The length of the wire used for the anode choke should not be longer than 1/4 of the wavelength the oscillator is working on. You can used a small neon to 'see' the high voltage nodes and low voltage but high current nodes along the choke if it is too long. At high power, those nodes will be hot as well. Also I agree with another TH-camr's comment that maybe a core shouldn't be used in this application cause it will induce a lot of parasetic capacitance.
It's an honor to have you here, thank you. I didnt think the choke would be the problem, last time i made this i used a plastic piece to wind stuff on, i suppose i will make a different one witb a large amount of turns. should i try removing those turns in thr air first?
@@janno288 Yeah, I think those wires in the air won't do too much help and ypu can just remove them. By the way, what valve did you use in your previous successfully operating pushpull oscillator? In my opinion 20pF feedback is way too large.
@@uecmitsuimarinedieselengin873 the value of the choke? I didnt calculate it, i had a piece of plastic and i would a random amount of turnd on there, maybe 100 turms or so, i still have it, i can put it into one of those testers and see what comes out. that oscillator worked at least and was able to light up fluorescents and neons, but only well above 300V supply input if i can recall. How should i change feedback capaciors with frequency and griddrive? I plan to make a big push pull tesla coil with two gu-81 thbes which hasnt been done before. the frequency is 160kHz. for the GU-81 the feedback resistance is equivalent to a 150W 230V tungsten lamp. I was thinking of converting that resistance to the impedence of the capacior at that frequency. I based 20pF on @PosthumanKindergarten push pull oscillator, he had a 5pf and the impedence was 300ohms so i also picked my impedence to be 300ohms at my frequency. the capacitor in push pull oscillators is still a mystery to me
@@janno288 Sorry I didn't make it clear, I mean what value of the feedback capacitos did you use for your previous oscillator? In your video that oscillator works good, judged by the not too hard ignition of that CCFL in that kind of a long coil.
@@uecmitsuimarinedieselengin873 I used 80pF, those small things are adjustable capacitors from 20pf to 80pF and seem to br high performant stuff rated to high voltage, i found them when i bought a lot of Polypropyle capaciors. The previous one had a 20kOhm gridleak and 80pF feedback capacior. i used the same capaciors and resistors and added another 20kOhm in parallel, so the gridleak is like 10kOhm now It was a CFL not a CCFL
i was looking for some explanatory comment but my comment is first and non explanatory :) - i dont recall what the push-pull oscillator like - will wiki it but some scheme would be usefull - 27 MHz sounds like a VGA pulsing (25,1) - maybe its just some intrinsic self oscillation - oh! 27 MHz is the 10/11 meters wave of CB radio - maybe you pick up some transmission, but usually there are many. - on some other videos people radioamators point out that LED lights on the street can emit 27MHz but in narrow channels - eg between 27,51 and 27,55 MHz
- No VGA Monitor powered up in the house - I am in the basement, radio signals dont go in or out - No LED lights in the basement, nor is any LED light on inside or outside 27MHz makes sense, the self resonant frequency of that foil is around 35MHz and together with the g3 to anode capacitance of 50pF per the the frequency drops to 27MHz which makes sense, its just a weird coincidence that it lines up with the frequency of another one i made that was able to excite neons and fluorescent bulbs
@@janno288 OK - there is a schematic at 2:49 - so the equivalent (mystery) inductance is 279 nH i have refreshed my memory about Collpits type of amplifier ( "coil - pits/pitfalls 😢 ") i knew it as the differential type of amplifier but i dont know anything about tubes (unfortunatelly :) i can point that 38/27.5 is roughly a Square_root(2) :) but it is pointless coincidence for me. It is an amlpifier so it could amplify anything that it finds useful as far as i understand The shape looks like a saw tooth - it reminds me of a crossover distortion that looks like a saw in the 0V region The self resonance makes the coil behave with inductor characteristics below the resonant frequency and as a capacitor above it Differential voltage amplifiers have some special phase shifting properties - they shift the phase by PI/4 in the working region and inverse the phase (PI/2 shift) in the above region i believe it is some distant CB radio interference :)
@@TymexComputing the sawtooth happens due to the scope probe being away and due to the large metal plate, when i being the scope close its a near perfect sine wave, either the probe or the scope, i have no way of telling, thr scope loop is not a perfect loop either. but as the weirdness happens you can see that it becomes more sinudoidal due to the higher power, overcomming some looses and radiating more for the scope to pick it up, then weirdness happens.
