SSTC probably the simplest design
Mates, Sun Oct 21 2007, 10:09PMHi guys,
I originally built this circuit for driving Fly-Back but it can be easily used for TC. I have not seen design of SSTC like this so it can be small inspiration for you TC-freaks…
This circuit has apart from the simplicity few other advantages. First you can drive the gate directly from TL494 (ideally followed by UCC372). Next you can play with power of the system by simply adding another motor caps (those can be over 70uF each) and finally you can parallel the transistors…
The only weak points are the TVS which have to be dipped inside small oil tank to be cooled. Also the rectifying diodes get pretty hot. Next week I’ll buy few more caps and let’s see how much more power I can get…
The resonant freq of the secondary is around 350KHz. The driver for the IGBT is based on TL494 (check
followed by UCC3721P and UCC3722P which are directly connected to the gate. here is a short movie, the TC in action
Comments and questions welcomed!
Re: SSTC probably the simplest design
Dalus, Sun Oct 21 2007, 10:18PM
It sure is nice and simple but it's not really efficient as you're wasting almost all of your inductive energy. Non the less great sparks for such an simple circuit. Keep them sparking.
Dalus, Sun Oct 21 2007, 10:18PM
It sure is nice and simple but it's not really efficient as you're wasting almost all of your inductive energy. Non the less great sparks for such an simple circuit. Keep them sparking.
Re: SSTC probably the simplest design
Shaun, Mon Oct 22 2007, 01:31AM
It is one of the simpler SSTCs I've seen, really nice arcs for its size.
However, the problem we run into with set-frequency drivers is that the Fres of the secondary changes depending on its surroundings, a phenomenon known as drifting. If you get too close to the coil, it could detune. Then primary current will rise and...let's just say you're dissipating enough heat already.
Shaun, Mon Oct 22 2007, 01:31AM
It is one of the simpler SSTCs I've seen, really nice arcs for its size.
However, the problem we run into with set-frequency drivers is that the Fres of the secondary changes depending on its surroundings, a phenomenon known as drifting. If you get too close to the coil, it could detune. Then primary current will rise and...let's just say you're dissipating enough heat already.
Re: SSTC probably the simplest design
Mates, Mon Oct 22 2007, 06:25AM
There is a kind of feedback in the system (check the circuit that's richie burnett concept), but it is true that the detuning could be solved in much better way.On the other hand it is cool to change the sounds of the coil just by moving your hand around the TL494, it is almost like some crazy Theremin...(you can hear it in the movie, without my moving hand in the driver vicinity the sound is clear 50Hz and the sparks are litle bit boring)
According to the efficiency: I don't think I'm wasting too much of energy. The point is that most of the energy which is not transformed gets back into the caps...The only losses are in the didodes and litle bit in TVS. The transistor also heats up, but is not critical and better driver design should improve this a lot (the shape of the driving wave is not ideal squere at the moment). I have no AC-Amp meter so I can't tell you what's the power consuption is, but I'll try to measure it on my flat power meter...
Mates, Mon Oct 22 2007, 06:25AM
Shaun wrote ...
It is one of the simpler SSTCs I've seen, really nice arcs for its size.
However, the problem we run into with set-frequency drivers is that the Fres of the secondary changes depending on its surroundings, a phenomenon known as drifting. If you get too close to the coil, it could detune. Then primary current will rise and...let's just say you're dissipating enough heat already.
It is one of the simpler SSTCs I've seen, really nice arcs for its size.
However, the problem we run into with set-frequency drivers is that the Fres of the secondary changes depending on its surroundings, a phenomenon known as drifting. If you get too close to the coil, it could detune. Then primary current will rise and...let's just say you're dissipating enough heat already.
There is a kind of feedback in the system (check the circuit
that's richie burnett concept), but it is true that the detuning could be solved in much better way.On the other hand it is cool to change the sounds of the coil just by moving your hand around the TL494, it is almost like some crazy Theremin...(you can hear it in the movie, without my moving hand in the driver vicinity the sound is clear 50Hz and the sparks are litle bit boring)According to the efficiency: I don't think I'm wasting too much of energy. The point is that most of the energy which is not transformed gets back into the caps...The only losses are in the didodes and litle bit in TVS. The transistor also heats up, but is not critical and better driver design should improve this a lot (the shape of the driving wave is not ideal squere at the moment). I have no AC-Amp meter so I can't tell you what's the power consuption is, but I'll try to measure it on my flat power meter...
Re: SSTC probably the simplest design
Marko, Mon Oct 22 2007, 12:33PM
It makes me sad when I see a circuit with 20 stacked transzorbs. So much neater ways to do the same!
No, that is possible only with bridge topologies. Considering the amount of TVS's and heatsink size you used, your efficiency actually looks very poor. How hot do they get?
Apart from that, IGBT is very slow compared to mosfets for same power level. Your circuit is similar to original steve W. micro SSTC which is also very inefficient.
When you attempt to interrupt the primary current, it's like you are trying to brake a large flywheel. Faster you do it it will put more force on the brake (voltage on the switch) and you must convert all of it's energy into heat to stop it.
