Loading ...
Sorry, an error occurred while loading the content.
 

Introduction

Expand Messages
  • _lr_
    Hi, folks, I m a new member, located in Cape Cod, MA. I m just climbing the learning curve on EVs, and the idea of something that was engineered in a factory
    Message 1 of 20 , Oct 3, 2012
      Hi, folks,

      I'm a new member, located in Cape Cod, MA. I'm just climbing the learning curve on EVs, and the idea of something that was engineered in a factory instead of someone' garage is pretty appealing, so Solectria is at the top of my list. Goals are to educate myself about options and eventually take ownership of one, after which I'll surely be looking for help in keeping it safe and roadworthy.
    • electricporschedotca
      Hello everyone, I recently converted an 86 Porsche 944 to electric drive as well as helped a friend upgrade his Solectria Force to Lithium (56x CALB CA60FI)
      Message 2 of 20 , Nov 22, 2012
        Hello everyone,

        I recently converted an '86 Porsche 944 to electric drive as well as helped a friend upgrade his Solectria Force to Lithium (56x CALB CA60FI) He still has some small issues with the Force as well as another Force in need of a lithium transplant. I will be helping him and learning more about the car. So far working on it has been a pleasure and the care was very well designed and built!

        I look forward to reading a lot and posting what we learn, there seems to be a wealth of information in this group.

        Robin Wainwright from Calgary Canada
      • keith
        Hello All, My name is Keith and I am from Chattanooga, TN. I recently purchased a 96 Force from my local electric co-op. It has the 110v chargers, 156v Deka
        Message 3 of 20 , Mar 20, 2013
          Hello All,

          My name is Keith and I am from Chattanooga, TN. I recently purchased a 96 Force from my local electric co-op. It has the 110v chargers, 156v Deka Gel PBA's, and only 7k miles. I have been working on getting this car operational and am very close. The information in this group is invaluable. I will post when I get her on the road. Keith
        • ilya_666
          Hello everyone. I have just acquired a DMOC 645 and DMGI 645 controller/inverter currently configured as a grid tie inverter and I want to drive a Remy HVH 250
          Message 4 of 20 , Aug 15, 2016
            Hello everyone. I have just acquired a DMOC 645 and DMGI 645 controller/inverter currently configured as a grid tie inverter and I want to drive a Remy HVH 250 in a hybrid conversion. The person I got the controllers from said there are some smart people on here who might just point me to exactly what I need rather than me figuring it all out from scratch so here goes! Have a tentative plan for mechanicals and battery. Happy to share results. Any help or pointers gratefully received. Thank you for your time.
          • Wolf
            Is your HVH 250 motor AC induction or permanent magnet? Resolver or encoder? Also, I am interested in the grid tie inverter if you are not going to use it. :)
            Message 5 of 20 , Aug 15, 2016

              Is your HVH 250 motor AC induction or permanent magnet?

              Resolver or encoder?

              Also, I am interested in the grid tie inverter if you are not going to use it. :)

              Thanks,
              Wolf


              On Aug 15, 2016 11:38 AM, "ilya_666@... [solectria_ev]" <solectria_ev@yahoogroups.com> wrote:
               

              Hello everyone. I have just acquired a DMOC 645 and DMGI 645 controller/inverter currently configured as a grid tie inverter and I want to drive a Remy HVH 250 in a hybrid conversion. The person I got the controllers from said there are some smart people on here who might just point me to exactly what I need rather than me figuring it all out from scratch so here goes! Have a tentative plan for mechanicals and battery. Happy to share results. Any help or pointers gratefully received. Thank you for your time.

            • Collin Kidder
              It seems likely that the motor would be PMAC and use a resolver but you would have to confirm that. PMAC with resolver is certainly the default configuration
              Message 6 of 20 , Aug 16, 2016
                It seems likely that the motor would be PMAC and use a resolver but you would have to confirm that. PMAC with resolver is certainly the "default" configuration for those motors. They are available as induction motors upon special request. I kind of doubt such a thing occurred though. 

                So, assuming the guess of PMAC with resolver is correct - good luck. The DMOC645 doesn't support that configuration at all. The main control board in the DMOC645 has a place to put components to read a resolver but they're not populated and I don't know which exact parts they had thought to put there. The firmware flashed onto the controller also doesn't seem to look for a resolver at all and was never meant for PMAC. 

                This leads down the road to replacing the main control board in the DMOC with something else. Eldis, in theory, is some day going to release a motor controller board that is universal. You could remove the DMOC main board and replace it with that. Two issues - there is NOT a lot of room in a DMOC. The capacitors inside basically touch the connectors on the top of the main board. You don't have any extra vertical room at all to add a replacement board. So, hopefully his board is thin enough. Secondly, you'd need to figure out which pins do what in the DMOC so you can wire them up to the replacement Eldis board. Luckily I have basically all of the pin outs. The reason I have them is that I built my own motor controller board that I put into a DMOC645 in order to drive a PMAC motor that has an encoder and three hall effect sensors. So, I actually have a replacement board that will directly interface with the DMOC power section and all the connectors. But, that doesn't completely work for you as I'm using an encoder (begrudgingly) and haven't done any work to be able to read a resolver. Eldis' board presumably can read a resolver. Neither board seems to publicly exist right now. That doesn't exactly help you either, sorry.

                But, both Wolf and I know quite a bit about the internals of a DMOC so hopefully we can help you make this a reality. I guess the question is, what is your skill level and what do you feel comfortable doing? Can you take pieces and make your own motor controller out of the parts? Do you need a completed solution you can just drop in? 

                If you want a completed solution look for Rinehart Motion controllers. A PM100 supports PMAC with a resolver. It should work fine for you and probably Rinehart has already worked out the proper parameters for driving that motor of yours. Sit down before you ask them the price. It's kind of expensive but it's done and it's very nice.

