Starter questions
Stuart G., Sat Dec 09 2006, 11:15AMHi
I've been attempting to design a small sstc
After looking at the full bridge configuration, i thought that it might be too complicated for a first coil.
Therefore I have designed a driver based on the push pull topology.
First question, can anyone see a problem with my proposed driver circuit?
Second, do I need to limit the current entering the primary?
I would be using the cmos version of the 555
I've attached my scematic, its an ltspice file and I've also attached a picture of the schematic
]1165662951_495_FT0_schematic.zip[/file]
Schematic:
MOD Edit: changed the schematic to a linked image instead of an image embedded in the post, as it messed up the forum format (and a 3MB BMP might be a killer for the people still using dialup)
Re: Starter questions
Marko, Sat Dec 09 2006, 11:53AM
Oush pull drive will just be a source of trouble.
Unless you turn it into class E you'l be able to un on very low powers, few tens of watts at most.
SInce coupling between two primaries is low you are practically operating it like a double-ended flyback converter, resulting in large voltage spikes (many times supply voltage). 10ohm resistors in series with priamry will do nothing but be enormous heaters!
Your gate drive is also overly complicated; Simple driver chips that you can get sampled (UCC's, 442x's) proven to beat any discrete-component drive yet.
Half or full bridge with GDT is actually less tricky thing than this.
I would also recommend an antenna for self-resonant operation; when bridge coil is always in tune there's practically no need for fast diodes across mosfet's.
Marko, Sat Dec 09 2006, 11:53AM
Oush pull drive will just be a source of trouble.
Unless you turn it into class E you'l be able to un on very low powers, few tens of watts at most.
SInce coupling between two primaries is low you are practically operating it like a double-ended flyback converter, resulting in large voltage spikes (many times supply voltage). 10ohm resistors in series with priamry will do nothing but be enormous heaters!
Your gate drive is also overly complicated; Simple driver chips that you can get sampled (UCC's, 442x's) proven to beat any discrete-component drive yet.
Half or full bridge with GDT is actually less tricky thing than this.
I would also recommend an antenna for self-resonant operation; when bridge coil is always in tune there's practically no need for fast diodes across mosfet's.
Re: Starter questions
Avalanche, Sat Dec 09 2006, 07:15PM
Yeah I would have to agree, and say you go for a halfbridge for your first coil, as opposed to push-pull. You can still get a decent performer with a halfbridge. If you really don't want to go to the hassle of sourcing some dedicated gate driver chips for a first project, the easiest approach would probably be to use a push-pull arrangement similar to the above - but to drive a gate drive transformer, with a center tapped primary and two secondaries for the halfbridge fets.
Feedback is helpful to get the coil to stay in tune when drawing arcs, etc. You might want to look at Steve Ward's coils, where he uses a 555 to generate the initial signal, but then once the coil is running the antenna 'swamps out' the signal from the 555. You can leave out the antenna until you have everything working, then add it last to make your coil self tuning.
Avalanche, Sat Dec 09 2006, 07:15PM
Yeah I would have to agree, and say you go for a halfbridge for your first coil, as opposed to push-pull. You can still get a decent performer with a halfbridge. If you really don't want to go to the hassle of sourcing some dedicated gate driver chips for a first project, the easiest approach would probably be to use a push-pull arrangement similar to the above - but to drive a gate drive transformer, with a center tapped primary and two secondaries for the halfbridge fets.
Feedback is helpful to get the coil to stay in tune when drawing arcs, etc. You might want to look at Steve Ward's coils, where he uses a 555 to generate the initial signal, but then once the coil is running the antenna 'swamps out' the signal from the 555. You can leave out the antenna until you have everything working, then add it last to make your coil self tuning.
Re: Starter questions
Tom540, Thu Dec 14 2006, 05:50AM
That circuit you made is very similar to the one I tried out before I found 4HV. If anyone wants to try this circuit I made go for it. I was trying to brush up on my schematic capture skills for a new job I just got. The schematic needs work but its useable. There are a few caps with no values yet and the bypass caps on ic's and gate drivers are not shown. Let me know how it works or if it doesn't. Set up the LM567 caps as shown on the datasheet. Then twist the pot until the freq lock led lights. You can also use it with an antenna.
]1166075316_487_FT18810_drsstc2.pdf[/file]
Tom540, Thu Dec 14 2006, 05:50AM
That circuit you made is very similar to the one I tried out before I found 4HV. If anyone wants to try this circuit I made go for it. I was trying to brush up on my schematic capture skills for a new job I just got. The schematic needs work but its useable. There are a few caps with no values yet and the bypass caps on ic's and gate drivers are not shown. Let me know how it works or if it doesn't. Set up the LM567 caps as shown on the datasheet. Then twist the pot until the freq lock led lights. You can also use it with an antenna.
