Solar Install; the on-going saga

Pulled the PSU board out to start repairs..had a look at the high current traces and well...they were a bit iffy so I bolstered them. One or two more to do now that I'm looking at it.



This is not the same impact done the rectifier. This traffo is broken on a 4 corners of the base. Death from above, the rectifier was an abrupt stop from a g-force or two.



E-poxy that, she'll be grand!

Rectifier passed a health check.



Replacing it anyway, it's brittle now.

I reckon it's about 60/40 new kit to repurposing old.
I haven't spec-ed an actuator yet. One with built-in limit switches would be nice and easy if they work.
Probably go UPVC window friction slides again they work pretty well.

New bus bars landed.



The 100A is actually 100A; 30mm² bar.
The 250A is closer 200A being very generous; 50mm² bar.

The thing you'll find with ELV electrics for mobile installations is it's the wild west out there! There's no regulations. Most things are ridiculously over-rated and rely heavily on the users lack of imagination and current limiting installations to prevent lawsuits...actually works quite well as a marketing strategy. Bigger numbers sell.

I derate Bosch relays by 50%, cigarette lighter plugs by 80%, leisure battery capacity by 15%, thin wall cable I use but ignore the ratings and use IEE instead (so 40% derating), I don't use any motorhome/marine specific switchgear, stock chargers I don't rate at all.

The "250A" bus bars are handy because they've 4 x M8 and 3 x M4 but the M4s on the positive must go directly to fuses because it's unlikely you'll be using that stud with 200A cable.

I heart DIN rail.



I think I might even use the right colour coding for this new install. :eek: Ah I'm sure they're be plenty of mains cable thrown in inappropriate places because I have loads and I like to wind up sparkies.

All stranded flex though because that solid core's muck!

Today's top tip; Shrinking holes in yer conductors.

Really easy do. Say you need to Ctrl+Z a hole you drilled in your copper flatbar, or you want to use a smaller bolt in a shunt, or you want to recycle a busbar...

All you need is a smaller hole or scrap copper, solder and some tools.

Here's a hole I made from copper Pipe.



Make yerself a hole press...fancy name for "run a bolt with some washers an a nut on through it.



Get yer weapons of choice out...





Run the bolt down so it compresses the smaller hole into the bigger hole and leave the bolt in for now.

Then use a flame thrower and solder to fill in any gaps.

You can quench cool it, but you'd be better off air cooling.

Run the bolt out and file or grind it down.

Presto shrank hole!



You can polish it if you want but the electrons won't care.

This is a side project I'm working on which will never pay because it's got too many skilled hours and hi-tech gizmos than Joe Public will be willing to pay for. Alas I'm not compromising, this will be the best entry-level multi-source charger the market has seen or it'll never make the field.



Actually spent most of today ironing the wrinkles outtov this...

I've commenced soak testing and it's looking promising. Meanwell make a top class SMPS. Surpasses it's rating straight outtov the gates. Built-in current limiting, over-temp protection, potential free relay, 100VAC - 250VAC in, adjustable Vout up to 16VDC.

Mains works great, need to add thermal management and some protective diodes.

MC4 connections on top, we start PV testing in a few days.

Split charge 100A capable with 800A instantaneous....which is where today's adventure begins;

The Smartbank has a really neat feature, if you use an illuminated momentary switch it lights up without interaction every time intelligent split charge is active. That's what the little power symbol light switch is at the bottom of the enclosure. If you press the switch off-charge it's a start assist that parallels the batterys for a setable amount of time to crank then disengages or cranks the engine and does it's thing because alternator online. However...the bloomin' schematic is wrong...



Here's the right one I figured out..don't ask how long it took....



The other wrinkle was the ammeter I'm including wasn't working.
Now a lot of people ask me what's a good meter and I say SmartGauge or TriMetric.

Here's ApprenticeVolt's €300 attempt that's so good I had to put another €5 meter beside it so I can read the relevant information without leaving my seat. :rolleyes:



Here's a cross-section of some of the (relatively) cheap meters I have tried to find a suitable affordable alternative that have failed to cut it...don't do what Donny-Don't does.. that pile of poo woulda got me a BMV..get the good one first...



Where was i....oh right yeah...stoopid Ebay meters, 2 hours into trying to calibrate/wire/not defenestrate the yolk, after realising I was using an unsuitable shunt and marmaladed a new one together, I decided I'd had enough and it was RTFM time. No manual online...:rolleyes:
..ooo...K! Another hour and it dawned on me..it needs an isolated power supply

TriMetric
Smartgauge
Victron BMV
Don't go cheaper..what's your time and hair worth?

