Solar charge controller, what for?

[PNjunction]

I still have to integrate the HWS dump relay into the house wiring so that part of the controls is inactive.

Standard warning here for lurkers - if this is a grid-tie inverter connection it may be illegal depending on your location. I'm sure the Op has done his due diligence in this arena.

From my OWN safety standpoint, even though this is illegal where I live, I'm not counting on it during an outage. Aside from line workers getting zapped, I personally don't want to encounter this from the guy down the street who just doesn't know.

So, when there is an outage in my area, I don't make ANY assumptions that my lines are not hot, or that they will not become hot from a guy down the street later on!

Anyway, congrats on your system so far. From what I've seen, it looks like you are playing it safe.

[bungawalbyn]

Nothing went bang!

[SunEagle]

Nothing went bang!

Glad nothing at your end went bang. But my concern, is your equipment really safe and will nothing be sent down the line from your home if the grid goes down? Unless you have a factory made UL listed isolation device that keeps power on your property there is always the chance of something failing for the wrong reasons with dangerous results.

[J.P.M.]

Nothing went bang!

Yet.

[bungawalbyn]

Glad nothing at your end went bang. But my concern, is your equipment really safe and will nothing be sent down the line from your home if the grid goes down? Unless you have a factory made UL listed isolation device that keeps power on your property there is always the chance of something failing for the wrong reasons with dangerous results.

Hey guys, read the thread. I'm nowhere near the grid. I think maybe you've been led astray by PNJ's warnings to the "Lurkers"

So far the only problem showing itself is a voltage drop at the sensor etc bus bars compared to the battery terminals. Higher when charging, lower when discharging. Obviously some drop across components. Messes up the cell voltmeter a bit as cells 1 & 4 read the difference between bus bar and adjacent cell.
Once I get a chance to go over it all while operating I'll find the culprit if it is just one component, if it's accumulative, I may have to reroute some stuff, or get a good handle on the differences at various charge values and learn to live with it.
After all meters etc are only there to indicate what's going on in the system. Absolute values are somewhat unnecessary when familiar with it's operation as it is movement in voltage, proportionality, ie rapid increase or rapid decline, as the batteries approach the "Knee" that is the thing we are looking for and trying to avoid. A clock set on the wrong time, still accurately measures the passage of time.

If one gets an accurate reading, does that mean it's a Fluke?

[PNjunction]

I've noticed that also when measuring directly across cell terminal aluminum knuts, or directly across the bus bar bolt heads vs at the end or middle of a bus bar.

Something like the difference between say 3.55v across the cell terminals, and 3.56v on the bus bars themselves. Close enough for my work.

However, one thing I do recommend is upon receiving these cells is to lightly hit up the terminal connection points with a bit of scotchbrite pad to remove oxidation, and perhaps a *light* coating of noalox or penetrox. Inspect the bus bars too and give them a light scrub where you are going to make the connection. My bus bars are nickel based, and even though they looked clean, a bit of oxidation did come off. If you use a wire brush, do so lightly instead of scrubbing it hard - since that may tend to imbed steel into the aluminum. Always check the terminal screw holes to make sure no junk falls inside there either.

Still, I do notice slight differences at the bus bars - it all depends on how tight in spec you need to be. Remember that we are sometimes making several dissimilar metallic connections, like aluminum to nickel to steel to copper wire. Good to keep an eye on it all like you are doing.

One great thing about your cells is that they have multiple cell terminal connections for very high amperage use. But even if you are doing the solar "Sub-C" thing, having an extra set of bus bars in parallel would be cool to have connected up anyway to reduce resistance connections even further down.

[karrak]

So far the only problem showing itself is a voltage drop at the sensor etc bus bars compared to the battery terminals. Higher when charging, lower when discharging. Obviously some drop across components. Messes up the cell voltmeter a bit as cells 1 & 4 read the difference between bus bar and adjacent cell.

The usual approach to getting around these problems is to have separate wiring directly from the battery terminals to all the monitoring circuitry, volt meters etc. so you are not reading the voltage drops in the cabling as well.

Is the relay at the bottom on the back the LV disconnect relay and the relay towards the top on the left the PV disconnect relay?

What is the PCB on the right in the middle of the front, is it a cell balancing board?

What is the rating of the cheap circuit breaker on the front bottom left hand side?

A circuit diagram would be useful.

Simon

[bungawalbyn]

The usual approach to getting around these problems is to have separate wiring directly from the battery terminals to all the monitoring circuitry, volt meters etc. so you are not reading the voltage drops in the cabling as well.

Is the relay at the bottom on the back the LV disconnect relay and the relay towards the top on the left the PV disconnect relay?

What is the PCB on the right in the middle of the front, is it a cell balancing board?