                On Mon, Aug 15, 2016 at 12:38 PM, ilya_666@... [solectria_ev] <solectria_ev@yahoogroups.com> wrote:
                 

                Hello everyone. I have just acquired a DMOC 645 and DMGI 645 controller/inverter currently configured as a grid tie inverter and I want to drive a Remy HVH 250 in a hybrid conversion. The person I got the controllers from said there are some smart people on here who might just point me to exactly what I need rather than me figuring it all out from scratch so here goes! Have a tentative plan for mechanicals and battery. Happy to share results. Any help or pointers gratefully received. Thank you for your time.


              • ilya_666
                Thank you! I believe I looked it up and the resolver has essentially sine and cosine signals. The motor markings have A722 540 34 88 and H17AU10A - 085827. I
                Message 7 of 20 , Aug 17, 2016
                  Thank you! I believe I looked it up and the resolver has essentially sine and cosine signals. The motor markings have A722 540 34 88 and H17AU10A - 085827. I think I found somewhere to look these up and will report when I find it again! As for skill level I spent a misspent youth with a soldering iron writing machine code but these days time is at more of a premium but I have a friend coming over this weekend with more time and aptitude to help me play! I saw the rinehart controller which is about$3-4K (which is now about 50% more to me than a couple of months ago before a few fellow countrymen pointlessly decided to render our currency worthless!) which would almost be plug and play and I got the DMOC for £150 expecting to spend a bit of time and tinkering to get it to work. I got the motor for £400. I'm looking for a solution which is a cost/effort compromise! I naively hoped a bit of config and an arduino/raspberry PI with a bit of code and a little bit of analogue electronics might fill in the gap. I'll get back with more info after the weekend after friend and I have done our investigations (just back from hol last night, got to work on current project setting up some variable frequency drives to drive moving solid cover and pumps for my solar/biomass heated pool project!). I REALLY appreciate the input from you and Wolf, just what I was hoping for. Thank you both very much. Ilya
                • Wolf
                  If you have time rather than money, and you are up for a challenge... ;) Get a resolver to digital converter (quadrature). Find a AC induction motor .par file
                  Message 8 of 20 , Aug 17, 2016

                    If you have time rather than money, and you are up for a challenge... ;)

                    Get a resolver to digital converter (quadrature).

                    Find a AC induction motor .par file that is close to the voltage/current of your motor.

                    Use ccShell to set your DMOC to the same number of counts per revolution for your RDC chip.

                    Now the fun part...

                    Set the magantization current to zero, (your rotor already has a magnetic field).

                    It will jump suddenly on start up if you don't do this.

                    Find the parameter(s) that controls slip, you want that to be zero all the time with a permanent magnet synchronous motor.

                    Zero out all the gains that control field weakening.

                    You want to be in volts/Hz mode for as long as possible.

                    Spin the motor under load, adjust the gains such that it works smoothly up to 30-40 mph... Then torque will fall off a cliff when the back EMF voltage equals your pack voltage...

                    Before this point is the RPM you want to field weakening to start kicking in...

                    You will then need to adjust the gains such that you gradually counteract the magnets in your rotor, and thus the back EMF.

                    This will let you run higher RPM with a gradual decline in torque.

                    Simple. ;)

                    And here is the hard part, there is not much documentation on what the parameters in ccChell do... You have to guess based on the description of the varible, and most are abbreviated...

                    And keep in mind that adjusting one parameter will affect the others.

                    If the motor is not in a vehicle, then make sure it is bolted down so it can't jump of the work bench. ;)

                    It can be done, it will just take a bunch of trial and error time.

                    Hope that helps,
                    Wolf


                    On Aug 17, 2016 2:48 AM, "ilya_666@... [solectria_ev]" <solectria_ev@yahoogroups.com> wrote:
                     

                    Thank you! I believe I looked it up and the resolver has essentially sine and cosine signals. The motor markings have A722 540 34 88 and H17AU10A - 085827. I think I found somewhere to look these up and will report when I find it again! As for skill level I spent a misspent youth with a soldering iron writing machine code but these days time is at more of a premium but I have a friend coming over this weekend with more time and aptitude to help me play! I saw the rinehart controller which is about$3-4K (which is now about 50% more to me than a couple of months ago before a few fellow countrymen pointlessly decided to render our currency worthless!) which would almost be plug and play and I got the DMOC for £150 expecting to spend a bit of time and tinkering to get it to work. I got the motor for £400. I'm looking for a solution which is a cost/effort compromise! I naively hoped a bit of config and an arduino/raspberry PI with a bit of code and a little bit of analogue electronics might fill in the gap. I'll get back with more info after the weekend after friend and I have done our investigations (just back from hol last night, got to work on current project setting up some variable frequency drives to drive moving solid cover and pumps for my solar/biomass heated pool project!). I REALLY appreciate the input from you and Wolf, just what I was hoping for. Thank you both very much. Ilya

                  • Charles Bliss
                    Or if you have money and less time, consider getting a Curtis 1239e controller with OS 30 or one that can have the firmware upgraded to OS 30, which supports
                    Message 9 of 20 , Aug 17, 2016

                      Or if you have money and less time, consider getting a Curtis 1239e controller with  OS 30 or one that can have the firmware upgraded to OS 30, which supports PMAC motors.  It has an auto tune algorithm which makes the setup pretty easy.  OS 30 does not yet support field weakening but that shouldn't present a problem.  It will in the future as a firmware upgrade.  It expects a Sine/Cosine encoder.


                      On 8/17/2016 5:14 AM, Wolf wolf@... [solectria_ev] wrote:
                       

                      If you have time rather than money, and you are up for a challenge... ;)

                      Get a resolver to digital converter (quadrature).



