diy solar

diy solar

Using Super Caps after a DC-DC to supply inrush current on inverter

Archerite

New Member
Joined
Sep 22, 2021
Messages
102
Location
Netherlands
For experimentation and because it was earlier...I have a 12V battery bank about 20 meters away from a small inverter. To get more power over a 4mm2 cable at that length I have been using a victron orion (10A) to boost the voltage to about 24V (24.75). I use car sockets and car USB chargers to charge various portable devices and that works great! Even a small fridge that runs on 12V/24V works perfectly fine on this 24V supply.

I have been using a Victron Phoenix 24/500 (220V) on this 24V supply as well to see what would run on it. For the last 3-4 months I have been watching TV from this. Not a small portable TV but a 49" LG 4k TV (on low brightness as I prefer that) and it uses about 50 watts on 24V (and about 70 watts on 12V, so yeah there is a loss I know). I can also charge my 12V bosch power tool batteries from it and I can even charge two of them at the same time, using 150 watts.

What does NOT work is a PS4 game console, a UPS for my PC and NAS, the 18V 4A bosch power tool charger and maybe a few others that have a high inrush current when they power up. At first I used a 24V->12V DC-DC converter and a 12V victron phoenix on that and then even my TV had trouble staying on.

My idea for (solving) this issue is by using a supercapacitor bank with enough cells to handle the high voltage and enough capacitance to provide the inverter with power for just a few seconds. I have been reading up on how super caps works, and I already know a bit from my modeltrain hobby. There they are used to supply power when driving over dirty tracks or points. The way to charge the capacitor in a locomotive decoder is a resistor to lower the charge current and a diode to allow fast discharge. But when I look at some of the protection circuits of ready made modules it's limited to just 1A....which increases charging time to an extreme! I believe it was 20 minutes for a 500F 6 cell module!

I have bought a bunch of 5.5V 1.5F super capacitors but I can't find any protection circuits for those! The idea was to put 5 or six in series to make it safe for 24V but I started reading about voltage imbalance.....and started thinking about some sort of BMS! Either an actual BMS meant for lithium cells and configure that for the voltage range of the super caps, or some other type of circuit that checks the voltage and redirect it somewhere once the limit is reached.

As I said this is al experimentation and I know this is an uncommon way to connect an inverter. But i still think this might work! The high current is only required for a few seconds...maybe even less than a second. Does anyone know about a protection and charge circuit that is required for these supercaps at a higher than 2.5V and more than 1A?

Thanks in advance. :)
 
 
Using Super Caps after a DC-DC to supply inrush current on inverter

In my case a particular set of BMSes (12v) goes to sleep when faced with the inrush, but it appears now that it's been solved. It needed 4 of them though in parallel, because 2 couldn't handle it. Also they're attached to the busbars with equal-length wires and in a diagonal configuration.

But I'm still wanting to explore using super/ultra caps.
 
Depending on your surge time/length, might be able to get away with just a parallel bank of standard electrolytic caps.

Common practice to keep standard caps balanced in series is to use resistors across the terminals.

Also keep in mind when you series caps, the voltage handling doubles, but the capacitance halves. 1F in series with another 1F = 500mF
Not sure if this applies to super caps, as I've never played with them.
 
Depending on your surge time/length, might be able to get away with just a parallel bank of standard electrolytic caps.

Common practice to keep standard caps balanced in series is to use resistors across the terminals.

Also keep in mind when you series caps, the voltage handling doubles, but the capacitance halves. 1F in series with another 1F = 500mF
Not sure if this applies to super caps, as I've never played with them.
I am not sure how large or long the surge is actually. All I know is as soon as I turn on one of the mentioned devices the inverter goes into a low voltage alarm and shuts off the AC output. Usually it retries every 10-30 seconds and loops until the high load is removed.

The PlayStation 4 works fine on my EcoFlow River and draws about 100-150 watts. But the inrush happens so quickly it's not visable on the display. So it must be really short....and thats why I think a small buffer of 2 seconds must be enough.

