Which device have you been using that you like and that is apparently already controllable?garygid said:
It may be a basis to get started and/or use as a template.
Please also list the answer to the obvious question: why not use that device?
Mini-QC Rapid-Charger (RC) Project for LEAF QC Port
- Thread starter garygid
- Start date
Help Support My Nissan Leaf Forum:
It may be a basis to get started and/or use as a template.
Please also list the answer to the obvious question: why not use that device?
garygid
Well-known member
It is not isolated, so not wise to encourage use by the general public.
I have explained the basic requirements in a number of posts.
Please read the earlier parts of the thread.
However, thanks for the suggestions.
Since these are mostly made in China, I had trouble finding
documentation, even the basic voltage and current specs
of some similar items.
These are generally designed to be constant-voltage
suppiles, and controlling the current output of 36 or so
in series might be difficult.
Please let us know what you find inside.
I have explained the basic requirements in a number of posts.
Please read the earlier parts of the thread.
However, thanks for the suggestions.
Since these are mostly made in China, I had trouble finding
documentation, even the basic voltage and current specs
of some similar items.
These are generally designed to be constant-voltage
suppiles, and controlling the current output of 36 or so
in series might be difficult.
Please let us know what you find inside.
Agree.garygid said:
I was not asking for the requirements, I understood from the thread that you look for a portable, isolated, cheap, high power fast charger using Chademo to convert 240VAC to the 360-400VDC of the Leaf pack at up to 12kW or so.garygid said:
My question is how do you want to control it, as you indicated that the control and communication is mostly done.
All I see on this thread is that there are CAN bus loggings made and you have reverse engineered some of the messages.
This means that you are still in the research phase, not yet designing the last bits of the code to control the charger....
It is possible to design some DAC circuits to drive from the software that is interpreting the CAN bus commands and drive the expected output from the charger. Interface to the charger should not be that difficult.
It is not difficult to control these supplies and for a charger, they do not all need to be controlled. If the voltage changes from say 350 to 400V then only a handful need control, the rest can be dumb supplies, I did that with my earlier charger.garygid said:
Luckily, the RC guys have found out most, I just needed to follow their lead. One thing that is not much documented is the load sharing input, so I plan on experimenting with that to see how much it can turn the supply down, because that determines how many supplies need control and how many can be dumb. Also, the output voltage can be tweaked using the "sense" inputs.garygid said:
OK, tonight I experimented with a server supply to see how much it can vary its output from the available external inputs.
It appears that this supply (nominal 12.2V output) can vary its output not much more than about 0.5V upward, both when using the "sense" input as well as for the load share input. (Apparently the load share works by raising the output voltage when the load share is high, so the sharing can reduce the control voltage to reduce the load on this supply)
The sense input works in the opposite way: it checks the input for being equal to the reference (12V) and if lower, it increases the output until it reads the sense feedback equal to the reference.
Since both inputs have a very limited control range, it was necessary to open the supply to find a better way to control the output voltage. Inside are two trim pots (glued to fixed position). One had no obvious effect, so I expect that this is the Overvoltage protection level adjustment. The other one changed the output voltage, but again only +/- 0.3V or so.
This lead me to unsolder that pot and try some fixed resistors. I found that replacing the 1k pot with a 5k resistor reduced the output voltage by about 40% (from 12.2V to 7.4V). I also measured that there was another 4k resistor in series with the pot, so I expect that reducing that resistor will allow increasing the output voltage, up to the OverVoltage protection.
Since the power supply needs to be opened and modified, I no longer think that it is a good idea to use 12V supplies to build a 400V controllable power supply, as modifying 32 power supplies will be a heck of a job, but there are also 48V switching server supplies which would require only 8 cases to be opened and some soldering done inside, so that is still a reasonable amount of work.
