Kapanadze Cousin - DALLY FREE ENERGY

[d3x0r]

Re: the tl494 push-pull....


shouldn't be a 50% duty cycle, would think more like 60% maybe less... because, once you drop one side, the voltage/field will drop on its own for quite a ways, really want to catch that as it nears zero on its own to push it further down; same on the up swing, where a dead time between pushing and pulling will limit the current draw...


Well; maybe since it's so slow, it won't have a nice sine wave anyway....
or... maybe it's actually at 1/4 the rate but 25% cycle so it looks like 4khz?

[verpies]

Do you have more info on your tube amp? I'm a hobbyist builder of tube amps and always wondered how well mains transformers would work as output transformers but have not had time to try it. would be very interested in your test results but do not want to derail this thread. Is your work posted anywhere?

It is not posted anywhere.
The amp is built around the 6N13S (6H13C) double triodes rated at 500V but running at 380V.  It also uses the ECC83S double triodes and the EM84 (6DH7) "magic eyes" for audio level output indicators.  The biggest central transformer is for the power supply and the thick winding you see on top of it provides the low voltage for the tube heaters.  The other two smaller toroidal transformers are step-down units for the final output stages. Their cores are tapewound (they came from a halogen-lighting shelf in a supermarket )  and their pri/sec windings are litzed and multilayered and sandwiched.  The amp does not use any semiconductors except for silicon rectifiers in the power supply section.   
The photo you've seen shows a 90kHz sinewave test signal at input and an unattenuated undistorted sinewave at output. (at 130kHz it attenuated by 6dB) with speakers still connected.

This amp is relevant to this thread because it illustrates that tapewound-cored transformers can transfer 90kHz however that might be possible only because the primary and secondary windings are still able to couple due to their sandwiching even when their tapewound cores expel the magnetic flux due to the diamagnetic action of the eddy currents in their conductive cores.

[itsu]

@ all,

thanks for your comments and advices, allthough some of it add's to my confusion as they are contradicting and make my head spin, like:


"I suspect there is sine wave on push-pull output"     
No sine wave from this TL494, right?

"see if you got 2 primaries right. They should go into opposite winding directions".
I thought most agreed to have the primaries as going into the same direction, like CW.

"shouldn't be a 50% duty cycle, would think more like 60% maybe less."
Need to know the best duty cycle of the TL494 to produce clean alternating square wave on the secondary
to a) "drive the L1 coil" and b) "to rectify it to produce 200V DC for the nano-pulser".

"Dally inverter as I see it is primarily a low frequency (around 50Hz) 150V PSU., thus the iron cored transformer."
I understand from the squealing sound the x-former runs at 4.6Khs, also confirmed by the sound analysis done
Also one of the diagrams (post #12 on page 1) states about this x-former:
"ТрансÑ,,ормаÑ,ор на Ñ,,ерриÑ,е оÑ, Ð'п ATX" which translate via google into: Ferrite transformer from the PSU ATX

"the mains transformer you are using (in reverse) is probably not a ferrite core"
Then: "But don't despair I have seen these regular silicon steel-tape transformers operate at 90kHz"

(By the way verpies, great looking classic amplifier, and i love the FG in the background).



So i need to digest all this and decide which route to take.
Presently i am planning to continue with the 4.6Khz approach, so need a suitable ferrite transformer first.


Regards Itsu

[verpies]

"I suspect there is sine wave on push-pull output"     
No sine wave from this TL494, right?

Right, TL494 outputs only rectangular pulses.  However by the time those pulses get to the gate of MOSFETs they may become distorted, although I don't think they will distort into a sine wave.

"see if you got 2 primaries right. They should go into opposite winding directions".
I thought most agreed to have the primaries as going into the same direction, like CW.

That is so relative! because there are 2 ways to serially connect these two primaries, 2 ways to wind them and 2 ways of supplying them with current.  Please use my method of "phasing" these primaries as outlined in pt.2 in Reply #57 - it cuts through all the confusion..

"shouldn't be a 50% duty cycle, would think more like 60% maybe less."
Need to know the best duty cycle of the TL494 to produce clean alternating square wave on the secondary
to a) "drive the L1 coil" and b) "to rectify it to produce 200V DC for the nano-pulser".

