gray connector.
There will be a diagram for this once I draw one up based on the photo with green and purple lines that Kiwifiat provided in a previous post, temporarily included here:
Next, battery measurement input to the brain board.
BAT is the voltage measurement input pin. It is NOT the power input to the brain board. This is pin 14 of the double-row header on the brain board. See schematic section at end of post. A resistor, R2 ( R_meas_battery ), needs to go between actual battery positive output and this pin. According to the schematic, the formula to determine it's resistance is
R_meas_battery = 500 * (highest battery voltage - 5) Round up.
So for my 14s system, the highest battery voltage is really 57.7v, but I'd just round that to 58v.
So that would be 500 * (58v - 5) = 500 * 53v = 26500, or 26.5kohms, which rounds up to 27kohm for the nearest standard value resistor. This can be any regular 1/4watt or even 1/8watt type (etc) resistor, or set of them that adds/etc to this value if you don't have the standard-value one handy.
One end of this resistor is wired into Pin 14 of the 24pin connector at the edge of the brain board. The other end goes to actual battery voltage in the controller, but I have not yet located where that is on the board itself (obviously it is available on the main P bolt-on terminal when the controller is assembled, but I'd rather run this from somewhere on the PCB itself). Once I find it I'll update these instructions to note where to run the wire.
Brain-to-Powerstage wiring requirements:
Thin, 22awg or smaller, insulated wires should work for all of the interconnects; all of these go to the 24 pin connector on one end of the brainboard.
The IGBT bridge gate driver wiring will need three sets of three wires twisted together, one set for each of the three bridges:
Gate High
GUA (pin8)
GUB (pin6)
GUC (pin4)
Gate Low
GLA (pin7)
GLB (pin5)
GLC (pin3)
Ground for each group
All the grounds are joined at one single point offboard, and then connect to a single ground (pin9) on the brain board to avoid ground loops.
I'm going to use sets of wires from an old many-conductor test equipment cable (you can also buy this stuff online new), which has color-striped wires. This means that I can use say, Yellow base wire color for GxA, so that GUA is Yellow with a white stripe, GLA is Yellow with a black stripe, and GxA Ground is just the plain Yellow wire.
Then I can do the same with Green for the GxB wires, and Blue for the GxC wires. (and so follow the "convention" of typical ebike hubmotor/controller stuff, to make it easier for me to remember what these are for).
The wires for each of these will run to the respective gate drive inputs on the IMA inverter PCB, which will be shown in a diagram below (once I draw it up, based on the photo with green and purple lines that Kiwifiat provided in a previous post, temporarily included here.
View attachment 5
The current sensor wiring will need three sets of two wires twisted together, one set for each of the three current sensors:
IA (pin 10)
IB (pin 11)
IC (pin 12)
Ground for each
I'm using pairs of wires color and color stripe, as with the gates above, to keep track. I don't have a convenient set of similar colors to the above in this cable, so I'll be using Violet and Violet-Yellow for IA, Black and Black-white for IB, and Pink and Pink-black for IC, with the striped wire as signal, and solid as ground (to follow the same convention as the gates).
The wires for each of these will run to the respective current sensor outputs on the IMA inverter PCB, which will be shown in a diagram below (once I draw it up, based on the photo with green and purple lines that Kiwifiat provided in a previous post, temporarily provided here).
View attachment 5
Serial Communication (for PC setup and programming):
Only RX, TX, and Ground are needed.
So for my use case, I'll probably use the Grin Tech USB-serial cable I already have, which uses a 3.5mm TRS phono plug for the serial end, and regular USB-A for the computer end. Then I can wire a panel mount jack or a cabled phono jack as a pigtail (of which I have plenty in old audio cables) to the brains' comm port (see attached PCB pinout at end of post). I can water-resist the jack by simply inserting a rubber plug into the hole. The jack can mount in the plastic housing of the IMA once I figure out the layout and spacing.
But most people will probably want to use a standard DB-9 panel mount connector for this. If so, then the wiring is as follows:
Comm RX pin = DB9 pin 3
Comm TX pin = DB9 pin 2
Comm GND pin = DB9 pin 5
That's what I will also do, though using a DB9 pigtail as I have many old serial cables here to make one from, and just use an adapter from phono to DB9. This will allow me to also run the SET and RES switch lines out on the unused DB9 pins, and make a "box" with the adapter stuff in it that mounts the switches, and that will leave the switches unconnected and the SET and RES lines open circuit when not actually plugged into the computer for programming.