From my experience with this and other switching power supplies, I expect that it is not much work to modify the 48V power supplies to give them:
- an isolated power on/off input (using 8 opto-couplers to switch the power supplies out of standby mode)
- output voltage control. Assuming the experiment with the 12V server supply is representative, I expect that a string of 8 series 48V supplies can be controlled to deliver from approx 240V to 440V which should be enough to cover all possible voltages that a Leaf battery can take. If I am not mistaken, a Leaf has 96 series cells, so I do not expect that it will ever need to be charged much over 400V and I also do not believe that the BMS will allow it to discharge below 300V.
The output voltage control can also be employed to control the current (into the Leaf battery pack) by reading back the charging current, using a shunt and amplifying the shunt voltage drop, reading it back with an ADC and calibrating it to be able to satisfy the Leaf's request for a certain current level up to the max the supply can deliver.
I believe it would be possible to build a low cost prototype of this using 8 such server supplies that can deliver 30A each and modifying them for this purpose as outlined above. I am inquiring with a local computer surplus store to get 8 units plus a spare for experimenting, so that it can actually be built and tested on a Leaf (I do not have one, but here in Silicon Valley they are plentiful). The size of a bundle of 8 supplies should be less than a cubic foot, air (fan) cooled and 12kW. This means that it is not an outdoor rated supply and needs to be weather protected (like a desktop computer), but I believe that this meets the design requirements and can probably be made for less than $300 in parts.
However, I do not know if the path chosen matches the ideas of control that resulted from the CAN bus measurements that other have done, so please give me feedback to tell if this is a viable path to an isolated DC Fast charger, especially the area of how we want to communicate over the CAN bus between charger and car. Is there a controller or a PIC that can support this, so it can talk to the car and use some DAC/ADC pins to control the power supply of the charger in response to the CAN control messages?
It appears that this supply (nominal 12.2V output) can vary its output not much more than about 0.5V upward, both when using the "sense" input as well as for the load share input. (Apparently the load share works by raising the output voltage when the load share is high, so the sharing can reduce the control voltage to reduce the load on this supply)
The sense input works in the opposite way: it checks the input for being equal to the reference (12V) and if lower, it increases the output until it reads the sense feedback equal to the reference.
Since both inputs have a very limited control range, it was necessary to open the supply to find a better way to control the output voltage. Inside are two trim pots (glued to fixed position). One had no obvious effect, so I expect that this is the Overvoltage protection level adjustment. The other one changed the output voltage, but again only +/- 0.3V or so.
This lead me to unsolder that pot and try some fixed resistors. I found that replacing the 1k pot with a 5k resistor reduced the output voltage by about 40% (from 12.2V to 7.4V). I also measured that there was another 4k resistor in series with the pot, so I expect that reducing that resistor will allow increasing the output voltage, up to the OverVoltage protection.
Since the power supply needs to be opened and modified, I no longer think that it is a good idea to use 12V supplies to build a 400V controllable power supply, as modifying 32 power supplies will be a heck of a job, but there are also 48V switching server supplies which would require only 8 cases to be opened and some soldering done inside, so that is still a reasonable amount of work.
From my experience with this and other switching power supplies, I expect that it is not much work to modify the 48V power supplies to give them:
- an isolated power on/off input (using 8 opto-couplers to switch the power supplies out of standby mode)
- output voltage control. Assuming the experiment with the 12V server supply is representative, I expect that a string of 8 series 48V supplies can be controlled to deliver from approx 240V to 440V which should be enough to cover all possible voltages that a Leaf battery can take. If I am not mistaken, a Leaf has 96 series cells, so I do not expect that it will ever need to be charged much over 400V and I also do not believe that the BMS will allow it to discharge below 300V.
The output voltage control can also be employed to control the current (into the Leaf battery pack) by reading back the charging current, using a shunt and amplifying the shunt voltage drop, reading it back with an ADC and calibrating it to be able to satisfy the Leaf's request for a certain current level up to the max the supply can deliver.
I believe it would be possible to build a low cost prototype of this using 8 such server supplies that can deliver 30A each and modifying them for this purpose as outlined above. I am inquiring with a local computer surplus store to get 8 units plus a spare for experimenting, so that it can actually be built and tested on a Leaf (I do not have one, but here in Silicon Valley they are plentiful). The size of a bundle of 8 supplies should be less than a cubic foot, air (fan) cooled and 12kW. This means that it is not an outdoor rated supply and needs to be weather protected (like a desktop computer), but I believe that this meets the design requirements and can probably be made for less than $300 in parts.
