Victron and Hub-1 Questions(and a few answers)

[phillipwhittaker]

Hi Guys,

I am currently busy preparing for an installation, gathering all the equipment etc. I bought the equipment based on the hub-1 assistant, and was wondering if anybody here has this working yet?

Equipment:
Victron Multiplus 3000/48
Victron Bluesolar 150/70
Victron Color Control
VEBus to VECan interface cable
4x260AH omnipower gel batteries

Link to assistant details:
http://www.victronenergy.com/blog/2014/03/07/hub-1-available/
http://www.victronenergy.com/upload/documents/Hub-1-System-Layout-V3-4-8-2014ADR-.pdf

Here are the issues I have run into so far.
1.) The cable needed to update the firmware on the bluesolar is stupidly expensive. In the region of 6k, is this something a victron installer should have?
2.) The VEBus to VECan cable is also more expensive than what it could be(3k). Victron have said this will come down with about 30% but still! It also has a firmware version, and I don't know how to update this? Anyone have experience with this?
3.) The assistant requires an Anti Islanding Device. What is this, where do I find one in SA? As I understand it's to prevent feedback if the AC power source has a power failure or outage?
4.) MJ Lorton had mentioned he would be doing a tutorial on the victron range, however I could only ever find some battery monitor videos and some general ones about solar(quite a few). Are there any more like this that specifically cover the details of the multiplus and configuration using MK2 to usb cable ect?

Can anyone tell me what the advantage of this assistant is over a different setup with a gridtie inverter?

Thanks guys, sorry for the WOT!

[MJLorton]

Hi Phillip,

Victron do have some nice equipment but the need for those cables, the complications and costs are not great. Several years ago they led the way with some of the functionality in their inverters. But these days folks like Outback have caught up and make updating firmware etc far easier with a simple SD card...as long as you have purchased their Mate 3. Not cheap but they have their own tutorials on YT.

I stopped working with Victron kit once I left South Africa as Victron are not that well known here in the US.

A Victron installer should have all the necessary cables, firmware etc to get your system up and running correctly. If not...I would not use them.

If you search this forum I think you will find someone that has found a hack to get around one of the Victron cables / interfaces...not sure if that will help you.

I would also ask the installer to refer you to an Anti Islanding Device and confirm that they know how to install it correctly i.e. have they done this before...get them to provide another client as a reference...

Cheers,
Martin.

[kibi]

Hello Philip,

The cables themselves are standard CAT5 or CAT6 network cables like you would use for computer networks. Lekker and cheap.

As far as programming goes, I do use a Victron USB-MK2 adaptor to access my inverter, not sure if the MPPT can use the same adaptor but most programming for the MPPT is done via the front panel.

I hope this helps.

[phillipwhittaker]

Hi,

Thanks for the responses. I have made some progress with this, perhaps this information may be useful to others.

After speaking with Andre at Victron, he suggested that since feedback is disallowed, it would be far simpler to use the virtual switch "Dedicated Ignore AC Input".

This allows you to configure at what load the grid should be used to assist, as well as, at what load to ignore the grid. Additionally there is an option to not ignore AC based on the battery voltage. I believe this will allow me to use solar power during the day, use the grid when needed due to high load, and never discharge the batteries past say 80% SOC. I hope this works out, it's not perfectly effecient, but until the hub-1 assistant releases a "Keep batteries charged" with a do not feedback option, there is not much I can do. This method has the added bonus of not having to flash my inverter. I was rather nervous to do this myself.

[phillipwhittaker]

Hello Philip,

The cables themselves are standard CAT5 or CAT6 network cables like you would use for computer networks. Lekker and cheap.

As far as programming goes, I do use a Victron USB-MK2 adaptor to access my inverter, not sure if the MPPT can use the same adaptor but most programming for the MPPT is done via the front panel.

I hope this helps.

Thanks kibi,

Afaik, you won't be able to program the mutliplus without the mk2 cable. The mk2 cable as a port for the rj45 on the one end, and usb on the other. There is circuitry with uses the cat cable to emulate a com port. I believe this is done because of the com interface used internally by the multiplus. The VEConfig software wants a com port, so baud rate, parity are all required for comms I believe.

The RJ45 port on the mppt is a VEBus I beileve, these are required to be terminated, and are not for programming. They seem to be for comms between other mppts or comms with a color control.