If there was no TVS's to burn it transistor would avalanche and dissipate this heat.
People don't use bridges and ZVS topologies for nothing. Just for the cost of TVS's you could have build a much better driver circuit... Just try it
Don't get mad on me, but a bridge driven coil is probably best inspiration for anyone getting into these things.
Marko, Mon Oct 22 2007, 12:33PM
It makes me sad when I see a circuit with 20 stacked transzorbs. So much neater ways to do the same!
The point is that most of the energy which is not transformed gets back into the caps...
No, that is possible only with bridge topologies. Considering the amount of TVS's and heatsink size you used, your efficiency actually looks very poor. How hot do they get?
Apart from that, IGBT is very slow compared to mosfets for same power level. Your circuit is similar to original steve W. micro SSTC which is also very inefficient.
When you attempt to interrupt the primary current, it's like you are trying to brake a large flywheel. Faster you do it it will put more force on the brake (voltage on the switch) and you must convert all of it's energy into heat to stop it.
If there was no TVS's to burn it transistor would avalanche and dissipate this heat.
People don't use bridges and ZVS topologies for nothing. Just for the cost of TVS's you could have build a much better driver circuit... Just try it
Don't get mad on me, but a bridge driven coil is probably best inspiration for anyone getting into these things.
Re: SSTC probably the simplest design
Dr. Dark Current, Mon Oct 22 2007, 02:02PM
Well.. if you want a really simple SSTC that would requie one chip and no GDT and would be for sure more efficient than yours, check the IR2111 chip (or similar ones- I'm sure there are faster versions). It's a half bridge driver with integrated dead time. Just connect it to your bridge, slap an antenna on the input pin and you're set.
Dr. Dark Current, Mon Oct 22 2007, 02:02PM
Well.. if you want a really simple SSTC that would requie one chip and no GDT and would be for sure more efficient than yours, check the IR2111 chip (or similar ones- I'm sure there are faster versions). It's a half bridge driver with integrated dead time. Just connect it to your bridge, slap an antenna on the input pin and you're set.
Re: SSTC probably the simplest design
Mates, Mon Oct 22 2007, 04:58PM
Let's disscuss this closer:
I´m not an expert in this but I think that most of the energy stored in the caps goes through the transistor long before it close (foil caps and the primary have almost no resistance). Only the rise time is critical point which makes the transistor hot (BTW the heat sink is something like my working table, much smaller would do the job). The TVS are necessary for clamping the inductive voltage (which otherwise kill the transistors) and I agree it's a pitty it must be like that. But it does not mean they dissipate all the energy! It would destroy them immediately. The energy just go back into the caps and it is again the rise time of the TVS (which is not optimal) which makes them hot. (they are rated only 5W each!). BTW I run only four doublets of TVS (that's not even 20W and no heat sink) before without oil and they survived 2-3 minutes...
Next, I do not understand why you think that the energy is not going back into the caps. I can easilly bypass the transistor with piece of wire (looks like short-circuit) and the power consuption goes to minimum. The energy just circulate in the circuit and the small loses are taken from 220V AC...That's also the reason I do not need fuse, because the short circuit is impossible...
However, I'm not saying that this coil is an efficiency competition winner
Cheers
BTW the price of the one way TVS is quite low...
Mates, Mon Oct 22 2007, 04:58PM
Marko wrote ...
It makes me sad when I see a circuit with 20 stacked transzorbs. So much neater ways to do the same!
No, that is possible only with bridge topologies. Considering the amount of TVS's and heatsink size you used, your efficiency actually looks very poor. How hot do they get?
Apart from that, IGBT is very slow compared to mosfets for same power level. Your circuit is similar to original steve W. micro SSTC which is also very inefficient.
When you attempt to interrupt the primary current, it's like you are trying to brake a large flywheel. Faster you do it it will put more force on the brake (voltage on the switch) and you must convert all of it's energy into heat to stop it.
If there was no TVS's to burn it transistor would avalanche and dissipate this heat.
People don't use bridges and ZVS topologies for nothing. Just for the cost of TVS's you could have build a much better driver circuit... Just try it
Don't get mad on me, but a bridge driven coil is probably best inspiration for anyone getting into these things.
It makes me sad when I see a circuit with 20 stacked transzorbs. So much neater ways to do the same!
The point is that most of the energy which is not transformed gets back into the caps...
No, that is possible only with bridge topologies. Considering the amount of TVS's and heatsink size you used, your efficiency actually looks very poor. How hot do they get?
Apart from that, IGBT is very slow compared to mosfets for same power level. Your circuit is similar to original steve W. micro SSTC which is also very inefficient.
When you attempt to interrupt the primary current, it's like you are trying to brake a large flywheel. Faster you do it it will put more force on the brake (voltage on the switch) and you must convert all of it's energy into heat to stop it.
If there was no TVS's to burn it transistor would avalanche and dissipate this heat.
People don't use bridges and ZVS topologies for nothing. Just for the cost of TVS's you could have build a much better driver circuit... Just try it
Don't get mad on me, but a bridge driven coil is probably best inspiration for anyone getting into these things.
