[subcooledheatpump]

You mean if I had to change something what would it be? and would that be in refrence to the controller or the whole project?

About the van in general; I'm in the process right now of getting a bigger motor, since the van I have is extra heavy (almost 8000 lbs?) that should give me better acceleration. also more batteries

The controller; I need a controller with torque mode+vector mode. The one I have now only does speed control with sensorless vector. Speed control works to a certain degree. I can't however, step on the hydraulic brake, then start back on the accelerator with the van still rolling. That creates a synchronization problem since the motor must coast during hydraulic braking to avoid a conflict of the motor fighting the hydraulic brakes. Regenerative Braking serves to elminate most of those problems though. Just letting off the accelerator and letting Regenerative Braking stop the van seems to be adequate for normal driving situations.

So I should have converted a controller with torque mode and vector mode so I can use Regenerative Braking to assist the hydralic brakes. Then resume acceleration even if the van is still moving.


Side note: those astro fuel pumps are a pain. My dad had one he used in a business. The thing went through 6 fuel pumps before he realized what I told him was correct; the stock wiring harness to the fuel pump causes problems and has to be cut.

You're converting it though so it's not an issue.
I also played with an old astro once, put an electric motor in it. It worked pretty well but it was in bad shape so I scrapped it.

[DDDvvv]

i meant as the conversion in general . thanks for the reply.

im having the same problem in sensorless vector control in my tests (ac 240v on unconverted vfd). if i take motor to high speed, then stop the vfd, i have to wait until the motor comes to a stop, before trying to reapply power, else i get delays and vibrations and overload faults.

ive been fighting with this for the last week, and im beginning to think that the solution is a vfd with true torque control also. ofcourse, this would mean a different vfd. thats why im holding off on the conversion, until i can figure this out.

yes, astros are a pain. there's whole forums dedicated to their problems out there. this one weighs around 5000#, no rust at all. previous owner sprayed that protective stuff on underside and inside all panels. and since it was headed to the junk yard, its like a free donor car.

i just ordered a brand new baldor motor for $495 on ebay (free shipping). its 20hp, ~1800 rpm and its the lightest 20hp i could find out there, weighing 220#. i plan on opening it up, to hook it up to delta, like those aussies, and running it at 120hz , to double the nominal horsepower rating.

still beats spending $2000+ on those warp motors that everyone is trying to lick....

[JRoque]

Hi. Does your VFD have an auto motor tune function and have you run it? I've had good results with a Hitachi SJ300 VFD. I can run tests in sensorless mode and set the amount of regen current to stop it. I can slow it down and speed it up without issues. For EV applications, you'll likely need full vector mode and that's why I was asking if you had found a sensor board for your VFD and encoder for your motor.

JR

[aeroscott]

I have been wondering about taking any good VFD and changing brain ( vector chip etc) .

[DDDvvv]

Originally Posted by JRoque

Hi. Does your VFD have an auto motor tune function and have you run it? I've had good results with a Hitachi SJ300 VFD. I can run tests in sensorless mode and set the amount of regen current to stop it. I can slow it down and speed it up without issues. For EV applications, you'll likely need full vector mode and that's why I was asking if you had found a sensor board for your VFD and encoder for your motor.

JR

i cant find any autotune features on this vfd's. i also think that the encoder route is the solution for ev applications (for this drive). luckily, encoder boards are available. im in the process of ordering one. on the other hand, i could just use decelleration. thanks for the advise. i just read up on that hitachi drive. lots of goodies in it.

I have been wondering about taking any good VFD and changing brain ( vector chip etc) .

its doable, but you might need full schematics, for the donor vfd, play around with desodering high pin count surface mount chips, and 16 or 32 bit programming resources - a major headache for me. this particular vfd uses a 16 bit 84 pin chip, which integrates all the functions.

[DDDvvv]

@ Jroque

about the encoder, there's no specific encoder needed. the encoder board will accept anything from 1 ppr to 4096 ppr, quadrature or single pulse. just specify in the encoder setup what you got. encoders are not that hard to source. the only hurdle will be adapting it to your motor shaft. im planning on doing this with my lathe.


good news; tuning problems finally solved. the problem was that the acceleration rate was set at 10 seconds (drive default; time it takes to get motor from 0 to 60 hz) and this was really confusing to the sensorless module. took that all the way to zero, and drive runs like a dream. stopping, running and resuming at any speed is all possible. no overload faults anymore. i tried the full reverse test several times ; take drive to maximum, then hit the reverse button to get drive to maximum speed on the other direction. took 3 seconds, and a screeching sound. (you will never get to do this on an ev, even on emergency Regenerative Braking).

moving on to the power module construction.

[mizlplix]

I am interested in the motor side mods.

Kevin in Topeka runs a HUGE Baldor motor (400#) with the Curtis controller used in AC50 drives. It is in a full sized pick up. It uses the 4 speed auto trans and runs well.

I was wondering about a maybe-future upgrade for the AC50. It would need to be somewhere in-between (AC50 and that giant Baldor).

Could you guys elaborate on motor selection and mods, please?

Miz

[DDDvvv]

that ac50 motor looks great, 50hp (peak) @ 115 pounds. i think it uses a curtis 1238 controller. but for me, it costs a fortune; $4000+ for the motor alone.

im not equipped to give ac motor advice as of now. the motor mod idea is gained from the australian ev forums, who are waaay ahead of everyone in the ac industrial motors for ev game. from what ive read, majority of ac motors are wired for star configuration. the trick is to find the center of the star, or wye (virtual neutral), bring out the wires, and rewire for delta. current rating doubles, voltage required is halved, speed remains the same. with voltage halved, you have the option to run the motor at double the frequency = higher power outputs.

i dont know what horsepower ratings you are interested in, but they suggest a 2-pole motor ( ~ 3500rpm) if using a transmission, and a 4-pole (~1800rpm) if going direct drive. this is a fact they have tested.

please read up on it to confirm (and correct me if im wrong) before making a decision. im going with the 4-pole direct drive idea. my new motor has not yet been delivered. i will document any mods on the motor.

[rochesterricer]

Originally Posted by DDDvvv

that ac50 motor looks great, 50hp (peak) @ 115 pounds. i think it uses a curtis 1238 controller. but for me, it costs a fortune; $4000+ for the motor alone.

Actually, that price is not for the motor alone, it is the package price for the motor and Curtis controller together. Example:

http://www.evparts.com/products/stre...ors/mt5615.htm

[subcooledheatpump]

Thanks for posting your findings, DDDvvv.

I was thinking of doing the same thing with my van, set "ACC" time to 0, that way I can just floor the pedal after braking and the drive will instantly spit out a high frequency.

I believe you are right about the aussie guys and their motors as well. They all say you can attempt to do that yourself, mod the motor, or get it rewound at a shop.

However, if you've got a 480 volt drive and a 240 volt motor, there is no need for any modifications as long as your drive can output the current required for peak torque.

The only real downside is that going to half the voltage on a motor will require twice the current to make full torque.