                    • Collin Kidder
                      That s a really awesome, geeky idea. But... I m not sure whether that ll totally work. The problem is that a resolver to encoder converter has no means of
                      Message 10 of 20 , Aug 17, 2016
                        That's a really awesome, geeky idea. But...

                        I'm not sure whether that'll totally work. The problem is that a resolver to encoder converter has no means of conveying absolute position unless it also outputs an index pulse. So far as I know the DMOC has no prevision for input of an index pulse and doesn't use it. Because of this you won't immediately have a good idea of where exactly the motor really is. The DMOC has as input A and B encoder channels. So, all you can really do is tell it how far the shaft has rotated, not where it truly is in the 360 degrees of rotation possible. Thus, upon start up it is entirely possible to be far off on the rotation angle. A PMAC motor needs the output waveform to be in quadrature to the magnetic field of the permanent magnets. Chances are, the DMOC firmware doesn't have any code to attempt to rectify this situation. On the bench with the wheel up in the air this isn't much of an issue. If you're off a little the wheel could "jump" to the proper position and be OK. If you're off 180 degrees the motor will run backward. 

                        The DMOC does output Iq and Id I think. If you try the idea of making the stock firmware attempt to drive the motor then monitor Id and Iq and see if you're getting much Id. Ideally you will have zero or very low Id and Vd and all the power will be in Iq and Vq. I think you'll find that depending on where the motor last stopped and if anyone spun it with power off you'll get very different results. 

                        But, hey, it might be worth a shot! Maybe I'm wrong about how tough it'll be to keep the shaft angle correct.

                        Here's a very fast run through of field oriented control and the terminology to help you find the variables you need:

                        FOC starts by figuring out the current at each phase A, B, and C. You only need to measure 2 and you can figure out the third so that's what the DMOC does. From those currents you generate Ia and Ib which are quadrature currents in the time variant domain. Ia and Ib are rotated according to the current shaft angle to yield Iq and Id which are time invariant. Id = Direct current (magnetizing) and Iq = Quadrature current (torque producing) Those two currents are compared against reference values that we wanted them to be. Those references could be called IdRef and IqRef. Those are the targets. The targets are fed into a PID controller for each. PID (proportional, integral, differential) controllers have three parameters: P, I, and D. Usually FOC does not use D so it might either be missing or set to zero. Generally you'll have P be a much larger value than I. The PID controllers output Vd and Vq which are quadrature voltages that will get fed back to tell the inverter how much power to provide in order to keep Id and Iq where the reference values are. The goal is to keep Id and Iq steady. The reference values don't change very rapidly. Anyway, with Vd and Vq we now can rotate backward to get to Va and Vb which are time variant quadrature voltages. Va and Vb are then transformed back to three phase voltages A, B, C that will get fed to the inverter's output transistors to power the motor.

                        So, Iq, Id, Ia, Ib, Va, Vb, Vd, Vd and their references. It's possible that there are more PID loops and you could even find IdRefRef or something of the sort just to add to the fun. In that case the reference is also generated by its own PID loop with its own control coefficients. 

                        As mentioned by both Wolf and I, what you want to do is minimize Id and Vd while you're not in field weakening mode. Field weakening a PMAC motor actually involves generating a reverse magnetizing current (well, 180 degrees out of phase) that counteracts the permanent magnets. Otherwise the back EMF will literally be so high that you're unable to accelerate as the motor is generating pretty much the same voltage you're feeding it but 180 degrees out of phase which cancels to 0 and you get no power. Because of that you'll never actually reach 0 volts in practice - just a low enough voltage to hold you with as much torque as it takes to keep going that same speed.

                        The DMOC wants parameters like stator resistance, inductance, etc. Those are for calculating the proper slip angle for an induction motor. I don't think those parameters are going to mean anything to you. But, it also has a million other parameters that I don't know the meaning of. It's probably going to be a lot of fun to play with all of those. Make sure you save the parameter file before screwing around and then frequently as you're working.

                        On Wed, Aug 17, 2016 at 8:14 AM, Wolf wolf@... [solectria_ev] <solectria_ev@yahoogroups.com> wrote:
                         

                        If you have time rather than money, and you are up for a challenge... ;)

                        Get a resolver to digital converter (quadrature).

                        Find a AC induction motor .par file that is close to the voltage/current of your motor.

                        Use ccShell to set your DMOC to the same number of counts per revolution for your RDC chip.

                        Now the fun part...

                        Set the magantization current to zero, (your rotor already has a magnetic field).

                        It will jump suddenly on start up if you don't do this.

                        Find the parameter(s) that controls slip, you want that to be zero all the time with a permanent magnet synchronous motor.

                        Zero out all the gains that control field weakening.

                        You want to be in volts/Hz mode for as long as possible.

                        Spin the motor under load, adjust the gains such that it works smoothly up to 30-40 mph... Then torque will fall off a cliff when the back EMF voltage equals your pack voltage...

                        Before this point is the RPM you want to field weakening to start kicking in...

                        You will then need to adjust the gains such that you gradually counteract the magnets in your rotor, and thus the back EMF.

                        This will let you run higher RPM with a gradual decline in torque.

                        Simple. ;)

                        And here is the hard part, there is not much documentation on what the parameters in ccChell do... You have to guess based on the description of the varible, and most are abbreviated...

                        And keep in mind that adjusting one parameter will affect the others.

                        If the motor is not in a vehicle, then make sure it is bolted down so it can't jump of the work bench. ;)

                        It can be done, it will just take a bunch of trial and error time.

