Necessity of toroidal cores in TCs?
Dr. Shark, Sun Jun 04 2006, 03:31PMPretty much every design of a DRSSTC uses ferrite toroids for the GDT as well as for current feedback. Now, I would not change a working design, but since I don't really have a lot of toroids, but all the more double-E cores, I wonder if these can be used or not. For example in a ATX power supply, both the GTD and the OCD transformers are double-E's, so why would they not be suitable for the same job in a DRSSTC? Sure they have a slighly higher leakage inductance, so one might argue that this could be a problem with the higher frequencys used in a TC. But is it really that bad? I have been using double-E's for GDTs in the past and they worked quited well, now I am trying to get a cascaded feedback current transformer done from E cores as well, but is seems to cause problems. I wonder if these problems are likely to come from somewhere else, or if using E-cores really is that bad.
Re: Necessity of toroidal cores in TCs?
HV Enthusiast, Sun Jun 04 2006, 03:56PM
You really can't say one core is better over another for a given application without knowing the specifications of the core, and the requirements for the design.
Although a toroidal core can yield a transformer which much higher coupling than one of an E-core, that gain in coupling may not be necessary for a given design. You need to look at all parameters when making a comparison.
An e-core if properly designed could work equally well for a GDT design. I think the biggest reason we use toroids in our GDTs is that they are simply easier to wind than that of an E-core.
Remember, just because something hasn't been done before (ie ecores over toroid cores) doesn't mean it can't work. Most of the things we do as amateurs in our hobby is due to ease and availability, not necessity. Winding e-cores can be a bitch, while toroids are simply easier to wind.
HV Enthusiast, Sun Jun 04 2006, 03:56PM
You really can't say one core is better over another for a given application without knowing the specifications of the core, and the requirements for the design.
Although a toroidal core can yield a transformer which much higher coupling than one of an E-core, that gain in coupling may not be necessary for a given design. You need to look at all parameters when making a comparison.
An e-core if properly designed could work equally well for a GDT design. I think the biggest reason we use toroids in our GDTs is that they are simply easier to wind than that of an E-core.
Remember, just because something hasn't been done before (ie ecores over toroid cores) doesn't mean it can't work. Most of the things we do as amateurs in our hobby is due to ease and availability, not necessity. Winding e-cores can be a bitch, while toroids are simply easier to wind.
Re: Necessity of toroidal cores in TCs?
Marko, Sun Jun 04 2006, 04:02PM
My CT's are all such.
I use E cores from the filter chokes.
They are much harder to wind and hold together after that, but that's only disadvantage.
I also found out that it is very easy to damage enameled wire on sharp ferrite edges, and it's hard to protect the core fully.
Dipping in epoxy or using thinnest possible PVC insulated wire (I couldn't fit even closely enough turns with standard telephone wires, and ended using tiny wire out of computer mouse cable).
Maybe you could also use existing former to wind with enameled wire, or ewen use existing winding (I found some of such chokes having just about 33 turns
).
That could be good for CT's but I guess former would cause too low coupling for a GDT.
If it doesn't work CT's are probably shorted, or there is malfunction somewhere in electronics (you can try antenna there)
Marko, Sun Jun 04 2006, 04:02PM
My CT's are all such.
I use E cores from the filter chokes.
They are much harder to wind and hold together after that, but that's only disadvantage.
I also found out that it is very easy to damage enameled wire on sharp ferrite edges, and it's hard to protect the core fully.
Dipping in epoxy or using thinnest possible PVC insulated wire (I couldn't fit even closely enough turns with standard telephone wires, and ended using tiny wire out of computer mouse cable).
Maybe you could also use existing former to wind with enameled wire, or ewen use existing winding (I found some of such chokes having just about 33 turns
).That could be good for CT's but I guess former would cause too low coupling for a GDT.
If it doesn't work CT's are probably shorted, or there is malfunction somewhere in electronics (you can try antenna there)
Re: Necessity of toroidal cores in TCs?
