Solar Install; the on-going saga

That's what I would have thought M. but when I emailed Morningstar previously they are very concerned about being within spec of the controller. The manual advises to only load it to 80% capacity with Isc PV to allow for edging spikes and low temp. up-rating.
That said the manual also says it'll take 130% under FET duress.
I'm suspicious that being that it's tied to a current source it's designed to to grab what it can and run, unlike traditional voltage sources that sit in and around what current they demand. ...correct me if I'm wrong I haven't researched it a whole lot tbh.


It has a wide input range but it's just a big buck regulator really and MPPT is a big buck/boost regulator with compensation for battery SOC.

It'll happily power up at 9V and idly twiddle it's thumbs until it sees a voltage higher than the batteries then buck everything it deems fit up to 125V.

Supply needs to be higher than my temp. compensated equalisation set-point inc. wiring and circuit losses.

Computer PSU max out at 12V? Can any wizardry be done with the 5v and 12v rails to make a 17V output?

Whadya make of this?



Do you think I could parallel two 10A outputs for 20A? Or is that a manufacturer question?

PS. My 3.8A charger dieing may have had more to do with overwork than the TS. Usually solar controllers read battery chargers as a charged battery and back off.

3.8A charger back online.
I read in some forum hitting it might help (tv fix :rolleyes:).
So after repairing broken solder legs on the the mains LC from brutalisation (hitting solid state lekytronics rarely does much good :pac:), it was just the mode switch gave up. Swapped it out and she's good to go.

Tristar; 2 - Cheapo mains chargers; 0 - Sir Liamalot; 1

I've now got ~150% more electrons flowing from my alternator to my house batteries per hour in the Summer, probably more in the Winter because my fridge is now running on a thermostat on the AC element from the inverter, which I've hard-wired through half my "ring main" on self-discriminating mains/inverter sockets (including the one I use for battery charging :pac:).
I can turbo boost my fridge when driving 12v + 230V = 225W :cool:-ing. Soon as fridge is down to temp. thermostat kicks out and fridge runs on 12v only.
The fridge can now feasibly run from PV on selective occasions.
I didn't convert all my sockets because I didn't fancy taking the walls and roof apart and in an effort to leave my dual source powered, triple circuit, 2-way switched overhead lighting well enough alone (here be dragons) there was no easy way to isolate them from 4 sockets. Eh. I've 2 dedicated inverter sockets. 5 auto-selecting and 4 dedicated mains sockets. Close enough.


This is a potentially hazardous circuit for a few reasons.
If you wire it incorrectly it'll neutralise your inverter (easy do with 8 poles). Depending on the inverter isolation I guess you might be able to explode your batteries.
If you break the relay while running an inductive load at full capacity (2300W) I reckon you could get an arc to bridge the contacts and neutralise your inverter.
If your connections/crimps fail it could neutralise your inverter...etc..
I needed to extend a core-set to the relay position. Normally I just pull out the old cable and run a new cable instead. I hate connector blocks. In this case that meant pulling out half the conversion and a third of the roof...as needs must I made the exception and butt crimped this one with a real crimping tool and heat shrink.

I don't really care for solder for a vibratey camper application, I prefer the flexibility of crimps, I think solder makes things rigid and more inclined to break...how and ever I've never seen a solder joint fail on a vehicle all the same...not that I have many.


Everything tested and working!

PS. If you're testing a MSW inverter on a non-RMS multimeter you'll get some funny numbers off the dirty sine wave ~150V-190V AC.

This is kindov impressive for a mains charger (it's a low bar ...small, expensive and Spartan by solar controller standards).

It goes into storage mode after < 12hours (all the manual says is "It is therefore essential to reduce the voltage even further when the battery remains connected to the charger during more than 48 hours.")

Looks like the float current is higher on the "High" algorithm. It's ~ 500mA @ 13.8v. Turning on and off loads doesn't effect the charging stage.
Load Compensation is soft rise, soft fall. Battery fills the gaps to avoid hard switching I guess.

Few more tests to run but they'll be a while.
Equalise current and voltage.
Heat sink temp. at full load.
Who quits absorption earliest TriStar or Victron (my money's on Victron)
I might test this 94% efficient claim if I could be bothered, although if they wanted 95% they could've just made the (12 AWG) charging leads fatter imho.

