Solar Install; the on-going saga

28-01-2018 04:22AM

My motor has swollen arteries!

Fuseboard as it currently stands.

28-01-2018 04:27AM

This week's tip;

Rubber insulating cable boots. Yes you need them and gorsh are they a pain in the hoop.

Given that any you see in the real world are pushed down the cable because whoever removed them wasn't bothered to reinstate them after service..my advise is this..

Cut the boot. Sure you are compromising it slightly but you'll find yourself using them when they are easy to move and the chances of catastrophic failure are a lot slimmer than when it's riding 0.25m up the cable from the connection point.

07-02-2018 02:35AM

There's many different ways to tackle an overheating contactor. I can't recommend this as one other than I'm stubborn and a cheapskate (all mods built from salvage/stocked gubbins).

May I say in my defense this is why I do things the hard way.

Look at these contacts! Silver alloy chunkers for reasonably lossless PV throughput. :cool:

You could eat the not hugely significant contact resistance and use a smaller contactor/relay.
You could PWM the contactor coil and run it at a much lower duty cycle (switching noise)
You could use a contactor with better cooling or a greater thermal mass.
You could use a bi-stable contactor and a magnetic latching contactor coil controller.
You could make a latching contactor coil controller.
You could run a dedicated mains charger.

.....or you could do what I do....:pac:
Active, vented coil cooling!:D

Chop out some vents. Add a fan on a 2A flyback diode protected DC - DC buck converter switching the load voltage of 15.5V down to 10V, then use a ~ 90Ω resistor downstream of that to limit the 12V coil fan down to 6V where it's barely audible.

I've reduced the coil consumption from 30W to 8W(including fan and buck converter) and the external saturated continuous operating temperature from 65°C to 20°C. :cool:

Sleeve bearing fan, with a duct made from hot glue...of course I found a brushless 10mins after the mod was mounted.

I abandoned the notion of heat sinking it and shielding after I saw the discolouration of the coil after about 8 hours operation. Passive cooling would only address the symptoms not the root problem because I could only vent radiated heat not generated.

Regarding the delatching issue I'm adding a series diode on the PSU output. I decided to mount it to the same heatsink as my flyback diodes to save space and so I could insulate them as a set.

Here we are ready for reinstatement. Heat sink and terminals are salvage from things I've previously blown up.

07-02-2018 02:40AM

If you find yourself running outtov USB ports you can add more by through charging using dual output power banks. The power bank will charge last and end loads first. As an added benefit your power banks will generally be fully charged when you need them as a result.

16-02-2018 06:53PM

Can't have nice things me. The buck regulator didn't survive the installation. Keyhole surgery after losing recalibrating the SW61 contact return spring proved too much for the SMD trimpot.

Had to resort to a more primitive rugged coil powah reducer.

SW61 mod 2.
Vented Coil,
Active Cooled with
6W 3.6Ω Leckytronic Dry-ifier© Coil Load

12volt shore power lecky heater. Harvesting heat to keep the leckytronics dry.

Mains facilities are back online.
Solar restored.

200A 12V SubMain Distribution.
45A PWM Charge Controller.
16A @ 16V Mains Supply / Charger
Solar / Mains Automatic ChangeOver
80A Earth Bus
Smartbank Standard on 20A Relay
100A DC Consumer Unit: PV Isolator & Breakers
12-Way 100A Blade Fuse Block
2 x 40A Flyback Diodes
Spare 120A Shottky Diode
Leckytronic Dry-ifier©

Sustainable refrigeration made me do it!

20-02-2018 07:21PM

I recuperated some of Mk-11's losses. I think they rest lies in the impedance of the Master Fuse and the improvements made to the B1 to B2 conductor. ie. I've improved charge to the Engine Battery...ho, hum...
I suspect a lower actual than recorded state of charge due to the upheaval, the battery hasn't been brought right up tippy top for months.

Peak Nominal Alternator Charge Powah
45% DOD

20-02-2018 07:33PM

Cable ties left like this snag, scrape and cut.