Dr. Shark, Sun Jun 04 2006, 04:23PM
Funny, I seem to be in the minority here. Given that you have a bobbin, which is always the case if you butcher a transfomer, I find it much easier to wind E cores. No threading through the hole, just go round and round.
Firkragg, you are probably right, my CT might have internal shorts. It is quite sloppily wound, and the wire has very thin enamel, so I better redo it.
Dr. Shark, Sun Jun 04 2006, 04:23PM
Funny, I seem to be in the minority here. Given that you have a bobbin, which is always the case if you butcher a transfomer, I find it much easier to wind E cores. No threading through the hole, just go round and round.
Firkragg, you are probably right, my CT might have internal shorts. It is quite sloppily wound, and the wire has very thin enamel, so I better redo it.
Re: Necessity of toroidal cores in TCs?
Marko, Sun Jun 04 2006, 04:51PM
The 'bobbin' of the transformer could be a problem for a GDT since they love very tight coupling.
And I don't know if you could fit all the turns in there, but for CT's they are fine.
I wound my CT windings separately, dipped in urethane, core was taped and I still managed to damage the wire making one CT useless.
You should use PVC wire or maybe leave the bobbin with original winding.
I'm also in need of some real ferrite toroids, maybe a good sample from fairrite or somewhere.
Marko, Sun Jun 04 2006, 04:51PM
The 'bobbin' of the transformer could be a problem for a GDT since they love very tight coupling.
And I don't know if you could fit all the turns in there, but for CT's they are fine.
I wound my CT windings separately, dipped in urethane, core was taped and I still managed to damage the wire making one CT useless.
You should use PVC wire or maybe leave the bobbin with original winding.
I'm also in need of some real ferrite toroids, maybe a good sample from fairrite or somewhere.
Re: Necessity of toroidal cores in TCs?
HV Enthusiast, Sun Jun 04 2006, 05:29PM
CWS Bytemark sells ferrite tores for about $2.00 each.
HV Enthusiast, Sun Jun 04 2006, 05:29PM
CWS Bytemark sells ferrite tores for about $2.00 each.
Re: Necessity of toroidal cores in TCs?
Steve Conner, Sun Jun 04 2006, 10:00PM
I made an E-I core GDT, and it worked fine at all the frequencies I tried it at, up to 250kHz. As long as you use the multifilar winding technique, I think the coupling is fine. I twisted 5 wires together into a bundle and then wound that round the centre leg of the "E".
I think toroid cores are somewhat nicer for CTs though, they seem easier to Faraday shield. Now I use toroid cores made of Fair-rite type 78 material for just about everything, since I can get them for about $2-$3 each. Instead of using hundreds of turns on CTs, or cascades, I tend to use 30 to 100 turns and just use a smaller burden resistance.
Steve Conner, Sun Jun 04 2006, 10:00PM
I made an E-I core GDT, and it worked fine at all the frequencies I tried it at, up to 250kHz. As long as you use the multifilar winding technique, I think the coupling is fine. I twisted 5 wires together into a bundle and then wound that round the centre leg of the "E".
I think toroid cores are somewhat nicer for CTs though, they seem easier to Faraday shield. Now I use toroid cores made of Fair-rite type 78 material for just about everything, since I can get them for about $2-$3 each. Instead of using hundreds of turns on CTs, or cascades, I tend to use 30 to 100 turns and just use a smaller burden resistance.
Re: Necessity of toroidal cores in TCs?
Marko, Sun Jun 04 2006, 10:04PM
I did the same, but with even more wires. I used 15 thinner wires and bundled them together with a lego motor. Then I used 3 togeher as single wire.
I don't kno wif this was necessary but coupling couldn't be better.
Marko, Sun Jun 04 2006, 10:04PM
I did the same, but with even more wires. I used 15 thinner wires and bundled them together with a lego motor. Then I used 3 togeher as single wire.
I don't kno wif this was necessary but coupling couldn't be better.
Print this page