                        Hope that helps,
                        Wolf


                        On Aug 17, 2016 2:48 AM, "ilya_666@... [solectria_ev]" <solectria_ev@yahoogroups.com> wrote:
                         

                        Thank you! I believe I looked it up and the resolver has essentially sine and cosine signals. The motor markings have A722 540 34 88 and H17AU10A - 085827. I think I found somewhere to look these up and will report when I find it again! As for skill level I spent a misspent youth with a soldering iron writing machine code but these days time is at more of a premium but I have a friend coming over this weekend with more time and aptitude to help me play! I saw the rinehart controller which is about$3-4K (which is now about 50% more to me than a couple of months ago before a few fellow countrymen pointlessly decided to render our currency worthless!) which would almost be plug and play and I got the DMOC for £150 expecting to spend a bit of time and tinkering to get it to work. I got the motor for £400. I'm looking for a solution which is a cost/effort compromise! I naively hoped a bit of config and an arduino/raspberry PI with a bit of code and a little bit of analogue electronics might fill in the gap. I'll get back with more info after the weekend after friend and I have done our investigations (just back from hol last night, got to work on current project setting up some variable frequency drives to drive moving solid cover and pumps for my solar/biomass heated pool project!). I REALLY appreciate the input from you and Wolf, just what I was hoping for. Thank you both very much. Ilya


                      • ilya_666
                        That is very helpful. Sorry for delay in acknowledging, discovered rather painfully earlier today that the capacitors in the filter for another variable
                        Message 11 of 20 , Aug 17, 2016
                          That is very helpful.  Sorry for delay in acknowledging, discovered rather painfully earlier today that the capacitors in the filter for another variable frequency drive I'm configuring at the moment hold their charge longer than I anticipated!

                          Sent from my iPhone; please excuse typos and arbitrary spelling "corrections"

                          On 17 Aug 2016, at 13:14, Wolf wolf@... [solectria_ev] <solectria_ev@yahoogroups.com> wrote:

                           

                          If you have time rather than money, and you are up for a challenge... ;)

                          Get a resolver to digital converter (quadrature).

                          Find a AC induction motor .par file that is close to the voltage/current of your motor.

                          Use ccShell to set your DMOC to the same number of counts per revolution for your RDC chip.

                          Now the fun part...

                          Set the magantization current to zero, (your rotor already has a magnetic field).

                          It will jump suddenly on start up if you don't do this.

                          Find the parameter(s) that controls slip, you want that to be zero all the time with a permanent magnet synchronous motor.

                          Zero out all the gains that control field weakening.

                          You want to be in volts/Hz mode for as long as possible.

                          Spin the motor under load, adjust the gains such that it works smoothly up to 30-40 mph... Then torque will fall off a cliff when the back EMF voltage equals your pack voltage...

                          Before this point is the RPM you want to field weakening to start kicking in...

                          You will then need to adjust the gains such that you gradually counteract the magnets in your rotor, and thus the back EMF.

                          This will let you run higher RPM with a gradual decline in torque.

                          Simple. ;)

                          And here is the hard part, there is not much documentation on what the parameters in ccChell do... You have to guess based on the description of the varible, and most are abbreviated...

                          And keep in mind that adjusting one parameter will affect the others.

                          If the motor is not in a vehicle, then make sure it is bolted down so it can't jump of the work bench. ;)

                          It can be done, it will just take a bunch of trial and error time.

                          Hope that helps,
                          Wolf


                          On Aug 17, 2016 2:48 AM, "ilya_666@... [solectria_ev]" <solectria_ev@yahoogroups.com> wrote:
                           

                          Thank you! I believe I looked it up and the resolver has essentially sine and cosine signals. The motor markings have A722 540 34 88 and H17AU10A - 085827. I think I found somewhere to look these up and will report when I find it again! As for skill level I spent a misspent youth with a soldering iron writing machine code but these days time is at more of a premium but I have a friend coming over this weekend with more time and aptitude to help me play! I saw the rinehart controller which is about$3-4K (which is now about 50% more to me than a couple of months ago before a few fellow countrymen pointlessly decided to render our currency worthless!) which would almost be plug and play and I got the DMOC for £150 expecting to spend a bit of time and tinkering to get it to work. I got the motor for £400. I'm looking for a solution which is a cost/effort compromise! I naively hoped a bit of config and an arduino/raspberry PI with a bit of code and a little bit of analogue electronics might fill in the gap. I'll get back with more info after the weekend after friend and I have done our investigations (just back from hol last night, got to work on current project setting up some variable frequency drives to drive moving solid cover and pumps for my solar/biomass heated pool project!). I REALLY appreciate the input from you and Wolf, just what I was hoping for. Thank you both very much. Ilya

                        • majorwilson
                          Ouch! Sent on the new Sprint Network from my Samsung Galaxy S®4 ... From: ilya_666@yahoo.com [solectria_ev] Date: 08/17/2016
                          Message 12 of 20 , Aug 17, 2016
                            Ouch!



                            Sent on the new Sprint Network from my Samsung Galaxy S®4


                            -------- Original message --------
                            From: "ilya_666@... [solectria_ev]" <solectria_ev@yahoogroups.com>
                            Date: 08/17/2016 3:55 PM (GMT-07:00)
                            To: solectria_ev@yahoogroups.com
                            Subject: Re: [solectria_ev] Introduction

                             

                            That is very helpful.  Sorry for delay in acknowledging, discovered rather painfully earlier today that the capacitors in the filter for another variable frequency drive I'm configuring at the moment hold their charge longer than I anticipated!

                            Sent from my iPhone; please excuse typos and arbitrary spelling "corrections"

                            On 17 Aug 2016, at 13:14, Wolf wolf@... [solectria_ev] <solectria_ev@yahoogroups.com> wrote:

                             

                            If you have time rather than money, and you are up for a challenge... ;)

                            Get a resolver to digital converter (quadrature).

                            Find a AC induction motor .par file that is close to the voltage/current of your motor.

