MMC arcing, blew a cap
crashstudio, Mon Sept 21 2015, 08:43PMI just added another transformer and redesigned the mmc from 2 strings of 20 to 3 strings of 16 with 3 other taps at 15, 14 and 13 for an LTR capacitance of 28nf, 30nf, 32nf and 35nf. The first power up with a varaic it started to arc somewhere then the Terry filter safety gap kicked in. I repeated this until I realized it was the mmc, I had to pull it out of the drawer to see it happen. It looked like one of the caps was blown but after removing it from the circuit it tested good.
I'm wondering if maybe the caps are too close to each other and maybe I need to insulate them? If I do have to insulate do you have any suggestions? I was very careful wiring them. Tested everything along the way. I am sure they are wired perfect, nothing crossed. 3 series strings all paralleled.
Power two 15kv 60ma NST Parallel
Static gap
Terry Filter
Four 80uF PFC caps (2 in parallel for each NST) total 160uF each
1 A/C line filter before the PFC's
Primary is 1/4" copper refrigeration tubing 50'
Secondary 4" x 23" PVC 26awg aprox 1200 turns 20" total height
Toroid 4" x 17" aluminum ducting wrapped in aluminum tape
Here is a link to all pics...
Thank you
Re: MMC arcing, blew a cap
Ash Small, Tue Sept 22 2015, 10:19AM
I'd get a plastic container, boil up some cheap cooking oil to remove any moisture, and just immerse the whole MMC in oil.
If this solves the problem I might then try to construct something more elaborate
Ash Small, Tue Sept 22 2015, 10:19AM
I'd get a plastic container, boil up some cheap cooking oil to remove any moisture, and just immerse the whole MMC in oil.
If this solves the problem I might then try to construct something more elaborate
Re: MMC arcing, blew a cap
crashstudio, Tue Sept 22 2015, 08:04PM
It functioned perfect with the one NST but now I'm thinking the extra current from the second one is causing it to arc right? If I construct something different then how much space will I need between contacts to avoid an arc?
The thing that causes me to question the cause is every image I see online of an mmc they are always mounted so close to each other, some of them much closer than mine. Or is that most people aren't pushing 120ma through their mmc?
crashstudio, Tue Sept 22 2015, 08:04PM
It functioned perfect with the one NST but now I'm thinking the extra current from the second one is causing it to arc right? If I construct something different then how much space will I need between contacts to avoid an arc?
The thing that causes me to question the cause is every image I see online of an mmc they are always mounted so close to each other, some of them much closer than mine. Or is that most people aren't pushing 120ma through their mmc?
Re: MMC arcing, blew a cap
Ash Small, Tue Sept 22 2015, 10:16PM
I've not built an MMC myself yet (except with parallelled 'hand rolled' caps) but I imagine you could reduce the ohmic resistance on the back of the board.
I'd try shortening the connections between the caps to reduce resistance, before adding additional strings to deal with the extra power.
Reducing resistance to a minimum is always going to improve performance anyway.
I'd still put it under oil first, then see if reducing resistance improves things
Just my opinion, see what others say.
Ash Small, Tue Sept 22 2015, 10:16PM
I've not built an MMC myself yet (except with parallelled 'hand rolled' caps) but I imagine you could reduce the ohmic resistance on the back of the board.
I'd try shortening the connections between the caps to reduce resistance, before adding additional strings to deal with the extra power.
Reducing resistance to a minimum is always going to improve performance anyway.
I'd still put it under oil first, then see if reducing resistance improves things
Just my opinion, see what others say.
Re: MMC arcing, blew a cap
Hazmatt_(The Underdog), Wed Sept 23 2015, 01:04AM
Just adding another transformer for "more power" is not enough, the capacitor needs to match the impedance of the transformer so that it is in a sense, a constant current charging source.
Your Z = 125 K ohm
Your Resonant C= 21.2nF
your LTR C = 30nF
So that all checks out, I agree it's probably a corona issue then.
I have had my MMC's flash over, even a distance of 3".
You might also consider getting a rotary gap going so you can better control how fast the energy is "removed" from the system.
Hazmatt_(The Underdog), Wed Sept 23 2015, 01:04AM
Just adding another transformer for "more power" is not enough, the capacitor needs to match the impedance of the transformer so that it is in a sense, a constant current charging source.