What is the rating of the cheap circuit breaker on the front bottom left hand side?

A circuit diagram would be useful.

Simon

Hi Karrak , You've got all that right except the board on the right is a cell monitor not a balancer.

I figured putting the sensors on a more direct line would reduce/ alleviate the Vdrop but I wanted all drain to come after LV cutout contactor. The cheap circuit breaker is rated at 300amps. I have to recheck, but this component may be part of the VDrop problem.

So far I only have half my panels hooked up to it, when charging at max amps, 35, the difference . 0.25V less at battery terminals than bus bars. If this is still the case when I attach the rest of the panels I may not bother with it and just adjust the cut off/on points on the controllers. As it stands it just makes those points more conservative.
I've been away a bit the last couple of days so I'm still to see it operate over a full day cycle.

I've left half of the old system functioning while I put the new setup through it's paces. It's interesting to compare the 2. Old system MPPT controller on FLA does squeeze a couple of amps extra early and late, but at peak sun it's back down to absorb/float and pulsing around 5 to 10 amps while the LFPs just keep on sucking up the 35amps. This is exactly why I've chosen to go LFP


PNJ, I cleaned all the cable lugs and other contacts, everything is soldered, If I can't find a particular component, I can live with it.

Early days, but it's all going well.

[karrak]

Hi Karrak , You've got all that right except the board on the right is a cell monitor not a balancer.
I figured putting the sensors on a more direct line would reduce/ alleviate the Vdrop but I wanted all drain to come after LV cutout contactor. The cheap circuit breaker is rated at 300amps. I have to recheck, but this component may be part of the VDrop problem.

Hi Bungawalbyn,

It is a little difficult to see from the pictures but I think the size of the cable you are using is too small for the current you are putting through it. Here is a simple online calculator to calculate your voltage losses. http://www.calculator.net/voltage-dr...res=1&x=89&y=9

Regarding cable size, the fuses/circuit breakers take time to break, your cable and connectors may have to cope with the full short circuit current that the battery will supply until this happens. I would think your 1000Ah battery could supply maybe up to 10,000 amps if shorted. If it takes 0.1 of a second before your fuse/circuit breaks your cable has to dissipate a huge amount of power without melting. This make cable sizing a tricky subject and subject of much discussion. An absolute bare minimum should be large enough to safely continuously carry the current the safety device is rated at, in your case at least 300A.

Most good quality fuses/circuit breakers will specify a maximum current that they will safely cope with, for fuses this is called the burst current. If you put more current through the device than this specification in might explode or not disconnect the load in a timely manner. Just something to think about.

I am a little concerned that you seem to only have one panel disconnect relay. If this relay fails closed you could quickly damage your battery.

If you are concerned about the monitoring electronics drawing power after the LVD relay had disconnected the load you could put a small disconnect relay in the separate circuit.

I hope I am not appearing to be too negative about your efforts. I think you are on the right track but there some things that I think could be improved or need to be thought about.

Simon

[bungawalbyn]

Hi Karrak
The cables are pretty heavy on the battery to inverter out, a little less from Panels to same, and less still to meter etc bus bars.

Nothing is getting warm.

The 300amp circuit breaker is backed up by a 400amp T class fuse that's supposed to break 20x faster and interrupt 100,000 amps. (Not in photo)

As for panel disconnect There is the high voltage cut off 150 amp relay, operated by the hi V sensor. A manual contactor on both + &- circuits and a 400amp manetic latching contactor on the battery in + , operated by the battery monitor to cut for both hi & lo V.
Is this insufficient?

As to the voltage loss from meter bus bar to battery terminals, if an easy fix is not available, then it's easy enough to get an idea of how that behaves and adjust settings accordingly. Fortunately the drop, in it's effect on safety in charge /discharge, actually brings the settings back from the extremes.

[bungawalbyn]

I think I am starting to be impressed by the efficiency of these batteries.

After running half the house circuitry and half the panels through them, yesterday they hit the knee. They stayed at around 13.2 to 13.3 for 2 days then at first slowly rose to around 13.4 then fairly quickly went up in V from 13.5 to 13.7 in about 15 minutes. This is at 35 amps. This tripped the HWS relay (Not yet wired up to house) So that is working. Will have the HWS relay wiring finished today so it will be amusing to see how that all works.

After it ran up to the high V I switched the whole house over to the LFPs and this morning they were at 13.24V. This seems to be about where they sit happily taking and giving heaps of charge.

We are only 5 weeks from shortest day, so I'm wondering if I will need all my panels to power this setup.

