question

Ed Baak avatar image
Ed Baak asked

CerboGX update

Not sure if this is the right forum to ask, i.e. if Victron SW Engineering is listening in as well, but here we go: Are there any plans to finish the CerboGX (UX) and bring it to production quality or is it forever going to be a half finished product based on a 5 year old QML version?

Examples: With the 'motorhome mode' (or whatever it's called exactly) enabled half of the components on the UX don't show any values. Swiping left gives the same contents but in an even less useful display. Small boxes at the bottom of the screen for generator and such don't work and are useless, temperatures aren't shown on screen, remote functionality doesn't work, documentation is half finished, UX is poorly configurable, the list goes one. I am very impressed with Victron devices in general but the Cerbo is an exception and an almost useless piece of gear for such an expensive device.

cerbo gx
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

4 Answers
nickdb avatar image
nickdb answered ·

Rants aren't generally a good motivator for help on a community forum, particularly on a product that is widely sold and generally seems to work. Most complaints tend to come down to how it is implemented/setup, more often than not, unhappiness is down to things the user has done with/to it.

What version is your cerbo currently?

What is the rest of your setup, how is it all connected?

3 comments
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Ed Baak avatar image Ed Baak commented ·
Yes I totally agree that rants aren't a good motivator, but there is a point where 'enough is enough' and some open and frank conversation is needed. I have a SW engineering background myself so I am quite comfortable working with and setting up the CerboGX. I have ssh'd into it and found my way to the QML based UX code, and would in principle be able to modify it myself. But that's not the point is it? This is a finished product. It's sold in Australia for $875 (combined with the Touch50) and according to the marketing fluff it's a great product and blah blah blah. It's not until you have paid for and started using the system that you realise that it's actually quite limited and worse that there is a lot of black magic happening that is totally undocumented. So, just like some other comments below by other folks, I'm not keen to have to spend a monumental amount of time to dive into a half documented code base, hacked together by a bunch of - I have no doubt very capable - open source contributors that have no incentive to take that development to production quality level in terms of bug fixing, documentation, and support. They develop features to a point that it works for their specific use case, and that's the end of the story.


Such a way of working is great for something like a RaspberryPi because everyone KNOWS that it's an experimental DIY platform, but it's NOT OK to sell a commercial product, pretend it's a finished product, and then come after the purchase and say "Oh well, if you actually want this product to work you will have to wait for an undefined timeline or fold up your sleeves and invest half a year of your spare time and do it yourself".

I appreciate that there is a new development going on with a VenusOS large, but that sounds like it is way over the top for what I need. It also sounds like VenusOS Large has chosen a different (browser based) UX implementation. So what do I do? Invest time in polishing the QML version, or do I need to invest time in the new code base? Which one will survive and is worth investing my time in? How do I develop the UX further and can be sure that a future OS update isn't going to wipe away and invalidate my changes? How can I be sure that I am not going to brick my CerboGX?

IF we're talking about a DIY development environment, then at the very least would I have expected a super detailed documentation of:

* how do I setup my development environment

* which Qt/QML versions should I used and where do I get them from (qt.io or Victron)

* what is the UX architecture design

* how can I find the API to talk to interfaces such as sensors

* examples of how to implement features

* and most importantly a suggested way of working for how to share code and collaborate

I know there is some documentation existing and I have read through it (multiple times). But there is a lot missing, and it looks and feels out of date.

1 Like 1 ·
Ed Baak avatar image Ed Baak Ed Baak commented ·
My system is build for a camper, so totally off-grid and off-line when traveling. VRM on-line is useless to me when traveling.


I have 400Ah lithium composed from individual cells and managed by a REC-BMS active that is connected to the CerboGX. For charging I have a Victron AC charger, and a Victron Solar charger. Both are Smart versions that talk to the CerboGX and finally I have an Orion (DC-DC) that dissapointingly doesn't talk to the CerboGX (never have heard a good explanation why that Smart connection couldn't be included). Last but not least I have a Victron 3000 VA Smart inverter and a Battery protect module. For sensors I have 3 tank sensors and a temperature sensor. I recently bought two wireless RUUVI ones that I still need to figure out how to connect them (yet another example of something that is poorly documented).

The CerboGX does seem to properly communicate with the REC-BMS and does seem to properly manage the various chargers, but it's not transparent in how it does that, or how I can modify settings to change behaviour. Again, documentation is missing.

0 Likes 0 ·
Ed Baak avatar image Ed Baak Ed Baak commented ·
And just for the odd chance that I haven't explained my frustration well enough yet. I'm not just upset about the fact that the CerboGX is underwhelming. I have bought into a Victron system of components, with a total worth of a couple of thousand bucks, and that system is supposed to be integrated into one collaborative system by using the GX as the brains. Without that, the components are single units that don't talk to each other and hence there is a risk that my battery charging profile is sub-optimal.


Using separate components for DC/DC, AC and Solar charging has been a deliberate choice because it gives me redundancy as well as the benefit of combining solar with DC/DC when I am driving, but without the GX it is an inferior system. If I can't get the GX to work like I believe it should be then that means I need to replace the whole system, not just the GX!


For all intends and purposes, the CerboGX appears to be considered a 'finished' product from Victron R&D perspective and Victron is focusing it's probably scarce R&D budget on other areas. What I am saying is that the CerboGX is not at all finished and that the marketing has been misleading and that I am now stuck with a system that is dead in terms of development (I haven't seen a SW upgrade for the CerboGX in quite a while) and a fluffy and aspirational roadmap that is focusing on bloating the system further, rather than making the things that are there already better.