It's all abit annoying, and there is not of information about this. The guys over at Victron are not the most helpful, though Andre did give me some good advice, I suspect this was more roll of the dice luck on my part than anything else.

The guys are finishing my installation today, so we will see how it all works out. MJ mentioned finding a reputable installer, but I have no idea where to find a victron installer! Even the supplier I used was uncertain about alot of this stuff, and they are the ones who supposedly get training!

Cheers

[phillipwhittaker]

Hi,

So I have got things mostly sorted, thought I would share what worked for me:

1.) Updated the Multiplus firmware to the xxxx3xxx series. This allows me to use the assistants, and disables virtual switch. The advantage of this, is I can use the external aux1 relay on the multiplus to drive whatever logic I want.
2.) Connect my CCGX relay and BMV700 relay in parallel to the aux in. In parallel, the relays or OR'd so if either relay is closed, the aux1 will be seen as closed.
3.) Set the BMV700 relay to close @80SOC, open at 90% - This allows precise control over the SOC vs using voltage as a setpoint. This is the fundamental flaw I found with using the virtual switch stuff.
4.) CCGX relay I can use as a sort of manual override. I plan on having a 3rd relay in parallel which I will drive with an arduino, or something similar. This can do things like connect to grid at night, or 1 hour before load shedding or whatever I want really. The plan is to be able to drive this remotely with a website or something.
5.) Setup the assistants in VE.Config. From what I have seen, and read the assistants run from the top down. The basic flow is:
a)  programmable relay - Set the relay on if Auz1 is open for 5 seconds. This is just to reset the General flag.
b) Generator stop start, setting the general flag instead of a relay(set gflag to false to connect to ac, or open relay in terms of the assistant). This is useful because the "Generator start stop" can now be substituted with "Don't ignore AC". This assistant allows you to configure when to connect to the grid, when watt is in excess of, and when watt is below xWatt for xTime ignore again.
c)After this assistant, we have the programmable relay. This will overwrite whatever the generator assistant did to the general flag, so we only allow connecting to AC here(setting general flag to false). This is based on the state of the aux1 relay. When it closes set the flag to false. This is the assistant that is driven from the BMV and CCGX ect
d)Lastly we have the general flag assistant, which we use to ignore AC.

I have this up and running now, and it seems to be working so far. I post back if I have any issues with this setup, but it looks promising. Far more so than the horrid VS stuff.

[MJLorton]

Hi Phillip,

Thanks for the update. It's frustrating to hear that trained installers still struggle with these systems.

If you are willing, it would be great to see some pictures of your installation.

Cheers,
Martin.

[ptomli]

4.) CCGX relay I can use as a sort of manual override. I plan on having a 3rd relay in parallel which I will drive with an arduino, or something similar. This can do things like connect to grid at night, or 1 hour before load shedding or whatever I want really. The plan is to be able to drive this remotely with a website or something.

Not sure if it's possible, but I gather the CCGX firmware being somewhat open source, you might be able to produce the functionality you're looking for without an Arduino.

5.) Setup the assistants in VE.Config.

...

I have this up and running now, and it seems to be working so far. I post back if I have any issues with this setup, but it looks promising. Far more so than the horrid VS stuff.

Did you end up installing an anti-islanding device?

I'm also having some issues, though mine are theoretical at the moment (I've not installed VEConfigure yet, nor do I have PV panels installed).

All the documentation I've seen about the "Hub-1" setup indicates needing to connect the MultiPlus to the MPPT via the VE.Bus/VE.Can interface. When I asked the supplier about this (because it wasn't installed), he said that it's only required when feeding back into the grid. That's not what I'd read, so I queried with Andre, who confirmed what the supplier had said. Granted, Andre is/was in Germany at Intersolar, so may have been a bit distracted.

So my install currently has the MultiPlus 48/5000 talking VE.Bus to the CCGX, the MPPT 150/85 talking VE.Can to the CCGX, and the BMV 700 talking VE.Direct to the CCGX.

My desire to get the "Hub-1" badge, so to speak, is because the Hub-1 assistant is the documented way to prioritise solar power over grid. (http://www.victronenergy.com/live/dc_coupling:prioritize_solar).

