Hi Keith. I thought that I was agreeing with you on your estimate. "
You were! I was asking; "Do you think there is any cool-down need for ~5kW screamers?"
Nice work on this Keith. Probably lps time.
Thanks Bill! (Getting closer to the LPS needed for the stuffed teddie bear.)
"Crate it up and ship it to you"...are you kidding me? No way I'd pass up the chance to play with solar panels and VFD's!!! Seriously though, thanks for the offer!
Damn.
Re: Cable length - The campus and proposed tank site is located approximately 3,000 linear feet from the proposed well site. It's approximately 130' higher in elevation. The initial system design resulted in a pump discharge head of ~170'. This may get closer to 130' as we extend run times with the solar option.
OK. Dang. That truly sucks - literally! I was hoping your 3,000ft was some mickeymouse length of all the wires in all directions combined.
Thanks for the clarification.
The front end rectifiers simply conducting rather than converting AC makes sense, as long as the drive isn't going to fault because it isn't seeing it the power it's expecting - which often happens at some of the plants/pump stations I've worked with when the incoming 3-phase gets a bit imbalanced or experiences voltage fluctuations. I’m remembering some liberal fault settings on the AB’s and ABB’s that typically are installed around here, but they do have limits.
Most definitely some drives would freak out about it. I don't believe this drive does because it is only a single phase input drive. I'll also say, most drives don't care about the input because that would require them to monitor all the inputs for current and that would increase the drive's cost by more than a nickle so they don't do it usually. It's likely done on bigger drives, I don't know - say 25HP and up. Just a guess.
Maybe we could wire some sort of solid state voltage monitor to the drives DI's to send start/stop commands soon as the voltage dips or rises on either side of a set level before the line voltage goes dead?
Yes, that's what I was thinking. In this setup it might help to include some sort of cycle limiter that would count the cycles and delay 10 minutes if two cycles happened in less than 5 minutes.
I wonder if we could add a small load to the circuit up-line of the drive to prevent an open circuit voltage condition? Otherwise the need for the braking resistor is escaping me (unless it’s related to DC power)…otherwise what would be the difference here than a conventional pump application where the pump simply winds down upon shutoff and more than ample power is still connected and available on the line side of the drive?
It sounds as if the drive would switch to the braking resistor even, or primarily, in a "motor stop" state, correct?
Waross has explained that very well. I'll just add that normally the circuit that deals with the DC BUS over-voltage is very dumb. It just looks at the voltage on the capacitor bank and if it is above a certain voltage triggers and starts PWMing the bus voltage to the brake resistor. As the voltage continues to rise it will increase the PWM duty cycle. There are limits to the current/power that can be shed this way. The 1HP version I mention has only an internal resistor clamped to the chassis heatsink served by the chassis fan. However, the chassis heatsink is really there for the output stage semiconductors. If the chassis heat sink gets too hot it's game-over as the drive will notice and shutdown. Hence, the use of external brake resistors.
Back to the PWMing circuit: As I mentioned it's dumb to the extent that the drive doesn't even know it's happening and the PWM doesn't know why there's a high voltage present, it just tries to dealwithit. The drive only looks at the peak voltage to decide when to trip - a voltage above where the PWM should've been able to cover it.
An issue is that as far as the drive is concerned, the only place that high voltage condition can come from is the motor being over-driven by its load. (Picture a hoisted load being lowered or a lathe's massive chuck being stopped suddenly.) The drive protects itself by disconnecting itself electrically from the load(motor) e.i. shutting off its output switches which causes the lathe chuck to friction-down for the next 2 annoying minutes or the hoist load being jerked to a stop by the safety brake.
In our case shutting off the output switches changes nothing since that's not where the over-voltage would be originating from. Hence, it falls on us to make sure either the "braking circuit" would be up to the task or we need to provide our own external PWM-like solution. Were I to do it, I'd de-box the drive and using a temp-gun monitor the PWM transistor to see what it thought about this. An external resistor can always be made to dump what's needed and live with it. Sometimes they design the brake power to be capable of only about 20% of the drive rating since
on average they don't expect the drive to be braking that much. It's a reasonable assumption. Though we would be in the 100% realm if the pump is not running. If a drive is going into a regenerative application a "regenerative" drive is used which includes a second separate inverter designed to drive the excess energy back into the power source/mains/network.
If we see a DC overvoltage condition due to no or light load on the panels, the VFD will pull the resistor in. I got that part.
As soon as the resistor is pulled into the DC solar circuit the voltage would then return to normal, so would the drive then disconnect it, only to see the high open circuit voltage again and continuously repeat?
Probably answered above but to clarify: The intent of the PWM circuit is to always prevent an actual over-voltage trip.
I met with one of the more mechanically inclined team members that is leaving for Haiti today. He’s planning to meet with the well driller to discuss his capabilities. With any luck we can move the well closer horizontally to the campus, but all of that is contingent upon his drilling depth capabilities.
Great! And understood. Interesting constraints. A 200ft well that put all this on campus would normally be ranked just above trivial. That would open all sorts of options, like using a standard solar inverter running the pump normally with a VFD and being able to use the array for all sorts of other stuff when not pumping. Alias.
The location I’ve found is a sure bet for getting water from a shallow well. Regardless of where it’s drilled, I think this overall concept is the way to go; we may just be able to eliminate the need for the excessive cable run. He’s also going to scout the area and discuss with the locals what we could do in the form of a small, secure, masonry building to house the VFD at the well site.
As far as I can see there is no solution if the VFD isn't at the well-head. An adequate 3000+ foot wire run is just too expensive, not to mention too great a theft target. He should scope out having the array there too. Possibly protected with its layout and a security system. That could include horns, bells, strobes, lighting, voice, campus annunciation, or a nite watchman(totally valid in some economies - may cost a fraction of what gasoline would've cost.)
For no more than the brake resistors cost I'd absolutely include one, but just thinking everything through here: To avoid the potential for the above mentioned scenario, could we maybe install a separate resistor on the line side of the VFD with a contactor and wire the coil through one of the drives relay outputs so it would pull it in when the drive is called to stop? I suppose that's where the home system testing / de-bugging is invaluable.
That has a lot of merit. My thought would be to put that at the array and switch it to doing something useful rather than just dumping energy via a resistor at the well. Realize the voltage the DC bus would be seeing is the voltage we're delivering from the array since it's DC. Q.E.D we know what's on the capacitor bank (sans voltage drop) at the array. Food for thought. And yes part of the point of debugging the whole thing.
Keith Cress
kcress -