I decided not to do it, but I've also considered building the USB-serial adapter directly into each HI-Lebowski unit, using a pair of those adapters that have the USB-B style connector, like these (but for RS232-TTL instead; the one linked is for something else and won't work):
https://www.amazon.com/dp/B07ZK2QJVG/?coliid=I1U8EJE2ZNYXUO
and then just using a standard USB device cable (A to B type) from the computer to connect each time. AFAICT that should work, as the adapter won't be powered except when plugged into the USB, and so shouldn't cause issues with the brains when not connected. I can also make or get rubber plugs for USB-B style holes, or even use gorilla-tape to cover the area around the plug and plug itself, though the tape will need to be inspected and replaced periodically.
SET and RES jumpers for setup and programming
These pads on teh board must be connected in a particular order (listed in the Lebowski software manual for the version running on your particular chip/brainboard) to start the setup process. Or you can use just the RES jumper to "reboot" the brain, if necessary during troubleshooting.
As noted in the previous section, you can use the unused DB9 wires to run to switches externally.
But you could also mount the switches on the IMA plastic casing, using panel-mount toggles, rockers, pushbuttons, etc., as long as they are a latching on/off, you only need SPST.
Or you can run the pins from the board's pads out to header pins mounted externally to the IMA, and jumper them there (most likely you won't want to open the IMA up every time you want to change settings).
More to be added (hall sensors, throttle, ebrake, reverse) after dinner. EDIT: probably not till tomorrow or later...important doggie business came up, so I was helping them with that (you know, sticks, scritches, etc) and now it's bedtime.
Prepping the IMA boards for wiring:
The conformal coat (clear gunk) over the boards precludes just soldering wires to them. First the CC has to be removed. Scraping is tedious and presents the possibility of physical damage, so using that degreaser/solvent previously pictured is the best bet, to remove the CC from the areas around the points wires must go, such as gate drive and current sensor pads.
Using a q-tip or cotton ball dipped in the solvent to wipe the CC off only takes a few seconds to dissolve it, then wipe it away. Leaves a gooey ring around the area for a few minutes until that re-dries. I don't recommend using paper towels or the like, as there will be fibers everywhere stuck to things. I do recommend wearing gloves to keep it off your skin, and doing it outside or in a well-ventilated area.
Power supply to the brain board:
BPOW is the "battery positive" input pin, to power the brain board. This is pin 24 of the double-row header on the brain board;
the two pins direclty next to it above and to the side are not used for anything else to improve isolation. See schematic section at end of post. The schematic shows that a resistor, R1, needs to go between actual battery positive output and this pin, but that is not necessary if powering the board from 12v, with the board's own 15v DC-DC removed (like mine are, as supplied from Kiwifiat, to prevent potential problems since I am not going to use it anyway), then this resistor is not needed.
(if you are doing something different, you'll have to figure out what this resistor is supposed to be, as I haven't yet found any documentation telling me how to determine it, and the schematic says "TBD"--if anyone knows, post it, and I'll include that here).
Since I will be using a 12V DC-DC to provide gate drive power, I will also use that 12V directly connected to the BPOW pin instead.
Since I will be running two brains, each in it's own IMA inverter, they will each use one of the DC-DC's separate isolated 12v outputs.
So for this type of setup, on the brain board itself, pin 24 BPOW goes to the 12v pad of the not-installed capacitor at "C5" on the IMA inverter board, which at the same edge of the board as the large gray connector (that you won't be using), across that "blank space" at the edge, past the screwhead, and just next to a pair of large can-style capacitors (C195, C238), and one small SMD capacitor (C2). The pad you need is the one closest to the edge of the board.
The other pad the same size but inboard of it is the ground pad, and you'll wire that to the group of grounds from gates and current sensors (see later in this post, down at that section of wiring instructions) that then goes to pin 9 of the 24pin connector on the edge of the brainboard.
That is also where I will be soldering the wires from the 12vDCDC pigtail, since it is easy to stress-relieve that by ziptying that cable to the PCB edge rigth there, in the unused hole next to the gray connector.
There will be a diagram for this once I draw one up based on the photo with green and purple lines that Kiwifiat provided in a previous post, temporarily included here:
Next, battery measurement input to the brain board.