Final word of warning, if you don't have an expensive meter to calibrate a cheapo with then don't even consider a cheapy, such a thing could cost you a battery.

I've got the adjustable CC SMPS all fixt and ready to rock, well...after I grind down those mounting brackets that is.

I added a screw to prevent further board damage from a vertical mount in a vibratey situation. There wasn't really an ideal isolated location for it, so I just drilled through the PE trace and I'm callin' it a feature..."secondary chassis bonding"



Glued all the caps and the two damaged traffos for the same reason. That should be silastic but I used hot snot because that's all I have.



New rectifier mounted better than a bought one.



If you look at the displacement of the legs compared to the original it would indicate that the damage was probably done in assembly when it was clamped, my legs have an extreme contour where the original was more fire and forget.

I also went around and bolstered all the rest of the high current traces
but never took a photo...:o

I should probably stress that having your hands inside a SMPS like this can kill if you don't discharge the bulk filter caps. You can do this by either turning it on while it's unplugged or shorting them with an insulated tool. I'm not one for woe-betiding, but it is something to be cognisant of working with these things.

I built myself a test rig from 2 car headlamps with both filaments wired on both bubbles.



230W-ish Test Load
Careful these can get a bit melty...attended appliance folks.



I do like it when gizmo gadgets exceed their ratings.
[EDIT: It does I promise, I've just uploaded the wrong photo...and never took the right one. ]





Stabilised External Heat at FET Mounting Screw @ Full Load.

Hrmmm, those black jacket SamXon electrolytic caps won't live long in there...meh.
I'd routinely replace them for something better for someone else's setup. For me, eh they'll be grand I'll wait until they die of heat stroke before I replace them. I rarely use shore power, not like some installations where it'd be expected to run 24/7.
I could throttle it; 0 > 20.2A adjustability. :cool:
or add cooling...we'll see, I doubt I'll bother though.



PWM Wave



Pretty clean DC (yellow)

Low Ripple AC (blue), there's probably more o-scope lead attenuation than true ripple in there.

She's a keeper.

I got my hands on the fridge Mwahaha!

Hey look it's got an automatic T-stat...like one for grown-ups!



I was mistaken previously, there's no mains kit...again, don't think I'll bother. Easy to retrofit down the road anyways.






What are they doing? Using the freezer to cool the fridge? No insulation what so ever on the freezer compartment and massif gaps



hmmm...12v compressor...that's what I call efficient onna 12V system.





It came with one of these horrible things on it. I'll be deleting that.

"20A" :rolleyes: Cigarette lighter plug deleted. Temporary prototyping spades installed.

Fridge & freezer loaded with thermal mass (water) to stabilise internal temperature for testing. Roomy in there. Only a few mm bigger than my current lemon and feels like twice the storage...deeper I think.



Inrush & running load.



Operating Temps at Continuous Duty



I can improve this...I think I'll reappropriate my 4 x 120mm abortion absorption fridge cooling low RPM PC case fans fed from the switched side of the T-stat for this.
I know some of them work because I hear them spin up every now and again, can't say if all of them work, they've been in the elements for 3 years.



This is what it uses in the first 11.5 hours getting down to temperature. Minus 5.2Ah already discharged at commencement of test run & minus 200mA * 11.5 hours = 2.3Ah (meter calibration discrepancy):
38.4 - 5.2Ah - 2.3Ah = 31Ah to operating temp first 11.5 hours

How am I running all this stuff in the back room of my house?
Oh I'm glad you asked. :pac:

Let me introduce you to the Sir Liamalot prototyping laboratory of electron enhancement:

This is a small corner of a much bigger project.

Most of this gear is scrap, ex-service, was bought broken or untested was donated.

The leftmost battery here powers my 12Volt supply. This is refilled with sunlight daily.



Here's the beating heart of that system. A humble MorningStar ProStar 30 PWM and distribution. The meter's annoying it automatically cycles PV Current In / Battery Voltage / LVD Current Load so you have to pause yourself and wait for the info you want.
The charge-controller ought to be in battery-land not interface-land because of voltage drop mitigation. I don't recommend solar controller meters they're pointless. This one was bought used and cheaper with the meter.




Here's my 12-ish volt lab power, wired from the LVD terminals of the ProStar.