Trim cable ties to the ratchet with a flush cutters.

20-02-2018 08:20PM

Gold star if you guessed what the other Bsense fuses were for!

Huzzah. Another meter!!

I acquired a Schlicktron BMV. I thought I'd use it on the house but the display is so small it'd just annoy me. So low I decided it's a poor monitor but a decent gauge..it's even 52mm. Compatible with 500A 50mV shunts!

Problem is Schlicktron like to make things easy to install instead of rugged & durable. So instead of wiring a shunt, they put this thing there.

131689620.4yhkQpdJ.jpg

UTP cable. Fragile thing like that in the land of the weighty conductors, gassy batteries and heavy accumulators. I can't be having it. Apart from all the questionable mechanical, electrical and chemical design decisions I already have a standard shunt for a better monitor and I want to use that. Two shunts are outtov the question.

I could use the Schlicktron PCB with my existing shunt if I wanted, and if it fit and if it didn't offend me.

I made this instead that lives behind the meter...ahh easy to install...couldn't even get that right..:rolleyes:

Schlicktron Meter to Don't-Reinvent-the-Wheel Normal Shunt Standard Interface©

It's wired like a TriMetric, without the option of using thicker wiring for long runs. There's a polyfuse between the power in and return B+ sense and a 100nF Capacitor between B+ and Shunt Load Sense. They use twisted pair but the power is not a twisted pair in the stack. :rolleyes:

Anyways I fixt it!

My Model is a 600s, I wanted a 602s.
Enter my selector switch.

A-B, Common Out Selector Switch
Meter B+ from B1sense or B2sense
The meter can read either battery voltage and not lose it's memory. :cool:

What's it do?

All this stuff!

The calibration is bang on agreeing with the TriMetric.
The TriMetric is smoothed, the BMV isn't. This is good in a monitor and better not in a gauge.
The BMV doesn't display watts. They make a neat pair.

21-02-2018 11:54PM

Finished shoring up the new Alternator B+. I De-it's-just-temporarified it.
I was just a little paranoid about it shearing at the post and starting an electrical fire. So I splashed out on some Stainless P-Clips For mechanical support.

B+ to Service Battery.

200A Bus-land, dual sensed shunt.
You know what'd be good in there? A PCB...anna phone cable! :P

Master Fuses

Clockwise.

25A: 16V PSU Maxi-Blade
150A: ANL Alternator to B1
80A: 600VA Mollified Square Waiver Maxi-Blade
150A: ANL Habitation Master

not shown 100A Maxi Blade B2 to Alternator...I don't like it, I think I'll go 200A ANL.

Fused alternators across redundancy starting conductors...:eek:
Unfused 7m cable runs with multiple junctions. :eek::eek:

TriStar Temperature Sensor (€35 NTC thermistor in lug)
& Ground Sense.

My Artificial 260W Solar Panel / Redundancy PSU / Shore Power Onboard Supply.
Derated to 16A (pending it's-never-gonna-happen 20A cooling mod) Constant Current, Constant Voltage Adjustable 16V Supply
Shown supplying 12.8A combined load demand + charge powah.

22-02-2018 12:56AM

This gauge doesn't know what it is.

It's got a blue backlight.
..because blue power?

It's not, none of the power is blue, I checked that. Had some, didn't like it. Blue power detector false positive.

So I keep thinking my headlights are on. It lost power under crank and lost memory. It's too small to be a habitation monitor, it can't be read from far away. It's not a monitor because it's tiny, has no power reading and has an SOC ceiling. It's not a very good gauge either.

The backlight either fades after 30 seconds since interfacing so it's unreadable night-time driving unless you press 15 buttons to enable always on backlight and 15 buttons after you stop the engine to turn it off again.

I can't ignition switch the backlight without losing the SOC function..

I can ignition switch my ApprenticeVolt BTM backlight...but you'd never fit that huge unit with a tiny display in a console.