Your Z = 125 K ohm
Your Resonant C= 21.2nF
your LTR C = 30nF
So that all checks out, I agree it's probably a corona issue then.
I have had my MMC's flash over, even a distance of 3".
You might also consider getting a rotary gap going so you can better control how fast the energy is "removed" from the system.
Re: MMC arcing, blew a cap
crashstudio, Wed Sept 23 2015, 05:22AM
OK so if I have to dissemble the mmc and redesign it I'm totally up for it but not sure how to match the impedance. I've been following the specs from TelsaMap which only says 34.3nF for optimum capacitance, then I just provided 3 additional tap points. Could the wires be the problem? I'm using the 15kv high voltage.
I obviously still have much to learn so I really need someone to point my in the right direction please.
Here's the pics of my TeslaMap specs.
crashstudio, Wed Sept 23 2015, 05:22AM
OK so if I have to dissemble the mmc and redesign it I'm totally up for it but not sure how to match the impedance. I've been following the specs from TelsaMap which only says 34.3nF for optimum capacitance, then I just provided 3 additional tap points. Could the wires be the problem? I'm using the 15kv high voltage.
I obviously still have much to learn so I really need someone to point my in the right direction please.
Here's the pics of my TeslaMap specs.
Re: MMC arcing, blew a cap
crashstudio, Thu Sept 24 2015, 09:00PM
TeslaMap shows that if I use a rotary gap my mmc would need to be 55.3nF. Is this right? Besides for now I would really like to stick with the static. My coil worked great with it powered by one NST. I already added 8 caps and rewired it from 2 strings of 20 to 3 strings of 16. And how do I go about calculating resistance in the mmc? If all caps were closer together that would reduce it right? I designed it with them so far apart in hopes of avoiding any flash over.
I just really want to avoid the many hours it will take to reconstruct this without first fully knowing what I'm doing.
crashstudio, Thu Sept 24 2015, 09:00PM
Hazmatt_(The Underdog) wrote ...
You might also consider getting a rotary gap going so you can better control how fast the energy is "removed" from the system.
You might also consider getting a rotary gap going so you can better control how fast the energy is "removed" from the system.
TeslaMap shows that if I use a rotary gap my mmc would need to be 55.3nF. Is this right? Besides for now I would really like to stick with the static. My coil worked great with it powered by one NST. I already added 8 caps and rewired it from 2 strings of 20 to 3 strings of 16. And how do I go about calculating resistance in the mmc? If all caps were closer together that would reduce it right? I designed it with them so far apart in hopes of avoiding any flash over.
I just really want to avoid the many hours it will take to reconstruct this without first fully knowing what I'm doing.
Re: MMC arcing, blew a cap
Hazmatt_(The Underdog), Fri Sept 25 2015, 01:05AM
You can do what you want, but I run 1.4x Cres with rotaries all the time and it works fine.
I did the simulations, just like Terry Fritz did ages ago and it calculates out.
What you are looking for is the capacitor to be charged in a constant current mode, and to load the transformer to the appropriate terminal voltage and current, that is all.
If you really want to know what is going on, characterize your transformer, and do a simulation, it takes time and patience but it is worth it to know what is really going on.
Hazmatt_(The Underdog), Fri Sept 25 2015, 01:05AM
You can do what you want, but I run 1.4x Cres with rotaries all the time and it works fine.
I did the simulations, just like Terry Fritz did ages ago and it calculates out.
What you are looking for is the capacitor to be charged in a constant current mode, and to load the transformer to the appropriate terminal voltage and current, that is all.
If you really want to know what is going on, characterize your transformer, and do a simulation, it takes time and patience but it is worth it to know what is really going on.
Re: MMC arcing, blew a cap
crashstudio, Fri Sept 25 2015, 11:42PM
I've been doing some reading online and came to the same conclusion that I need a rotary gap. The static gap can't handle the amount of current I have now. Now I need to learn exactly how to design one. Or do I just slap one together with a variable speed control and wing it?
Also, according to my TeslaMap info I will need to alter my mmc to 55.3nF. Do you agree with this?
crashstudio, Fri Sept 25 2015, 11:42PM
I've been doing some reading online and came to the same conclusion that I need a rotary gap. The static gap can't handle the amount of current I have now. Now I need to learn exactly how to design one. Or do I just slap one together with a variable speed control and wing it?