I'm very hard on my Victron 15A charger, mostly because I think it was an over-hyped waste of money that's not fit for porpoise, and yes you can argue this is not a fair appraisal of it's performance because you are not supposed to put chargers on the low voltage disconnect terminals of a charge controller nor should you try to operate it while said charge controller is also operating.



Here we can see it floating my lab 12v battery, after reaching 62% apparent SOC after the initial fridge load discharge test. It had dropped to 2.4A float which is what happens when a solar charge controller and a mains charger try to work in disharmony. One at a time! ..having said that the ProStar knew what to do and was still hard at it.

After I isolated the 110Wp solar panel the Victron charged the battery to 90% SOC hours and hours later before I gave up on the choccy-pot and resumed fridge testing.



15A Charger "@ Full Load" on LVD charge controller terminals.

While I'm wittering I might as well give out about this BTM III too, the cal has drifted -200mA since I set it (that could be an efficiency compensation setting), the display is too small to read unless you go fiddle with it, the SOC leds are too broad because nobody sensible uses the lower than 50% SOC of a LA battery (so I changed the battery capacity to 50% plated so they work better..apparent energy not actual), it reads time until discharged but can't manage time until charged, it doesn't display watts, it doesn't register > 100% charge.
Don't buy one, waste of money, you can do much better for two thirds the price.

I'm presently running a 24hr energy consumption test at normal operating temp. in 15°C nominal atmospheric.



There's some heat ingress, I may add external insulation but I'm already tight on space.



Normal Operating Refrigerant Temps.



Hmmm..not a lot to be gained with active cooling.

If a customer asked you for a 6pk1990 belt which of these two would you think it most appropriate to offer them 4 times?


Still have the belt they wouldn't issue a refund without the invoice that's on their system in my hand. I nearly lifted the manager across the counter with it when he tried to convince me the 5mm was insignificant.

Gates make a better belt. More supple, more sticky.

Load tested to 50A.
No squeaks.
First time ever the battery light turns off without an engine rev after cranking.




Not that I have any particular reason for saying this but use pneumatic impact wrenches for swapping pulleys. Electric ones might not get you all the way there.

..just bring the alternator and new pulley to a tyre shop. I'd be very surprised if they charge you.

24hr cooling load in 15°C nominal atmospheric.

38.4 - 8.9 - (0.2 * 24) = 25Ah p/d. Bone Stock.
Shall we shoot for 20Ah? Place yer bets!



Being the obliging sortov chap I am I gave the Victron another chance to prove it's worth, and besides my lab service battery needs some love.

It can push 14.9A for about 5 seconds. What does that prove?...well technically they're not lying about the rating, more importantly it has nothing to do with being on the LVD bus.



If I was rating that I'd call it a 12.4A charger...but I guess I'm odd like that.
Now that's a Victron, while I consider that pretty poor and unsuitable for fitment if you compare that to a Calira or Electroblok....it's the dogs danglies...:rolleyes:

I hope you good folks are realising the importance of a battery monitor by now, maybe understanding why I'm so cynical pragmatic about the industry taking the market for a ride on it's magical fairytale of deceit and misnomers.

The world is crying out for new and better batteries that can outperform the previous generation. Well shucks if we actually had equipment that was fit for purpose at a fair price and installers who knew what they were doing there'd be no dispute what chemistry is the most capable.

Maintenence charging every week only applies to chargers like that Victron that do half-bottomed jobs. I run up to 6 week deep cycles and have no deterioration according to discharge and electrolyte testing because the charging I use can and does reverse negative chemistry 100%.

autumnalcore

Is that an old shoreline fridge?
25Ah seems a lot at 15C - 1.08MJ about the same energy to cool 25l of water from 15c to 5c insulation probably very poor.
At that rate a modern r600 mains mini fridges plus inverter losses might be better - new husky mini fridges are 84kwh/annum at 25C should be a lot less at 15C.

Dunno, haven't looked it up. She's old alright, flash rust aboot. Metalwork needs a respray.
Yurp I second the insulation is poor I've been poking it and I reckon a 10mm hi-tech jacket and some caulk will do her the world of good. Any recommendations of an impermeable celotex/kingspan type substance?
Refrigerant sounds bubbly too, might be low, do people still re-gas these things or just chuck them away?

Looks like the latest generation here.
Same Volume give or take. Rated 1.1Ah ...probably means drop the h because numpty.

In the interest of fairness I just rewired the Victron direct to the battery terminals and nope...still junk.

