• Have you tried out dark mode?! Scroll to the bottom of any page to find a sun or moon icon to turn dark mode on or off!

diy solar

diy solar

From office UPS to whole house... my journey

There are youtube vids that show how to make the lights work without the shunt. It requires making a cable. Looked pretty simple.

Yep. Alas the inline transformer is only good for 12V & 24V systems. A 48V->5V transformer is a lot bulkier. I'll be using my Cerbo GX as a direct 5V supply via one of its USB ports. No transformer needed (hopefully it will work). Of course, I'll post my results either way.
 
Yep. Alas the inline transformer is only good for 12V & 24V systems. A 48V->5V transformer is a lot bulkier. I'll be using my Cerbo GX as a direct 5V supply via one of its USB ports. No transformer needed (hopefully it will work). Of course, I'll post my results either way.
I'm guessing you are also aware that you can't get Cerbo monitoring of the fuse status without using the lynx smart BMS unit, even with the power hack for physical light indication.
 
2) I used 1/0 (150 amp) DC cables connected between my Lynx Distribution box and the MPPTs. The Lynx Distribution is fused with 125Amp fuses, so hopefully they are the weak point. I had to really wrestle with them to get the lugs over the bolts. Suspect 2 AWG with smaller fuses would have made life easier.

3) These MPPT controllers need 300mm (about 12 inches) of open space above and below them, thus the spacing. The top one is 12 inches from the ceiling. The bottom one is 18 inches from the top one - so hopefully won't feed too much heat into the top one. I really didn't want to go much lower on the bottom one given all the MPPT feed cables that will be hanging from there.
What do you mean about wrestling the lugs over the bolts? My 125A megafuse is the same size as my 300A, they easily fit over both bolts.

18" should be plenty of space my 450/100 only provides gentle warm output at full power.
 
What do you mean about wrestling the lugs over the bolts? My 125A megafuse is the same size as my 300A, they easily fit over both bolts.

18" should be plenty of space my 450/100 only provides gentle warm output at full power.
Sorry, wasn't clear: Wrestling the 1/0 cable lugs over the 450/100 lugs. Lynx Distribution was, as you indicated, a piece of cake (other than a tight socket fit on the negative bolts)
 
Sorry, wasn't clear: Wrestling the 1/0 cable lugs over the 450/100 lugs. Lynx Distribution was, as you indicated, a piece of cake (other than a tight socket fit on the negative bolts)
Ah yes, I know what you mean. the entry oval hole in the bottom isn't fully aligned with where the battery cables drop vertically down, for the positive.

Screenshot_20240725_192445_Gallery.jpg
 
Sometimes things are just TOO easy:

usb.jpg

What are you looking at? I needed a 5V power supply for the Victron Lynx Distributor. Some folks with 12 and 24 volt systems have built a cable using an inline transformer to do the job and feed it off the Lynx input terminals. I'm running 48V and haven't found any small transformers down to 5V. So... 5V... hmmm.... that's USB voltage... and my Cerbo GX has spare USB ports...

So I grabbed an ancient USB Printer cable figuring I aught to be able to find the power wires. Oh... red, black... bet those are them. Yep!

Now just waiting for RJ10 connectors. Thought the Lynx used a standard RJ9, but no such luck. Should get them in a few days.
 
RJ10s (4 pack from Amazon) arrived today. Bought a new crimping tool as well, just to make sure I had the right one.

About 1 minute after I unpacked everything, the crimping was done. Another minute later my Victron Lynx Distributor LED indicated card was powered up and working!


IMG_1828.jpg

That said, there wasn't a way to route that thick of a USB cable and get the cover back on. Found a more modern, thinner cable, and it routed just fine:

Lynx USB.jpg
 
When I started this project, one of my goals was to provide peak shaving capability both at the office (current system) and the future house system.

Spent a few hours today playing with ESS. I was rather excited to see the new menu items appear on my Cerbo after I added the assistants to both of my units. Alas, under the "Peak Shaving" menu, it insist on having an external energy monitor - it doesn't like having the MP IIs monitor the power. So far, I've been unable to find a Victron Cerbo compatible energy monitor that supports Split Phase. Everything listed on the Victron site is single phase or 3 phase only - including the Victron Energy Monitor.

Anybody find one that works?
 
Well... came across this thread this morning, after doing yet another search for a Victron compatible, reasonably priced, energy monitor:

https://community.victronenergy.com/questions/265288/vm-3p75ct-split-phase-any-issue.html

Apparently the Victron 3P75CT energy monitor (about $300) can indeed handle split phase. One just sets it up as a 3 phase monitor and nulls out the third phase. Ordered one today in hopes it will enable Peak Shaving in ESS. Guess I'll know in a week or two.

Of particular interest was the response from Victron to the post.
 
