Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Flatpack2 output response to changes
#1
Read a few posts and threads on other forums and struggling to find an answer...

With the flatpack2 units the output voltage or current limits can be set, however I'm struggling to find anyone who says how quickly (shorter than 10 seconds the only answer so far) the output changes after issuing a command.

Basically I'm tempted to use the flatpack2 as a charger that would then track the net grid demand for the house and charge when I have a solar excess to reduce the net grid demand to zero. This would allow a larger grid tie solar inverter to be used which would also be more efficient for supplying power to house loads..... only if the flatpack2 can actually change output fast enough.

I guess the units are technically capable of very fast change tracking as they have units setup for wind and solar controllers (you will stand in a steaming pile of rocking horse s... before seeing these on eBay !) and wondering if the "walk-in" 5 seconds output change is a 5 second delay in the PID output setting ?

Anyone with flatpack actually know how fast the output changes to a control command ?

Walk in time message :
0x05FF4000 = ? seconds / invalid ?
0x05FF4001 = ? seconds / invalid ?
0x05FF4002 = ? seconds / invalid ?
0x05FF4003 = ? seconds / invalid ?
0x05FF4004 = 5 seconds (default)
0x05FF4005 = 60 seconds (long option)

From the PowerSuite (356807.063_mfm_2010-07) pdf, page 189 (Rectifier Walk-in Time) there are only two options listed and is a linear ramp-up/down over 5 seconds, however this is for the voltage level.... and guessing it may also apply to current settings or not ?
If you can't quantify how much they cost, it's a deal, I'll buy 5 of them for 3 lumps of rocking horse ......
Reply
#2
Bit of an update, after a lot of reading and searching around (no testing yet) I have figured out that the load following will be a slow ramp (1-3 seconds), unless I can identify the protocol message to change to less than 5 seconds or remove the walk-in delay for the control loop. When I get some time (and some more RS485 ports) I will end up wiring up the 3 units I have to control them and play around ahead of next winter.

The primary aim is to use them for charging the pack with cheap grid power, although the ideal aim is also to use them to charge the pack when I have a grid tie solar inverter connected up and then increase the charge to eliminate any exports.

If anyone is controlling thier flatpacks for netting grid demand to zero (netting off solar) I'd be very interested Big Grin
If you can't quantify how much they cost, it's a deal, I'll buy 5 of them for 3 lumps of rocking horse ......
Reply
#3
Been looking at doing something similar to this. Instead of using the flatpack2's inbuilt current control adding my own to the output. Believe this could get the faster response time required but still working out the details. Should give much faster changing time though there could be a lot I don't know.

What kind of response time are you looking for? Can't really see a need to go sub second you will lose a bit of solar and draw from the grid a bit but not that much.

Liking the idea.
Reply
#4
Sub second.

The combination of cloudy day solar output variations and load changes from things like a washing machine make for a larger than expected imbalance over time. Lights, overn, etc. are all ok with 5 second imbalance tracking.

Where solar output changes at cloud edges the rate of change in output can be over 1kW per second and frequent load changes from a washing machine motor are insant 300W on-off events.

Think of it this way, if a washing machine on time for the wash motor is 4 seconds on and 4 seconds of and your tracking delay is 2 seconds you will end up with a proportionately large constant imbalance.

Remember as well that electricity meters aggregate the energy at over 100Hz sampling rate so there is no buffering of any import/export swings or lead/lag in matching.

Electronically the flatpack should be capable of a near instantanious output change so tracking sub second should not be an issue at all. What does become a time delay issue is sample time -> network -> computer -> network -> output device/inverter.

Network being RS485 and ethernet for me.
If you can't quantify how much they cost, it's a deal, I'll buy 5 of them for 3 lumps of rocking horse ......
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)