I suppose an ignition switched normal tungsten, or warm white illumination feature in your 52mm gauge would be another wire for poor old me to worry about connecting somewhere useful.

Which of these is more accurate? If the monitor says charged when the charger says charged then it's a charger monitor not a battery monitor.

Why would a battery charger manufacturer have a ceiling on their meters?

Is this not the same as offsetting the charge efficiency?
Glad you asked! :pac:
No because the battery efficiency is non-linear so if a disproportionate amount of time was spent in the efficiency belt, micro-cycling like a typical user for a week between 65% & 85%, calculated efficiency offsetting will cause drift.

A charge cycle may be 80% efficient SOC 100% - 50% - 100%
but lead acid battery can be 98% efficient between 50% to 80%.

I'm considering demoting it to a voltammeter.

air · 22-02-2018 09:23AM

I think all your experiments are going to show the folly of lead acid charge state estimation. Time to move into the 21st century and lithium batteries I think!
Far, far, far simpler to own and operate, lighter and longer cycle life, state of charge at a glance with a $2 voltmeter or from whatever BMS you're using.

Jaden · 22-02-2018 09:45AM

air wrote: »

I think all your experiments are going to show the folly of lead acid charge state estimation. Time to move into the 21st century and lithium batteries I think!
Far, far, far simpler to own and operate, lighter and longer cycle life, state of charge at a glance with a $2 voltmeter or from whatever BMS you're using.

This is gonna get good *grabs popcorn*.

22-02-2018 10:05AM

air wrote: »

I think all your experiments are going to show the folly of lead acid charge state estimation.

My TriMetric works very well. Efficiency set to 100%. Capacity derated 20Ah. Average discharge C35. Coloumbic efficiency set to 1. Manual 100% SOC reset.

air wrote: »

Time to move into the 21st century and lithium batteries I think!

Any examples you can show of a functioning system for a motorhome?

air wrote: »

Far, far, far simpler to own and operate, lighter and longer cycle life, state of charge at a glance

How much simpler? I just have to reprogram the TriStar for a custom li-ion charge algorithm add a BMS and I'll be fashionable.
[Edit] ...after making my alternator compatible.

Any data on the cycle life claim?

Nothing wrong with the 20th century technology and 21st century equipment. My biggest problem is badly designed products or devices not meeting specification, not SOC estimation.

air · 22-02-2018 10:23AM

There are umpteen different lithium battery motorhomes detailed extensively on Youtube and elsewhere.
I'm not denying that your trimmetric does the job but it's an expensive piece of kit. With all the various gubbins you've tried which have failed and 19 diffrerent revisions you've gone through you could have set up a lithium system a few times over I suspect.
Used cells from Nissan Leafs and other vehicles are coming available at better and better prices all the time.
In the past 6 months I've picked up 17kWh of 10C rated water cooled lithium cells myself for buttons.

In a vehicle / motorhome scenario, all you need for a vehicle charge is a $15 500W DC boost converter (or several in parallel) at the alternator feeding the bank at 48V or whatever your nominal is and a charge controller / BMS for peace of mind.
No fussing about infinitesmal connection resistances, losses in huge 12V cables or any other faff.

I've been running a 2kWh 8S Nissan Leaf system in a van for 3 years with no issues, haven't even gotten around to fitting a BMS yet, it's capacity is so large I just manage charge manually for now.

One pack before I removed the 4 x 2kWh 12S modules:

Lead Acid works, no denying that, but things have moved on and it has a very limited future in any mobile applications.

22-02-2018 10:44AM

air wrote: »

There are umpteen different lithium battery motorhomes detailed extensively on Youtube and elsewhere.

Yeah seen one set fire because they didn't insulate the battery and left it on charge sub-zero. :pac:
It's just not my thing air they're not worth the hassle bang for buck, I guess this we will always disagree on.

air wrote: »

I'm not denying that your trimmetric does the job but it's an expensive piece of kit.