[bungawalbyn]

I finished wiring in the 12V to 240V relay to operate the HWS. I had to isolate it from it's original circuit and run a new line through the bathroom back to the control board.
This afternoon the voltage started the rapid rise again and the relay kicked in at the right level. Owing to the voltage drop between battery and V sensors the gap between high and low is artificially reduced, so I had to drop the LV point to stop it cycling. This is on half my panels charging at 35 amps.
Tomorrow I will hook the rest of the panels up to the system and see what it does. I expect the batteries will top up in about half the time.
When it comes to switching on the HWS load dump, the voltage drop should be lower as the total panel output is a fairly close match for the HWS element. It SHOULD work much better, and run the HWS mainly from the panels. If that's the case I'll bring the on off points closer together again.

[PNjunction]

Re the LVC:

Even though the graphs for Winstons show them going down to 2.5 or 2.7v do not be tempted to go that low, especially since you are not discharging at massive current levels like an EV. 3.0 to 3.1v per cell is what I'd shoot for as an lvc, and of course does not mean you have to go there all the time.

Other cells like A123 seem to be able to survive going that far or further, but we are not sure if the Winstons will handle a similar regime with long life. Despite their testing down to 100% DOD, what we DON'T know is how they charged them back up.

Did they just hammer them at deep discharge, or did they wisely control the climb out of the steep knee with a *shallow* charge current until the cells climbed out of the knee, to maybe 3.1/3.2v and THEN followed up with normal current levels?

This is what I caution people about, and wonder if EV'ers using prismatics do the same? I doubt it. Intercalation on recharge while in the steep part of the knee is inefficient compared to the rest of the normal slope, so if you don't want / have the ability to control current to about .05C or less while coming out of the deep part of the discharge slope, then a 3.0 / 3.1 / 3.15v (take your pick) is a good place to stop for an LVC.

These are rested voltages of course, so take into account the amount of current you are pulling when approaching the lvc. If high, then the lower 3.0 may be appropriate. (EV'ers choose ones even lower). If you are tickling them at .1C, then perhaps stop at the higher voltage.

Surprisingly, if you look at it from a 100-80% SOC level, then stopping at approx 3.0v yields 80% of the cells rated capacity. But, at the top end we don't want to go that high on a regular basis, so chop 10% off the top, and now you have about 70% rated capacity.

This is just me approaching it from a conservative standpoint, and also taking into account that prismatics may not be as robust as an A123 cell.

[bungawalbyn]

So far PNJ, my problems seem like they are more in the likelihood of overcharge. 3.0V is 12V in the pack, that's way low.
At present current usage in the last couple of days, I've gone down to about 13.24 rested overnight (except for fridge +inverter). And run up to cut out at about midday. That's on half my panels, 6 weeks from the shortest day. You seem to be able to get a lot of power out of these with little voltage drop when you get to around 13,2V.
I have yet to take them below that (rested) or about 13.15 under light load.

I'm thinking of leaving one of my old FLA sets at the house doing what they like best, being an ornament on float, for emergency bad weather or what ever other reason. I'll give them a panel to keep them floating.

I drive my systems carefully

[PNjunction]

So far PNJ, my problems seem like they are more in the likelihood of overcharge. 3.0V is 12V in the pack, that's way low.

Actually, you are correct for a low-current application like ours! So much talk of high current EV that even I get it wrong.

To be totally conservative, a value of just under 12.8v is when the discharge slope starts to dive - shallow, but if you like 12.75 would be fine with our low current application.

On the small side, consider that a smart lifepo4 charger, the Tecmate Optimate Lithium TM-291 12v/ 5A charger considers 12.8v to be the 80% DOD *rested* value, and anything lower than that gets the .01C "climb out of the discharge knee slowly" effect, until about 12.8v is reached whereupon it applies full current.

I tested that on both powersports and my own larger prismatic banks (12.75v lvc), and from a 100% SOC charge (which I don't normally do!), I was able to get 80% of the rated value from the bank when measured on a West-Mountain CBA-IV computerized analyzer setup - and compensating for the test-leads too with my Fluke 87V as the final say.

Small potatoes, but I was glad to see that 12.75v was as low as I'd ever like to go for maximizing cycle dod's. The low-current revival when under this value was an eye-opener which makes total sense.

My analogy (love those even if far-fetched) is like scuba diving - go too deep and you better come out slow or you get the bends.








At present current usage in the last couple of days, I've gone down to about 13.24 rested overnight (except for fridge +inverter). And run up to cut out at about midday. That's on half my panels, 6 weeks from the shortest day. You seem to be able to get a lot of power out of these with little voltage drop when you get to around 13,2V.
I have yet to take them below that (rested) or about 13.15 under light load.

I'm thinking of leaving one of my old FLA sets at the house doing what they like best, being an ornament on float, for emergency bad weather or what ever other reason. I'll give them a panel to keep them floating.

I drive my systems carefully[/QUOTE]