A few hours thinking on my part, and I simply can't see how an MP and an MPPT can both charge the same battery bank without there being a (data) connection between the two. The VE.Bus/VE.Can interface allows the two chargers (MP and MPPT) to be/talk on the same bus, and thus communicate their status. I don't think the CCGX acts as a VE.Bus/VE.Can bridge, and the CCGX documentation certainly doesn't suggest that.

[iburger]

I googled into this thread this morning, specifically because I am presently trying to decide whether I should spend money on the BlueSolar 150/70, because I had this long term idea of eventually ending up with a hub-1 system. I am wondering if -- outside of hub-1 considerations -- it is worth the money. That's the broad outline of my query.

More specifically, I presently have a much cheaper charge controller, a Microcare MPPT controller. It's recently developed a memory problem, (more details below) and they have offered to repair it free of charge, except I have to ship it to their offices in Port Elizabeth. So I will be without a charge controller for a few days.

So for a second I was wondering if this is not a perfect opportunity to install the Victron MPPT controller, and sell the MC unit on when it comes back from repairs. For the moment I would not add the CAN-cable, just run it as a normal charge controller.

So now I'm wondering if it makes sense to spend this much money given some factors,
a) that hub-1 is not certified by CoCT (City of Cape Town) to a allow grid connection,
b) that it needs a separate anti-islanding device of which only one is certified for SA, and that thing costs on the other side of 4000ZAR to import,
c) that no home-owner in his right mind is going to spend money on that expensive CAN-USB cable, so you will always have to go back to the installer for firmware updates,
d) that the VE.Can<->CE.Bus cable is also rather stupidly expensive, and
e) this leaves me in a similar position with a single point of failure, better to have two strings each with their own CC in future rather than one large CC

So I'm just wondering if those who have a Victron MPPT controller like it so much that they would advise me to spend the money... or if my thinking is more or less on track. :-)

Re memory issue on the MC controller: I learned it is common for these controllers as they repeatedly write the same sector in EEPROM to update stats and the memory just wears out eventually. To their credit (much as this strikes me as a design fault), Microcare offered to replace the microcontroller chip free of charge, and the problem has been fixed in newer versions of the firmware (by writing less often, and by spreading over three sectors).

Re backup support on this controller: The guys I bought it from two years ago simply stopped stocking these controllers, citing quality issues. I had the case open and although there's one bit of circuitry sort of "hacked on" with a bit of hot glue, and though the PCBs are not laminated (so this is no good on a boat :-) ), the quality isn't that bad at all. It appears to be a standard Buck converter with a big Toroidal choke in the back, a Mosfet and a diode on a heatsink, and a microcontroller with an LCD in the front. Even if this is to break again in a few years, it would make a brilliant test rig for building something arduino-controlled :-)

[PieterN]

I seem to have a similar bridge i am about to cross. Have a Victron Multi 3000 and due to not so good advice bought a Muptiplus MPPT 60 A and 6x250w Remisola(?) panels. Now have two entities powering my batteries and no sunlight harvesting other than when using batteries. At a point where i am considering upgrading to a 150/70, but the literature on setting it up and getting it to work is far and few. Seems it might just be easier setting up the current configuration to run with battery priority to 80% and have no intelligence between my inverter and my mppt. This is not first prize but would allow me to use the current setup to its potential.
Any advise would be welcome

Sincerely

[jose]

Hi guys. I have a similar setup with a Multiplus 5000 and a 150/85 MPPT. I also got a Color Control, GX which is connected to the MPPT via VE.Can and the Multiplus via the VE.net. I did some tests using the Hub-1 and it did export excess solar power to the grid without any problems but I can't continue doing this as I don't have an anti-islanding device.
I also don't have a BMV yet and am a bit confused about how the Multiplus and the MPPT control battery charging in the hub-1 config. My Multiplus shows that it is in float but the MPPT shows that it is in bulk. The SOC is also not correctly shown as I think the voltage supplied by the MPPT is confusing the Multiplus. I also think that these 2 devices need to talk to each other to sort this out and this is where the VE.can to VE.bus cable comes in but it is really expensive in the region of R3.5K.

Incidentally I saw in the writeup for the new Color Control firmware v1.28 that this update has improvements to allow the future updating of MPPT firmware which will be a great improvement and will make the MPPT cable redundant.

[iburger]

I forgot that I posted here... :-)

I ended up buying the Victron charge controller, but the story had some twists.