]1166075316_487_FT18810_drsstc2.pdf[/file]
Re: Starter questions
Stuart G., Sun Dec 17 2006, 06:29PM
any suggestions for a driver chip for a half/full bridge?
would I need anyhing other than pwm for current liming?
Stuart G., Sun Dec 17 2006, 06:29PM
any suggestions for a driver chip for a half/full bridge?
would I need anyhing other than pwm for current liming?
Re: Starter questions
Marko, Sun Dec 17 2006, 07:31PM
Just use one UCC37322 and one 37321 in a 'H bridge' for gate drive, and one 74HC14 to square up your feedback signal.
Practically nothing more is needed for a sucessful SSTC, and this approach is probably most used so far.
Steve ward's mini SSTC is basically that, the NE555 is interrupter that actually isn't necessary.
Bridge can be half or full, depending on your wish.
Marko, Sun Dec 17 2006, 07:31PM
Just use one UCC37322 and one 37321 in a 'H bridge' for gate drive, and one 74HC14 to square up your feedback signal.
Practically nothing more is needed for a sucessful SSTC, and this approach is probably most used so far.
Steve ward's mini SSTC is basically that, the NE555 is interrupter that actually isn't necessary.
Bridge can be half or full, depending on your wish.
Re: Starter questions
Stuart G., Tue Dec 19 2006, 12:44AM
what are all the inverting buffers for?
Stuart G., Tue Dec 19 2006, 12:44AM
what are all the inverting buffers for?
Re: Starter questions
Tom540, Tue Dec 19 2006, 07:42AM
The buffers are used to clean up the signal comming from the current transformer or antenna so that they are useable by the driver chips.
On my circuit I have a few extra ones which are used to open and close the transmission gates according to wether or not the circuit detects the right frequency comming from the CT or antenna. If the incomming freq is off then the circuit ouputs the oscillator frequency. I tested the circuit I posted tonight and it does work with a signal generator. The LM567 should be setup to have a very narrow bandwidth. the cap on pin2 should be greater than 1uF.
Tom540, Tue Dec 19 2006, 07:42AM
The buffers are used to clean up the signal comming from the current transformer or antenna so that they are useable by the driver chips.
On my circuit I have a few extra ones which are used to open and close the transmission gates according to wether or not the circuit detects the right frequency comming from the CT or antenna. If the incomming freq is off then the circuit ouputs the oscillator frequency. I tested the circuit I posted tonight and it does work with a signal generator. The LM567 should be setup to have a very narrow bandwidth. the cap on pin2 should be greater than 1uF.
Re: Starter questions
ragnar, Tue Dec 19 2006, 08:24AM
You only need one inverting buffer... I think Steve just put six there because there are six in the chip.. that doesn't mean you have to use them all.
Just make sure any free inputs are tied to ground and the outputs left open.
ragnar, Tue Dec 19 2006, 08:24AM
You only need one inverting buffer... I think Steve just put six there because there are six in the chip.. that doesn't mean you have to use them all.
Just make sure any free inputs are tied to ground and the outputs left open.
Re: Starter questions
Marko, Tue Dec 19 2006, 10:02AM
Seriesing all the inverters makes no use at all, except it may be handy to do on pcb and ensure none of gates are floating. I guess it was just Steve Ward's passion.
Maybe a better idea would be to paralell remaining 5 drivers, to speed up input switching times for UCC's. (although I doubt it would also make any noticeable difference).
Steve did this in his first DRSSTC driver, IIRC.
Marko, Tue Dec 19 2006, 10:02AM
Seriesing all the inverters makes no use at all, except it may be handy to do on pcb and ensure none of gates are floating. I guess it was just Steve Ward's passion.
Maybe a better idea would be to paralell remaining 5 drivers, to speed up input switching times for UCC's. (although I doubt it would also make any noticeable difference).
Steve did this in his first DRSSTC driver, IIRC.
Re: Starter questions
Steve Ward, Wed Dec 20 2006, 03:00AM
Its funny how often this comes up. At the time i did it, i didnt really consider the gate delays, and it was easy and didnt leave any gates floating. I would do it differently now of course using only 1 or 2 gates in series, or even none at all. The last SSTC of that "type" i built doesnt even use the buffers, i just tie the "antenna" right into the gate driver inputs. Seemed to work fine when there were only 2 UCC3732X drivers to supply signal to.
Steve Ward, Wed Dec 20 2006, 03:00AM
I think Steve just put six there because there are six in the chip.. that doesn't mean you have to use them all.
Its funny how often this comes up. At the time i did it, i didnt really consider the gate delays, and it was easy and didnt leave any gates floating. I would do it differently now of course using only 1 or 2 gates in series, or even none at all. The last SSTC of that "type" i built doesnt even use the buffers, i just tie the "antenna" right into the gate driver inputs. Seemed to work fine when there were only 2 UCC3732X drivers to supply signal to.
Print this page