Well, 49% is the maximum duty cycle the TL494 can output.  The duty cycle can only be decreased from 49% using the error amplifiers (pins 1,2,16,15) and the dead time input (pin 4).

"Dally inverter as I see it is primarily a low frequency (around 50Hz) 150V PSU., thus the iron cored transformer."
I understand from the squealing sound the x-former runs at 4.6Khs, also confirmed by the sound analysis done
Also one of the diagrams (post #12 on page 1) states about this x-former:
"ТрансÑ,,ормаÑ,ор на Ñ,,ерриÑ,е оÑ, Ð'п ATX" which translate via google into: Ferrite transformer from the PSU ATX

Yes, the sound analysis of the Dally device suggests low kHz switching frequency.  It does not matter if the core is made out of a ferrite or steel tape as long as it can be magnetized and demagnetized quickly enough.
Yes, ATX computer power supplies routinely use ferrite cores. (EI or toroidal)

"the mains transformer you are using (in reverse) is probably not a ferrite core"
Then: "But don't despair I have seen these regular silicon steel-tape transformers operate at 90kHz"

Yes, ferrite has superior high frequency characteristics but apparently steel-tape transformers are not much worse than ferrite at tens of kHz. At least the one I used in that tube amplifier was able to transfer 90kHz. 
One advantage of steel over ferrite is that steel has much higher saturation point than ferrite, thus steel can handle more current in the primary before it saturates.

(By the way verpies, great looking classic amplifier, and i love the FG in the background).

The background used to have my girl in it.
The Arbitrary Function Generator has 2 independent channels that can be phase-synchronized and max frequency of 160MHz (5ns rectangular pulse rise time and 10ns minimum rectangular pulse width).  It can be bought here.
I did not use my digital oscilloscope because the amplifier's output voltage has dangerous spikes in it when the speakers resonate.

So i need to digest all this and decide which route to take.
Presently i am planning to continue with the 4.6Khz approach, so need a suitable ferrite transformer first.

As illustrated by my amp, that steel tapewound core of your present transformer might not be so bad.  Just "phase it correctly" as outlined in pt.2 in my Reply #57

...and for God's sake, don't try to transform DC as explained in pt.1 of Reply #57 !!!

[Hoppy]

@ all,

thanks for your comments and advices, allthough some of it add's to my confusion as they are contradicting and make my head spin, like:


"I suspect there is sine wave on push-pull output"     
No sine wave from this TL494, right?

"see if you got 2 primaries right. They should go into opposite winding directions".
I thought most agreed to have the primaries as going into the same direction, like CW.

"shouldn't be a 50% duty cycle, would think more like 60% maybe less."
Need to know the best duty cycle of the TL494 to produce clean alternating square wave on the secondary
to a) "drive the L1 coil" and b) "to rectify it to produce 200V DC for the nano-pulser".

"Dally inverter as I see it is primarily a low frequency (around 50Hz) 150V PSU., thus the iron cored transformer."
I understand from the squealing sound the x-former runs at 4.6Khs, also confirmed by the sound analysis done
Also one of the diagrams (post #12 on page 1) states about this x-former:
"ТрансÑ,,ормаÑ,ор на Ñ,,ерриÑ,е оÑ, Ð'п ATX" which translate via google into: Ferrite transformer from the PSU ATX

"the mains transformer you are using (in reverse) is probably not a ferrite core"
Then: "But don't despair I have seen these regular silicon steel-tape transformers operate at 90kHz"

(By the way verpies, great looking classic amplifier, and i love the FG in the background).



So i need to digest all this and decide which route to take.
Presently i am planning to continue with the 4.6Khz approach, so need a suitable ferrite transformer first.


Regards Itsu

Itsu,

I've just finished building and testing the nano VCO and get 4.6KHz at around centre pot adjustment with 100nF timing caps. It looks as if this is the source of the 4.6KHz. The Dally schematic shows 1nF  caps and these produce around 500KHz! I have seen a circuit element of just the nano VCO and pulser for the Dally version and this shows 100nF timing caps. Now onto the nano pulser.

Regards
Hoppy