question

VDC ripple rules

Good evening,

I see so many questions about DC-ripple that I had to write a simple test for installers and/or DIY-ers to check your installation if it works as should be when it charges, either from the grid or via the AC output of an inverter.

There are 2 formula, 1 for bulk (and absorption) and 1 for float.

The ripple should not be more than 1.5% times bulk (and absorption) voltage.

The ripple should not be more than 0.5% times float voltage.

This, as the ripple can overheat batteries and can change their chemistry inside, less with Lithium but more with lead acid.

So, conclusion, below values are maximum values as said above.

A 12V system, float ripple 0.07V and bulk (and absorption) ripple 0.22V, not more!

A 24V system, float ripple 0.14V and bulk (and absorption) ripple 0.43V, not more!

A 48V system, float ripple 0.27V and bulk (and absorption) ripple 0.87V, not more!

That alarms and shut-offs are set way higher is due to when you use power from an inverter.

Also then there is a ripple and mostly way higher than the charging ones, depending mainly on the load of the inverter and the size of the battery.

If you follow the primary guidelines from the Victron instructions, then all should be in these limits.

Easy to test, just switch the charger on and check the ripple values at bulk (and absorption) and at float.

Of course, the above values are calculated for bog standard installations, calculate your maxima yourself with your slightly different charge voltages, specially the 48V systems.

And, the inverter at idle, there is also a ripple, this is about, 0.04V or less for a 12V system, 0.10V or less for a 24V system, 0.15V or less for a 48V system, these values can be used for commissioning before putting on the load.

Of course there are slightly small differences between a single phase and a three phase (etc) system, but the values above can be used for all types of system.

With the best regards, Jeroen.

Visit me at, vrm.victronenergy.com/installation/48664

1 comment

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

·

Hello @Jeroen Kuijf ,

Thanks for researching this topic. I have a couple of questions to improve my understanding.

1. Where did you get those maximum ripple figures?

2. Does the frequency of ripple matter?

3. Are the figures Peak to Peak, Peak or RMS?

My experience has been that an active BMS on a battery can casue "DC Variation" and not "Ripple" as ripple has a fundamental frequency (for 50Hz system it would be 100Hz) and DC Variation is simply variation in sympathy with the BMS actvity. Both types of variation are interpreted as DC Ripple by the Quattro. Would you agree with that or do you think that is not necesarily correct?

A good topic that has caused me grief in the past.

Regards, Trevor

0 Likes 0 ·

Good afternoon,

I measured all I could in a 24/8000 quattro, 11000Wmax = about 450A.

26 capacitors parallel over the DC bus, 50V each, 0.2132F and 83.2A capable ripple in total are inside the inverter.

2000AH/24V lead acid battery with a 1000A BMV712 and a 36V/32.5F supercap parallel on the DC bus.

The test is done with a couple of induction heaters and water pumps to simulate normal load (with the heaters) and overload (starting a pump against its pressure).

Chart 1 shows clearly the difference between charging, taking load and overload in ripple voltage.

Chart 2 shows clearly that the 5% calculated Ipp current and the real normal load Ipp measured are way below the maximum current the capacitors of the inverter can handle, 4 times more on average.

Chart 1.

Chart 2.

Conclusion, this inverter is running very nice and the ripples are all in the limits.

I hope that to make it more visible there is more understanding of this fenomena called ripple.

With regards, Jeroen.

visit me at, https://vrm.victronenergy.com/installation/48664

1714825441523.png (35.1 KiB)
1714825474525.png (38.1 KiB)

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

Good afternoon,

Further testing on the above mentioned installation, there are 3 quattros in 3-phase.

I mounted on each inverter a 220mF Kemet 63V capacitor on the battery connection as close as possible with silver cables instead of copper.

This I did on the 22-06-2024, now a month later the results are clearly visible.

At idle and very low loads there is no difference, but at high load it can be clearly seen that there is a significant reduction in ripple.

I measured the currents produced by these capacitors and sometimes they give a healthy extra boost of about 30A each, reducing the overload warnings by 10 fold and keeping the 230V output voltage more stable.

If I would try to eliminate the ripple to a minimum (0.10-0.11V) at a high load, than I should mount about 2.2F per inverter.

That's about 10 times the capacitors mounted inside the inverter but that's not needed as these heavy loads are just very short at start-up.

Instead, I will now make a choke/capacitor circuit for high current and place that directly on the battery side terminals instead of in the inverter.