The more I think of insulating it the more I think I need to rebuild the cabinet, the more I think I could move the water tanks and fit more battery...jaysus I'll be buying a new van altogether before I'm done...:rolleyes:

gctest50

Top loading 12v fridge be better, all the cold air doesn't pour out when you open up the door

I keep the cold inside by loading it with coolant, ie, spare water. Close the door again and the cold bottles cools the fresh air.

Yurp top loading is better but I like the access convenience of side opening and seeing as I've gone and bought one I'll carry on.

Don't fancy the slides and eternal rummaging of a top loader for this I will compensate with powah.

autumnalcore

Sir Liamalot wrote: »

Dunno, haven't looked it up. She's old alright, flash rust aboot. Metalwork needs a respray.
Yurp I second the insulation is poor I've been poking it and I reckon a 10mm hi-tech jacket and some caulk will do her the world of good. Any recommendations of an impermeable celotex/kingspan type substance?
Refrigerant sounds bubbly too, might be low, do people still re-gas these things or just chuck them away?

Looks like the latest generation here.
Same Volume give or take. Rated 1.1Ah ...probably means drop the h because numpty.

In the interest of fairness I just rewired the Victon direct to the battery terminals and nope...still junk.

The more I think of insulating it the more I think I need to rebuild the cabinet, the more I think I could move the water tanks and fit more battery...jaysus I'll be buying a new van altogether before I'm done...:rolleyes:

For insulation kingspan is usually 20-25mm minimum. A couple of layers of foil backed camping mat can usually be had this time of year for a couple of euros. Attach with high temperature spray adhesive.

For regassing you can use a £5 self piercing saddle valve to do a DIY regas but tis a bit dodgy I wouldn't unless all else fails.

autumnalcore

autumnalcore wrote: »

At that rate a modern r600 mains mini fridges plus inverter losses might be better

I doubt it, inverters always have inflated efficiencies and I'd rather not have one. No need in an installation you can touch opposite walls from the same position, just adding losses to the system imo.

Any data from yours?
The only way I'd consider it is if it's cheaper because it's less reliable and my shiney new danfoss was a (relative) steal.

autumnalcore wrote: »

For insulation kingspan is usually 20-25mm minimum. A couple of layers of foil backed camping mat can usually be had this time of year for a couple of euros. Attach with high temperature spray adhesive.

You can get celotex in 12mm but that's 24mm I don't have without major renovations. Under, over and behind are no problems, the fridge walls are already 25mm thick. What's the R-value of 12mm birch ply?

Canal boat forums are suggesting blue board because it's nonabsorbant.

The way I'm going I don't think I'll be finished this for months I can't help but want to fix everything.

autumnalcore wrote: »

For regassing you can use a £5 self piercing saddle valve to do a DIY regas but tis a bit dodgy I wouldn't unless all else fails.

I'd say it's fine, not worth the liability. let's assume that informative "datasheet" rating is indicating the fridge runs @ 1.1A per hour average, 24hr = 26.4Ah...let's call mine within design tolerances.

air

Sir Liamalot wrote: »

I doubt it, inverters always have inflated efficiencies and I'd rather not have one. No need in an installation you can touch opposite walls from the same position, just adding losses to the system imo.

Any data from yours?
The only way I'd consider it is if it's cheaper because it's less reliable and my shiney new danfoss was a (relative) steal.



You can get celotex in 12mm but that's 24mm I don't have without major renovations. Under, over and behind are no problems, the fridge walls are already 25mm thick. What's the R-value of 12mm birch ply?

Canal boat forums are suggesting blue board because it's nonabsorbant.

The way I'm going I don't think I'll be finished this for months I can't help but want to fix everything.




I'd say it's fine, not worth the liability. let's assume that informative "datasheet" rating is indicating the fridge runs @ 1.1A per hour average, 24hr = 26.4Ah...let's call mine within design tolerances.

I've a newer fridge with a similar compressor, bubbling and running water type sounds are totally normal, it's just the refrigerant circulating. If the current draw is right and it's cooling effectively I'd leave well enough alone.

gctest50

Sir Liamalot wrote: »

Any recommendations of an impermeable celotex/kingspan type substance?
............

Aerogel if you've no room to play with ? :

http://www.proctorgroup.com/sector/custom-aerogel-insulation


http://www.oldbuilders.com/insulation%20centre/Aerogel/1_aerogel_01.htm






/

Aerogel....thanks...but...I think I'll bodgineer someat. That stuff will probably cost more than my fridge.