I'm a little underwhelmed with this Victron in the end.
Just a little though. I was very disappointed that after cycling from 40% DOD and then restarting an new charge cycle it called it a day @ SG 1.250. Then I put the batteries on the TS and it went straight into float too.
That's actually saying a lot though because previously the TS has always outperformed my cheapos.
I got 5 cells up to SG 1.275 last week after 4 days on PV float (13.5V) the last cell was 1.265 and having none of it (yup I spent a week looking for my last 1.6% charge :rolleyes:).
I tried a C20 discharge in an attempt to re-balance the cells after equalisation did nothing and them 6v's are bang on 235Ah.
I also swapped around the batteries so the weaker cell isn't in the middle anymore. Moved the temp sensor to the 0v post that gets the most sun (instead of the previous +6V post )

It's all a bit too overcautious and tight-fisted with the electrons.

Equalisation is set at 15V (25°C); 1 hour :rolleyes: (set-point not listed in the manual just a max of 16.5v)...kinda pointless; my requirements are 15.5V, 3 hours.
I can't just equalise on it either, I have to bulk first and let it run through it's stages. So running it 3 times is just ridiculous. :mad:
Basically just a hot cycle then.

The manual says the EQ charge current is 1.2A and

Victron wrote:

"Some battery charger manufactures recommend pulse charging to reverse sulfation. However, most battery experts agree that there is no conclusive proof that pulse charging works any better than constant voltage charging. This is confirmed by our own tests."

I metered it as a slow current pulse averaging 1.4A.

Temp comp. is 2mV below standard recommended.

Victron wrote:

Ultra high efficiency “green” battery charger with up to 95% efficiency, these chargers generate up to four times less heat when compared to the industry standard.....
...
...Durable, safe and silent
- Low thermal stress on the electronic components.

Victron wrote:

Protection against overheating: the output current will reduce as temperature increases up to 60°C, but the charger will not fail.

Looks like storage kicks in after ~3 hours.

Storage load compensation sits exactly 0mA if you give it long enough on good batteries with a constant/no load.

That's understandable.
From what I've seen discounting Sterling on reliability issues (but not warranty or customer service). You can get twice the charger for half the price with a solar controller and an external current limited supply. That said I've still to make it work.

It's a steady 58% at full load. By the time it reaches 14.6v it's already switching to reduce internal heat so higher volts doesn't improve efficiency.

Well spotted, I was just zero-ing the meter it was reading 4mA open circuit with the meter tied to the charger iirc.
Seeing as you asked the Trimetric is ~30mA high of cal too (that's my wiring's fault though not the unit).

The 87 auto-powered off a few times while I was watching it, I didn't delta it for the rest of the results, the Victron is still 58% up to 14.1V.

The upper middle is a cheapo Chinese "GopherT" ran into the TS on solar inputs via C/O as it's predecessor. Current limited (5A) 32v variable power supply. It's a neat little do-fer. Passive cooling, seems well able to manage full load. Low noise, I can tune out the buzz that was previously plaguing me. It's non-existent under 17V. I kept the frankinductor because it makes a great battery restraint and it'll smooth the ripple.

The GopherT is my high float, load compensating to 5A, Winter equalising box and emergency back-up/universal transformer. I'm well happy with it. (Don't worry I'll post pics if this goes down in another smoldering death :pac:).
Coincidentally it hits the current limit at a perfect equalising and float set point voltages in stand-alone, compensating for diode drop (on positive lead to protect the output from the battery just in case). Efficiency listed 88%, actual 50% @17v, through the tristar, schottky diode and frankinductor; ~20% when switching.

The Victron not so much, I'm just using it for bulk & absorption charge. Sure, storage mode is great, but not when it goes into it at 25% DOD (probably my own meddlesome fault in fairness....)
A 5A + 15A charger does not make a 20A charger, one just interferes with thuther and screws up the absorption timers on both. Especially when the set points and temp-comp. are different.

I did a bit of experimenting on combining chargers, after 3 weeks on mostly just PV and flirting with 40% DOD....it doesn't work. Everything I had was signing off on a job well done at SG 1.225 per cell after all the confusion. In the end I had to run the PSU; diode protected, current limit 2.5A and voltage @ 14.9v for 12 hours straight to battery without the TS regulating after all my chargers had had enough, this earned me SG 1.250 then a 4 hour TS equalise and a day's high float to get 3 cells up to 1.275, I expect the rest to make up the 0.015 difference after another 3 days floating and I reckon I've boiled off a litre of electrolyte. I'm no longer concerned with short-cycling...the blighters are never bloomin' charged anyways. :rolleyes:

I need to tidy that mess a bit more, It's getting to the point I ought to be considering some heavy trunking. I also need to refit the 230v outlet for a double socket.

