High Voltage Forum discussion thread: highvoltageforum.net/index.php?topic=2980.0 Documentation, pictures and schematics: kaizerpowerelectronics.dk/pulse-power/3500-joule-pfc-mkp-capacitor-bank/
I have 3x of the same Maxwell 1.8kV 2500uF caps you have , when i got them way back in 2010 around one came damaged in shipping the other 2 were fine. The terminals were broken on one, i cut the top open by running a grinder around the welds, managed to get it open nice and clean, i then added 2 heavy duty binding posts to replace the damaged ones, put the top back on and welded it closed, filled it back up with oil, its working good ever since. Inside those caps there is a big flat aluminium strips with a hole that is bolted to the bottom of the terminals. The lid has about a 10mm folded flange all aroud which helps with keeping inside clean during grinding and makes it really easy to weld closed again without burning the contents inside. You should try repairing your capacitor if you still have it.
I had the exact same feeling. I asked the kids out in the garden, they could not hear it by the neighbor. Wife could however hear the house "rattle" a bit :)
I think you were right in your initial idea that you get ~166uF by shorting across one element and then using the other two in parallel for DC. The reason why you get >100uF measuring between two terminals instead of 83 is because the other two elements are also in series with each other across the terminals, so that string is paralleled with the single element between the terminals. I think total capacitance of the series/parallel bank is ~830uF for ~2.6kJ energy at 2500V.
You are absolutely right! Thanks for catching this error. This also tell a tale about just how scary fast MKP capacitors can discharge. This 2.6 kJ is magnitudes more BOOM, than 4 kJ on my latest single Maxwell capacitor.
@@KaizerPowerElectronicsDk yeah they are low ESR and ESL so they put out a lot of current and almost all of the energy is deposited in the load. I thought those Maxwell caps were either MKP or oil/paper too though? Either one should be very fast, in theory. I bought 60 used electrolytic capacitors on eBay that were each 5800uF 300V. That’s almost 16kJ but if you put them in series the short circuit current would be way less than yours and the voltage across the actual load would be way less. Most of the energy would be used making the capacitors warm. My plan is have them all in parallel and then you can get high peak current, good efficiency into pretty low impedance loads (hybrid armature railgun, maybe a 3mm^2 copper bridge wire inside a watermelon etc) but you still get a long pulse without huge peak power.
I unplug the charger first, so stick is de-energized when handled. I disconnect it from bad experience of damaging my diodes on earlier capacitor banks. I will change it for some solenoid with isolated rod instead, but maybe I am fixing a non-issue due to old experience? :)
Yes definitely beef up the wires for the capacitors themselves. Super awesome capacitor bank though! I would be a menace if I had that many capacitors!
@@KaizerPowerElectronicsDk 😂Gotcha! I was just thinking you annihilated one of those wires with the can crusher, the rest of the circuit is at max 2 of them. 😬 Can't wait for V2!
A couple suggestions: 1. Hook up a lightbulb in series with a decent resistor across the capacitors, that is always connected. It'll take longer to charge, and a parasitic loss after disconnect, but a visual indicator whether there's charge in the bank, and a means of no charge accruing is a must. 2. Swap out the copper wire for aluminium bus bars. It's cheaper, more easily formed/machined for purpose, and due to the skin effect should allow higher peak currents.
1: Safety first is always a great idea, I skipped it here, but on another bank I made, with 35 electrolytics in series for 14kV, I had LEDs everywhere for status (and balancing the long string). 2: The wiring is fast, cheap and something I just had as scrap. This was the first test, a rebuild was planned, now that they proved themselves :)
Lol awesome! As much as I'd like to play around with stuff like that I'm afraid of answering the door and explaining to officers that I don't in fact have a firing range in my living room.
Also was curious if you had a video I'm not seeing in regaurds to the maxwell cap that failed? Seems pretty rare those things fail, even when you find 20 year old ones they often still have many thousands of discharges left at their full ratings, to which I'd never expect to see/hear about a terminal on one failing especially. I picked up a couple that are rated for 50KJ and I plan to basically never go over 25KJ so they're more/less last me 'indefinently' since they're pulse caps rated for dead-shorting. Kinda surprised/impressed to hear you had a maxwell cap fail, just never seen/heard of it myself unless someone did over-volting and had hundreds of shots in @ over-volted. Be curious to see more on what happend there!