I did think about regular caps before but I figured they had to be really big for this idea to work.
I have read about the capacitence being halved in series connection, but I don't understand why yet. Not sure if it's important enough to know the 'why'...but it's interesting.
While I have not actually used them much I do know for sure the same rules apply to super caps.

During my research I did find out there are supercap charge and monitor IC's from Texas Instruments. Not exactly a BMS but it's close enough and supports 10A charge current. But I need to look into it a bit more. I'll post the chip numbers later
 
I am not sure how large or long the surge is actually. All I know is as soon as I turn on one of the mentioned devices the inverter goes into a low voltage alarm and shuts off the AC output. Usually it retries every 10-30 seconds and loops until the high load is removed.

The PlayStation 4 works fine on my EcoFlow River and draws about 100-150 watts. But the inrush happens so quickly it's not visable on the display. So it must be really short....and thats why I think a small buffer of 2 seconds must be enough.

I did think about regular caps before but I figured they had to be really big for this idea to work.
I have read about the capacitence being halved in series connection, but I don't understand why yet. Not sure if it's important enough to know the 'why'...but it's interesting.
While I have not actually used them much I do know for sure the same rules apply to super caps.

During my research I did find out there are supercap charge and monitor IC's from Texas Instruments. Not exactly a BMS but it's close enough and supports 10A charge current. But I need to look into it a bit more. I'll post the chip numbers later

Looks like the BMS goes into sleep or haywire hence the LV alarm, same with my case.
 
Sonds like what you really want is this on AC side to control inrush current.
What I really want is for the inverter to just stay on :ROFLMAO:;)...but would the inverter not have something this simple build in? I mean....it's not a cheap no name brand inverter.
Looks like the BMS goes into sleep or haywire hence the LV alarm, same with my case.
I do not have a traditional/common BMS. I have 11 7Ah batteries that each have their own BMS that I can't acces/read a status of. Wired in parallel to make an 88Ah battery bank at 12V.

To this I have the victron Orion DC-DC that makes 24V and sends that over a 20 meter 4mm2 wire. This is the end where the inverter is connected. And what I think is that either the voltage drops to low or more than 10A at 24V is drawn....making the Orion drop the output to 0V to protect itself.
It comes back nearly instantly, but is down just enough to give a huge drop in voltage that the inverter does not like and shutsoff with a low voltage alert.

This is how I got the idea for using super caps to give it just enough time to recover from that drop....or even prevent it since the current will come from the caps instead.

I think I will just have to give it a try and see if they make any difference at all. For a short test I am not that concerned about voltage differences between the caps.
 
It is not ,why doesn't the inverter have this built in. It would only work once for the first thing plugged in.

It is, why doesn't your other electronics have this built in. This is designed for the first current surge of the plugged in electronics as the capacitors go from zero charge to full voltage. Many have this built in to protect its own electronics from surge damage. There are also external modules called soft starts that use a fixed resistor and then a relay shorts it out after a set time. I have this little RIDGID 100W inverter I got from an electronics lot and one 80W switching power supply brick I wanted to use it on. With nothing plugged into the brick, this inverter instantly goes into fault. I would have thought a brand name could have handled that. It is the same principle everyone here is using to precharge inverter capacitors.

I like capacitors and I have them in my system. But, hundreds of dollars to solve a $10 problem. There are ones much cheaper than this

 
Be careful, there are a lot of Chinese counterfeits with poor ESR performance. If price is too good, it is likely counterfeit. Actual Maxwell supercaps are quite expensive, which creates the incentive to counterfeit them.

Another common con is six packs claiming 500F that are actually made of six 500F caps in series for 83F net.
 

Attachments

  • 2.7V-3000F-Cell_20210406.pdf
    2.4 MB · Views: 9
Last edited:
I have finally got some time to test this idea and put together this little breadboard:
20221002_120333_resized.jpg
I know these wires are WAY to thin for any high currents, but I did not think it was really needed anyway. It's just a really short drop in voltage that I needed to prevent on the inverter. Those thicker red and black wires go into the main 24V line though a breaker (disconnect only). Not sure if the breaker is supposed to have current go both ways....probably not since it was kind of cheap. But it's low current and an experiment anyway.