I decided to do the "responsible thing", which is NOT to spend more money on solar toys, and rather repair the old controller. So I bought a cheap chinese PWM controller to tide me over for a few days, and shipped the MC unit off. It took two weeks before it came back. When it came back it didn't work, and what I discovered inside was more broken bits. It is unclear how this happened and when it happened, but I found a very hot MOSFET and a burnt transformer inside.

At this point I berated myself for not doing "the right thing" (as opposed to the "responsible thing"), which is to buy the Victron controller. Funny what emotion does. I went out and bought it.

Then I set about trying to figure out what went wrong with the MC controller, and I got so hooked on this that I decided not to ship it back for repairs again. I know what is wrong, and I know how to fix it... just made a few mistakes and trial/error cycles along the way so it's still hanging, but I will fix it.

The story begins here: http://izak.github.io/2015/08/04/the-microcare-sage-begins/

I hope to conclude the story soon.

[PieterN]

IBurger, does the hub1 setup actually work with your current setup? do you actually transfer excess power through the mppt onto your load??

[SeanB]

The story begins here: http://izak.github.io/2015/08/04/the-microcare-sage-begins/

I hope to conclude the story soon.

Izak, I can see why the diode failed, and the reason for it. The isolated power supply shown is simple, it transfers energy stored in the primary side to the secondary side with every cycle, so if the load is low then the stored energy has to go somewhere, thus appears as a high voltage ( much higher than the turns ratio of the transformer, even 1:1 will generate high voltage at low load with a flyback converter) at low load and in the high load condition it will be roughly the turns ratio.

The problem is that the designer did not think it through, the diode will dissipate 1W if and only if you heatsink it, and there is no big copper pad on the board to do so, so even with 1W it will eventually cook itself short circuit.

The primary side capacitor was likely added because the initial design would cook the mosfet short from the turn off spike, so they added a apacitor to clamp it. This then dissipates energy in the transformer ( and the diodes, we will get to that later) as ringing with every turn off cycle. Thus around half the stored energy will be going in the transformer as heat from losses in the ferrite and copper windings. The other half will be dissipated in the secondary load, where they originally likely used real genuine 1N4007 diodes, which are good diodes, but not at 40kHz. They likely were absorbing 0.5W as heat in 2 of the diodes ( the 2 which had the flyback pulse imposed on them) as reverse current each cycle as the reverse recovery time of the diode took energy to turn off. They ran warm, but within ratings, so the zener probably was only needing to dissipate 0.5W which it could do bare with only it's glass and short leads. Production got a batch of diodes with improved specs ( still meets the worst case spec of a 1N4007 in reverse voltage, max current, leakage, but with greatly improved reverse recovery spec, just because of better process controls) so now they run cooler, leaving the zener diode to handle all the excess power.

Solutions aside from your one of a proper converter in a brick are to use a shunt regulator, any TO220 packaged transistor ( any current from 500mA to 5A, any voltage over 20V, NPN or PNP) with a 15V zener diode from base to collector, so it will safely handle any excess current. Will do 2W without anything more than the bare tab. Otherwise a proper SMPS controller chip ( cheap, any simple SMPS driver, at this power even ones with integrated FET will work as you only need 2W) and a transformer with 2 isolated secondaries, one for the load and the other for voltage sensing. Bonus, you use a smaller transformer as well, cheaper, and you use thinner wire to wind in in any case.

I will not go into the use of a 25V capacitor in a circuit where it regularly will be running at over 27V ( dynamic impedance and tolerance of the zener increasing it), when a 35V capacitor would have been better.

[ZuluAlpha]

Hi all,
I'm joining the conversation because I also have a Victron setup (150/70 MPPT; Multiplus 3000; CCGX; BMV702) and the lure of going Hub1 is big, but daunting.

Coincidently, I resell Victron equipment (mostly smaller off grid kit), and did go through a training session, but I did so before owning my own Victron kit to experiment with, and it's a LOT of info to remember. After having played with the kit I really should do it again.

I can tell you this: The local Victron guys are hard at work to get a default "suitable for South Africa" setting, which should make all our lives easier. The legal issues with grid feedback is quite a hamper to all of our enthusiasm  :-(

For the moment I'm gearing up to configure the Virtual Switch, and will let Hub1 be for a while.
I'm eagerly following this thread! :-)
Neil