Check my site (see below), go to advanced, set to "last 90 days", scroll down, there you can see a graph with the 3 "ripples" and a graph with the "voltage loss" over the battery cables, and slightly higher the "warnings/alarms" graph as well.

The voltage loss is of course the same, but the ripples are way lower and overload warnings are way less since a month.

Also the sounds of the inverters are way less when starting a heavy load, this is the first thing we noticed very well.

With regards, Jeroen.

https://vrm.victronenergy.com/installation/48664

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

Good evening,

Why did you place your question in the first part of this topic.

Did you check my VRM site mentioned below and see the results over the last 90 days?

Well, I answer anyway, ripple for most people is "abracadabra", but can destroy your inverter caps and the batteries in a couple of years if not checked at commissioning.

Therefore I showed a clear DIY test in this topic.

This info is widely available from all inverter manufacturers (read their manuals), I will not write all these brands.

We do ships (IMO-number) and deal with those issues quite often.

All pro-inverters are made to run at mostly 0.10V to maximum 0.15V ripple in idle, beside lab-power supplies who can do near zero, as you know very well seen your background.

That means, when charging or load is applied the ripple goes up, normal, how much, is the amount of capacitors build in, in the inverter.

Well, a start heavy load is very short, so every inverter brand keeps their capacitors to a minimum or a medium, to pass a 10-15 year use of that inverter, nothing else, normal, all brands do!

Mounting capacitors on a BMS or at your inverter will definitely improve your ripple.

Do Not mount a set on a BMS and on the inverter, as then you can get a swaying effect between the capacitores due to low resistance in the cabling.

So, mount these on 1 side only, or use a choke, then it is possible to use on either side capacitors...

Normally you can easily mount 5-10 times the capacitors total F inside the inverter to reduce ripple to a minimum (never zero as that is a lab supply, under/critical/over damped you must calculate), I believe you can, seen your background...

Besides, read first the book (from Victron) "Energy unlimited" in detail and do what it says, then all is working perfectly within all inverter limits normally!

My system, as you should have paid attention is running on 300Hz as I have 3-phase, it does not matter, you add per inverter capacitors and tread them as a stand-alone, therefore it does not matter at all what configuration you have (some very small deviations beside) , see my point?

With the best regards, Jeroen.

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

Hello @Jeroen Kuijf

Many thanks for your considered response. I will try to answer sequentially.

2. I placed it at the first part of the topic because my questions specifically related to the first part, not the others. Sorry if that has concerned you.

4. I don't really know what "abracadabra" means. If that means people don't understand ripple I think you will find a high level of understanding to many readers of this forum.

5. As this is a Victron Forum, in this forum I am interested in Victron equipment. It is great to have information from other manufacturers and that helps us but we are on a Victron forum so Victron specifications are important.

6. The "swaying effect" I am assuming relates to a resonance between the capacitors with the inductance of the supply cable. Is that what you mean? Surely the impedance of the supply cable will be very low and would effectively be putting those capacitors in parallel.

7. I'm not sure where a choke could be used in this type of system to reduce ripple, given the design current we are trying to achieve. Perhaps you could elaborate?

7. The Wiring Unlimited information is great but relates to simple supply ripple we would expect to experience if the supply cable resistance is too high. It doesn't mention some of the issues sometimes outlined in this forum relating to Lithium Battery BMS activity, which although are not really DC Ripple manifest as a DC Ripple problem on the equipment.

A couple of questions if you dont mind.

1. I am keen to know where you got the ripple figures from? Victron seems to show higher numbers than you have stated and I wonder where the figures came from as maybe I need to be looking elsewhere within the Victron documentation.

2. It would be customary to quote ripple figures in RMS values. That makes it easy to measure with a simple multimeter instead of an oscilloscope. Your first post doesn't advise if the levels are RMS and your graphs show Peak to Peak. Rather than work it out I thought I would simply ask the question.

I am not concerned about DC Ripple as caused by high resistance supply cables. That is simple to understand and easily resolved. I am more interested in the DC variation caused by certain Lithium Battery BMSs causing problems with some Victron Inverter chargers.

Many thanks,

Trevor

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

Good afternoon Trevor,

https://vrm.victronenergy.com/installation/48664/dashboard

This should work for you and others as a guest.

Other, VRM World, search, www.quintacasadafonte.com in Portugal.

Both only with a computer, not a phone...