It has paid for itself as a diagnostic tool.

air wrote: »

With all the various gubbins you've tried which have failed and 19 different revisions you've gone through you could have set up a lithium system a few times over I suspect.

Some of that is the learning curve, it'd be the same for lithium. My system does both it's fully programmable that's one of the major design considerations

air wrote: »

In the past 6 months I've picked up 17kWh of 10C rated water cooled lithium cells myself for buttons.

Sweet.

air wrote: »

In a vehicle / motorhome scenario, all you need for a vehicle charge is a $15 500W DC boost converter (or several in parallel) at the alternator feeding the bank at 48V or whatever your nominal is and a charge controller / BMS for peace of mind.
No fussing about infinitesmal connection resistances, losses in huge 12V cables or any other faff.

This is not a lead acid problem this is a one alternator two battery problem.
I actually have 3 x 150W boost regulators for exactly that purpose Constant Voltage Charging. I haven't opened them, copper to me looks like a much more reliable lower profile solution.
The 12 versions of alternator charge is more me being stubborn and won't pony up for a battery to battery charger before a cable upgrade. I've no assurance battery to battery chargers are any better than unbridled. I've asked for data and been ignored.
Sterling do big ones, low build quality and expensive.
ProNautic expensive.
Everything else is throttled to 20A or 30A if yer lucky.

The moral of the story is if I did it right the first time that would be all there was to it but the way that works is in contravention to conventional knowledge. Hence I decided to document a solution.

air wrote: »

I've been running a 2kWh 8S Nissan Leaf system in a van for 3 years with no issues, haven't even gotten around to fitting a BMS yet, it's capacity is so large I just manage charge manually for now.

Start a thread, g'wan!
*Sits beside Jaden with popcorn*

air wrote: »

Lead Acid works, no denying that, but things have moved on and it has a very limited future in any mobile applications.

Ah I think motorhome hab electrics is a semi-static application.

22-02-2018 10:55AM

If I was going for a battery upgrade which I'm not because mine is a trooper.
I'd go Oasis Firefly AGM. Clicky.

air · 22-02-2018 10:57AM

What you don't seem to be getting is that there is immensely less hassle in running a lithium based battery than a lead acid one.

No off gassing so easier to accommodate, much lower explosion risk
Less volume per kWh so easier to accomodate
Less weight so easier to accommodate, easier to carry more storage, better vehicle fuel economy, increased payload for cargo
No complicated / expensive charge management systems - there are expensive BMS systems but if you use high quality cells balancing issues are very rare indeed. The discharge rates in a "semi-static" application as you call it, are a fraction of the design loads for these vehicle cells.

They also lend themselves to higher voltage packs - 48V would be my recommendation. This allows for higher powered inverters and cheap accessories, since a huge proportion of real world DC systems are 48V.

I'm sure you'll find poor lithium installations detailed online, as you will for lead acid and every other type of technology. It's no reflection on the equipment though.

It's not even worth my while starting a thread on my 8S system, it's so simple.
An 8S pack in a wooden box feeding 32V and 12V distribution rails.
The 32V rail feeds a heater, the 12V (down converted from the 8S pack) powers everything else.

I generally charge it from a voltage limited mains charger once a week (about €35). The other option is the DC boost converter which auto starts when the vehicle is running (adjustable startup voltage) and charges the 8S bank via a 6sq cable.

https://www.aliexpress.com/item/1200W-20A-DC-Converter-Boost-Step-up-Power-Supply-Module-IN-8-60V-OUT-12-83V/32706357666.html

My "Trimetric charge estimation device" is a $2 7 segment LED voltage display.

I'm currently working on a 6S system for a small vehicle, I'll be using a bluetooth enabled BMS on that and will likely retrofit one of those to the 8S pack afterwards.
Will give me bluetooth access to all cell voltages, coulomb counting (yipeee), charge and discharge voltage limits at the pack and cell level, etc etc.
Total cost - about €50 delivered.

air · 22-02-2018 11:01AM

Sir Liamalot wrote: »

If I was going for a battery upgrade which I'm not because mine is a trooper.
I'd go Oasis Firefly AGM. Clicky.