BAT is the voltage measurement input pin. It is NOT the power input to the brain board. This is pin 14 of the double-row header on the brain board. See schematic section at end of post. A resistor, R2 ( R_meas_battery ), needs to go between actual battery positive output and this pin. According to the schematic, the formula to determine it's resistance is
R_meas_battery = 500 * (highest battery voltage - 5) Round up.
So for my 14s system, the highest battery voltage is really 57.7v, but I'd just round that to 58v.
So that would be 500 * (58v - 5) = 500 * 53v = 26500, or 26.5kohms, which rounds up to 27kohm for the nearest standard value resistor. This can be any regular 1/4watt or even 1/8watt type (etc) resistor, or set of them that adds/etc to this value if you don't have the standard-value one handy.
One end of this resistor is wired into Pin 14 of the 24pin connector at the edge of the brain board. The other end goes to actual battery voltage in the controller, but I have not yet located where that is on the board itself (obviously it is available on the main P bolt-on terminal when the controller is assembled, but I'd rather run this from somewhere on the PCB itself). Once I find it I'll update these instructions to note where to run the wire.
Brain-to-Powerstage wiring requirements:
Thin, 22awg or smaller, insulated wires should work for all of the interconnects; all of these go to the 24 pin connector on one end of the brainboard.
The IGBT bridge gate driver wiring will need three sets of three wires twisted together, one set for each of the three bridges:
Gate High
GUA (pin8)
GUB (pin6)
GUC (pin4)
Gate Low
GLA (pin7)
GLB (pin5)
GLC (pin3)
Ground for each group
All the grounds are joined at one single point offboard, and then connect to a single ground (pin9) on the brain board to avoid ground loops.
I'm going to use sets of wires from an old many-conductor test equipment cable (you can also buy this stuff online new), which has color-striped wires. This means that I can use say, Yellow base wire color for GxA, so that GUA is Yellow with a white stripe, GLA is Yellow with a black stripe, and GxA Ground is just the plain Yellow wire.
Then I can do the same with Green for the GxB wires, and Blue for the GxC wires. (and so follow the "convention" of typical ebike hubmotor/controller stuff, to make it easier for me to remember what these are for).
The wires for each of these will run to the respective gate drive inputs on the IMA inverter PCB, which will be shown in a diagram below (once I draw it up, based on the photo with green and purple lines that Kiwifiat provided in a previous post, temporarily included here
. View attachment 5
The current sensor wiring will need three sets of two wires twisted together, one set for each of the three current sensors:
IA (pin 10)
IB (pin 11)
IC (pin 12)
Ground for each
I'm using pairs of wires color and color stripe, as with the gates above, to keep track. I don't have a convenient set of similar colors to the above in this cable, so I'll be using Violet and Violet-Yellow for IA, Black and Black-white for IB, and Pink and Pink-black for IC, with the striped wire as signal, and solid as ground (to follow the same convention as the gates).
The wires for each of these will run to the respective current sensor outputs on the IMA inverter PCB, which will be shown in a diagram below (once I draw it up, based on the photo with green and purple lines that Kiwifiat provided in a previous post, temporarily provided here).
View attachment 5
Serial Communication (for PC setup and programming):
Only RX, TX, and Ground are needed.
So for my use case, I'll probably use the Grin Tech USB-serial cable I already have, which uses a 3.5mm TRS phono plug for the serial end, and regular USB-A for the computer end. Then I can wire a panel mount jack or a cabled phono jack as a pigtail (of which I have plenty in old audio cables) to the brains' comm port (see attached PCB pinout at end of post). I can water-resist the jack by simply inserting a rubber plug into the hole. The jack can mount in the plastic housing of the IMA once I figure out the layout and spacing.
But most people will probably want to use a standard DB-9 panel mount connector for this. If so, then the wiring is as follows:
Comm RX pin = DB9 pin 3
Comm TX pin = DB9 pin 2
Comm GND pin = DB9 pin 5
That's what I will also do, though using a DB9 pigtail as I have many old serial cables here to make one from, and just use an adapter from phono to DB9. This will allow me to also run the SET and RES switch lines out on the unused DB9 pins, and make a "box" with the adapter stuff in it that mounts the switches, and that will leave the switches unconnected and the SET and RES lines open circuit when not actually plugged into the computer for programming.