I've hard-wired my inverter on 42mm² conductors (just cos I had it not that I needed to) with crimped ring termination instead of the stock 8mm² on croc. clips now that I have a use for it (fridge). I just don't like croc. clips for contact surface area or their spring grip for a mobile application. I've a set of jump leads that I've cut the croc's off and stripped back the conductor so I put the stranded conductor straight onto the battery terminals and use the croc. clip to hold it in place. Learned that from an arc welder.

All the cable that came with the Chinese jobbie was junk. Australian plug with an everything-in-the-world to UK BS adapter that was too small for BS sockets and the contacts looked glued on besides. The DC leads were 0.4mm²
Never got around to re-capping it either...might-do down the line. 7 CapXon and 6 SamXon...worst capacitors in the world I hear. I'd rather swap for better.

moodrater wrote: »

You'd get a quality used 750w-1000w gaming psu for very little on ebay. I have one here that will put out 100A on 12v rail for quad graphics cards; at 45A it wouldn't even break a sweat cost £35.

I figured this one out. Two identical 300W current limited PSU's in series with the earth lifted on the second and the casing earthed to the first and isolated from the second would do the job as a 24V 25A supply.

Ambient 12°C = 15% De-rate
Overwatering = 10% De-rate
48Hr rate = 255Ah
255 x 0.75 = 190Ah
190 @ C48

180Ah Aprox. Relative Capacity

Other news: Trimetric SOC is finally calibrated. :pac: ...years after installing it.
It's only out by ±8Ah after a week. I've set the auto reset off.
Efficiency: 100% (actually batteries are ~85% efficient and discharge is ~C100/C50...it works out).
Capacity: 85% (-weather, -too much water, +reduced discharge curve).
It's also subtracting 5Ah a week that aren't actually used due to to disputing conductors.

Didn't get to see the split charger working today, ran out of photons before I had a chance to power up. My shunt was giving me grief probably from loom manhandling. Claiming I had a 7A load on with the system offline. So I pulled it apart and gave it whatfor. Removed the kelvin connections, put new ring terminals on them with solder + crimp, hit the shunt with a wire wheel and ground the ring terminals down to the copper. 15mA high of cal. now. :cool:

Ten days I've had the deep cycles on the Victron. S.G.: 1.260. It's not a patch on the TriStar I tells ya.

I've a weak cell in my crowns.
One's always 0.010 SG below the rest.

Instructions:

• Take off alternator, put new one on, simples.
• Adjust belt as required...In my case with an automatic adjuster I just used a shorter one I have in the roadside repair drawer. Check service manual for tension setting. If it squeaks the alignment is out or there's a bad bearing in the mix.
• Reconnect all electrical connections before starting engine.

Now I just needed to upgrade my split charge relay and grounds. Grounds are easy enough one for the 0V deep cycle bank terminal to shunt and the other shunt to chassis.
Most of the rest of the job was done by my 95mm² leftovers. All I had to do is go from the end of the B2B link into a new upgraded relay and across to my deep cycles.
For now I've upgraded the B2B fuse for a 80A from the spares drawer. If I need bigger later I'll go bigger later.
The primary PWM split charge relay circuit via solar controller is still a humble 40A and is current limited by MCB.

The secondary is big and dumb: battery to battery.
In the end I couldn't commit my 250A relay...that thing is a monster...15W coil!! My laptop uses less on a good day..ahem well maybe not..but close!

Next contender is that lervely 200A latching relay.
Permanent magnets, fires like an unloaded gun. Pulse on, reverse polarity pulse off.
Uses less lekytrons than my piddling 10A relay.

Rolls Royce of Relays, P.
Albrght relay with a Merlin Badge.
That seller comes by them from time to time braking ambulances.

datasheet

6W flat mount this morning. 55W tilted 60°.

Now that my window slides are a bit weathered they hold the module fine without clamps even in 3m/s wind.

I'm wise to your mischiefs P.


I've a 1.5kW array to stick on the gaff that's more pressing.

The police crash response jeeps have an interesting LED sign deployment mechanism on their roof. :pac:

It's only vital tilt for the cusps of Winter season and they're the ones most campers are indoors or only wanting trickle.