                            Use ccShell to set your DMOC to the same number of counts per revolution for your RDC chip.

                            Now the fun part...

                            Set the magantization current to zero, (your rotor already has a magnetic field).

                            It will jump suddenly on start up if you don't do this.

                            Find the parameter(s) that controls slip, you want that to be zero all the time with a permanent magnet synchronous motor.

                            Zero out all the gains that control field weakening.

                            You want to be in volts/Hz mode for as long as possible.

                            Spin the motor under load, adjust the gains such that it works smoothly up to 30-40 mph... Then torque will fall off a cliff when the back EMF voltage equals your pack voltage...

                            Before this point is the RPM you want to field weakening to start kicking in...

                            You will then need to adjust the gains such that you gradually counteract the magnets in your rotor, and thus the back EMF.

                            This will let you run higher RPM with a gradual decline in torque.

                            Simple. ;)

                            And here is the hard part, there is not much documentation on what the parameters in ccChell do... You have to guess based on the description of the varible, and most are abbreviated...

                            And keep in mind that adjusting one parameter will affect the others.

                            If the motor is not in a vehicle, then make sure it is bolted down so it can't jump of the work bench. ;)

                            It can be done, it will just take a bunch of trial and error time.

                            Hope that helps,
                            Wolf


                            On Aug 17, 2016 2:48 AM, "ilya_666@... [solectria_ev]" <solectria_ev@yahoogroups.com> wrote:
                             

                            Thank you! I believe I looked it up and the resolver has essentially sine and cosine signals. The motor markings have A722 540 34 88 and H17AU10A - 085827. I think I found somewhere to look these up and will report when I find it again! As for skill level I spent a misspent youth with a soldering iron writing machine code but these days time is at more of a premium but I have a friend coming over this weekend with more time and aptitude to help me play! I saw the rinehart controller which is about$3-4K (which is now about 50% more to me than a couple of months ago before a few fellow countrymen pointlessly decided to render our currency worthless!) which would almost be plug and play and I got the DMOC for £150 expecting to spend a bit of time and tinkering to get it to work. I got the motor for £400. I'm looking for a solution which is a cost/effort compromise! I naively hoped a bit of config and an arduino/raspberry PI with a bit of code and a little bit of analogue electronics might fill in the gap. I'll get back with more info after the weekend after friend and I have done our investigations (just back from hol last night, got to work on current project setting up some variable frequency drives to drive moving solid cover and pumps for my solar/biomass heated pool project!). I REALLY appreciate the input from you and Wolf, just what I was hoping for. Thank you both very much. Ilya

                          • Collin Kidder
                            Yikes... It s pretty standard to have around 10,000uF worth of capacitance and right around 360-400V. If my quick calculations are correct, that s in the
                            Message 13 of 20 , Aug 17, 2016
                              Yikes... It's pretty standard to have around 10,000uF worth of capacitance and right around 360-400V. If my quick calculations are correct, that's in the neighborhood of 650 joules of stored energy. In theory, should you drop a wrench and short such a capacitor bank and it discharges in 10 milliseconds you'd be getting only an average of 65 kilowatts of power over that 10 millisecond period. That's an exciting hundredth of a second for your wrench. Heaven help you if the discharge time can creep down to a millisecond or less.

                              That's why most every controller I've ever seen has a bleeder resistor in it (around 10k ohms) that bleeds off the voltage at a reasonable rate. But, the RC time is something like 100 seconds (10000uf * 10k ohms) so it takes 300 seconds or 5 minutes to completely drain off. What I do is keep a spare precharge resistor handy. I think I'm using a 40 or 50 ohm resistor. That precharges rapidly (10000uf * 50 ohms is 1/2 second RC time which means a full charge is basically 1.5 seconds) which means it also decharges rapidly. If you need to render an inverter safe then disconnect it from high voltage and short the + and - terminals with a precharge resistor. But, if you're going to be holding onto that resistor then get a nice, big 100w rated one. You don't want to be holding onto a 4 watt resistor when you do this.

                              Anyway, hopefully you're OK and the bright spot in your vision has faded by now.

                              On Wed, Aug 17, 2016 at 6:55 PM, ilya_666@... [solectria_ev] <solectria_ev@yahoogroups.com> wrote:
                               

                              That is very helpful.  Sorry for delay in acknowledging, discovered rather painfully earlier today that the capacitors in the filter for another variable frequency drive I'm configuring at the moment hold their charge longer than I anticipated!

                              Sent from my iPhone; please excuse typos and arbitrary spelling "corrections"

                              On 17 Aug 2016, at 13:14, Wolf wolf@... [solectria_ev] <solectria_ev@yahoogroups.com> wrote:

                               

                              If you have time rather than money, and you are up for a challenge... ;)

                              Get a resolver to digital converter (quadrature).

                              Find a AC induction motor .par file that is close to the voltage/current of your motor.

                              Use ccShell to set your DMOC to the same number of counts per revolution for your RDC chip.

                              Now the fun part...

                              Set the magantization current to zero, (your rotor already has a magnetic field).

                              It will jump suddenly on start up if you don't do this.

                              Find the parameter(s) that controls slip, you want that to be zero all the time with a permanent magnet synchronous motor.

                              Zero out all the gains that control field weakening.

                              You want to be in volts/Hz mode for as long as possible.

                              Spin the motor under load, adjust the gains such that it works smoothly up to 30-40 mph... Then torque will fall off a cliff when the back EMF voltage equals your pack voltage...

                              Before this point is the RPM you want to field weakening to start kicking in...

                              You will then need to adjust the gains such that you gradually counteract the magnets in your rotor, and thus the back EMF.

                              This will let you run higher RPM with a gradual decline in torque.