I decided not to do it, but I've also considered building the USB-serial adapter directly into each HI-Lebowski unit, using a pair of those adapters that have the USB-B style connector, like these (but for RS232-TTL instead; the one linked is for something else and won't work):
https://www.amazon.com/dp/B07ZK2QJVG/?coliid=I1U8EJE2ZNYXUO
and then just using a standard USB device cable (A to B type) from the computer to connect each time. AFAICT that should work, as the adapter won't be powered except when plugged into the USB, and so shouldn't cause issues with the brains when not connected. I can also make or get rubber plugs for USB-B style holes, or even use gorilla-tape to cover the area around the plug and plug itself, though the tape will need to be inspected and replaced periodically.
SET and RES jumpers for setup and programming
These pads on teh board must be connected in a particular order (listed in the Lebowski software manual for the version running on your particular chip/brainboard) to start the setup process. Or you can use just the RES jumper to "reboot" the brain, if necessary during troubleshooting.
As noted in the previous section, you can use the unused DB9 wires to run to switches externally.
But you could also mount the switches on the IMA plastic casing, using panel-mount toggles, rockers, pushbuttons, etc., as long as they are a latching on/off, you only need SPST.
Or you can run the pins from the board's pads out to header pins mounted externally to the IMA, and jumper them there (most likely you won't want to open the IMA up every time you want to change settings).
More to be added (hall sensors, throttle, ebrake, reverse) after dinner. EDIT: probably not till tomorrow or later...important doggie business came up, so I was helping them with that (you know, sticks, scritches, etc) and now it's bedtime.
Attachments
I received this PM, and am answering it here so that anyone else also needing this info will know the status of the project:
However, since the project is not completed, I do not yet know this information, as Kiwifiat will be helping me with that part after wiring up of brain board to inverter is completed, and I have verified I can connect to the brain from the computer, etc.
When the project is completed, it will have all of the information that is available directly in the thread. At present, all information I have available is already posted to the thread.
At this point I have not even gotten the brain wired to the inverter, so those connections are not documented yet (other than Kiwifiat's basic diagram that doesn't include everything). Between being sick and working, I haven't had the energy/time to sit at the workstation and get the wiring/etc done.
Once the connections are done, then documentation of the initial setup and testing, including all parameter values in the brain menus, will be posted up.
However, since the project is not completed, I do not yet know this information, as Kiwifiat will be helping me with that part after wiring up of brain board to inverter is completed, and I have verified I can connect to the brain from the computer, etc.
When the project is completed, it will have all of the information that is available directly in the thread. At present, all information I have available is already posted to the thread.
At this point I have not even gotten the brain wired to the inverter, so those connections are not documented yet (other than Kiwifiat's basic diagram that doesn't include everything). Between being sick and working, I haven't had the energy/time to sit at the workstation and get the wiring/etc done.
Once the connections are done, then documentation of the initial setup and testing, including all parameter values in the brain menus, will be posted up.
kiwifiat
100 W
Tom de Bree published an investigation of the Honda IMA inverter over on DIYElectriccar.com back in 2015. Tom advises that the current sensors are 5.8mV/amp.
Anybody wanting to set up a Lebowski controller from scratch should read the manual published by Lebowski for the firmware version loaded on the controller. Lebowski also kindly put up a youtube video going through the procedure step by step. The video is for an older firmware never the less there are no significant changes in procedure.
[youtube]57kRSoOrTxw[/youtube]
Anybody wanting to set up a Lebowski controller from scratch should read the manual published by Lebowski for the firmware version loaded on the controller. Lebowski also kindly put up a youtube video going through the procedure step by step. The video is for an older firmware never the less there are no significant changes in procedure.
[youtube]57kRSoOrTxw[/youtube]
Been sick a while, so no news on this yet.
Even though I've been stuck home much of the last few weeks, I've been too worn out to even sit at the worktable and wire things. For the last week, too tired to even post on ES, mostly dozing/waking and sometimes letting youtube science videos play when I could concentrate. Sometimes poking around on Amazon and Ebay for wierd things I might be able to use. (didn't find much). The days I have made it to work I coudln't do much, and went home early a fair bit.