Please, check there first, and see the results the last 5 weeks against previous time.

In VRM you can make a widget and read/log your own ripple, as you can see on my 3 phases...no need for a scope etc...pending inverter.

The normal page is RMS, by clicking detail than you can see top-bottom, I've left that widget in detail already.

When a BMS is "bypassing" 1 lithium cell (as it is already "full"), then the supply voltage to the BMS should act accordingly.

If that doesn't happen then there might be a conflict in the communication to the inverter.

Read in the VE configure how to set that and/or in the BMS software.

You asked how I know above mentioned specific things, directly from installers and vendors of Victron, we do "IMO-number" ships, and so we see many other brands, for us all basically the same due to classification specs.

But, have a bit patience for when I put these chokes between the batteries and inverters.

You would ask why I do this, easy, the first cell of the positive side of a battery is failing first in 90% of the cases since it gets this ripple first the highest and then it's gradually smoothing out a bit over the remaining cells, we've seen too many cases the last 30 years.

So, I thought to try to do something about the root of the issue and the first test as can be seen is a success.

We gained about 25-27% less ripple at high load than before just by installing the same amount of caps as is installed inside the inverter.

Our goal/test is now to go critical to hopefully near zero vpp in idle, 0.12-0.19 vpp with load and only 0.3 vpp in overload.

We will use bog standard of the shelf parts so that DIY-ers can make it themselves also, and yes for high ampere also, say about 150A constantly, 200A up to 1 hour, 450A just 2-3 minutes, to start with, later we can make it as big as needed.

As you will notice, I have 2 batteries 24V, 2383Ah and 1830Ah, the 1830 one through 600A MOSFETs (PID voltage regulated) in parallel with it's bigger brother, and yes 3 BMVs, 1 for each battery and 1 common, just to show that it is indeed possible to run 2 different batteries on an inverter/charger, this now since 2014 (check the midpoint deviations on my site), I use air injection and catalyzers.

Also we use the frequency limiting for our solar inverters, to heat our water, also PID with 5000W choppers since 2016, 1-1.5MW a year gained.

We cool our solar panels with water, there we gain on hot days easily 20% more, yearly 1.5-2MW gained.

Check the photos on my site.

Any questions, let me know.

With the very best regards, Jeroen.

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

jeroen-kuijf

Thanks Jeroen-Kuijf,

Good afternoon Trevor,

https://vrm.victronenergy.com/installation/48664/dashboard

I still have no access rights to this site

This should work for you and others as a guest.

Other, VRM World, search, www.quintacasadafonte.com in Portugal.

Interesting site but nothing to do with Victron DC Ripple...I'm not sure what I should see but it is a guesthouse in Portugal?

Both only with a computer, not a phone...

Please, check there first, and see the results the last 5 weeks against previous time.

In VRM you can make a widget and read/log your own ripple, as you can see on my 3 phases...no need for a scope etc...pending inverter.

The normal page is RMS, by clicking detail than you can see top-bottom, I've left that widget in detail already.

When a BMS is "bypassing" 1 lithium cell (as it is already "full"), then the supply voltage to the BMS should act accordingly.

I suspect you mean it will drop by one cell voltage.

If that doesn't happen then there might be a conflict in the communication to the inverter.

I am not sure what "communication to the inverter." means. What communication?

Read in the VE configure how to set that and/or in the BMS software.

I am well aware of the VE Configure functions. Are you referring to Victron BMS software or is that a general comment?

You asked how I know above mentioned specific things, directly from installers and vendors of Victron, we do "IMO-number" ships, and so we see many other brands, for us all basically the same due to classification specs.

So your comments do not relate to Victron equipment but a general knowledge from other manufacturers?

But, have a bit patience for when I put these chokes between the batteries and inverters.

Are you saying you havent used chokes yet? If that is the case I suggest you may struggle to make that work.

You would ask why I do this, easy, the first cell of the positive side of a battery is failing first in 90% of the cases since it gets this ripple first the highest and then it's gradually smoothing out a bit over the remaining cells, we've seen too many cases the last 30 years.

If you are talking abpout using chokes to reduce ripple the internal resistance of the choke will do more harm than the AC resistance to the Ripple.

So, I thought to try to do something about the root of the issue and the first test as can be seen is a success.

We gained about 25-27% less ripple at high load than before just by installing the same amount of caps as is installed inside the inverter.

It is perfectly understandable that if you add more bypass capacitors to a DC circuit you will reduce ripple. The question that needs to be asked "why would you?".