I do not think I ever made a video or even took pictures of the damage. Here at 60 seconds, you can see a black smoke puff in the left side and in the slow motion, you can see the terminal flying into frame: ruclips.net/video/Z4UBXoG-Aj8/видео.html
I love these, I have 20 caps to make my own, 196uf, 2300vdc from cardiac defibrillators....I'm terrified to change it past like 100v as full charge is 520j per cap or like 10,000 if I wire them all up. I have a question for anyone, when discharging like this do the caps go reverse voltage? How do you prevent this or does it matter? I cant find a data sheet for the caps aerovox ql232ew196v23a
Defibrillator capacitors are also some beasts! 10kJ is way up in the scary end of the scale. Some of the best Maxwell pulse discharge capacitors are only rated for 50-70% voltage reversal, but must still be limited to some certain dV/dt. Voltage reversal matter a lot. It is the most likely killer of pulse discharge capacitors. To avoid it, I make my work piece / work coil weak in a point where I am sure it blow open circuit. Ensuring it with diodes can be less cost effective. Check out the pulse power folder here for some papers that might inspire you for a solution: kaizerpowerelectronics.dk/theory/file-archive/?drawer=application_notes
I've got my hands on 120 caps of 500V 3900uF, could build an enormous bank but I'm afraid to start doing something with them. Close to 60kJ. Thx for the tip on voltage reversal.
Oh boy, reminds me of when I built 10mF 400V capacitor bank for DRSSTC and accidentally shorted the terminals with fully charged capacitors. Suffice to say, my ears were ringing for good half an hour.
I did that with smaller capacitor bank long, just wanted to make sure it was shorted, before touching it, still had enough charge to ensure some ringing in the ears :) Ear defenders stay on since that day.
PFC capacitors can take a lot of abuse. Used a few of these for a laser power supply that needed a fast pulse discharge. They can take 1800v net a piece. Charge the elements with plus and minus HV and the third one being neutral. ❤
Shouldn’t peak DC voltage rating be ~1.4 times higher than AC RMS voltage not 2 times? Also that’s the 2 second test voltage. My guess after looking at those specs, is you start getting little breakdown points in the polypropylene insulation somewhere a bit over 1000VDC, which will gradually eat away at the capacitor.
I did not supply the information, but in Tesla coil community we gladly just look at AC ratings and say oh that lovely swing from negative to positive, that's the same as DC, so twice the AC is DC rating :) I said 60 seconds, but its 2 seconds in datasheet. I also got breakdowns at 1350+ VDC, that's where I heard crackling/popping, its the second test in video.
Jes!!! Thats 2 times the energy of a rifle round like .308 For shure that emits at least as much acoustic power as the muzzle blast of such a round, probably even more.😆👍
Gun shots are more sharp and high pitched, this is so deep and hollow. The surface of the foil is also something like 20 cm long and 5 cm wide, so a large area going into boom :)
Ive done 2KJ off a .maxwell pusle cap, and when you stand outside the barn across the yard, it literally sounds exactly like a pistol being shot. Thankfully in my case its country..so people shoot in their backyard all the time. Very sharp crack!
-1 pant 😂. Still nice experiments , also you could look into producing graphene with this method , just add carbon instead of whatever you are frying in the middle 😂👌
What do you think could make a capacitor bank like this last longer than 200 discharges ? ( other than discharging them slower ) . I’m also thinking of building a smaller capacitor bank for experiments but would like to use scrap capacitors just for starters , in case I get something wrong 😂
High Voltage Forum discussion thread: highvoltageforum.net/index.php?topic=2980.0
Documentation, pictures and schematics: kaizerpowerelectronics.dk/pulse-power/3500-joule-pfc-mkp-capacitor-bank/
I have 3x of the same Maxwell 1.8kV 2500uF caps you have , when i got them way back in 2010 around one came damaged in shipping the other 2 were fine.
The terminals were broken on one, i cut the top open by running a grinder around the welds, managed to get it open nice and clean, i then added 2 heavy duty binding posts to replace the damaged ones, put the top back on and welded it closed, filled it back up with oil, its working good ever since.
Inside those caps there is a big flat aluminium strips with a hole that is bolted to the bottom of the terminals.
The lid has about a 10mm folded flange all aroud which helps with keeping inside clean during grinding and makes it really easy to weld closed again without burning the contents inside.
You should try repairing your capacitor if you still have it.
I never gave it a thought to repair it, I stood with oil everywhere after it popped and really just wanted to get it out of the workshop :)
Wow. Watching you do it means I never will! Thanks for sharing 👍 🇬🇧
If it can comfort you, I can find lathes a bit scary :)
I'm surprised your neighbors didn't call the cops on you with those sounds - great video!