Also, according to my TeslaMap info I will need to alter my mmc to 55.3nF. Do you agree with this?
Re: MMC arcing, blew a cap
Hazmatt_(The Underdog), Sat Sept 26 2015, 01:48AM
Rotary gaps are not all that hard to design and build, but you do need at least a drill press. I built this gap mostly on my Atlas milling machine.
I don't have a proper lathe, but I have a chuck I can put on the Atlas and do the lathe work semi-accurately.
I run 1.4x Cres with rotaries. That 55nF number is for a synchronous gap. My gaps are not synchronous, they are variable speed Dayton
type series motors, which are fine for the job if they are not worked too hard. ^^
Hazmatt_(The Underdog), Sat Sept 26 2015, 01:48AM
Rotary gaps are not all that hard to design and build, but you do need at least a drill press. I built this gap mostly on my Atlas milling machine.
I don't have a proper lathe, but I have a chuck I can put on the Atlas and do the lathe work semi-accurately.
I run 1.4x Cres with rotaries. That 55nF number is for a synchronous gap. My gaps are not synchronous, they are variable speed Dayton
type series motors, which are fine for the job if they are not worked too hard. ^^
Re: MMC arcing, blew a cap
crashstudio, Sat Sept 26 2015, 05:32AM
OK so I stick with the same capacitance I had with the static gap then? It's getting confusing because everywhere I look all I see is values for static gap or sync rotary but nothing for just a rotary. Sync is the one that's synced to the frequency of the A/C input right? I'm not even going to attempt that one.
Did you actually vacuum seal your spark gap? That looks cool as hell. And it looks like you're using knob caps. They work better for you then cornell dubillier?
crashstudio, Sat Sept 26 2015, 05:32AM
OK so I stick with the same capacitance I had with the static gap then? It's getting confusing because everywhere I look all I see is values for static gap or sync rotary but nothing for just a rotary. Sync is the one that's synced to the frequency of the A/C input right? I'm not even going to attempt that one.
Did you actually vacuum seal your spark gap? That looks cool as hell. And it looks like you're using knob caps. They work better for you then cornell dubillier?
Re: MMC arcing, blew a cap
Hazmatt_(The Underdog), Sat Sept 26 2015, 09:31PM
Synchronous you would have to machine the rotor of an AC motor.
Asynchronous will work with almost any motor, drill, angle grinder, whatever.
The gap isn't sealed, its just closed up for noise reduction, it took a while to build, but most anyone here could build one too.
I use doorknobs because they save space and have high voltage ratings, plus they don't get hot and die, they're really rugged.
Hazmatt_(The Underdog), Sat Sept 26 2015, 09:31PM
Synchronous you would have to machine the rotor of an AC motor.
Asynchronous will work with almost any motor, drill, angle grinder, whatever.
The gap isn't sealed, its just closed up for noise reduction, it took a while to build, but most anyone here could build one too.
I use doorknobs because they save space and have high voltage ratings, plus they don't get hot and die, they're really rugged.
Re: MMC arcing, blew a cap
crashstudio, Sat Sept 26 2015, 11:52PM
So as far as the capacitance, what formula do I go with since everywhere I look all I see is only 2 values, 1 for a static and another for a sync rotary. I'm at a loss here.
crashstudio, Sat Sept 26 2015, 11:52PM
So as far as the capacitance, what formula do I go with since everywhere I look all I see is only 2 values, 1 for a static and another for a sync rotary. I'm at a loss here.
Re: MMC arcing, blew a cap
Hazmatt_(The Underdog), Sun Sept 27 2015, 01:24AM
If you want to go with what the software tells you, you can.
I was more interested in the actuals, like you are trying to find out, but you will find that if you scratch the surface it gets
complicated very quickly.
In Short, I use 1.4x Cres.
Here is the work to prove why:
Hazmatt_(The Underdog), Sun Sept 27 2015, 01:24AM
If you want to go with what the software tells you, you can.
I was more interested in the actuals, like you are trying to find out, but you will find that if you scratch the surface it gets
complicated very quickly.
In Short, I use 1.4x Cres.
Here is the work to prove why:
Print this page