Our goal/test is now to go critical to hopefully near zero vpp in idle, 0.12-0.19 vpp with load and only 0.3 vpp in overload.

That is a nice goal but most installers just want the system to work using the manufacturers equipment and recommendations. Most installers don't want each job to become a science experiment.

We will use bog standard of the shelf parts so that DIY-ers can make it themselves also, and yes for high ampere also, say about 150A constantly, 200A up to 1 hour, 450A just 2-3 minutes, to start with, later we can make it as big as needed.

As you will notice, I have 2 batteries 24V, 2383Ah and 1830Ah, the 1830 one through 600A MOSFETs (PID voltage regulated) in parallel with it's bigger brother, and yes 3 BMVs, 1 for each battery and 1 common, just to show that it is indeed possible to run 2 different batteries on an inverter/charger, this now since 2014 (check the midpoint deviations on my site), I use air injection and catalyzers.

I can't see anything - I have no access to your VRM and the website you sent me is for a guesthouse in portugal.

Also we use the frequency limiting for our solar inverters, to heat our water, also PID with 5000W choppers since 2016, 1-1.5MW a year gained.

Slightly off topic but I don't know what "frequency limiting" is with regard to heating water and the only reference to PID I know about is Proportional, Differential, Integral with respect to PID control systems. I suspect you are not referring to that.

We cool our solar panels with water, there we gain on hot days easily 20% more, yearly 1.5-2MW gained.

Check the photos on my site. I tried but failed

Any questions, let me know.

With the very best regards, Jeroen.

Kind regards, Trevor

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

Good morning Trevor,

You used a phone!.....as said above....use a computer, https://vrm.victronenergy.com/installation/48664/dashboard.

Then....you clicked on our site directly, no...read carefully....VRM world, search, www.quintavadafafonte.com in Portugal, also only with a computer, do you know what is VRM world?

I see in your comments that you are not really familiar with Victron products, a bit with the ones you purchased, did you do all exams online at the Victron site?

I asked you to be a bit patiënt for the results, you speculate a lot without any calculated background, I have very good contacts with Victron directly and have a very high component knowledge of the parts inside (only from the bigger inverter/chargers and chargers, as these are used on IMO-number ships), do you know what IMO is?

We maintain, repair or modify pro-inverters if needed and give feedback to the makers if we find parts for improvement, we see inverters sometimes more than 25 years old...still going well, yes, also Victron ones.

We and others cool PV panels already for more than 10 years with great results.

We are going off topic...

A high current dc choke or multiple with capacitors for us is childrens play, easy to calculate and to install, as we do that a lot to protect the battery or BMS, as cabling of emergency systems on the ships bridge and control room can be quite long and so ripple increases.

I've seen all the issues in the community about ripple the last years and more people started to ask more and more, so I opened this topic to show on a live installation (one of my sites, see above links) what can be done to reduce ripple, instead photos from the professional installations on board which do not have VRM and the owners not willing to make their ship public as can be understood...

This installation is indeed running fine bog standard, as mentioned above very well, it does not need any improvement (commissioned in 2013 and grown since to what it is today)..we use this installation (as it is semi-big) just for the visibility to the community nothing else...

Please be patient, and wait for the next results.

Questions about other related modifications we do, also on above mentioned site, please open a topic, e.g. water cooled PV for example, if you want to know all ins and outs that do work, or better, search the community better seen your question about communication between an inverter and BMS...lots to find and many solutions and solved issues, read the manuals, if not, open a topic.

With the very best regards, Jeroen.

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

Good evening,

The results are in as I visited the site, see below.

 month max avr high low low 2 1,06 0,31 0,07 3 1,02 0,34 0,07 4 1,09 0,41 0,07 5 1,02 0,38 0,07 6 1,06 0,36 0,07 7 0,73 0,27 0,07

Month 2-6 bog standard operation, month 7 with the capacitors.

Maximum about 30% improvement, average high about 25% improvement, low low no change.

At the moment we are in the progress of engineering and purchasing the choke/capacitors we will hopefully test in September/October.

Check the results yourself at https://vrm.victronenergy.com/installation/48664/advanced and/or https://vrm.victronenergy.com/installation/48664/dashboard or go to VRM world, https://vrm.victronenergy.com/world/, search, www.quintacasadafonte.com in Portugal.

Use for above a computer not a phone.

With the best regards, Jeroen.

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