Because I have set the DC-DC output to 24.5v and these caps are only rated for 5.5v and don't have any protection I decided to go with six cells to spread the voltage. Overall it worked and most of them were around 3.60v - 3.90v with only the last one going higher to 4.8v. Not very balanced but still within the specs of the caps 5.5v so I did not mind it for the experiment. If I would put this into actual use permanently I'll have to look into protection circuits that keep them below 5v at all times.

The good news is that with these caps on the 24v rail and plugging in my Bosch 18V 4A charger the inverter stayed on! And the charger kept running normally until the battery was charged. I also tested the PS4 to see if it would even turn on...and at first just plugging it in gave a fault on the inverter. But after about 2-3 cycles the inverter stayed on again. Then I powered the PS4 on...and it booted! ?

Both the Bosch charger and PS4 draw 'only' about 100-120 watts which is well within the limits of my DC-DC converter. And now my test shows that the super caps do provide just enough juice to prevent a low voltage alarm on the inverter so I can use these devices with the current setup. Now I have a few options to protect the caps from over voltage:
- 1: Use the readily available 2.5v protection circuits and waste half of my capacitors.
- 2: modify those 2.5v circuits to work at a higher voltage (if possible)
- 3: get all new 2.7v super caps and the protection circuits
- 4: get a ready made super cap module
I am not sure which one is best right now. I have a bunch of these 5.5v 1.4F capacitors now...that cost less than half the price of even a single pre-made module! It seems to work great for now to initially power on the bosch charger...and I did test disconnecting the caps to see if it would stay on. and it did!

This gives a lot of options when all you need is a short high power boost but your battery (or DC-DC converter) can't supply it. But it's important to use protection circuits and the right size of wires between the caps so they can actually deliver these high currents safely. Just wanted to share a successful test for those interested in something similar.
 
Last edited:
Nice work!

Does the protection board supports plugging and soldering the caps directly, or would you still need those wires (or a busbar)?

Also it would be nice to see how much inrush current this bank could withstand, which the Victron couldn't.
 
Nice work!

Does the protection board supports plugging and soldering the caps directly, or would you still need those wires (or a busbar)?

Also it would be nice to see how much inrush current this bank could withstand, which the Victron couldn't.
Thanks. The protection boards I have seen are either round and meant to solder directly on the cap...so that's a yes to your question. But those are 2.5v only! The other type is a PCB with the protection circuits for 6 super capacitors...and also 2.5v! I have seen some video's and articles explaining how these protection circuits work...but I only understand the bigger picture of what they do. It's kind of an analog voltage compare with internal reference voltage and then an op-amps (i think) driving a transistor/FET pushing current through a resistor.
Weirdly enough there is not a power-in/power-out on these circuits and they go straight on the positive and negative of the capacitors. Which is part of the reason I do not fully understand how they work...or modify one to work at 5v actually.

I would really like to know to how much current it actually needs in this brief period too...but none of my tools are fast enough to capture it. It really is just a fraction of a second...and I do see the voltage drop on a small monitor to something like 20 volt and then it goes back to 24.5v.
While pulling almost 180 watts I did see the voltage drop to 23.5v but that is expected for the 20m 4mm2 wire that is between the inverter and DC-DC converter.

The victron orion 12/24 can deliver 10A at the output and if it's the orion that shuts the output of due to over current I am guessing it would have to be above 10A. I do actually have a second orion with the same specs and intended to connect them in parallel for 20A at 24.5v....but I never got around to hooking it up near the battery. only used it for a couple of other experiments.
The idea with them in parrallel is to both have more power available but also to share the load between them so it won't get as hot. When pulling above 7-8A it get's REALLY hot actually and during summer I put an external fan above it to keep it cool enough. I never had it shutdown from overheating though so I guess it's still "normal" for it to run that hot. I do keep an eye on the power I pull from it and what is on at the same time of course.

I also have no idea how fast or how much current this capacitor bank can deliver. I do think they have some internal resistance since it takes about 30 seconds at 5v to charge them until they no longer pull any current. I got enough of them to make a second row and even put in a 7th to lower the voltage per capacitor a little bit. Maybe that will be enough to provide the power for the PS4 when plugging it in.