I'm still very tired, and fairly weak--just hobbling around the yard with the dogs for a few minutes can put me back in bed for hours. But I am having longer periods of activity-capable time, more often, so I might accomplish something "soon" (next week or two). Just not completely sure of myself in delicate work like the wiring and such still to be done. (I find myself dozing off randomly sometimes, too, which is a problem when soldering and the like).
As there are other non-controller-related things I may accomplish (or start) with the trike and some other household projects, I'll post a version of this in those threads too.
Now, for some of the stuff still to be done on this:
--have to finish wiring the first brain up to the first inverter. Still have to locate the point on the inverter board where main battery power is available, so I can wire it to the voltage sense resistor/pin on the brain.
--test and verify brain boots up and connects to pc via serial, and setup the basic stuff in the brain menus.
--modify plastic inverter housing for waterproofing. Most
--add fiberoptic strands for each of the status LEDs on the brain board into a "cable" that extends far out of the case (a few feet) so I can see the LEDs while testing and operating the controllers on the trike. This means making a little "cable" of FO strands for each brain, consisting of one strand for each LED, and then going out to a little mounting box to put where i can see it (on the handlebars if the light will go that far (probably not), on the side of the trike near the controllers pointing upwards so it can at least be seen looking down and back if that's all I can get. The mounting box would need a marking on it to show which strand is for which LED, and maybe a lens-like magnifier to spread the light out beyond the tiny dot the FO strands make.
Could also use "optical pickups" at the end of the FO to detect if there's light and then wire their outputs to separate LEDs at the handlebars.
Then I can test the first controller on the trike, and see if this is all worth it.
Even though I've been stuck home much of the last few weeks, I've been too worn out to even sit at the worktable and wire things. For the last week, too tired to even post on ES, mostly dozing/waking and sometimes letting youtube science videos play when I could concentrate. Sometimes poking around on Amazon and Ebay for wierd things I might be able to use. (didn't find much). The days I have made it to work I coudln't do much, and went home early a fair bit.
I'm still very tired, and fairly weak--just hobbling around the yard with the dogs for a few minutes can put me back in bed for hours. But I am having longer periods of activity-capable time, more often, so I might accomplish something "soon" (next week or two). Just not completely sure of myself in delicate work like the wiring and such still to be done. (I find myself dozing off randomly sometimes, too, which is a problem when soldering and the like).
As there are other non-controller-related things I may accomplish (or start) with the trike and some other household projects, I'll post a version of this in those threads too.
Now, for some of the stuff still to be done on this:
--have to finish wiring the first brain up to the first inverter. Still have to locate the point on the inverter board where main battery power is available, so I can wire it to the voltage sense resistor/pin on the brain.
--test and verify brain boots up and connects to pc via serial, and setup the basic stuff in the brain menus.
--modify plastic inverter housing for waterproofing. Most
--add fiberoptic strands for each of the status LEDs on the brain board into a "cable" that extends far out of the case (a few feet) so I can see the LEDs while testing and operating the controllers on the trike. This means making a little "cable" of FO strands for each brain, consisting of one strand for each LED, and then going out to a little mounting box to put where i can see it (on the handlebars if the light will go that far (probably not), on the side of the trike near the controllers pointing upwards so it can at least be seen looking down and back if that's all I can get. The mounting box would need a marking on it to show which strand is for which LED, and maybe a lens-like magnifier to spread the light out beyond the tiny dot the FO strands make.
Could also use "optical pickups" at the end of the FO to detect if there's light and then wire their outputs to separate LEDs at the handlebars.
Then I can test the first controller on the trike, and see if this is all worth it.
thoroughbred
100 W
hope you are feeling well enough to make a little progress on this :thumb:
Physically I'm pretty much better, but life things (trying to get work to let me come back before I run out of money and have to get a different job) have gotten in the way. At the moment, it is now a waiting-on-responses game.
So back to this project:
I've been measuring all over the PCB (not every single point, but at least a hundred so far, most of them unlikely but ran out of likely ones), and still have not found a place I can connect to for battery voltage so the brain board can measure the battery voltage. Either I've just missed it, or I'm doing something wrong.
So at this point I'll need to setup a large ring terminal on a thin wire to go to the battery voltage measuring pin (BAT), via the Rmeas resistor, on the brain board, and then that ring terminal has to go outside the casing to be bolted along with the battery positive wire to the POS terminal on the outside of the casing. I was really hoping to avoid this.