So received my (29) Hyperion 400W bifacial panels yesterday. Not seeing any shipping damage, so will have a few extra to use on different projects. All good there. One wrinkle... I was planning on mounting these horizontally, using brackets on the 45" side instead of the 68" side where the wholes are drilled... screwing directly into the 6x6 post. (Sorry, vertically mounting these first one, facing east/west, stacked 2 high on their sides.)

Looks like it would be easy enough to drill new wholes in the frame, presuming I back stop the bit to prevent it touching the glass on the other side. Is there any wind strength difference if I do that vs. using a set of the existing holes along the long side? I was going to have 2x6 cross boards between each post, I could screw into them instead... I am presuming 4 points per panel is sufficient? The panels have 4 hole per side, I presume to allow for some mounting flexibility? Any reason to select the innermost 2 vs. the outermost 2?

The mounts I bought, Z-bracket style, came with 1.5" self taping metal screws - which would be great if I was mounting on my tin roofs, but wondering if I should replace those will wood screws for a better grip?

IMG_1835.jpg

One removed for planning purposes, blue tape is on me... tried to cover the electrical stuff with plastic in order to avoid the horses finding it.
 
Last edited:
Just for documentation purposes, here is the field the vertical east/west panels will be going into:

IMG_1836.jpg

Planned on putting them about 12' from the edge of the driveway, but took out my iPhone compass and discovered the driveway is about 15 degrees off of true north. Squaring it up would put the other end right at the driveway - and that won't do (winter snow plowing can easily result in a 6 foot off driveway skid). Might have to live with that 15 degrees - Wife does care what this looks like, will show here later today and get a reaction.

Screws up mowing as well since I have a 6' brush hog. Although the horses have keep the grass parts pretty low this year.

FYI - those are not weeds. Over half of our farm has been replanted in "native" prairie. A couple of weeks ago that whole field was in bloom and quite pretty.
 
Wow... can I use the Can bus??? Ends up its a good thing the Cerbo GX has two VE.can buses supported. The SOK batteries, via the special SOK to Victron cable, need to be run in CAN-bus BMS LV mode (500 kbits/sec). My 450/100 MPPT controllers appear to prefer the 250 kbits/sec as does the Victron Energy Monitor.

Found this link which further states that the battery BMS cannot be connected alongside other can bus devices. Probably explains the 69% packet error rate I was seeing... The link also talked about the speed differences.

So... oddly know I can see all my PV charge controller, the inverters, and the current meter on Can 1, but the batteries don't show up on Can 2. That said, everything, including the batteries are showing up under Devices, but not the current meter. Still playing...

Enabled Mod-Bus TCP/IP since the Energy Meter setting in the GX suggested it. Hooked up the Ethernet as well.

The energy meter is showing up as a Mod-bus device, but its not staying connected. Time for bed.
 
Last edited:
OK - bed while I have a problem just doesn't work for me...

So... progress. While scanning articles on the Energy Meter I saw people running 1.05 firmware and I recalled mine being at 1.01. Took awhile to realize one has to use VRM to update firmware - there doesn't appear to be any way to do that locally, except for the GX itself.

Once that was updated (did my 450/100 PV charge controllers too), the energy meter showed up. Bit odd though: Its showing total watts, but not by phase which is just "--". Numbers feel about right.

... an hour later ...

Fighting various weird situations, like VRM reporting the Energy Monitor at firmware 1.06 but Remote Console VE.can 1 devices showing it at 1.01 (and randomly disappearing from the high level Device list). Rebooted the Cerbo GX and many things look much better, like the phases and amperages on each leg are now showing up instead of being "--".

So far, ESS hasn't worked as hoped, although AC Input was "--" until I did that last reboot. So far, when I set it to Peak Shaving, it just goes into passthru mode, even when I exceed the power limits. oh, I did upgrade my Multiplus IIs to the current firmware. One more try...
 
Last edited:
Hmmm... maybe. The inverters are bouncing around between normal, passthru, and inverting mode. This has hope. I'm watching the Energy Monitor via Remote Console -> Devices and have not seen the two phases exceed 10 amps yet. One or the other will briefly exceed the 4.0 amp input current setting defined within the Multipluses, so unless that is just a "goal", not a hard setting, they do not appear to be ruling the day. Under ESS setting, I have the grid setpoint at 950W, yet I've seen, again briefly, over 1100W being pulled per the Energy Monitor. Multiphase Regulation is set to "Total of all phases". Under Peak Shaving, maximum system input current (per phase) is set at the minimum, 5Amps - yet I briefly see a leg jump to 8 amps. Time will tell. Overall goal is to keep my 15 minute average Peak below 1KW.

