Ed,
Have you checked the plate swing over the full load range (intensity) of the CRT circuit?
Yes. From zero to about mid intensity, the plate (and grid) swing increases slightly, while the screen voltage goes from about 85 to 105V (far too high). At that point the regulation circuit goes out of range, and if I increase the intensity further, the swing at plate and grid starts decreasing, and so does the HV. In this situation, the 150k resistor limits screen current to about 2mA, screen voltage to about 105V, and the tube is in current limit. If I had more primary turns, I could get more power from the same current. That's why I was considering winding #3 for a higher primary turns number, relative to secondary turns.
Also, you may want to consider and experiment with the value of the primary tank capacitor. Since the HV winding capacitance now seems to dominate the resonance, maybe less on the primary would help.
Or rather, MORE would help! That's because I'm using relatively thick wire for the primary, and thin for the secondary, relative to the capacitive currents they are carrying now. If I reduce the frequency by drastically increasing the capacitor, the primary will take some more current, which it can handle easily at low loss, while the secondaries will take less current, reducing their presently too high loss.
But the transformer will sing!
reducing the air gap would be a better way for reducing the frequency and improving effiency.
Right now there are two (or three) tanks that are coupled to an unknown extent, each with its own resonant frequency. If you have equipment to measure inductance at low frequency (far below resonance), you can check the individual winding and leakage inductances, which would help to understand what's going on.
Yes, I could do that.
You can experiment by driving the plate (with power off of course) at low voltage with a function generator and some series R, while measuring various points with a scope, to see if there are any strange or multiple resonances, and the effect of the primary cap value. There shouldn't be enough power available to light anything up, so it should remain linear.
Yes. I can try that.
Stefan,
You can adjust the gap easily by hand with some fine grit sandpaper.
I never thought it was possible to shape ferrite using sandpaper! It's pretty hard stuff. The only grinding I have done so far on ferrite cores was by means of a Dremel tool, using the green stones, intended for hard material. But the Dremel of course isn't suited for getting a flat surface.
Will try it. And if it doesn't work, I have another option: I have several cores of the same size. Each consists of one half with a shorter center leg, and one with a full length center leg. So I can simply use two "full" halves, to get zero air gap in the ferrite, and then gap it with paper.
It is easier to put spacers in between (paper etc.) and live with half
of the gap on the outer legs, even if that is not ideal.
It works well enough. Should there be too much flux leaking out, a shorted turn done with foil around the whole transformer (all three legs) cures that pretty well.
I keep an assortment of pre-ground cores in certain steps, and often
grind the gap smaller for correct inductance.
I have so far only adjusted my gaps by putting paper and plastic sheets between the halves.
Thanks for the idea. I will see if I can really sand down these cores, with a reasonable investment of time!
Chuck,
If you have the 5642 filament voltage (1.3V/T) right on the first two
transformers, and they are running around 25KHz, keep your primary as you
have it.
No! That's a mistake. The filament voltage in this circuit depends on the _secondary_ turns number, not the primary! This is because the regulator circuit will keep the voltage on the HV cathode secondary constant, while the primary voltage is free to change as required. So the volt/turn for the transformer is finally set by the turns number of the cathode secondary, and the setting of the regulator circuit. And not by the primary turns!
Cutting down the air gap is very easy. The Chinese have blessed us with very
cheap diamond dust knife sharpening "stones" They are about 5cm x 15cm in size,
very, very flat, and will cut ferrite like it is chalk. Use water to keep the
stone from clogging up, and use a thumb-forefinger grip on the sides of the core
to avoid curving its already flat faces, and have at it.
I will see if I can get such a stone. It should be better than sandpaper on a flat surface. The only sharpening stone I have is roughly of that size, with medium cut on one side and fine on the other, Brazilian-made, but it's not diamond. Probably softer than the ferrite!
Use the 0.001uf resonating capacitor to get in the ball park with the primary inductance.
Yes, that's right, I can reverse-engineer the turns number by calculating the inductance needed to resonate at 25kHz with the .001uF. But there is a factor of uncertainty introduced by the stray capacitance, specially that of the secondaries.
John,
If you need any other measurements for comparison let me know.
Thanks. If I need something, I will ask!
With that in mind, I reviewed all the usual suspects for manuals Bama,
Ebaman, Elektrotanya and W140, which look to me to be all the same version
with the schematic for the transformer circuit labeled "CRT CIRCUIT" "700
thru 760". I could not get the Manoman copy of the manual.