you sound desperate for such an irrelevant project :D anyways yeah, the choke can be the problem, i wound my chokes like small tesla coil secondaries but with slotted windings, no ferrites no matter how high is their maximum working freq. and i dunno where your ferrite is coming from, if it's a random one it hardly goes so high in f. hence you remain with just those turns of copper, that is like nothing in inductance, i see very few turns also, such choke can be basically a short the way you made it... i calculate the inductance of the choke in two steps, i chose a Z value for it that is sensibly higher than the impedance of the tubes and coil, like 3X more (or even more), then calculate the inductance form the F and Z so... like 2kohms of the tubes times 3 = 6k .... 6000/(6.28 * 27'000'000) = 35 microhenries .... that (according to online calculators) is like a pipe of 3cm of diam. with 50 turns of wire (organized in like 3 groups) for a length of like 5cm, you can add more turns to be even better, DC resistance is irrelevant that's what i suggest you, apart this it may be a busted tube or the layout that is "secretly" not good for some reason, this happens at hi f.
I didnt think the choke could be that critical, but yes i see, it does make sense. I should wind one instead on something non inductive, you're right. Thanks a lot for commenting, i will also try making that Tesla coil on it, because i havent seen something like this myself and i love the idea of it.
Hah yoh wouldnt belive it! I found out the reason with someone who was uneducated on tubes, he checked the datasheet and found the example schematic that had g2 decoupling, but had a 100ohm resistor between the grid 2 and the bypass. I out such resistors in, they smoked, i out higher power ones in, flawless operation together with the new RF choke, i can light uo CFLs but strangely only part of it lights up. Remmeber when you had g2 issues, this couldve been your fix
@@janno288 these things happen, honestly i saw those resistors there and also G1 on audio amps, but original schematics never had those, only amateur ones, so i guessed that the cap should go tight on G2 and G1 can be without, never had problems myself also i said it could be the choke cos maybe 27MHz isn't too high for a ferrite but who knows, i'd not trust random ones
@@PosthumanKindergarten if you look at some tube datasheets they sometimes show a resistor in series with g2 from the bypass capacior to the tubes g2. (PL519 datasheet)
@@janno288yeah, i see... first time i see that datasheet, i don't have any 519 or 6P45S, didn't need to see that... honestly it's incredible that you solved such a big problem with this simple thing, MAYBE this is mostly for the PL519.... in my case i doubt this would solve a thing, i strongly believe that putting an SCR on a pentode VTTC is prone to fuck ups, or it's just the GU50 or i dunno, maybe some day i'll work on it again
The length of the wire used for the anode choke should not be longer than 1/4 of the wavelength the oscillator is working on. You can used a small neon to 'see' the high voltage nodes and low voltage but high current nodes along the choke if it is too long. At high power, those nodes will be hot as well. Also I agree with another TH-camr's comment that maybe a core shouldn't be used in this application cause it will induce a lot of parasetic capacitance.
It's an honor to have you here, thank you.
I didnt think the choke would be the problem, last time i made this i used a plastic piece to wind stuff on, i suppose i will make a different one witb a large amount of turns.
should i try removing those turns in thr air first?
@@janno288 Yeah, I think those wires in the air won't do too much help and ypu can just remove them. By the way, what valve did you use in your previous successfully operating pushpull oscillator? In my opinion 20pF feedback is way too large.
@@uecmitsuimarinedieselengin873 the value of the choke?
I didnt calculate it, i had a piece of plastic and i would a random amount of turnd on there, maybe 100 turms or so, i still have it, i can put it into one of those testers and see what comes out.
that oscillator worked at least and was able to light up fluorescents and neons, but only well above 300V supply input if i can recall.