                              Simple. ;)

                              And here is the hard part, there is not much documentation on what the parameters in ccChell do... You have to guess based on the description of the varible, and most are abbreviated...

                              And keep in mind that adjusting one parameter will affect the others.

                              If the motor is not in a vehicle, then make sure it is bolted down so it can't jump of the work bench. ;)

                              It can be done, it will just take a bunch of trial and error time.

                              Hope that helps,
                              Wolf


                              On Aug 17, 2016 2:48 AM, "ilya_666@... [solectria_ev]" <solectria_ev@yahoogroups.com> wrote:
                               

                              Thank you! I believe I looked it up and the resolver has essentially sine and cosine signals. The motor markings have A722 540 34 88 and H17AU10A - 085827. I think I found somewhere to look these up and will report when I find it again! As for skill level I spent a misspent youth with a soldering iron writing machine code but these days time is at more of a premium but I have a friend coming over this weekend with more time and aptitude to help me play! I saw the rinehart controller which is about$3-4K (which is now about 50% more to me than a couple of months ago before a few fellow countrymen pointlessly decided to render our currency worthless!) which would almost be plug and play and I got the DMOC for £150 expecting to spend a bit of time and tinkering to get it to work. I got the motor for £400. I'm looking for a solution which is a cost/effort compromise! I naively hoped a bit of config and an arduino/raspberry PI with a bit of code and a little bit of analogue electronics might fill in the gap. I'll get back with more info after the weekend after friend and I have done our investigations (just back from hol last night, got to work on current project setting up some variable frequency drives to drive moving solid cover and pumps for my solar/biomass heated pool project!). I REALLY appreciate the input from you and Wolf, just what I was hoping for. Thank you both very much. Ilya


                            • ilya_666
                              Hi, Friend arrived with Solectria. We tried to run motor off a spare 22kw unidrive he had but that was faulty, got as far as getting fans to power up and
                              Message 14 of 20 , Aug 20, 2016
                                Hi,
                                Friend arrived with Solectria. We tried to run motor off a spare 22kw unidrive he had but that was faulty, got as far as getting fans to power up and supply to DC bus then turned to Solectria using half the DC bus to power it.  Just looking for wiring info for Solectria, seen it needs 12v power enable on 35 pin connector, best we can find are "typical" pinouts!  Can you point us to pinouts, will try to hook up RS232 to 8 pin and start to program.
                                Have thought that if we get rpm right and ramp up voltage/current we'll just get a hiccup as it hooks up even if we don't sort resolver. Or we can attach a digital optical encoder to shaft.

                                Sent from my iPhone; please excuse typos and arbitrary spelling "corrections"

                                On 17 Aug 2016, at 13:14, Wolf wolf@... [solectria_ev] <solectria_ev@yahoogroups.com> wrote:

                                 

                                If you have time rather than money, and you are up for a challenge... ;)

                                Get a resolver to digital converter (quadrature).

                                Find a AC induction motor .par file that is close to the voltage/current of your motor.

                                Use ccShell to set your DMOC to the same number of counts per revolution for your RDC chip.

                                Now the fun part...

                                Set the magantization current to zero, (your rotor already has a magnetic field).

                                It will jump suddenly on start up if you don't do this.

                                Find the parameter(s) that controls slip, you want that to be zero all the time with a permanent magnet synchronous motor.

                                Zero out all the gains that control field weakening.

                                You want to be in volts/Hz mode for as long as possible.

                                Spin the motor under load, adjust the gains such that it works smoothly up to 30-40 mph... Then torque will fall off a cliff when the back EMF voltage equals your pack voltage...

                                Before this point is the RPM you want to field weakening to start kicking in...

                                You will then need to adjust the gains such that you gradually counteract the magnets in your rotor, and thus the back EMF.

                                This will let you run higher RPM with a gradual decline in torque.

                                Simple. ;)

                                And here is the hard part, there is not much documentation on what the parameters in ccChell do... You have to guess based on the description of the varible, and most are abbreviated...

                                And keep in mind that adjusting one parameter will affect the others.

                                If the motor is not in a vehicle, then make sure it is bolted down so it can't jump of the work bench. ;)

                                It can be done, it will just take a bunch of trial and error time.

                                Hope that helps,
                                Wolf


                                On Aug 17, 2016 2:48 AM, "ilya_666@... [solectria_ev]" <solectria_ev@yahoogroups.com> wrote:
                                 

                                Thank you! I believe I looked it up and the resolver has essentially sine and cosine signals. The motor markings have A722 540 34 88 and H17AU10A - 085827. I think I found somewhere to look these up and will report when I find it again! As for skill level I spent a misspent youth with a soldering iron writing machine code but these days time is at more of a premium but I have a friend coming over this weekend with more time and aptitude to help me play! I saw the rinehart controller which is about$3-4K (which is now about 50% more to me than a couple of months ago before a few fellow countrymen pointlessly decided to render our currency worthless!) which would almost be plug and play and I got the DMOC for £150 expecting to spend a bit of time and tinkering to get it to work. I got the motor for £400. I'm looking for a solution which is a cost/effort compromise! I naively hoped a bit of config and an arduino/raspberry PI with a bit of code and a little bit of analogue electronics might fill in the gap. I'll get back with more info after the weekend after friend and I have done our investigations (just back from hol last night, got to work on current project setting up some variable frequency drives to drive moving solid cover and pumps for my solar/biomass heated pool project!). I REALLY appreciate the input from you and Wolf, just what I was hoping for. Thank you both very much. Ilya

                              • ilya_666
                                So I have deciphered pinouts, got manuals for controller and ccshell. Now all I need is to find a place to get ccshell. Friend thinks he has a copy on old
                                Message 15 of 20 , Aug 20, 2016
                                  So I have deciphered pinouts, got manuals for controller and ccshell.  Now all I need is to find a place to get ccshell. Friend thinks he has a copy on old laptop. Any pointers links or zip files appreciated.  Also details of what the encoder needs and I will make a digital encoder for it.