How much do those run?

22-02-2018 11:12AM

Yeah I can't watch Utoob conversions different flavours of poorly executed and rarely metered.

Yours sounds great.
I don't like bluetooth...can't fix it with a spanner. Which is a thing I appreciate a lot in a gizmo.

Higher voltage means cheaper solar too.

Any way you get to a camper having a stand-alone sustainable power system is an achievement in my book. They're a rarity...should be normal no?
Extra points for Winter.

22-02-2018 11:12AM

air wrote: »

How much do those run?

Ah mega-buck$$!

Allegedly cheapest per kWh extracted.

air · 22-02-2018 11:20AM

Sir Liamalot wrote: »

I don't like bluetooth...can't fix it with a spanner. Which is a thing I appreciate a lot in a gizmo.

It has a UART port, comes with a bluetooth module but you can plug in with an FTDI adapter and use the PC application if you want. Only necessary for setup.

Sir Liamalot wrote: »

Higher voltage means cheaper solar too.

Exactly, cheaper panels, cheaper wiring, cheaper contactors, cheaper everything. 12V is only fit for USB, laptop chargers, lights, water pumps etc

My 8S system is ideally matched to the bog standard 250W panel with an MPP voltage of around 33-36V. I'd recommend 14S though, voltage range best suited to 48V inverters, use 2 x 250W panels in series then.

22-02-2018 11:31AM

12volt without an inverter works.
I recommend higher voltages for inverter systems.

Yes split voltage systems have a lot of potential. Double wound, dual output alternators would make that a lot more interesting.
I wouldn't series big panels like that. Shading losses will take out the array. I have two old 60Voc panels would be better suited on smaller per panel controllers.

air · 22-02-2018 12:11PM

Sir Liamalot wrote: »

12volt without an inverter works.
I recommend higher voltages for inverter systems.

Agreed, but it's mainly useful due to the ubiquity of 12V accessories. Considering their typical low power consumption it does the job fine.

Sir Liamalot wrote: »

Yes split voltage systems have a lot of potential. Double wound, dual output alternators would make that a lot more interesting.

Hardly worth the hassle unless you really need fast recharge. That's another benefit of lithium though, very fast recharge up to 85-90%, which is all you need.

Sir Liamalot wrote: »

I wouldn't series big panels like that. Shading losses will take out the array. I have two old 60Voc panels would be better suited on smaller per panel controllers.

Easy enough locate a vehicle to avoid shade on what is a very small array. I bought a similar panel myself though to allow single panel charge, was more for the size reduction vs 2 panels that I did it though.

22-02-2018 01:48PM

air wrote: »

Hardly worth the hassle unless you really need fast recharge.

Live-aboards need fast charge. Small batteries with big loads need fast charge.

22-02-2018 01:49PM

I've finished fixing your gauge now Victron!*
*Other than it having a brownout if you power it from a cranking battery and loses it's memory.
So I can't see cranking load voltage drop.
Which would be a nice feature in an automotive standard 52mm sized gauge.
A feature that the former usurped 52mm voltmeter could manage.

I've added a yellow powah detector, on the switched dash illumination line. It's a tasteful normal looking light that isn't confusing for a headlight indicator. It's common red, green and white of an RGBW led strip filtered through the retaining bezel an opaque bakelite diffuser.

Turns itself on and off in times of want, without me having to faddle with it.

Did you notice the backlit keys Victron?
Yurp, backlit keys Victron!

air · 22-02-2018 01:50PM

Sir Liamalot wrote: »

Live-aboards need fast charge. Small batteries with big loads need fast charge.

One more reason to go lithium then! Easy, efficient fast charging as standard!

23-02-2018 06:10PM

Brownouts eh? Not :cool:!

Cranking Voltage at meter power.

Hmm meter response time in question.

I have an idea...I could outsource the problem to a capable meter..

