diy solar

diy solar

Cinergi's 28 kWh / 4 kW Solar / 10 kW inverter RV build

cinergi said:
it can’t hurt to ask your distributor or the Victron forums. I’ll check mine after it’s back up and running.
Click to expand...
I'm just tweaking at the moment, trying to adjust to perfect values for my pack, but relating it to typical charge and discharge rates....great on the way down, doesn't take so long, but charging up...oh boy...lol. If i get seriously stuck, i'll try the victron community forum. Thanks.
I wont clog up your thread no more....so looking forward to your progress updates. (y)
 
It's alive!

I've got the system back online. I haven't done a thorough job yet (e.g. heat shrink or cleaning terminals). I put a modest 1500 watt load on it and 1300 watt charge. I measured all the doubled-up serial cables between groups of 4 and the amperage was sharing pretty well (worst case was off by 7% - quite possibly due to dirty terminals or lugs). BMS graph was normal (none of the cells had anything funky going on). You can see the BMS, relay, circuit breaker, etc mounted to a temporary board for now.

Those of you with an Orion and 3/0 wire ... I used a 2/0 lug with 3/0 wire (and some patience getting the wire into the lug) and the 2/0 lug passes through the hall effect current sensor!

See that blue thing next to the laptop? That's my Victron 250/70 SCC which I'm going to hook up to the 100 volt DC power supply on the right (which will simulate a solar panel) so I can make sure the SCC pays attention to the charge instructions that the BMS is sending as well as learn how to prioritize where power comes from and just get a feel for things before I go live with it in the RV.

IMG_9079.JPGIMG_9078.JPGIMG_9076.JPGIMG_9077.JPG
 
I am returning my Meanwell 120A power supply to Amazon. It still works but I don't like it.
And I have a few other things to return before January 31st
 
cinergi said:
... AND my DC power supply failed (and it's not the fuse). :( Fortunately I can still return it (Amazon).
Click to expand...

linear switching design?

"Kungber DC Power Supply Variable, 120V 3A Adjustable Switching Regulated DC Bench Linear Power Supply with 4-Digits LED Power Display 5V2A USB Output, Coarse and Fine Adjustments with Alligator Leads"

Switching power supplies are sometimes right on the edge.
I've had a number power supplies for products fail. Also problems with custom designs in my work.

It is possible the MPPT charge controller, as it moves around the I/V curve testing for changes in maximum power point, unduly stresses the power supply.


"Advanced Maximum Power Point Detection in case of partial shading conditions If partial shading occurs, two or more maximum power (MPP) points may be present on the power-voltage curve. Conventional MPPTs tend to lock to a local MPP, which may not be the optimum MPP. The innovative SmartSolar algorithm will always maximize energy harvest by locking to the optimum MPP."

It has to leave the MPPT if found and make excursions through I/V to look for other peaks.

Maybe a series resistor between power supply and SCC would reduce the tight coupling between these two switching power supplies, get smoother MPPT behavior.
 
Well, any proper bench power supply should be able to handle any kind of load without killing itself, including highly capacitive/inductive ones, fast/wildly changing ones and infinite short circuit.
 
Last edited:
BiduleOhm said:
Well, any proper bench power supply should be able to handle any kind of load without killing itself, including highly capacitive/inductive ones, fast/wildly changing ones and indefinite short circuit.
Click to expand...

"Proper"
"Should"

Good luck with that.

When I first encountered "active loads", I thought "Great! Now we can synthesize complex loads to automate testing of supplies."
Turns out they typically do constant resistance, constant current, couple things like that. And often go unstable with some of those options.

Hardwired test loads are a way to test some points in the complex plane.
How many off-brand off-shore no-name suppliers do any of that?

Now feed a switching power supply into a switching power supply and see what happens.

I have a stack of linear boat-anchors at ballast at the bottom of my test equipment rack.

6624A DC Power Supply, Quad-Output, 2 Outputs at 7V/5A, 2 Outputs at 20V/2A, GPIB [Obsolete]

The Keysight 40W, four-output 6624A is a versatile solution to help improve system accuracy and reduce test time in a 3U high package.
www.keysight.com www.keysight.com


In a custom switching power supply board design (10 outputs of various voltage and current, buck, boost, inverting), we had some that would spontaneously die. Only a percentage of them. I tried to test out some theories, stressing a control unit and one I modified, but couldn't kill either. Maybe it was 10% infant mortality of the IC at a particular stress level. Same component and circuit topology for a 17V and a 10V supply, and only the 17V one ever failed (several times at that.) All within data sheet specs.
 