For another experiment with generating energy from compressed air I got some analog voltage and amp meters. maybe they will respond quick enough to show what actually happens during those high current draws. The amp meter goes to 50A so that should give enough precision I think. The voltage meter can go to 100V so that is not very precise to see a few volts drop...but it should at least move enough to see that it does go below 20v maybe.

Sorry for talking too much...I'll be quiet now. :giggle:
 
Thanks. The protection boards I have seen are either round and meant to solder directly on the cap...so that's a yes to your question. But those are 2.5v only! The other type is a PCB with the protection circuits for 6 super capacitors...and also 2.5v! I have seen some video's and articles explaining how these protection circuits work...but I only understand the bigger picture of what they do. It's kind of an analog voltage compare with internal reference voltage and then an op-amps (i think) driving a transistor/FET pushing current through a resistor.
Weirdly enough there is not a power-in/power-out on these circuits and they go straight on the positive and negative of the capacitors. Which is part of the reason I do not fully understand how they work...or modify one to work at 5v actually.

I would really like to know to how much current it actually needs in this brief period too...but none of my tools are fast enough to capture it. It really is just a fraction of a second...and I do see the voltage drop on a small monitor to something like 20 volt and then it goes back to 24.5v.
While pulling almost 180 watts I did see the voltage drop to 23.5v but that is expected for the 20m 4mm2 wire that is between the inverter and DC-DC converter.

The victron orion 12/24 can deliver 10A at the output and if it's the orion that shuts the output of due to over current I am guessing it would have to be above 10A. I do actually have a second orion with the same specs and intended to connect them in parallel for 20A at 24.5v....but I never got around to hooking it up near the battery. only used it for a couple of other experiments.
The idea with them in parrallel is to both have more power available but also to share the load between them so it won't get as hot. When pulling above 7-8A it get's REALLY hot actually and during summer I put an external fan above it to keep it cool enough. I never had it shutdown from overheating though so I guess it's still "normal" for it to run that hot. I do keep an eye on the power I pull from it and what is on at the same time of course.

I also have no idea how fast or how much current this capacitor bank can deliver. I do think they have some internal resistance since it takes about 30 seconds at 5v to charge them until they no longer pull any current. I got enough of them to make a second row and even put in a 7th to lower the voltage per capacitor a little bit. Maybe that will be enough to provide the power for the PS4 when plugging it in.

For another experiment with generating energy from compressed air I got some analog voltage and amp meters. maybe they will respond quick enough to show what actually happens during those high current draws. The amp meter goes to 50A so that should give enough precision I think. The voltage meter can go to 100V so that is not very precise to see a few volts drop...but it should at least move enough to see that it does go below 20v maybe.

Sorry for talking too much...I'll be quiet now. :giggle:

I think these protection boards are mostly for audio or other electronics projects. The Maxwell caps which are for automotive, I've seen them in busbars; but with "protection board" and "balancer", only with a plastic wrapper, so I'm not really sure of those (also the genuine ones are expensive, and have higher farads).

The inrush can be measured by a DC clamp-meter with inrush, but they're expensive. But the good thing is that your setup is working, and you can build on that.

About my problem, I have already solved it by putting 4 batteries in parallel.
 
For experimentation and because it was earlier...I have a 12V battery bank about 20 meters away from a small inverter. To get more power over a 4mm2 cable at that length I have been using a victron orion (10A) to boost the voltage to about 24V (24.75). I use car sockets and car USB chargers to charge various portable devices and that works great! Even a small fridge that runs on 12V/24V works perfectly fine on this 24V supply.

I have been using a Victron Phoenix 24/500 (220V) on this 24V supply as well to see what would run on it. For the last 3-4 months I have been watching TV from this. Not a small portable TV but a 49" LG 4k TV (on low brightness as I prefer that) and it uses about 50 watts on 24V (and about 70 watts on 12V, so yeah there is a loss I know). I can also charge my 12V bosch power tool batteries from it and I can even charge two of them at the same time, using 150 watts.