However, I do not know if the path chosen matches the ideas of control that resulted from the CAN bus measurements that other have done, so please give me feedback to tell if this is a viable path to an isolated DC Fast charger, especially the area of how we want to communicate over the CAN bus between charger and car. Is there a controller or a PIC that can support this, so it can talk to the car and use some DAC/ADC pins to control the power supply of the charger in response to the CAN control messages?
JasonA
Well-known member
Just FYI if I can chime here.. Cor, this is something that I do alot and know alot about is modifying 12v & 48v server power supplies to put in series (or par) to get more power to charge BIG LiPo's on our PowerLab's and iChargers.
Issue with server PS's is that they have to be modified to be isolated. Opened up, case and board removed, ground removed from 120v line, and then PCB put back so it's not touching or grounded to the outside case. Technically floating but the 120v is isolated from the DC.
This needs to be done if groups are used in series or parallel. I have TONS of these in my garage right now (don't ask)
I have a 72v, 60a 4x4 Dell setup right now I use for charging, 2s2p config. Finding 48v server PS's is more difficult. Something even I've been on the lookout for along time. 12v- 1000+w units are the common.
Companies THROW THEM AWAY!! :shock:
If you need ref material, this thread on RCgroups is one of the best PERIOD for Server PS's.
http://www.rcgroups.com/forums/showthread.php?t=1292514
and I have thought about this.. I have enough units to make a Leaf charger... but it would be HUGE!
Also, different PS's can be controlled and have a wide swing, just need to look at RC groups and find which ones. But 48v in the end would work best
Issue with server PS's is that they have to be modified to be isolated. Opened up, case and board removed, ground removed from 120v line, and then PCB put back so it's not touching or grounded to the outside case. Technically floating but the 120v is isolated from the DC.
This needs to be done if groups are used in series or parallel. I have TONS of these in my garage right now (don't ask)
I have a 72v, 60a 4x4 Dell setup right now I use for charging, 2s2p config. Finding 48v server PS's is more difficult. Something even I've been on the lookout for along time. 12v- 1000+w units are the common.
Companies THROW THEM AWAY!! :shock:
If you need ref material, this thread on RCgroups is one of the best PERIOD for Server PS's.
http://www.rcgroups.com/forums/showthread.php?t=1292514
and I have thought about this.. I have enough units to make a Leaf charger... but it would be HUGE!
Also, different PS's can be controlled and have a wide swing, just need to look at RC groups and find which ones. But 48v in the end would work best
garygid
Well-known member
Thanks for the work, enthusiasm, and info.
While charging, battery voltages as low as 280 or so have
been reported for a Turtle-stopped battery pack.
A voltage of 420 plus or minus a bit might be required
for charging at higher currents.
However, during the "mating dance", it appears that the
output voltage must climb to the max (maybe 450 is sufficient,
but we do not know the exact number).
Also, the output voltage must fall to within a few volts of zero,
but again we do not know the exact number).
Presently we are using an AVR-CAN development board because
it has one CAN interface completely included, and we have used
this board in the GID-Meter and for Logging the CAN bus messages.
We watch the output voltage and current, and monitor internal
temperatures. Our primary control of the Power Supply is by
providing the Pulse Width Modulated (PWM) control signal to
the output regulating switcher stage of the Power Supply.
However, we need finer control of the PWM signal to better
"regulate" this particular supply. So, we are considering
moving to the new 84 MHz Arduino Due.
While charging, battery voltages as low as 280 or so have
been reported for a Turtle-stopped battery pack.
A voltage of 420 plus or minus a bit might be required
for charging at higher currents.
However, during the "mating dance", it appears that the
output voltage must climb to the max (maybe 450 is sufficient,
but we do not know the exact number).
Also, the output voltage must fall to within a few volts of zero,
but again we do not know the exact number).