When we talk about a cheap chinese one of course you can't expect much... That's why I said proper. And bench power supply, not other type of PS.
 
BiduleOhm said:
When we talk about a cheap chinese one of course you can't expect much... That's why I said proper. And bench power supply, not other type of PS.
Click to expand...

True, a bench supply is expected to handle a range of loads.
Colleagues have had various problems with them, including power-on voltage excursions which blew up the DUT.

$95 for 360W, including shipping (assuming it can deliver 120V and 3A at the same time.)

vs. $7000 msrp for a 150W supply from HP.
That's why I like to buy vintage commercial grade equipment for pennies on the dollar.

Sometimes the cheap stuff gets the job done. It makes these projects possible for many people given price and availability.
Same for the low cost inverters.
If they sometimes fail, we aren't surprised.
 
Playing with springs and a new saw blade and guide etc so I can make 16 plywood panels. Got nice new plywood (project stock at Lowe’s). I’m shocked at the fact that I’m unable to tighten the springs down all the way with the wing nuts. I’m going to double check my calculations. That said the nuts are still under the threshold of my torque wrench (5 Nm). But boy this creates a SOLID piece. I’m going to see what happens after I hit near 0% SoC. I’m really curious. Because if you’re NOT using springs I have a feeling your pressures are quite wrong after I’ve seen how this spring system works firsthand. Draining the cells will be the true tale.
I haven’t even tried the neoprene yet. Under 50% spring compression I can still move the plywood along the outer cell ... so I’m glad I’m going to try neoprene. I hope it works as well as I think.

89E66AA9-A683-4C9A-812F-3D56212F9583.jpeg
 
Today I'm killing multiple birds with 1 stone. And one big surprise (below).

  1. Full fixture prototype, including neoprene
  2. Measure spring length difference between 100% and 0% SoC
  3. Measure terminal difference between 100% and 0% SoC
  4. Bring cells to 2.000v per reseller's request (and charge back to full and redo my capacity tests)
  5. Heat 5 gallons of water ?

IMG_9102.JPG
IMG_9104.JPG
IMG_9103.JPG
IMG_9101.PNG

Big surprise: my 0.6" plywood is bending under ~650 pounds (175 each corner). The photo below captures it. As I thought about it, I think it makes more sense to move the 4 rods closer to the center (bring top rods down, move bottom rods up). That should help, but I really wonder about the pressure in the center vs. the edges. The new plywood I have (this is my old plywood) is 0.7" which should be better ... makes me wonder though!


IMG_9106.JPG
 
Yes .... wood will warp under pressure and it will get worse over time. This is why most wooden shelves will have an additional thickness of wood across the front.

It might be possible to put some angle iron across the top and bottom between the threaded rods.
 
Bob B said:
Yes .... wood will warp under pressure and it will get worse over time. This is why most wooden shelves will have an additional thickness of wood across the front.

It might be possible to put some angle iron across the top and bottom between the threaded rods.
Click to expand...

Yeah I was thinking of Unistrut but angle iron should do the trick too; and it will give me handles. Good suggestion, thanks!
 
cinergi said:
  1. Measure spring length difference between 100% and 0% SoC
  2. Measure terminal difference between 100% and 0% SoC
Click to expand...

Did you get that part done? What were the results?
In your pix I saw 99% and 27.24mm twice.
 
Hedges said:
Did you get that part done? What were the results?
In your pix I saw 99% and 27.24mm twice.
Click to expand...

In progress. SoC is around 60% now with about 3 hours to 0. I haven't noticed any change in spring closest to the camera yet.
 
You must log in or register to reply here.

Similar threads

C
How many KW now, what inverter to prepare for the future ? (Netherlands)
Replies
2
Views
137
CostOptimizer
C
A
4000 kWh usage per month
Replies
15
Views
414
wpns
W
T
14.4 KW solar, 40.96 kWh backup, 2 solark 15k inverters, off grid family home, doable or delusional?
3 4 5
Replies
83
Views
4K
Bop
B
J
EG4 6.5 Kw inverter max array voltage per PV input
Replies
8
Views
421
SignatureSolarJess
SignatureSolarJess
yabert
What 282 kWh CATL LFP look like
2 3 4
Replies
60
Views
3K
yabert
yabert
Back
Top