What does NOT work is a PS4 game console, a UPS for my PC and NAS, the 18V 4A bosch power tool charger and maybe a few others that have a high inrush current when they power up. At first I used a 24V->12V DC-DC converter and a 12V victron phoenix on that and then even my TV had trouble staying on.

My idea for (solving) this issue is by using a supercapacitor bank with enough cells to handle the high voltage and enough capacitance to provide the inverter with power for just a few seconds. I have been reading up on how super caps works, and I already know a bit from my modeltrain hobby. There they are used to supply power when driving over dirty tracks or points. The way to charge the capacitor in a locomotive decoder is a resistor to lower the charge current and a diode to allow fast discharge. But when I look at some of the protection circuits of ready made modules it's limited to just 1A....which increases charging time to an extreme! I believe it was 20 minutes for a 500F 6 cell module!

I have bought a bunch of 5.5V 1.5F super capacitors but I can't find any protection circuits for those! The idea was to put 5 or six in series to make it safe for 24V but I started reading about voltage imbalance.....and started thinking about some sort of BMS! Either an actual BMS meant for lithium cells and configure that for the voltage range of the super caps, or some other type of circuit that checks the voltage and redirect it somewhere once the limit is reached.

As I said this is al experimentation and I know this is an uncommon way to connect an inverter. But i still think this might work! The high current is only required for a few seconds...maybe even less than a second. Does anyone know about a protection and charge circuit that is required for these supercaps at a higher than 2.5V and more than 1A?

Thanks in advance. :)
I am working on something similar and bought two supercapacitor banks to put in series.
 
I think these protection boards are mostly for audio or other electronics projects. The Maxwell caps which are for automotive, I've seen them in busbars; but with "protection board" and "balancer", only with a plastic wrapper, so I'm not really sure of those (also the genuine ones are expensive, and have higher farads).
I don't know but searching this I also found many special automotive audio capacitors that look fancy and have build in volt meters. Mostly 1F - 4F...and one claiming 18F but in the comments on amazon it was revealed to be 0.18F I think. Like I said there are a few youtube video's about the topic and some guy analyzed the protection circuit and explained it. I just did not understand all of it...but I think above or around 2.5v it pulls the resistor to ground...I think! Then when voltage get's below 2.5v it shuts off again and that is how it keeps them around that voltage. It probably wastes a bit of energy this way until all caps are full.

The inrush can be measured by a DC clamp-meter with inrush, but they're expensive. But the good thing is that your setup is working, and you can build on that.
I have the popular red clamp multi-meter's from amazon costing about 50 euro's and they are great! Got three of them now so I can measure multiple wire's at the same time and see how much amps are going through. But they refresh at what feels like about a 1 second interval...and since it's a much shorter inrush I can't see it. Also don't have a big scope to visualize voltage drop's...

About my problem, I have already solved it by putting 4 batteries in parallel.
great! More batteries in parallel with good wires and fuses of course will share the load between them and you can run more powerful things. My batteries only do 10A with their internal BMS...but with 11 in parallel that should be "ok" for 110A. ? (fused the battery bank at 80A though)

So how much did this solution cost?
Ordered in a pack of 5 for 12,38 euro's with prime whipping on amazon. Roughly 2,50 per capacitor...having 6 in series...so about 15 euro's I guess. Already had the wires and breadboard laying around. If I would go for higher capacitance and modules then it would be more like 120 euro! Almost did that but found it a bit to expensive.
Now I am looking at some other higher voltage non-super capacitors to see if they give me the same results. Not sure what to get yet but I want at least 10,000 uF and supported voltage of more than 28V. Ofcourse those will be way bigger than the caps I have now...but I won't have to put them in series so it might end up smaller.

The other and probably much cheaper option is actually connecting that second Orion to see if that gives it enough stable voltage to solve the problem! I do think my 4mm2 wire is capable of 20A at 20 meters without heating up....at least for a short while.
I am working on something similar and bought two supercapacitor banks to put in series.
This is exactly the type of module I was talking about! They are a bit expensive...but if you have a similar problem than it should solve the issue. :)
 
Back
Top