Let's disscuss this closer:
I´m not an expert in this but I think that most of the energy stored in the caps goes through the transistor long before it close (foil caps and the primary have almost no resistance). Only the rise time is critical point which makes the transistor hot (BTW the heat sink is something like my working table, much smaller would do the job). The TVS are necessary for clamping the inductive voltage (which otherwise kill the transistors) and I agree it's a pitty it must be like that. But it does not mean they dissipate all the energy! It would destroy them immediately. The energy just go back into the caps and it is again the rise time of the TVS (which is not optimal) which makes them hot. (they are rated only 5W each!). BTW I run only four doublets of TVS (that's not even 20W and no heat sink) before without oil and they survived 2-3 minutes...
Next, I do not understand why you think that the energy is not going back into the caps. I can easilly bypass the transistor with piece of wire (looks like short-circuit) and the power consuption goes to minimum. The energy just circulate in the circuit and the small loses are taken from 220V AC...That's also the reason I do not need fuse, because the short circuit is impossible...
However, I'm not saying that this coil is an efficiency competition winner
Cheers
BTW the price of the one way TVS is quite low...
Re: SSTC probably the simplest design
Marko, Mon Oct 22 2007, 06:16PM
Your motor caps do nothing apart from building a reactively limited halfwave rectified power supply. For TC it's all just a voltage supply. You may be fine simply with a transformer giving similar output voltage and current to what coil happens to be using. That's what it has to do with your bypassing of the IGBT.
There is no reason why it would. The produced voltage is always such that it attempts to keep the current flowing in same direction, and adds up with supply voltage instead of canceling with it (which is what is needed to actually restore power back to supply).
That is why all true flyback converters which are designed for varying load absolutely need feedback to operate efficiently.
So you are dumping it all, and I don't know what you consider ''low'' efficiency - 100W of dissipation worth of TVS's under oil?
I bet anyone could easily build a bridge coil producing same sized sparks for maybe 2-3W of wasted power including control electronics...
Marko
Marko, Mon Oct 22 2007, 06:16PM
Your motor caps do nothing apart from building a reactively limited halfwave rectified power supply. For TC it's all just a voltage supply. You may be fine simply with a transformer giving similar output voltage and current to what coil happens to be using. That's what it has to do with your bypassing of the IGBT.
Next, I do not understand why you think that the energy is not going back into the caps.
There is no reason why it would. The produced voltage is always such that it attempts to keep the current flowing in same direction, and adds up with supply voltage instead of canceling with it (which is what is needed to actually restore power back to supply).
That is why all true flyback converters which are designed for varying load absolutely need feedback to operate efficiently.
So you are dumping it all, and I don't know what you consider ''low'' efficiency - 100W of dissipation worth of TVS's under oil?
I bet anyone could easily build a bridge coil producing same sized sparks for maybe 2-3W of wasted power including control electronics...
Marko
Re: SSTC probably the simplest design
Mates, Mon Oct 22 2007, 06:44PM
Mates, Mon Oct 22 2007, 06:44PM
Marko wrote ...
Your motor caps do nothing apart from building a reactively limited halfwave rectified power supply. For TC it's all just a voltage supply. You may be fine simply with a transformer giving similar output voltage and current to what coil happens to be using. That's what it has to do with your bypassing of the IGBT.
There is no reason why it would. The produced voltage is always such that it attempts to keep the current flowing in same direction, and adds up with supply voltage instead of canceling with it (which is what is needed to actually restore power back to supply).
That is why all true flyback converters which are designed for varying load absolutely need feedback to operate efficiently.
So you are dumping it all, and I don't know what you consider ''low'' efficiency - 100W of dissipation worth of TVS's under oil?
I bet anyone could easily build a bridge coil producing same sized sparks for maybe 2-3W of wasted power including control electronics...
Marko
OK Marco...
The coil disipate much more than 3W (my guess is around 30W). I'm not saying it is a substitution for the bridge design. I just like doing things differently...
Thanks, for comments my problem is I have normally nobody to ask. I'll think about what you wrotte...
Cheers mates
Your motor caps do nothing apart from building a reactively limited halfwave rectified power supply. For TC it's all just a voltage supply. You may be fine simply with a transformer giving similar output voltage and current to what coil happens to be using. That's what it has to do with your bypassing of the IGBT.
Next, I do not understand why you think that the energy is not going back into the caps.
There is no reason why it would. The produced voltage is always such that it attempts to keep the current flowing in same direction, and adds up with supply voltage instead of canceling with it (which is what is needed to actually restore power back to supply).
That is why all true flyback converters which are designed for varying load absolutely need feedback to operate efficiently.
So you are dumping it all, and I don't know what you consider ''low'' efficiency - 100W of dissipation worth of TVS's under oil?
I bet anyone could easily build a bridge coil producing same sized sparks for maybe 2-3W of wasted power including control electronics...
Marko
OK Marco...
The coil disipate much more than 3W (my guess is around 30W). I'm not saying it is a substitution for the bridge design. I just like doing things differently...
Thanks, for comments my problem is I have normally nobody to ask. I'll think about what you wrotte...
Cheers mates
Print this page