Project is going on hold.
Got fish elsewhere in need of fryin'

It occurred to me I've never officially introduced yee to the humble shunt.
Here's the one on the BTM which is actually the reason I bought the meter in the first place because it was being sold as "broken" which I just figured meant the previous owner didn't wire it right or realise the limitations of Ah counting. Turns out I was probably right. :pac:



Brass bolts are better conductors than stainless.
Don't use BZP on electrics that's just asking for electrolysis issues.
Shunts will come with appropriate bolts.

The extra black wire is my solar controller ground sense.


Shunts are precision resistors.They're far more accurate than hall sensors...clamp on jobs.
They do waste a little power which is why I over-spec them to mitigate this.

A 500A 50mV shunt has a full load rating of 500A and there's a 50mV drop between it's terminals at full load. So 500 x 0.05 = 25W waste heat at full load.
The cheapo chinesium jobs are 75mV.

They're easy to wire, don't mind them all-ya-need-is-a-voltmeter bunch, how many of them can back that up with an operational 7 year old battery?
I can read a voltmeter as a SOC meter, it took years of practice, and given the choice I will fit a real meter.

There's two very important things to remember wiring these.

The battery and

only the battery goes to one side of the shunt main terminal. Everything else (except a SmartGauge) goes to the load side.

• The sense wires must be the exact same resistance. Twisted pair is best.
  
  
  
  I used Cat V in the house and paralleled the conductors. I would not use this cable in a vehicle because it's too fragile and parallelling conductors is shoddy practice. The solid coloured conductors are twisted pair to the striped ones so this is how I paralleled them. All solids and all stripes.
  The meter calculates the power across the shunt by using these conductors to sense the voltage drop across the shunt blades.
  
  These are the only terminations on a vehicle I will solder that's not attached to a circuit board. This is because resistance of this termination takes precedence to mechanical fortitude.

autumnalcore

Sir Liamalot wrote: »

Any data from yours?

Nothing useful because we supplement with freezer blocks but we only plugged in once this summer after a grey rainy week down the bere peninsula sapped our solar.

autumnalcore

autumnalcore wrote: »

...we supplement with freezer blocks

I'd call that normal operating conditions. The reason I've shy'd away from the inverter road is I reckon it's grand for brief 2-4 day expeditions but hardly as sustainable as 12V for 3 week escapades in the Winter where I've already been feeling under-spec-ed on lead. Not yer typical use granted.

I've always felt that the more durable plumbing on a 12V compressor for vibration resilience was more bluster than substance and not a justifiable reason for a pricing disparity in the order of €350..or 6 fridges in lay-terms. It's more akin to just charging more because you can get away with it, the usual leisure market retailer lark.
My plumbing certainly doesn't look any more rugged than anything I've seen in white-goods town. My condenser isn't retained on the upper middle, the compressor has 2 outtov 4 bolt-holes bolted and loosely cinched.

My Zolotnik B2B is pewter grade. Two of three regulators committed hara-kiri on the test bench. One is still alive but not boosting, I let it live to tell the tale. All of them pulled >20A to push 2A the first two let the smoke out within minutes of inception. Not one of them reached target voltage. Garbage...
PNP anyways...not much use....not bothered resurrecting them...sending them back. :mad:

Meanwhile, I thought this was looking lonely and decided to foster it. It might change everything or nothing.

autumnalcore

Sir Liamalot wrote: »

I've always felt that the more durable plumbing on a 12V compressor for vibration resilience was more bluster than substance and not a justifieable reason for a pricing disparity in the order of €350.

The extra €350 is definitely not justified. I don't think its the visible mounting and plumbing that's the difference but the internal mounting and the composition and arrangement of the discharge tube that makes it more resilient to the continuous shock / vibration vs periodic shock/vibration in a stationary application. Or so you would hope although clearly not infallible:
http://www.nomadic-one.com/travel/index.php?option=com_content&task=view&id=71&Itemid=2

You can see the circuitous route the discharge tube takes vs the typical shock loop you see on a mains compressor.

Junk...why bother...honest question?...




Thermal paste not spread evenly on high current components.
Sketchy thermal pad placement. (Upper Right)



QC passed Dry Joints



I can't argue the off-brand electrolytics are outtov spec because they're within ±20% tolerance. However it's an uncanny coincidence that every one I tested is 10% under-rating.






Bench Test Fail.
Back of the Class!



www DOT waste-of-a-stamp DOT com

:cool:

Offset Crimping; more progressive flex compared to parallel butt connections, less line bulk.


'course nothing beats continuous runs.

Peek inside a danfoss 24v/12v motor controller.



Conformal Coating Galore.
She'll weather well.