Presently we are using an AVR-CAN development board because
it has one CAN interface completely included, and we have used
this board in the GID-Meter and for Logging the CAN bus messages.
We watch the output voltage and current, and monitor internal
temperatures. Our primary control of the Power Supply is by
providing the Pulse Width Modulated (PWM) control signal to
the output regulating switcher stage of the Power Supply.
However, we need finer control of the PWM signal to better
"regulate" this particular supply. So, we are considering
moving to the new 84 MHz Arduino Due.
klapauzius
Well-known member
Assuming that 40 x 12 v would do the trick, maybe except for un-grounding the boards, nothing more needs to be done on the power supplies. If you put them in series, you could vary output voltage in 12 v increments with a multiplexer and a bunch of IGBTs?
Hi Jason, thanks for chiming in here. I see that you are in Sylmar - I was visiting North Hills and Hollywood earlier this month!
Anyway - I read up a bunch of RC group discussions about modding these supplies already and have a couple of those 12V beasties, that is why I started with those. But finding that they need to be opened up anyway, I started searching for 48V units which are still common and cheap, also to reduce the physical size - as I said: about a cubic foot for 12kW.
Regarding grounding: for safety and other reasons I *strongly* prefer to keep the case grounded, so indeed that means that I have to sever the grounding from the negative output, but I see this as far safer (albeit more labor intense) than the way most RC guys recommend, to cut the AC ground (externally) and live with the cases at different voltage. Heck, I have even seen 3D-printed cases to keep the supplies from touching. I want none of that at 400V output!
I need to wait for the 48V units to comment further on how easy or difficult it will be to control and to isolate them.
Anyway - I read up a bunch of RC group discussions about modding these supplies already and have a couple of those 12V beasties, that is why I started with those. But finding that they need to be opened up anyway, I started searching for 48V units which are still common and cheap, also to reduce the physical size - as I said: about a cubic foot for 12kW.
Regarding grounding: for safety and other reasons I *strongly* prefer to keep the case grounded, so indeed that means that I have to sever the grounding from the negative output, but I see this as far safer (albeit more labor intense) than the way most RC guys recommend, to cut the AC ground (externally) and live with the cases at different voltage. Heck, I have even seen 3D-printed cases to keep the supplies from touching. I want none of that at 400V output!
I need to wait for the 48V units to comment further on how easy or difficult it will be to control and to isolate them.
OK thanks Gary, that helps and those numbers will be the target to include in the 12kW design.garygid said:
Note that it would be no problem at all to go another 50V higher just for the initial "mating dance"
as that would be solved with adding a small 48V supply that I have plenty of.
My intend was indeed to allow the power supply to go to full voltage and zero voltage (no current
in both cases) as well as the normal charging voltages, which I suspect are *entirely* determined by
the Leaf battery voltage and the charger is only *current* controlled.
I have not yet been able to decode the loggings you made from the CAN bus, so it is only a guesstimate.
That sounds like an interesting board. I see that Olimex offers it for 25 Euro while Mouser carries it for $35.garygid said:
I will need to look into that board to see if the 35 GPIO that it has also offer DAC and ADC to control the
power supply to build it into a charger without further additional controllers or interfaces, otherwise a
PWM output and comparator are needed to add the necessary glue to the power supply.
Owww, no wonder that you need a lot of processing power - that is a hard way to solve this problem.garygid said:
From your reference to the Arduino I conclude that you use it to control the power supply?
Or is the CAN interface an Arduino shield and controlled by the Arduino also?
I get the impression from the Olimex website that it is a stand-alone controller,
you are probably using it as a simple CAN-to-RS232 converter for logging the CAN bus,
but I think the logic to control the power supply should be simple enough to add that
code to the CAN bus board and make it a stand-along board, so you only have the
8 (or 9) 48V power supplies with this CAN board in a box and the only two wires are
the 240V AC input and the Leaf fast charge cord, which contains the DC bus to the
battery and the CAN bus. Possibly a user selectable max power setting in case he
does not have the full 12kW 240V 50A, but for example needs to run half power because
he only has a dryer outlet (30A) or plugs into a J1772 that only allows 24 or 30A.
