[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Tuning for Class-E Re: [TCML] Tuning SSTC to avoid losingoscillation when I try and draw discharge
The LT1222 from Linear Technology is very good for this sort of thing.
500MhZ GBW, dip package, +/- 12V supplies, very low noise, $6.50 from
DigiKey.
---Carl
-----Original Message-----
From: Andreas Hahn
Sent: Wednesday, September 05, 2012 3:33 PM
To: Andreas Hahn
Cc: tesla@xxxxxxxxxx
Subject: Re: Tuning for Class-E Re: [TCML] Tuning SSTC to avoid
losingoscillation when I try and draw discharge
I got the SSTC tuned to what is (hopefully) quite close to Class-E using a
current transformer on the primary lead.
(As per Terry Fritz's measurements --
http://deanostoybox.com/hot-streamer/TeslaCoils/MyPapers/modact/modact.html
-- the current and voltage track quite closely in the primary, and I
assume even more so in this case: it's a 1 turn SSTC primary.)
The coil ended up liking the 0.001uF capacitor with a one turn primary
best. Tweaking the vertical position of the primary changed things very
little.
I did note one thing that suggests further room for improvement: Though it
ends up hitting zero just as the MOSFET is turning on again, the
current decays much more quickly than in Steve Ward's simulated graphs
(http://www.stevehv.4hv.org/classEsstc.htm). Perhaps increasing the size
of the capacitor a bit would help, though going to .01uF killed the corona
entirely. I have a .0047uF cap around here somewhere...
Regarding the strategy of using a diode-protected opamp to shield the
'scope from any disastrous misfortunes in the primary driver, I learned
a few things. It turned out the MOSFET gate signal was rising faster than
the LM358 in
the diode-protected opamp scope probe buffer could follow; I swapped it
out for a spare OPA134 and got a
little more definition. I also learned the gate signal was clipping
courtesy the buffer's protection diodes, doubling the buffer voltage
helped a bit.
On Tue, 4 Sep 2012, Andreas Hahn wrote:
Update: Dropping the capacitor back down to 0.001uF got streamers flying
again, as did unwinding the primary until just two turns were left. I also
realized that what I was seeing on the scope screen was the 555-driven
30khz enabling/disabling of the MOSFET driver. The antenna was barely
picking up anything from the secondary! Extending the antennna to about
double its length fixed this.
However, class-E still appears out of reach. Here's how things look on the
scope right now:
- The MOSFET gate goes high (+3.5V*) in about 1uS. (*approximate as I have
a diode-protected opamp follower of unknown but probably low-ish input
impedance between probe and scope, and there's a slim chance I built a
voltage divider into it way back when I built it)
- Just before the MOSFET gate reaches the peak, the primary sees a .3uS
dI/dt spike (magnitude .6V, but this is horribly inaccurate due to the
homemade shielded loop being pretty crude), which reaches zero just as the
MOSFET gate reaches its peak.
- The MOSFET gate drops almost but not quite down to zero in approximately
0.6uS, then rises back up to 3.5V in about 1uS, and the process repeats.
On Tue, 4 Sep 2012, Andreas Hahn wrote:
I rebuilt the MicroSSTC's driver to Steve Ward's Class-E SSTC schematic.
Using a shielded magnetic loop (which gives output proportional to dI/dt)
scope probe on the primary lead suggests the primary is tuned way too
high... with the main spike lasting well under 1/10th of the MOSFET
turn-off / turn-on time. (the MOSFET gate goes HIGH for 10uS)
This seems odd, as I increased the capacitor value by a factor of 10 in
the rebuilding (to 0.01uF), so I figure it should be tuned too low
[capacitor taking longer to charge] if anything.
Tripling the number of turns on the primary and playing with coupling
didn't change much either.
Any thoughts?
_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla
_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla