Good afternoon Matt,
HellermannTyton 635 and 2515, pending busbar or cable.
Sometimes we just make similar ones ourselves or make a “lip” on the busbar.
With the very best regards, Jeroen.
So yesterday I fitted one capacitor per inverter. The results are below. Its not quite the same as we had a lot more sun today so the system wasn’t working as hard and the waterjet was only running this morning but still its a huge difference. Got another set of capacitors coming that I will install on Thursday then I will see how they change it again. The spikes are larger than the vrm show as not sampling quick enough but good enough as an example!
Thursday last week
Today
Good Morning Matt,
Indeed, a very good result already, did you get ripple warnings?
When you use a Fluke multimeter which can measure/record minimum and maximum, set on AC, connected to the master battery terminals, then you can see the spikes.
With the very best regards, Jeroen.
Hi Jeroen,
I should of said I had already done that, on Friday last week the peak I saw was 3.9v, I left it on over night and today and the peak is 0.65v. It even went from charging at 80kw to discharging at 70kw’s this morning, which normally would of made it unhappy and put an alarm out. So a huge difference! No Alarms since fitting the capacitors, even after a couple were installed the system had become so stable, probably going overkill now but I would prefer that.
Good evening Matt,
What you experienced during charging is indeed a positive side effect, the charge voltage is now way more stable and it will charge your batteries slightly more efficient and so in a shorter time.
This we have seen many times before, it will be around 5% or more faster charging than bog standard installs.
My farm 24V batteries are, bank 1 2283Ah and bank 2 1830Ah, install 2012, running like new…
Many people say a capacitor overkill can have negative results, in your case it will not due to the cable resistance you have, it will only have a positive result, but, costs are more, but your efficient use of power will raise significantly, putting your capacitor investment back in the pocket within 2-3 years, and, a very longer lifetime of the whole equipment.
For “normal” installs, yes, those people are correct, in that case it’s better to put my CLC as mentioned above, not to have a negative result with such overkill…
Can you measure with an AC clamp the capacitor current from the master inverter while in “rest” and when starting your Jet, that is for us very interesting to know, as we are also “learning” from this install…
Matt, we are already very happy with the results you obtained, the second capacitor you will mount will make all of us here on the forum speechless, as again, as I promote already for years, ripple is bad…
Only you and me, showed the results in and out about ripple, we can only hope that people start to understand what is ripple…and what to do with it…
I have to say one more thing, later, when the second install is done, we will give you another advice, but give us the data asked for above, we cannot say now what we have in mind, but, we will refer soon.
With the very best regards Matt, I go Super Bock (Portugal) today, I have to attend a 5 week project here.
Hi Jeroen,
In the grand scheme of things another £1600 for a second set of capacitors is nothing compared to the reliability and performance gains. Everything I do is normally overkill and long term it normally pays off like this will I think.
I will get the clamp meter on first thing tomorrow and measure what the amps are when the waterjet starts up and in rest. I’m interested to know now!
I’ll look forward to what your other ideas are.
Good night,
Yes, we need your update as of the second capacitors, for us, what you obtained already is a partly on its own, isn’t it?
Pending on the second set you do, 0.33F correct?
And then the results, show us please…
Really Matt, “nobody” on this forum (some yes) understand what we do…sorry guys, it is the truth, we can only say, keep the ripple down……protect your inverter and battery ….
With the very best regards Matt, we are looking forward to the second stage results…
Btw, we calculated, it will cut half the ripple again, more or less, as we said before from the start of this issue, start with one, and if needed mount a second array!
Morning Jeroen,
For me capacitors is the only viable fix. For the storage capacity needed and the room shape and size there was always going to be distance in the DC cabling, I spent about £15-20k on dc cabling and busbars (we made the busbars so that made it cheaper), if I went larger it would of added about £10-15k extra and you would of probably still had ripple.
I measured the amps on the L1 master inverter cap, It wasn’t perfect as we already had a decent amount of sun and one of the solar arrays goes into the board where the waterjet is so it takes a good load off but I saw a spike of about 3 amps. I have left the clamp meter on and the waterjet is being used this morning so will report back if I see any higher peaks.
Edit: It’s had a spike of 4.4 amps on the capacitor this morning, next set are all mounted now and will be wired in this afternoon.
So a little update, highest peak overnight whilst charging was 0.5, that would have been when starting charge, VRM didn’t show this as too quick, when charging it’s 0.15ish. When the waterjet and other equipment started this morning from cold there was a peak of 0.7, but now during operation it’s about 0.2 but peaks to 0.3-0.4 everytime the waterjet pumps which is about every 1 second. I will publish the VRM graphs later but a bit pointless as it’s too slow now and not detailed enough. The average ripple on the multimeter over the last 15 hours is 0.08v, this will now go up as we are pulling 100-140kw’s off the batteries this morning. It’s worse today as it’s raining and very overcast….again. As I said before if there is sun the solar inverters absorb some of the pulse in the ac side.
Pic’s of the install, next job is to make some plastic covers for the external capacitors.
Good Morning Matt,
We think you obtained a very good result, seen the behaviour of the installation now.
Piek when starting a heavy load cannot be taken with the capacitors completely, as your measurements show.
This is also a result of the batteries themselves, type, chemistry, size, temperature, etc.
We rarely use super capacitors directly mounted on the battery terminal of the batteries themselves, but did a couple of times with great results.
For example, this one can deliver 140A constantly and 1900A maximum, and way faster than any type of battery, for example Licap SM0125-064-ATH.
You have to know, super capacitors cannot handle ripple very well, so, now you have your ripple in the margin of it, first objective is done, the second objective could be mounting such super capacitor.
But, really, we are very happy with your results, we would monitor the installation some time now, to “learn” it’s behaviour in and out, and then decide if such a super capacitor is needed, one thing, it never hurts to mount one anyway.
Pending the size, it will take care of the last piek when starting your Jet.
With the very best regards, Jeroen.
Morning Jeroen,
Yeah I’m very happy with the results, thank you very much. With the super capacitor, can I mount that to the busbar at the top where the inverters are connected, Main reason is I can mount it nicely and I have a spare connection on the busbars. It would be a equal total length to all inverters and all batteries from that connection point.
Good morning Matt,
Thank you for the compliment.
No, also in this case the leads should be as short as possible to the battery, then it will have the most effect.
It helps the battery side to deliver a punch when you start the Jet.
See below 2 graphs we took from VRM yesterday, one in detail what shows piek and another which shows only average.
This you can switch on and off in VRM.
The detail one.
The no detail one.
The pieks you seen here are very big water pumps, switching automatic when the pressure drops, the raised ripple is due to solar charging when the frequency goes slightly up (50.08-50.20 Hz) to limit the panels but instead dumbs the extra load in our hot water tanks by choppers (PWM) and heating elements.
Phase 1 is slightly higher as on Phase 2+3 solar is connected and not to phase 1.
You can compare the difference with the graphs from page 1 of this topic, big differences.
With the very best regards, Jeroen.
Good evening Matt,
Did you have any alarms related to overload or ripple since you installed the capacitors?
What are your highest spikes of ripple now?
Did the system “calm” down further?
Can you sent us one measurement with the maximum starting amps (or power) the Jet does together with the voltage drop and ripple, all at the same time.
In VRM you can make such graph in AI, or by hand, very convenient btw.
We will make one as well at our example site in a couple of days, then I publish the graph and settings here, this, as then we can all “learn and show” to others with similar issues, don’t you agree?
Have a very nice weekend.
With the very best regards, Jeroen.
Hi Jeroen,
No more alarms and ripple is under 0.4 on the VRM and the largest I saw on the multi meter was 0.8. Under normal load peak is about 0.2 and constant load is about 0.1-0.15. I see this as a fix.
This is the last 30days of ripple for units 1-6, First set of caps went in on the 23rd then the second set were installed on the 26th. You can see a massive drop on the 23rd, then a smaller drop with the second set but smoothed the big spikes out more.
I’ll sort the graph out that you want when I get some time but the above tells you everything.
Good Morning Matt,
Indeed, very very nice, I will now ask opinions from my colleagues as well about it, and see with what we come up with.
I will have a meeting with some of them today.
Is this the detailed graph (I presume) or the normal graph from VRM?
With the very best regards, Jeroen.
Good afternoon Matt,
Your batteries should be able to deliver easily 8000A or way more.
Your inverters can pull easily 4000A, when starting the heaviest load, at 75% of the inverters capacity.
Meaning, you have an overkill of batteries in terms of starting current, as during such start 2-3 times I-nominal occurs, that’s optimum as your Jet is not pulling that much.
Our advice, the only spike you see once and a while is not so harmful as it is very very short, and so not really justifying to buy a super capacitor at this time, maybe in the future as you might add more machinery to your install.
But, as said before, such super capacitor never hurts to install already, as it will give a healthy punch and so reducing your last high spikes only, not the overall (RMS) reduction of the ripple.
With the very best regards, Jeroen.
Hi Jeroen,
That’s correct the batteries can do 8960 amps, I will be getting an extra 12 this year to fill the selves so then total output will be 12320amps, I want to keep it under 100 amps per battery ideally as it will extend the life of them, so when we have 44 batteries it can do 4400 at 100amps each. I keep the whole battery house at 15-18c constant too. The only constraint is the Victron can bus message can only accept 6553 for CCL and DCL but that should be fine as the inverters should never get to that figure.
If I fancy messing around later in the year then I might try a super capacitor but for now I just want the system to work and start paying for itself, its been a huge investment along with the solar but as we come into the lighter days its really starting to pay for itself.
Goed afternoon,
Good to hear, thank you for showing the community that ripple van be easily reduced drastically with a couple of quality capacitors on a bigger size install.
We can now only hope that people with a small install check their system and reduce ripple as it can be very destructive.
Matt, if any more questions in the future, we are here to help with pleasure, now, it is cashing time for you…
With the very best regards, Jeroen.
Interesting topic. There was a guy on this forum with flickering lights when his well pump kicked in. My advice was to install caps, but the guy kept on complaining about Victron and prices so I gave up.
I was a project lead at a major engineering company in the Netherlands where capacitor banks were just as normal as the rest of the PLC stuff that we installed. I remember my first ever Victron install… and running into some flickering lights. Yeah you guessed it. Bad Ripple voltage. Addressed properly with caps. Never seen that anymore.
Some people here even installed caps on their transformer less PV inverters to filter out the AC component. Yeah haha 
Good morning chief,
Ac capacitor banks will not improve ripple at the DC side, they will only improve the power factor of the AC side, and indeed, big factories with lots of 3-phase E-motors and lighting need these to save on the energy bill by correcting the power factor, but that’s an other topic.
Just a little then; therefore in big lighting installs you see a combination of tube lights and gaslights to correct it already without capacitors, with 3-phase E-motors each has a small capacitor bank which are bog standard for each type of E-motor and amount of power it uses, there are tables for that, so, nothing to calculate anymore.
I have seen these installed also on MW solar fields delivering to the net, mainly due to the long cabling they cannot avoid, but, also here, each such install is different and different banks must be calculated.
With the best regards, Jeroen.