BTW, Gary I am looking into the AT90CAN128 devt boards and they look promising in available GPIO, ADC and such.
Available memory should be no problem to make a self-contained charger controller.
How do you program them - via the serial port that the AVR-CAN board has or via JTAG?
I am not yet familiar enough with this controller to know what is possible and want to avoid buying a complete
devt system when all that is needed is a serial cable.
Regarding controlling the charger output: Is it accurate enough to control the charger current in steps of say 0.1A
and let the output voltage fall where the battery's SoC happens to be? I can easily read back the voltage with the controller's ADC but I doubt that it needs controlling, although either voltage and/or current is possible. The voltage control will likely be limited to: OFF (zero volts) and the range from about 240-440V or whatever upper limit is required.
Another question: isolation check does that mean that the DC output of the charger must be connected to *nothing* or is it OK to have the negative output connected to earth ground? In the final setup once the controller is stand-alone it makes not much difference, but if the Leaf requires that the DC is fully isolated, the debugging of the controller must also be floating which is a bit of a hassle (using a laptop and 2-prong cords, risking a shock whenever touching metal on the setup) but it is possible.
Tomorrow morning I will pick up a bunch of 48V supplies, so I can check if they can be modded for wide output control.
Available memory should be no problem to make a self-contained charger controller.
How do you program them - via the serial port that the AVR-CAN board has or via JTAG?
I am not yet familiar enough with this controller to know what is possible and want to avoid buying a complete
devt system when all that is needed is a serial cable.
Regarding controlling the charger output: Is it accurate enough to control the charger current in steps of say 0.1A
and let the output voltage fall where the battery's SoC happens to be? I can easily read back the voltage with the controller's ADC but I doubt that it needs controlling, although either voltage and/or current is possible. The voltage control will likely be limited to: OFF (zero volts) and the range from about 240-440V or whatever upper limit is required.
Another question: isolation check does that mean that the DC output of the charger must be connected to *nothing* or is it OK to have the negative output connected to earth ground? In the final setup once the controller is stand-alone it makes not much difference, but if the Leaf requires that the DC is fully isolated, the debugging of the controller must also be floating which is a bit of a hassle (using a laptop and 2-prong cords, risking a shock whenever touching metal on the setup) but it is possible.
Tomorrow morning I will pick up a bunch of 48V supplies, so I can check if they can be modded for wide output control.
garygid
Well-known member
I use WinAVR (portable) to program and compile to hex, then
avrdude (included with WinAVR) through a JTAG programmer
(about $15) for flashing the AVR-CAN card.
However, the PWM included in the AT91CAN128 does not have fine enough
control of the pulse width, the way we are using it, to control the output
current well. One bit change in the pulse width can vary the current by
several amps.
I expect that controlling the output current to plus or minus one amp would
be suficient, and that control to 0.1 amps is unnecessary.
The negative output grounded to earth is not isolated.
avrdude (included with WinAVR) through a JTAG programmer
(about $15) for flashing the AVR-CAN card.
However, the PWM included in the AT91CAN128 does not have fine enough
control of the pulse width, the way we are using it, to control the output
current well. One bit change in the pulse width can vary the current by
several amps.
I expect that controlling the output current to plus or minus one amp would
be suficient, and that control to 0.1 amps is unnecessary.
The negative output grounded to earth is not isolated.
JasonA
Well-known member
I actually got 4 of these PS's from a buddy yest. These are pretty nice as they are low-profile and put out a nice bit of juice! I'm finding out more and more about them but making a stack of them would be very easy..
Voltage swing is from 43-65v.. plenty of amps to start.. Like I find with all these units, they are conservatively rated and can pull more!
Voltage swing is from 43-65v.. plenty of amps to start.. Like I find with all these units, they are conservatively rated and can pull more!
Hi Jason,
Nice!
