Wireless Mains Adapter-Frequency etc.?

[boiss]

OK Guys,
You've given me some good ideas to try out. If I can reduce the power in and get 30-40% efficiency I think it will fit my requirements. I'll try these suggestions over the next day or two and get back.
Thanks
Mike

[boiss]

Ok, Started with adding small ferrite beads to gates, very little difference, changed for large tubular ferrite cores, definite improvement.added 100 then changed to 220 Ohm resistors in gate circuit, ringing eliminated but there are transitions still. Added ferrite beads to the drain leads, instant increase in current and hot beads!! added toroid to the supply, (3 turns on small toroid), current dropped from 2A to 0.4A, much more efficient, adding load - current increased to 1.4A with 6W DC into 50 Ohm, Fets very cool. In this connection there was an additional 0.33uF cap after the toroid. adding 1000uF after the toroid returned the current to 2A as before. This little toroid with the 3 turns has a major effect.
I have not tried inserting Rs in the source ccts yet.
These are the waveforms. There is a difference in the Gate waveforms?
Mike

[Skywave]

Hi Mike,
Earlier, I commented on the shape of the gate waveform and suggested parasitic oscillation. Upon reflection, I suspect that I was wrong about that. I now suspect that the cause of that distorted waveform is that the Zener diodes do not have an h.f. performance that is required in this cct.: the Zeners are causing that distorted gate waveform. So, try this: using two resistors in series, arrange for a bias of about 10 v. at each gate.

Al. / June 14 //

[boiss]

Al, I agree it's a bit strange, I can't see a reason for the ringing at about 5V on the gates, The transitions are quite evident and are the cause of the ringing. I will try what you suggest as follows;- If I take the 60V PP (30V pos)signal on the drains then seeing as I have 8K2 gate to deck then 18K in series with the 47nF and no zener ??
Mike

[Skywave]

Hi Mike: no.
I am thinking of a 1.5 kΩ resistor from +12 v supply rail to the gate and the existing 8.2 kΩ resistor from the gate to 0v. By my calculation, that puts a +ve bias on the gate of about +10 v.: same as with the Zener.
It's a easy change to implement and answers the Q. about the Zener's performance.

Al. / June 14 //

[boiss]

Hi Al,
I thought the Zener was there to limit the + gate voltage to 10V because the drain waveform is a 60 V peak+ half sine wave (more or less), in fact the gate volts go between zero and +10V.
Mike

[Skywave]

Ah! I see what you mean: a 60-v. amplitude pulse that needs to be 'cut down' to 10 v: hence the Zener. Perhaps the Zener is being driven too hard then?
Anyway, I regret to have to tell you that I've run out of ideas on this now - except to suggest that you re-configure the method into a separate oscillator followed by a separate push-pull power amplifier - as I mentioned earlier. At this point in the proceedings, that is the route I would now take.
With your existing arrangement, I can make no more (hopefully) helpful suggestions; sorry. Perhaps it's someone else's turn to 'pick up the gauntlet'.

Al. / June 14 //

[boiss]

Thanks Al,
The Zener is a 1 watt type and it gets hot, I've already destroyed one, I always watch the the colour of the diodes, the pink colour goes grey. I'll try reducing the feedback, I'm wondering if the strange points on the curve are due to some secondary resonance with the feedback caps, and I thought this was simple!!
I will try the Source resistance mods tomorrow and if there is no improvement in the wave shape I'll try inductive feedback coupling.
If I use an external oscillator I will still run into an efficiency problem due to non resonance of the PA coil unless I can tie the two together somehow. The whole point of the exercise was to arrive at a simple low cost solution. If I wait a year or two there will be microwave solutions to this but by then I will have spent 30 quid on batteries and postage.
I Like my Hacker and I'm not going to drill any holes in it
Thanks
Mike

[boiss]

Deeeeerrr, To coin a famous phrase.... "They don't like it up em Mr Mannering, Sir"
I've blown up my 3 legged fuses all S/C d-s, g-d-s. Disconnected the Zeners after cooking one (It became a flashing Red Led), Changed the gate resistors and Voila cooked fets. It will have to go on the back boiler now until I can get some more.
Mike

[boiss]

Ok Latest Status.
With the last remaining Fet I made a few mods to make a single ended Hartley.
this time the gate has 2 zeners back to back. It's a bit more efficient, with no load the power input is 2.5W. With a 50 ohm load and FW rect, the O/P is around 16V ie about 5W, Power I/P with load about 14W. The Fet runs cold and the coil is only warm. There is a snag though........................It won't start on its own !!!!!!!!!!!!!!!!!!!!!!!#$%^&
to start I have to touch another cap across the cap in the gate cct. I have tried putting other Resistors between the + supply line and gate (100K...to...15K) but it still won't start.
This was starting to look good !!! ANY IDEAS?
Pics are Drain waveform, and cct
Mike