I had the exact same feeling. I asked the kids out in the garden, they could not hear it by the neighbor. Wife could however hear the house "rattle" a bit :)
I think you were right in your initial idea that you get ~166uF by shorting across one element and then using the other two in parallel for DC.
The reason why you get >100uF measuring between two terminals instead of 83 is because the other two elements are also in series with each other across the terminals, so that string is paralleled with the single element between the terminals.
I think total capacitance of the series/parallel bank is ~830uF for ~2.6kJ energy at 2500V.
You are absolutely right! Thanks for catching this error. This also tell a tale about just how scary fast MKP capacitors can discharge. This 2.6 kJ is magnitudes more BOOM, than 4 kJ on my latest single Maxwell capacitor.
@@KaizerPowerElectronicsDk yeah they are low ESR and ESL so they put out a lot of current and almost all of the energy is deposited in the load.
I thought those Maxwell caps were either MKP or oil/paper too though? Either one should be very fast, in theory.
I bought 60 used electrolytic capacitors on eBay that were each 5800uF 300V. That’s almost 16kJ but if you put them in series the short circuit current would be way less than yours and the voltage across the actual load would be way less. Most of the energy would be used making the capacitors warm. My plan is have them all in parallel and then you can get high peak current, good efficiency into pretty low impedance loads (hybrid armature railgun, maybe a 3mm^2 copper bridge wire inside a watermelon etc) but you still get a long pulse without huge peak power.
Why disconnect the charging circuit? Would be safer to leave it connected to avoid the sketchy wire-on-a-stick method.
I unplug the charger first, so stick is de-energized when handled. I disconnect it from bad experience of damaging my diodes on earlier capacitor banks. I will change it for some solenoid with isolated rod instead, but maybe I am fixing a non-issue due to old experience? :)
i just love this,. im making a 400v cap bank,. in the process,. this gives me motivation,. nice thanks
What are you going to use it for and how do you plan to switch it?
Yes definitely beef up the wires for the capacitors themselves. Super awesome capacitor bank though! I would be a menace if I had that many capacitors!
The wiring is fast, cheap and something I just had as scrap. I expected to blow a cap before the wiring being a limit :)
@@KaizerPowerElectronicsDk 😂Gotcha! I was just thinking you annihilated one of those wires with the can crusher, the rest of the circuit is at max 2 of them. 😬
Can't wait for V2!
Ha that was cool, good work mate, 😎👍
"It just needs a nice box", the famous saying :)
A couple suggestions:
1. Hook up a lightbulb in series with a decent resistor across the capacitors, that is always connected. It'll take longer to charge, and a parasitic loss after disconnect, but a visual indicator whether there's charge in the bank, and a means of no charge accruing is a must.
2. Swap out the copper wire for aluminium bus bars. It's cheaper, more easily formed/machined for purpose, and due to the skin effect should allow higher peak currents.
1: Safety first is always a great idea, I skipped it here, but on another bank I made, with 35 electrolytics in series for 14kV, I had LEDs everywhere for status (and balancing the long string). 2: The wiring is fast, cheap and something I just had as scrap. This was the first test, a rebuild was planned, now that they proved themselves :)
Lol awesome! As much as I'd like to play around with stuff like that I'm afraid of answering the door and explaining to officers that I don't in fact have a firing range in my living room.
Not exactly a common cause for officers to knock a door in Denmark ;)
Also was curious if you had a video I'm not seeing in regaurds to the maxwell cap that failed? Seems pretty rare those things fail, even when you find 20 year old ones they often still have many thousands of discharges left at their full ratings, to which I'd never expect to see/hear about a terminal on one failing especially. I picked up a couple that are rated for 50KJ and I plan to basically never go over 25KJ so they're more/less last me 'indefinently' since they're pulse caps rated for dead-shorting. Kinda surprised/impressed to hear you had a maxwell cap fail, just never seen/heard of it myself unless someone did over-volting and had hundreds of shots in @ over-volted. Be curious to see more on what happend there!
I do not think I ever made a video or even took pictures of the damage. Here at 60 seconds, you can see a black smoke puff in the left side and in the slow motion, you can see the terminal flying into frame: ruclips.net/video/Z4UBXoG-Aj8/видео.html
I love these, I have 20 caps to make my own, 196uf, 2300vdc from cardiac defibrillators....I'm terrified to change it past like 100v as full charge is 520j per cap or like 10,000 if I wire them all up.
I have a question for anyone, when discharging like this do the caps go reverse voltage? How do you prevent this or does it matter? I cant find a data sheet for the caps aerovox ql232ew196v23a
Defibrillator capacitors are also some beasts! 10kJ is way up in the scary end of the scale. Some of the best Maxwell pulse discharge capacitors are only rated for 50-70% voltage reversal, but must still be limited to some certain dV/dt.
Voltage reversal matter a lot. It is the most likely killer of pulse discharge capacitors. To avoid it, I make my work piece / work coil weak in a point where I am sure it blow open circuit. Ensuring it with diodes can be less cost effective. Check out the pulse power folder here for some papers that might inspire you for a solution: kaizerpowerelectronics.dk/theory/file-archive/?drawer=application_notes
I've got my hands on 120 caps of 500V 3900uF, could build an enormous bank but I'm afraid to start doing something with them. Close to 60kJ. Thx for the tip on voltage reversal.
Sounds like electrolytic capacitors, they are even more prone to fail from voltage reversal.
@@axiom1650 yes build it and they will come,.lol
Awesome - & good fun
Can you use a voltage multiplier with capacitors and diodes to the the cap bank instead of MOT?
to charge the capacitor bank*
yes, you can use a wide variety of power supply methods. The charge current available will however determine the charge time.
Oh boy, reminds me of when I built 10mF 400V capacitor bank for DRSSTC and accidentally shorted the terminals with fully charged capacitors. Suffice to say, my ears were ringing for good half an hour.
I did that with smaller capacitor bank long, just wanted to make sure it was shorted, before touching it, still had enough charge to ensure some ringing in the ears :) Ear defenders stay on since that day.
PFC capacitors can take a lot of abuse. Used a few of these for a laser power supply that needed a fast pulse discharge. They can take 1800v net a piece. Charge the elements with plus and minus HV and the third one being neutral. ❤
So you have a center tapped high voltage power supply?
Shouldn’t peak DC voltage rating be ~1.4 times higher than AC RMS voltage not 2 times?
Also that’s the 2 second test voltage. My guess after looking at those specs, is you start getting little breakdown points in the polypropylene insulation somewhere a bit over 1000VDC, which will gradually eat away at the capacitor.
I did not supply the information, but in Tesla coil community we gladly just look at AC ratings and say oh that lovely swing from negative to positive, that's the same as DC, so twice the AC is DC rating :) I said 60 seconds, but its 2 seconds in datasheet. I also got breakdowns at 1350+ VDC, that's where I heard crackling/popping, its the second test in video.
I wonder if enough of these can/will trigger lightning detectors
- time to check DMI's Lyndata for freak strikes ;ø)
I would not want to try something can be detected by that, because that properly means I triggered lightning. RIP me :)
@@KaizerPowerElectronicsDk I would neither - keeping the loop small keeps one safe'errrr
/me imagining litze wire in a 2 meter loop blitzing 🌩
All the LPL's would say all Flukes are Flukes.
Should the Fluke be unfluked?
@@KaizerPowerElectronicsDk I bought 2 flukes in my life and both had the shortest life despite being the dearest!
Jes!!! Thats 2 times the energy of a rifle round like .308
For shure that emits at least as much acoustic power as the muzzle blast of such a round, probably even more.😆👍
Generally not a thing we are used to in Denmark, no one knows what a firearm sounds like. Except from the army, but that's mostly shooting outside.
Had to sound like a gun shoot, awesome 👍👍
Gun shots are more sharp and high pitched, this is so deep and hollow. The surface of the foil is also something like 20 cm long and 5 cm wide, so a large area going into boom :)
@@KaizerPowerElectronicsDk Ok, thanks for clarifying 😁
Ive done 2KJ off a .maxwell pusle cap, and when you stand outside the barn across the yard, it literally sounds exactly like a pistol being shot. Thankfully in my case its country..so people shoot in their backyard all the time. Very sharp crack!
@@PhxSt0rmz I can not trust anyone else with this, so I only heard it from the inside of the work shop :)
-1 pant 😂. Still nice experiments , also you could look into producing graphene with this method , just add carbon instead of whatever you are frying in the middle 😂👌
Gotta spend money to make money! ;)
What do you think could make a capacitor bank like this last longer than 200 discharges ? ( other than discharging them slower ) .
I’m also thinking of building a smaller capacitor bank for experiments but would like to use scrap capacitors just for starters , in case I get something wrong 😂