Lessons learned:

1) You can only update firmware via VRM.
2) Updating Multiplus IIs is a bit tricky, and they warn you. Basic configuration settings like Split Phase are saved across the update, but other settings are lost, like any specific absorption or float voltages you want to use.
3) Don't fool with ESS until you get your energy monitor fully working.
4) Peak Shaving setting under ESS (which was the whole goal) requires an external AC energy monitor such as the one sold by CurrentConnected.
5) No problem using the AC In lines on the inverters for the Energy Monitor clamps, they fit just fine
6) The Energy Monitor reports AC In data a LOT faster than the internal inverter one does.
7) When in doubt, reboot the Cerbo

A few tech details on the Energy Monitor setup:

I used 18 gauge stranded 2-pair wire, with metal tips added, to feed the N, L1, L2, and PE inputs on the Energy Monitor. One pair was connected to the L1 inverters AC hot and neutral wire, the other pair to the L2 inverters ground and hot wire. No L3 input since I have a split-phase setup.

Plug in all 3 phases, even if one is not in use, so the unused one won't float but read zero instead. I just store it in the bottom of L1 away from any power lines.
 
Last edited:
When I checked in on the remote console this morning I saw REALLY low AC Load numbers... like 30W with AC Input at 750W and the batteries discharging at 798W. Took awhile, but finally figured out that I fat fingered the "Inverter AC output in use" toggle.

BTW - Wife OK the 15 degrees off parallel with the driveway, so will be moving forward with true East/West alignment - or at least within a degree or two. Projecting my iPhone's 1/2" direction mark with 60 feet of twine isn't the most accurate thing to do in the world, but it will be a lot closer.

Still don't understand what is controlling my AC input current. Can't find anything I can change that makes a difference. Goal is simple: Draw up to 1000W from AC-In, supplementing with battery if AC-Load exceed that. Charge batteries if AC-Load is below that. Think its time to post on the Victron Community page.

ok.... posted: https://community.victronenergy.com...ols-ac-input-current-in-a-peak-shaving-s.html
 
Last edited:
More playing around: Apparently the Shore-Limit ->AC Current Limit set on each inverter is the governing factor.

A few tidbits when playing with this:

1) Don't forget to send any changes to "all devices", or manually do it on both
2) You need to be COMPLETELY out of VE.Configure3 and I think the VE.Bus Quick Configure tool before the changes will take effect
3) Give the changes a few minutes to engage
4) At least with my model, I've taken it as low as 4 amps and it appears to work. Remember that Wattage = Amps * Voltage, so 4 amps is close to 500W. Lower limit seems to vary, a LOT, between models. I've attached a zip of the excel spreadsheet, but note its dated December of 22, so things may have changed.
5) Reviewing that spreadsheet, I noticed there is a Minimum Powerassist amperage as well. That explains a few observations I think (like why Powerassist isn't kicking as soon as I expected).
 

Attachments

Just had a "Duh" moment - I need DC disconnect switches!

A bit of Googling later, I see they come in Switch format (typically round, 90 degree on/off style), Breaker format, and some come with lightning arrestors built in. Some are indoor mounted, others are outdoor mounted.

Any reason to pick one vs. the others? The Switch format ones almost always advertise they have built-in Arc protection, doubt if the breaker style has that. But do I really care? All appear to be built in fire proof or suppressing boxes.

Not a huge issue, but indoor would be more secure, and convenient. Outdoor would allow for fire department turn off, if I didn't have a security lock on it... and being rural, but the time they get here the building is toast. Vandalism is rare, but these days I could also see "kids" (of all ages), thinking it would be funny to turn off the panels. e.g. I'd sleep better with indoor ones. Added bonus, a quick visual when your about to do something needing a disconnect is possible, vs. running out to the array and checking the switch status.

The lightning arrestor seems like overkill. Panel frames will be locally earth grounded... lightning would follow that path (vaporizing the ground lead in the process, but the plasma would suffice for the duration of the bolt). On the other hand, if it actually worked to prevent a surge getting to the MPPT controllers it might not be a bad idea?

Any issue with using a dual string unit (since I'll have two strings in this first array, that would be convenient, especially if mounted outside)

Thoughts? Comments?

Amazon (e.g. Chinese)? ION at 2X+ the price?
 
Same product with fuses? Somehow feels like a bit of extra protection? Looks like they just provide the fuses as extra items in the package.

Isn't it a bit unusual to fuse both sides of the incoming power?
 
Same product with fuses? Somehow feels like a bit of extra protection? Looks like they just provide the fuses as extra items in the package.

Isn't it a bit unusual to fuse both sides of the incoming power?
I don't have mine fused. I will be putting combiner boxes in at the panels eventually and those have fuses and such. I changed over to the new switch vs the breakers because of the fire risk on the breakers for pv usage. No real danger of fires on a ground mount out from the house. Inside I want nothing that increases fire risks.
 

diy solar

diy solar
Back
Top