 Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me. |
Forum Index : Electronics : Choosing caps for inverter - which cap properties are most important?
| Author | Message | ||||
| mab1 Senior Member  Joined: 10/02/2015 Location: United KingdomPosts: 151 |
Hello all, My Chinese inverter board is here and following some good advice from this boards members I was looking into possibly getting some better caps for the power section (came with 6 x 3300uF, 100v). It's a 24v system and the space between the heatsinks will take a can up to 35mm dia. So 35v or up; 35mm or less, any height, and 10mm lead spacing; but putting these parameters into Farnel gives me 35 options between 8200uF and 22000uf. But do I go for max uF? or lowest ESR (most don't give ESR value)? or max Hrs at 105C? or, if I can get more Hrs at 85C is that better? Which properties are most important for inverter power boards? Or am i over-thinking this and should just get something a bit bigger uF without spending silly money? cheers M |
||||
| noneyabussiness Guru  Joined: 31/07/2017 Location: AustraliaPosts: 506 |
Im pretty sure oz and others went through the cap thing a while ago ( maybe do a search) however simply put, more the merrier within reason.. ESR along the same lines, if you have plenty of capacitance, your ESR will " automatically " be low.. lets say you purchase 85° jobs, they spec at 100000 hours at that temp, now we all know our system will never get to that temp let alone 100000 houra of it( we hope). The mosfets would most likely be smoke by then, those same caps might do 10x " the hours " at 40 - 50 ° that we generally run our systems at max temp. Make sense?? As long as voltage ratings are enough, the cheaper lower temp ones should be plenty. Hope this helps.... I think it works !! |
||||
| mab1 Senior Member Joined: 10/02/2015 Location: United KingdomPosts: 151 |
Hi Noneya, I probably am over thinking it - although i note that all the cheaper caps are rated at just 2000hrs @temp, but they shouldn't get that hot it's true. I did try a search, but haven't hit on the right search terms it seems - not yet anyway. cheers M |
||||
Haxby Guru Joined: 07/07/2008 Location: AustraliaPosts: 418 |
Yes, what is it with the capacitors that only have 2000hrs lifetime? That doesn't inspire confidence for a device that will be on 24/7. If someone can find the capacitor thread, can you post the link here? |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
How long they last is usually pretty much temperature dependent. The manufacturers have a problem specifying lifetime, because they have no idea what you are going to use the capacitor for, or how much you are going to stress it out. If you run everything right up to the absolute max, operating voltage, ripple current, and temperature, 2,000 hours would probably see many fail. But why would you do that, when you could use a more sensibly rated device that might last fifty times as long and cost only very slightly more ? Its the same kind of question as what mileage will a certain specific tire have on a vehicle ? A racing driver might be extremely happy if it lasted a full 500Km race. Granny might get a set to last her twenty years. How would you specify tire life ? If you were a manufacturer you might rate the tire for 150 Mph, 800Kg load, with a lifetime of 500Km or six hours. Otherwise he might get sued by the racing driver. So the capacitor suppliers say 2,000 hours to cover their arses. But unless you are really pushing everything to the absolute maximum ratings, the expected lifetime will be vastly more. Granny never goes over 30 mph, and her whole car is less than 800Kg so the tires are never stressed and will probably last forever. Cheers, Tony. |
||||
renewableMark Guru Joined: 09/12/2017 Location: AustraliaPosts: 1678 |
I think I have managed to blow up almost every kind of component apart from caps. Mount the unit vertically so it has natural convection and set your fans to come on at a fairly low temp. That will help to look after components. Cheers Caveman Mark Off grid eastern Melb |
||||
| Solar Mike Guru Joined: 08/02/2015 Location: New ZealandPosts: 1123 |
I would go for lowest ESR per uF against cost, low ESR generally means less internal heating. Unlike renewableMark, I have blown up caps, incorrect type used and bang!, case of cap nearly embedded in the plaster ceiling, not fun at close range. Mike |
||||
poida Guru Joined: 02/02/2017 Location: AustraliaPosts: 1388 |
The capacitance present on the Chinese sourced boards is not very important in my view. I have run an unmodified 4kW 12xHY4008 based board for about one year with as fitted caps. No problems from the caps. I only stopped using it when I built an inverter based on Madness's 6kW power board and my nanoverter control board. It ran my house, with an average power level of 500W with frequent peaks over 2kW. As Mark says, make sure you have a reliable fan that will run and cool the heatsinks properly. When the fan fails, your inverter will fail. The problems start from other issues or effects. I would spend a lot more time thinking about the choke, ensuring it does not ever saturate at any conditions it finds when running. I made a choke that can handle 90 Amps and is about 50uH without saturating. I also tend to use Ferrite core based chokes since they work well with the high frequency energy present with the fast, high current switching present in PWM inverters. I would also ensure the (likely IR2110 based) gate driver board will never see negative voltages coming back into the IR2110 IC's output pins. I do this with diode clamps located close to the driver board outputs. Insufficient capacitance results in higher than usual voltage swings on the input. Even with 6 x 10,000uF caps, my inverter still has a large voltage swing on the input and it's only running at 500W to 2kW peak. Big deal.Nothing will break. Lead acid batteries can have a shortened life when loaded with this fluctuating voltage load. But not a lot of life loss, a bit. We need to provide references to battery life verses the load's AC current level if we want to get all "shock-horror" over it. It is an effect but I ignore it. I have only ever seen caps blow up due to over voltage. Only ever over voltage. Never have I seen caps blow up due to too much ripple current. Other people here will have their horror stories for sure. I have seen them fail due to excessive ripple current. The failure is not explosive, it is a long term failure mode that can take months or years. And no bangs. I found that a $6AU 10,000uF 100V cap sourced from Aliexpres could not take 105 - 110V. It got hot, and made a small bang. In the house on the workbench when fed pure DC for 2 minutes. No AC current component. Overvoltage is the issue. The Chinese made inverter boards have caps rated (going by the capacitor case markings - and we need to always keep in mind it's Chinese..) well in excess of the expected operating voltage. In my case the 4kw board had 63V caps and was expected to be run with "48V" input. I used it with 56V input. wronger than a phone book full of wrong phone numbers |
||||
| mab1 Senior Member Joined: 10/02/2015 Location: United KingdomPosts: 151 |
All interesting views - but it sounds like I shouldn't get hung up on cap choice. just as well as i've already got some :) 22000uF, 85c - not low ESR as when i searched for 'low esr' it only came up with small sizes, although a few of the large size caps i found did include esr values, most didn't. Reading Poidas view on choke importance I'm glad to say i had the foresight to buy 4 of these along with the caps: E80/38/20-3C90 FERRITE CORE, E, 3C90https://uk.farnell.com/jsp/search/productdetail.jsp?sku=2103346&CMP=i-bf9f-00001000 i don't really know if they're suitable but they were big and relatively cheap (available 'til stock is gone) I don't have much knowledge of designing chokes, so I though if i couldn't make a suitable choke with 2 i could double them up. I did a trial with a thin wire and 2 cores and 11 turns gave me ~0.05mH ( resolution limit of my component tester) with a ~0.5mm cardboard spacer. Somehow I've managed to squeeze 11 turns of household 25mm2 7 strand cable into the core but haven't clamped it yet so haven't confirmed that it's 0.05mH yet. Whilst i understand in principal how to test for saturation using a scope and high current pulse generator (watch for where the rising current ramp turns steeper), I'm not certain I can make a 90A+ pulse generator without a high probability of blowing up mosfets - i was going to search the forum for test setups but need to finish clamping the choke first... |
||||
 Print this page |