So back to this project:
I've been measuring all over the PCB (not every single point, but at least a hundred so far, most of them unlikely but ran out of likely ones), and still have not found a place I can connect to for battery voltage so the brain board can measure the battery voltage. Either I've just missed it, or I'm doing something wrong.
So at this point I'll need to setup a large ring terminal on a thin wire to go to the battery voltage measuring pin (BAT), via the Rmeas resistor, on the brain board, and then that ring terminal has to go outside the casing to be bolted along with the battery positive wire to the POS terminal on the outside of the casing. I was really hoping to avoid this.
kiwifiat
100 W
amberwolf said:I've been measuring all over the PCB (not every single point, but at least a hundred so far, most of them unlikely but ran out of likely ones), and still have not found a place I can connect to for battery voltage so the brain board can measure the battery voltage. Either I've just missed it, or I'm doing something wrong.
So at this point I'll need to setup a large ring terminal on a thin wire to go to the battery voltage measuring pin (BAT), via the Rmeas resistor, on the brain board, and then that ring terminal has to go outside the casing to be bolted along with the battery positive wire to the POS terminal on the outside of the casing. I was really hoping to avoid this.
You can attach to battery positive from any of the three IGBT's within the case at the point shown. You will still need a thin ring terminal to make a safe and secure connection but the wire will remain with in the housing. Bear in mind that Lebowski has indicated that battery voltage feedback is not a necessity and only makes a small difference to the FOC calculations. Plainly a prerequisite if you want take advantage of voltage related inverter shutdown in cases where you are not using a BMS for that task.
I'm glad to see someone actually getting this done (since I still haven't quite finished one! 
).
Please post as many details as you can of your progress and setup/programming, and usage.
Which method are you using to disable the onboard CPU?
).Please post as many details as you can of your progress and setup/programming, and usage.
Which method are you using to disable the onboard CPU?
What is under that trace, in the other board layers?
In this photo coleasterling took of the back of the board it looks like there is a big enough gap between the components on the other side of the board and the white rectangles to make a cut. Just need to make sure not to go near the driver board underneath the main board. I will cut the two gate supply traces going to each 3 rectangles with a sharp razor blade to avoid damage though.thoroughbred
100 W
what is the best vintage of used IMA inverters to buy? most of the listings are for 2010-2014 cars which look different from what is documented here. I want small, powerful and simple
I don't know why I havent seen this before. Brilliant work. It appears the Honda hybrid IMA controller is probably the perfect size for a street motorcycle E-conversion.
I've been looking at conversions a lot recently, and the most common complaint is controllers. There's absolutely nothing available in the same class as a used Honda controller from salvage, using a lebowski brain.
The 28S voltage is pretty ideal for motorcycles. That is the stock Zero motorcycle voltage, and (I know this sounds arrogant, but) I think Zero missed an opportunity by not making the 28S pack out of two 14S sub-packs. There are off-the-shelf inverters that can provide mains power to your home in an emergency from 13S/14S (I have two if them, and three battery packs).
I've been looking at conversions a lot recently, and the most common complaint is controllers. There's absolutely nothing available in the same class as a used Honda controller from salvage, using a lebowski brain.
The 28S voltage is pretty ideal for motorcycles. That is the stock Zero motorcycle voltage, and (I know this sounds arrogant, but) I think Zero missed an opportunity by not making the 28S pack out of two 14S sub-packs. There are off-the-shelf inverters that can provide mains power to your home in an emergency from 13S/14S (I have two if them, and three battery packs).
thoroughbred said:what is the best vintage of used IMA inverters to buy? most of the listings are for 2010-2014 cars which look different from what is documented here. I want small, powerful and simple
AFAIK the version needed for this specific thread's info is "06-11 Civic Hybrid IMA Power Distribution Module Unit 1B300-RMX-0032" based on the notes on the tape and stickers on mine (see pics in various posts of thread). This link (which will eventually go dead) is for one of these:
https://www.ebay.com/itm/06-11-Civic-Hybrid-IMA-Power-Distribution-Module-Unit-1B300-RMX-0032/324436949602
(I've added this info to the first post, too)
But it is likely that if you couldn't find one of those that you could still do it with a different version, as long as that version has the proper current sensors and is for a three-phase motor of the general voltage and size you are going to use.
Failing that, the Prius, Tesla, and others have also been repurposed this way (see links in first post).
Similar threads