This morning I picked up the entire inventory of 1500W power supplies from a local refurbisher, who pulls these from servers in bunches of 4 of those beasts (Sun servers).
The power is so high that they are only allow 200-240V input, since it would make no sense to run them from 110V.
Output is 48V, 30A but can most likely be cranked up to at least 55V. I will find out soon.
Nice!
This morning I picked up the entire inventory of 1500W power supplies from a local refurbisher, who pulls these from servers in bunches of 4 of those beasts (Sun servers).
The power is so high that they are only allow 200-240V input, since it would make no sense to run them from 110V.
Output is 48V, 30A but can most likely be cranked up to at least 55V. I will find out soon.
garygid
Well-known member
Cor,
Can you locate a schematic for these supplies?
That would tell us a lot.
Maybe the refurbisher has a circuit diagram?
We await your discoveries.
Please be careful, these can be harmful, even deadly.
Cheers, Gary
Can you locate a schematic for these supplies?
That would tell us a lot.
Maybe the refurbisher has a circuit diagram?
We await your discoveries.
Please be careful, these can be harmful, even deadly.
Cheers, Gary
JasonA
Well-known member
Funny you say that, I used to toss old Sparc boxes and servers away! :lol:
Now I see HP and Dell's going bye bye all the time... but again, they are mostly 12v units.. I have soooo many I don't know what to do with them any more that I just stopped!
Now I see HP and Dell's going bye bye all the time... but again, they are mostly 12v units.. I have soooo many I don't know what to do with them any more that I just stopped!
FalconFour
Well-known member
Just curious, as I've noticed it hasn't been really mentioned here yet. What about the aftermarket Leaf charger add-on, a Brusa NLG513, from this thread? http://www.mynissanleaf.com/viewtopic.php?f=37&t=12323" onclick="window.open(this.href);return false; - that seems like the perfect power supply/charger, if just a little on the high end of the price scale. Interesting to note the "parallel connection options" noted on the specs page... seems very well documented!
edit: And judging from your sig JasonA, I guess you're familiar with the Brusa. :lol:
edit: And judging from your sig JasonA, I guess you're familiar with the Brusa. :lol:
garygid
Well-known member
At $2500 each, 4 of them would be $10,000 for about 12 kW,
which is about 5 times our hoped-for goal.
But, they might be sufficiently controllable to use in parallel.
I suspect that it is a lot more difficult to control power supplies
that are hooked in series.
which is about 5 times our hoped-for goal.
But, they might be sufficiently controllable to use in parallel.
I suspect that it is a lot more difficult to control power supplies
that are hooked in series.
FalconFour
Well-known member
Not for production use, but I think the Brusa would be good for testing the protocol and how to design the basic system. Then step up to a custom-made solution with much higher capacity once we've got the Leaf doing L2-rate charging with the QC port.
I highly doubt Nissan designed an actual lock-out into the software to prevent a QC from being operable below a certain rate, as there'd be no business reason to do so... if a QC tells the thing to go at 7 amps, it's got no reason to say "no thx"... this lower-rate charger could be used to develop the basic protocol, then maybe step it up with multiple Brusas in parallel for the full speed. It might not be cost-effective on the much larger (30kw+) installs, but it'd serve as a good development platform to start with...
I highly doubt Nissan designed an actual lock-out into the software to prevent a QC from being operable below a certain rate, as there'd be no business reason to do so... if a QC tells the thing to go at 7 amps, it's got no reason to say "no thx"... this lower-rate charger could be used to develop the basic protocol, then maybe step it up with multiple Brusas in parallel for the full speed. It might not be cost-effective on the much larger (30kw+) installs, but it'd serve as a good development platform to start with...
garygid
Well-known member
OK, let's investigate a Bursa-based micro-QC solution,
even if not terribly practical at the moment.
Let's begin by getting answers to some basic questions...
The Bursa has an isolated HV output, right?
240v in, and air cooled, with around 3.3 kW out?
The only control of the is CAN, right?
Or, are there other control lines as well?
Can the output voltage be ramped from zero to max
(or a few volts to around 450 volts), and then ramped back
down, under no-load conditions, using its CAN control?