                                  Sent from my iPhone; please excuse typos and arbitrary spelling "corrections"

                                  On 17 Aug 2016, at 13:14, Wolf wolf@... [solectria_ev] <solectria_ev@yahoogroups.com> wrote:

                                   

                                  If you have time rather than money, and you are up for a challenge... ;)

                                  Get a resolver to digital converter (quadrature).

                                  Find a AC induction motor .par file that is close to the voltage/current of your motor.

                                  Use ccShell to set your DMOC to the same number of counts per revolution for your RDC chip.

                                  Now the fun part...

                                  Set the magantization current to zero, (your rotor already has a magnetic field).

                                  It will jump suddenly on start up if you don't do this.

                                  Find the parameter(s) that controls slip, you want that to be zero all the time with a permanent magnet synchronous motor.

                                  Zero out all the gains that control field weakening.

                                  You want to be in volts/Hz mode for as long as possible.

                                  Spin the motor under load, adjust the gains such that it works smoothly up to 30-40 mph... Then torque will fall off a cliff when the back EMF voltage equals your pack voltage...

                                  Before this point is the RPM you want to field weakening to start kicking in...

                                  You will then need to adjust the gains such that you gradually counteract the magnets in your rotor, and thus the back EMF.

                                  This will let you run higher RPM with a gradual decline in torque.

                                  Simple. ;)

                                  And here is the hard part, there is not much documentation on what the parameters in ccChell do... You have to guess based on the description of the varible, and most are abbreviated...

                                  And keep in mind that adjusting one parameter will affect the others.

                                  If the motor is not in a vehicle, then make sure it is bolted down so it can't jump of the work bench. ;)

                                  It can be done, it will just take a bunch of trial and error time.

                                  Hope that helps,
                                  Wolf


                                  On Aug 17, 2016 2:48 AM, "ilya_666@... [solectria_ev]" <solectria_ev@yahoogroups.com> wrote:
                                   

                                  Thank you! I believe I looked it up and the resolver has essentially sine and cosine signals. The motor markings have A722 540 34 88 and H17AU10A - 085827. I think I found somewhere to look these up and will report when I find it again! As for skill level I spent a misspent youth with a soldering iron writing machine code but these days time is at more of a premium but I have a friend coming over this weekend with more time and aptitude to help me play! I saw the rinehart controller which is about$3-4K (which is now about 50% more to me than a couple of months ago before a few fellow countrymen pointlessly decided to render our currency worthless!) which would almost be plug and play and I got the DMOC for £150 expecting to spend a bit of time and tinkering to get it to work. I got the motor for £400. I'm looking for a solution which is a cost/effort compromise! I naively hoped a bit of config and an arduino/raspberry PI with a bit of code and a little bit of analogue electronics might fill in the gap. I'll get back with more info after the weekend after friend and I have done our investigations (just back from hol last night, got to work on current project setting up some variable frequency drives to drive moving solid cover and pumps for my solar/biomass heated pool project!). I REALLY appreciate the input from you and Wolf, just what I was hoping for. Thank you both very much. Ilya

                                • Wolf
                                  I have the ccShell installers on this page: http://www.wolftronix.com/motorControllers.htm You need a Quadrature signal. The number of lines on the encoder and
                                  Message 16 of 20 , Aug 20, 2016

                                    I have the ccShell installers on this page:
                                    http://www.wolftronix.com/motorControllers.htm

                                    You need a Quadrature signal.
                                    The number of lines on the encoder and direction needs to match what is in ccChell (you can change the parameter to match your physical encoder).


                                    On Aug 20, 2016 11:02 AM, "ilya_666@... [solectria_ev]" <solectria_ev@yahoogroups.com> wrote:
                                     

                                    So I have deciphered pinouts, got manuals for controller and ccshell.  Now all I need is to find a place to get ccshell. Friend thinks he has a copy on old laptop. Any pointers links or zip files appreciated.  Also details of what the encoder needs and I will make a digital encoder for it.

                                    Sent from my iPhone; please excuse typos and arbitrary spelling "corrections"

                                    On 17 Aug 2016, at 13:14, Wolf wolf@... [solectria_ev] <solectria_ev@yahoogroups.com> wrote:

                                     

                                    If you have time rather than money, and you are up for a challenge... ;)

                                    Get a resolver to digital converter (quadrature).

                                    Find a AC induction motor .par file that is close to the voltage/current of your motor.

                                    Use ccShell to set your DMOC to the same number of counts per revolution for your RDC chip.

                                    Now the fun part...

                                    Set the magantization current to zero, (your rotor already has a magnetic field).

                                    It will jump suddenly on start up if you don't do this.

                                    Find the parameter(s) that controls slip, you want that to be zero all the time with a permanent magnet synchronous motor.

                                    Zero out all the gains that control field weakening.

                                    You want to be in volts/Hz mode for as long as possible.

                                    Spin the motor under load, adjust the gains such that it works smoothly up to 30-40 mph... Then torque will fall off a cliff when the back EMF voltage equals your pack voltage...

                                    Before this point is the RPM you want to field weakening to start kicking in...

                                    You will then need to adjust the gains such that you gradually counteract the magnets in your rotor, and thus the back EMF.

                                    This will let you run higher RPM with a gradual decline in torque.

                                    Simple. ;)

                                    And here is the hard part, there is not much documentation on what the parameters in ccChell do... You have to guess based on the description of the varible, and most are abbreviated...

                                    And keep in mind that adjusting one parameter will affect the others.

                                    If the motor is not in a vehicle, then make sure it is bolted down so it can't jump of the work bench. ;)

                                    It can be done, it will just take a bunch of trial and error time.