Hmm not yet...but good idea isn't it Victron?

We'll build it a UPS and consider a boost regulator set to 9.5V for the future.

Voltage drop across the schottky. Global 0.3V reduction on B2 readings.

It's a fix not a solution. Meter isn't so sensitive to undervoltage...bless it's little heart

...I might fix the B2 offset by other means later.

24-02-2018 03:48PM

I finished my homework Victron!

I separated the 12V ins on the meter and suspicions correct we have Vsense and Vpowah.
Thank you Victron! That's a great feature I wasn't expecting and very professional. The meter doesn't measure it's own power feed.

I used the last spare core of the trailer control cable I ran to the centre console.
The meter is now powered from B1 the Service Battery. Doesn't start engines..often, hasn't ever seen the wrong side of 11V.
New bigger better UPS with no series diode losses.

The sensed lines are switched but they can lose power without affecting the meter memory. BMV reads 10.5V cranking voltage drop. No fussin'.

I like it when meters agree.

26-02-2018 09:56PM

I was trying to avoid running a sense line to the engine battery. TriMetric doesn't need it because the engine is off when that's of service. The SmartBank doesn't need to be that accurate.... but having the BMV on the fuse block and not the battery meant the auto-electrics are running on the same conductor (albeit 95mm²).
This meant I was measuring the fuse block and not the battery. :rolleyes:

+150mV of actual at full lighting load (powered from charger supply)

I really was not expecting this meter to be so accurate, in light of it's double decimal resolution and it's ability to monitor a voltage variance between linked batteries...

..I owed it another proper sense line. I buy good meters so I don't have to waste my time thinking about these things. I don't want to look at the thing and then compensate, that's aggravating and janky and not the point of a higher end instrument at all.

The sense line runs to meet the old at the fuse block then splits into the Trimetric and BMV.

Not worth losing sleep over ±150mV under certain load conditions..ya know! :pac:

21-03-2018 01:22AM

I changed my mind I'm keeping the TriStar MPPT in reserve...I have much better plans for that.

The industry would have you believe MPPT gives you a 30% gain over PWM. Well, it does, in contrived circumstances, namely very cold weather,
with badly spec-ed panel to battery voltage, unrealistically low battery SOC and zero cable losses.

I intend to do my own comparative analysis of parallel systems one of these days to gather real-world data. Let's suffice to say I expect a 10% gain from this upgrade.

Solar panels are current sources. If you spec-ed the panel right for the climate (32 cell in Éire) PWM can outperform MPPT because it has less quiescent!

Ahem....Current Source!

In my opinion, MPPT is the lazy, expensive and less reliable way to optimise a system.
Which doesn't justify the price difference in controller but then I didn't pay RRP for this (do I ever?

)

However it has a very considerable advantage for 12V. The biggest "12V panel" you can get is ~170Wp last time I checked and it's disproportionately expensive to higher voltage panels.
MPPT does enable you to use domestic panels that have a much pocket friendlier price per watt.
Which still doesn't justify the price difference in controller....

Behold the MorningStar

15A SunSaver MPPT with MODbus interface & FTDI USB to RS232.

First things first let's see what she's made of!

Shiney. :cool:

Next up pull the data...this is the best part of used controllers..You can see how their predecessors ran their system.

What the eff?! :eek:
That poor brutalised battery.:(

Controller Odometer 26280 operational hours.
3 years old...that's nothing for a MorningStar Controller

Ok so I'm putting 20A of PV and 20A of mains rectifier on a 15A controller?
Well..er yes but...It's flat mounted (usually) so derate 20% in Summer and say 40% - 60% depending on season and usefulness (ie. derate 700% in Winter
when it's not useful... 7 times eff all = eff all).
The rectifier I've already derated on over-temp, so no loss.
The controller has current limiting so I can over-spec the PV and get it to run at a higher nominal...honestly the losses when it actually makes a difference
I expect to be minimal.

Solar Install; the on-going saga

Comments