OK, so the transformer design sheets of those should give a good reference.
Having turned towards computers early in my electronic path during the
period of linear supplies I missed the use of inductors in my work.
A high percentage of current electronic engineers and technicians don't fully understand electromagnetism. This has come to such an extent that a well known professional textbook for electronic enginners says "The design of electromagnetic parts is highly complex, and is best left to specialized consulting companies." A clear indication that the authors of that book don't have a clue how transformers work! ;-)
So I got somewhat side tracked from my transformer project by looking at
inductors nature using a scope and signal generator/PA amp.
I think I was able to measure inductance and series resistance fairly well
on various inductors
But I did not get to the point to be able to measure loss in the transformer
due to the distributed capacitance or the limits of the core.
I need to publish my inductor meter on my web site. It's a little gadget based on a 555 timer, a high current MOSFET and a capacitor bank, that allows to test power inductors and transformers for inductance, saturation, and resistance. I built that thing several years ago, but still haven't put it on my site! It handles up to 70 amperes of pulse current.
It was not clear to me what the measurements could be made on an arcing
transformer that would be valid, or a brown epoxy lossy transformer.
Arcing transformers and lossy transformers are different animals. A lossy transformer will have lots of resistance appearing in parallel with its inductance. The inductance will probably be lower than it should, too. Instead and arcing transformer will measure perfectly normal at low voltage, but will go crazy at a voltage at which arcing starts.
This defines how they will behave in a circuit like the Tek's: If a transformer is very lossy, the oscillator won't even start. Or if it starts, the transformer will get hot fast, and the oscillator tube's screen voltage will be higher than normal, and maybe it will fail to reach the nominal output voltage. Instead an arcing transformer will start normally, but when reaching a certain voltage it will arc, and the oscillation waveform will show sudden break-ins, tending to limit the oscillation amplitude.
Of course one and the same transformer might both be lossy _and_ arc! Because the arcing degrades the insulation and makes it partially conductive.
Three thoughts, keep focused, do you have room for more insulation between
layers that might reduce capacitance?
Yes. First, I'm gaining some room by reducing the total turns number. Then, I can put less insulation over the primary - what I have there now is intended to survive the full 2kV plus headroom, which isn't necessary if secondaries are wound starting at the cold end. And then, if I wind the two secondaries together, either bifiliar or side to side, I can save half the insulation. The whole thing combined should allow me to use three layers of masking tape, instead of one, between wire layers. That, combined with the narrower with per winding and more layers, should allow me to make secondary windings with aroundf 8-10 times less capacitance than what I have now. That's pretty promising.
And third you might try cutting C705
to 75% or 50% of it's value to see what the difference in frequency will do?
That would probably be worse. The frequency would go slightly up, and the loss in the secondaries would also go up.
David,
I may have this wrong, but I think 2.6V/t on the center leg equates to 1.3V/t on the outer legs, because each outer leg conducts half of the center-leg flux.
It's right, but with a grain of salt: Unless a flux equalizing winding is used, the flux will very easily split up unevenly between the two side legs. So, you might get more than 1.3V/t one one side leg, and less on the other, depending on how well the core halves fit on each side, and how much load there is in those windings on the side legs.
If a flux equalizing winding is used, which I see specified for most Tek transformers, the flux is forced to divide very evenly between the two side legs.
The flux equalizing winding is simply one turn on each side leg (installed anywhere on the side loops, usually close to the center leg), connected in parallel with proper phasing to cancel any current if both side legs induce the same voltage.
Albert,
At first sight (at those 6AQ5 characteristic curves) it seems possible to increase the primary AC and obtain lower anode voltages during conduction.
What I wanted to do is the opposite!
BTW grid 1 operates at considerable positive voltage during part of the oscillator cycle.
No, not much! Due to the 4k7 resistor, and the fact that a positive grid conducts, it gets only a few volts above zero. Instead the negative excursion is large, to about -100 to -150V.
Most curve are for zero or negative grid 1 voltage.
The RCA datasheet gives curves for up to +5 or +10V on the grid. That's plenty!
If you want a safer way to test your transformer you can feed the screen with an external variable DC power supply. Then you don't depend on the feedback regulation and you can work at considerable smaller AC voltages.
Yes, that's a good idea.
Well, friends, I need to do a big job over the next days, which is unrelated to scopes... I will be back at the scope and transformer tuesday or wednesday!
Manfred
========================
Visit my hobby homepage!
========================