0 Likes 0 ·
johnsmith avatar image
johnsmith answered ·

Coming from a coding background I would say that developing new code around a dual core mobile phone processor 5 years old will be challenging to produce such a product. I am asking myself the same questions. It seems the GX Multiplus has a quad core cpu with only 512MB of RAM which for an upgrade from the cerbo seems to be a bit odd as the memoery requirements are for that CPU are more demanding. Maybe the new Quad core CPU they have used doesn't address more RAM, I don't know but to be honest I am sick of having to read deep into this Victron stuff when really the product should be retail ready which it is far from. Victron need to get some in house developers and some decent design engineers to improve their whole Ve.Can system. It's a shame as their invertor hardware is very robust and reliable, but the whole Ve.can range is riddled with bugs and problems when it comes to software and supporting hardware. Victron if you are listening times are moving forward and everything is about the software and user experience with GUI. This means get rid of the Atari style GUI's and make your programming interfaces modern, well documented and easily uinderstood for the end user. Developers and coders love blaming the "simple" end user for their incapacity to understand their software. When really the devlopers job is to make the user experience simple and still produce and product that performs, which is far more challengin than just writing code that does the job and leaves the end user in command line to try and figure stuff out. (sarcism)


Having said all this I will still buy Victron (until Huawei and such become more time tested) as it works and is built half decent, but Victron should be worried as fancy marketting won't beat the main stream global electronics giants with all the resources and unlimited budget for R and D to come in and steal the market, which juding by the increase in energy prices over the next few months will give them a reason to go into a previosuly risky market.

I refer to my post earlier today regarding the same subject with the Multiplus II GX which has no support for Ve.Can until they release the updated hardware which I am still waiting for.


I am reluctant to buy the cerbo GX with old Dual core CPU and 1GB of Ram when I can have the Multiplus II GX with a quad core, but I still don't understand the design decision to make a quad core package with 512MB RAM, this is 2022 not 2017.

In summary I feel your pain. Victron need to move with the software market as it moves FAST!

Upgrade the hardware every 2 years and consitantly have in house full time developers (big payroll but full controll and no excuses) with monthly roll outs for firmware updates. I myseld wouldn't mind buy a new Cerbo Gx every 2 years for $300 if I knew it was fast, did the job and has smooth code. Following Apple wouldn't be a bad idea. Writing good code in the first place saves you having to patch the crap out of everything like Microsoft evidently enjoy doing, but that's another story.





2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Guy Stewart (Victron Community Manager) avatar image
Guy Stewart (Victron Community Manager) answered ·

Hi @Ed Baak

VenusOS and GX family is a multi-track, multi-year and multi-team continuous-improvement development effort.

GX devices and the VenusOS that powers them is a project that Matthijs (managing director/CEO & son of the Victron founder Reinout) takes a personal and active role in developing. In part because of how central a good GX experience is to a good overall Victron experience for many customers.


While nothing to announce today, the road map ahead for GX range and Venus is exciting, progressing, and addresses a lot of what you're asking for and more.

Components (and most specifically microprocessors) are undergoing unprecedented supply constraints in 2021 and 2022, affecting all industries, and while I have no insight into those specifics, I can say with confidence that has not helped speed up the R&D timeline.



As an interesting anecdote, the current UI was contributed by a frustrated/inspired end user with design and software development skills who thought that he could do better, and then did. This is all possible as the necessary components were open source. He is now on the dev team.

I myself was also a frustrated/inspired end user who thought that the community discussion and product documentation could be improved at Victron, and thought I could even help make things better myself.

If anyone has a clear vision for improvement, I can speak from experience to say that Victron is a company that responds very positively to people who jump in to the ecosystem with good ideas, and capacity to help make things better for themselves and everyone else as well.

It is definitely part of the point of having this community site.

Also worthy to mention there are already users actively developing on the platform (within the existing frameworks, extending them, and also out into other systems entirely via the APIs endpoints).

There is infinite potential for individual customisation to show exactly the data the user wants to see, exactly how they want to see it.

Making those tools more accessible is a part of the roadmap and upcoming releases.



The 'official' Victron interfaces will always be an attempt at the best fit for the most users, and as easy as possible for even the non-technical.

Essentially a compromise to personal customisation - as there are just so many potential interactions and applications of Victron products that it is simply not possible to make it perfect for everyone. That is in part why there is a door left open for people to continue where we cannot and make it how they would like it.

For example - hat tip to @Mark who built this interface for his Victron system.image-4.png


I personally started with a raspberry pi, then a CCGX, then a Venus GX, and now a Cerbo GX and use it every day to control the systems that power my house, I have been suggesting improvements since 2018 - and find it extremely useful as it is, and good value for money in the context of a Victron power system.


image-4.png (838.9 KiB)
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

johnsmith avatar image
johnsmith answered ·

Hi,


In response to your explanation of the above, myself and my UK Distributer are still waiting on the answers from this unanswered question in your forum.

https://community.victronenergy.com/questions/125727/multiplus-ii-gx-updated-hardware-and-lynx-shunt-ve.html

I would be grateful if you could answer these questions regarding the GX as it would make a lot of frustrated customers happy and allow them to expand their system with forward compatibilty.

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.