How fast can the current be ramped up, under CAN control?
The car asks for 2 amps more each 1/10 second,
but the QC does not have to follow exactly.
However, the car does not tolerate too much delay.
Can the output current be shut off abruptly, with the
Output voltage falling quickly?
How well regulated is the output current?
What noise is present in the HV output.
Is there any reverse-polarity protection built in?
Other than HV and CAN, does it have any other outputs?
The CAN and any other control or status signals
are referenced to protective (earth) Ground?
How often can the Bursa be "commanded"?
If we get past these questions successfully, then
the details of commands and status are next,
followed by the needed connectors, cable,
and a QC plug.
Do we need two CAN buses, or would the car's
QC CAN bus work to connect to both our Controller
and the Bursa.
Yes, having two CAN buses would be best.
Right now, we just support one CAN bus, but we are
attempting to migrate to the Arduino Due, which
seems to be able to support two CAN buses.
However, the Due is so new, that it appears that
this will be a significant effort. More Due-players
are welcome.
Since monitoring and logging are a big part of
testing and making sure that the car is "satisfied",
we will want to include that function in the
design of the new controller, as we have done
presently with the AVR-CAN board.
If we had a satisfactory set of female contacts,
we could make an independent logger, a female in,
male out plug to insert into the car's QC socket,
to log the interface activity.
Yes, in monitoring the "internals" of the Power Supply
with our Controller, we can put useful data on a CAN
message on the QC bus (that the car appears to ignore),
so that an "interface" logger could get "Special Debug"
information, when enabled and available.
Tag, back to you all.
even if not terribly practical at the moment.
Let's begin by getting answers to some basic questions...
The Bursa has an isolated HV output, right?
240v in, and air cooled, with around 3.3 kW out?
The only control of the is CAN, right?
Or, are there other control lines as well?
Can the output voltage be ramped from zero to max
(or a few volts to around 450 volts), and then ramped back
down, under no-load conditions, using its CAN control?
How fast can the current be ramped up, under CAN control?
The car asks for 2 amps more each 1/10 second,
but the QC does not have to follow exactly.
However, the car does not tolerate too much delay.
Can the output current be shut off abruptly, with the
Output voltage falling quickly?
How well regulated is the output current?
What noise is present in the HV output.
Is there any reverse-polarity protection built in?
Other than HV and CAN, does it have any other outputs?
The CAN and any other control or status signals
are referenced to protective (earth) Ground?
How often can the Bursa be "commanded"?
If we get past these questions successfully, then
the details of commands and status are next,
followed by the needed connectors, cable,
and a QC plug.
Do we need two CAN buses, or would the car's
QC CAN bus work to connect to both our Controller
and the Bursa.
Yes, having two CAN buses would be best.
Right now, we just support one CAN bus, but we are
attempting to migrate to the Arduino Due, which
seems to be able to support two CAN buses.
However, the Due is so new, that it appears that
this will be a significant effort. More Due-players
are welcome.
Since monitoring and logging are a big part of
testing and making sure that the car is "satisfied",
we will want to include that function in the
design of the new controller, as we have done
presently with the AVR-CAN board.
If we had a satisfactory set of female contacts,
we could make an independent logger, a female in,
male out plug to insert into the car's QC socket,
to log the interface activity.
Yes, in monitoring the "internals" of the Power Supply
with our Controller, we can put useful data on a CAN
message on the QC bus (that the car appears to ignore),
so that an "interface" logger could get "Special Debug"
information, when enabled and available.
Tag, back to you all.
JasonA
Well-known member
This is the current pricing avail to us.. for the Air Cooled units..garygid said:
5-9 $1,945 each + 270 cables = $2,215
10-24 $1,881 each + 270 cables = $2,151
Going to that degree..and what we do, I would not recommend buying the factory cables and just make our own (pretty easy to do, I have some connectors and they're not hard to work with)..
They are BEAUTIFULLY controllable to almost the Nth degree via CAN.. and you can read almost anything that is going on with them as well..