                                    Hope that helps,
                                    Wolf


                                    On Aug 17, 2016 2:48 AM, "ilya_666@... [solectria_ev]" <solectria_ev@yahoogroups.com> wrote:
                                     

                                    Thank you! I believe I looked it up and the resolver has essentially sine and cosine signals. The motor markings have A722 540 34 88 and H17AU10A - 085827. I think I found somewhere to look these up and will report when I find it again! As for skill level I spent a misspent youth with a soldering iron writing machine code but these days time is at more of a premium but I have a friend coming over this weekend with more time and aptitude to help me play! I saw the rinehart controller which is about$3-4K (which is now about 50% more to me than a couple of months ago before a few fellow countrymen pointlessly decided to render our currency worthless!) which would almost be plug and play and I got the DMOC for £150 expecting to spend a bit of time and tinkering to get it to work. I got the motor for £400. I'm looking for a solution which is a cost/effort compromise! I naively hoped a bit of config and an arduino/raspberry PI with a bit of code and a little bit of analogue electronics might fill in the gap. I'll get back with more info after the weekend after friend and I have done our investigations (just back from hol last night, got to work on current project setting up some variable frequency drives to drive moving solid cover and pumps for my solar/biomass heated pool project!). I REALLY appreciate the input from you and Wolf, just what I was hoping for. Thank you both very much. Ilya

                                  • ilya_666
                                    Thanks! Just taking break to fix relays on hot tub to keep our respective other halves happy then back to play! Sent from my iPhone; please excuse typos and
                                    Message 17 of 20 , Aug 20, 2016
                                      Thanks! Just taking break to fix relays on hot tub to keep our respective other halves happy then back to play!

                                      Sent from my iPhone; please excuse typos and arbitrary spelling "corrections"

                                      On 20 Aug 2016, at 17:14, Wolf wolf@... [solectria_ev] <solectria_ev@yahoogroups.com> wrote:

                                       

                                      I have the ccShell installers on this page:
                                      http://www.wolftronix.com/motorControllers.htm

                                      You need a Quadrature signal.
                                      The number of lines on the encoder and direction needs to match what is in ccChell (you can change the parameter to match your physical encoder).


                                      On Aug 20, 2016 11:02 AM, "ilya_666@... [solectria_ev]" <solectria_ev@yahoogroups.com> wrote:
                                       

                                      So I have deciphered pinouts, got manuals for controller and ccshell.  Now all I need is to find a place to get ccshell. Friend thinks he has a copy on old laptop. Any pointers links or zip files appreciated.  Also details of what the encoder needs and I will make a digital encoder for it.

                                      Sent from my iPhone; please excuse typos and arbitrary spelling "corrections"

                                      On 17 Aug 2016, at 13:14, Wolf wolf@... [solectria_ev] <solectria_ev@yahoogroups.com> wrote:

                                       

                                      If you have time rather than money, and you are up for a challenge... ;)

                                      Get a resolver to digital converter (quadrature).

                                      Find a AC induction motor .par file that is close to the voltage/current of your motor.

                                      Use ccShell to set your DMOC to the same number of counts per revolution for your RDC chip.

                                      Now the fun part...

                                      Set the magantization current to zero, (your rotor already has a magnetic field).

                                      It will jump suddenly on start up if you don't do this.

                                      Find the parameter(s) that controls slip, you want that to be zero all the time with a permanent magnet synchronous motor.

                                      Zero out all the gains that control field weakening.

                                      You want to be in volts/Hz mode for as long as possible.

                                      Spin the motor under load, adjust the gains such that it works smoothly up to 30-40 mph... Then torque will fall off a cliff when the back EMF voltage equals your pack voltage...

                                      Before this point is the RPM you want to field weakening to start kicking in...

                                      You will then need to adjust the gains such that you gradually counteract the magnets in your rotor, and thus the back EMF.

                                      This will let you run higher RPM with a gradual decline in torque.

                                      Simple. ;)

                                      And here is the hard part, there is not much documentation on what the parameters in ccChell do... You have to guess based on the description of the varible, and most are abbreviated...

                                      And keep in mind that adjusting one parameter will affect the others.

                                      If the motor is not in a vehicle, then make sure it is bolted down so it can't jump of the work bench. ;)

                                      It can be done, it will just take a bunch of trial and error time.

                                      Hope that helps,
                                      Wolf


                                      On Aug 17, 2016 2:48 AM, "ilya_666@... [solectria_ev]" <solectria_ev@yahoogroups.com> wrote:
                                       

                                      Thank you! I believe I looked it up and the resolver has essentially sine and cosine signals. The motor markings have A722 540 34 88 and H17AU10A - 085827. I think I found somewhere to look these up and will report when I find it again! As for skill level I spent a misspent youth with a soldering iron writing machine code but these days time is at more of a premium but I have a friend coming over this weekend with more time and aptitude to help me play! I saw the rinehart controller which is about$3-4K (which is now about 50% more to me than a couple of months ago before a few fellow countrymen pointlessly decided to render our currency worthless!) which would almost be plug and play and I got the DMOC for £150 expecting to spend a bit of time and tinkering to get it to work. I got the motor for £400. I'm looking for a solution which is a cost/effort compromise! I naively hoped a bit of config and an arduino/raspberry PI with a bit of code and a little bit of analogue electronics might fill in the gap. I'll get back with more info after the weekend after friend and I have done our investigations (just back from hol last night, got to work on current project setting up some variable frequency drives to drive moving solid cover and pumps for my solar/biomass heated pool project!). I REALLY appreciate the input from you and Wolf, just what I was hoping for. Thank you both very much. Ilya

                                    Your message has been successfully submitted and would be delivered to recipients shortly.