How should i change feedback capaciors with frequency and griddrive?
I plan to make a big push pull tesla coil with two gu-81 thbes which hasnt been done before.
the frequency is 160kHz. for the GU-81 the feedback resistance is equivalent to a 150W 230V tungsten lamp. I was thinking of converting that resistance to the impedence of the capacior at that frequency.
I based 20pF on @PosthumanKindergarten push pull oscillator, he had a 5pf and the impedence was 300ohms so i also picked my impedence to be 300ohms at my frequency.
the capacitor in push pull oscillators is still a mystery to me
@@janno288 Sorry I didn't make it clear, I mean what value of the feedback capacitos did you use for your previous oscillator? In your video that oscillator works good, judged by the not too hard ignition of that CCFL in that kind of a long coil.
@@uecmitsuimarinedieselengin873 I used 80pF, those small things are adjustable capacitors from 20pf to 80pF and seem to br high performant stuff rated to high voltage, i found them when i bought a lot of Polypropyle capaciors.
The previous one had a 20kOhm gridleak and 80pF feedback capacior.
i used the same capaciors and resistors and added another 20kOhm in parallel, so the gridleak is like 10kOhm now
It was a CFL not a CCFL
i was looking for some explanatory comment but my comment is first and non explanatory :)
- i dont recall what the push-pull oscillator like - will wiki it but some scheme would be usefull
- 27 MHz sounds like a VGA pulsing (25,1) - maybe its just some intrinsic self oscillation
- oh! 27 MHz is the 10/11 meters wave of CB radio - maybe you pick up some transmission, but usually there are many.
- on some other videos people radioamators point out that LED lights on the street can emit 27MHz but in narrow channels - eg between 27,51 and 27,55 MHz
- No VGA Monitor powered up in the house
- I am in the basement, radio signals dont go in or out
- No LED lights in the basement, nor is any LED light on inside or outside
27MHz makes sense, the self resonant frequency of that foil is around 35MHz and together with the g3 to anode capacitance of 50pF per the the frequency drops to 27MHz which makes sense, its just a weird coincidence that it lines up with the frequency of another one i made that was able to excite neons and fluorescent bulbs
@@janno288 OK - there is a schematic at 2:49 - so the equivalent (mystery) inductance is 279 nH
i have refreshed my memory about Collpits type of amplifier ( "coil - pits/pitfalls 😢 ") i knew it as the differential type of amplifier but i dont know anything about tubes (unfortunatelly :)
i can point that 38/27.5 is roughly a Square_root(2) :) but it is pointless coincidence for me.
It is an amlpifier so it could amplify anything that it finds useful as far as i understand
The shape looks like a saw tooth - it reminds me of a crossover distortion that looks like a saw in the 0V region
The self resonance makes the coil behave with inductor characteristics below the resonant frequency and as a capacitor above it
Differential voltage amplifiers have some special phase shifting properties - they shift the phase by PI/4 in the working region and inverse the phase (PI/2 shift) in the above region
i believe it is some distant CB radio interference :)
@@TymexComputing the sawtooth happens due to the scope probe being away and due to the large metal plate, when i being the scope close its a near perfect sine wave, either the probe or the scope, i have no way of telling, thr scope loop is not a perfect loop either.
but as the weirdness happens you can see that it becomes more sinudoidal due to the higher power, overcomming some looses and radiating more for the scope to pick it up, then weirdness happens.
do you still play beta :D
No really....
you?
@@janno288nah just saw your video pop up 😂
you sound desperate for such an irrelevant project :D
anyways yeah, the choke can be the problem, i wound my chokes like small tesla coil secondaries but with slotted windings, no ferrites no matter how high is their maximum working freq. and i dunno where your ferrite is coming from, if it's a random one it hardly goes so high in f. hence you remain with just those turns of copper, that is like nothing in inductance, i see very few turns also, such choke can be basically a short the way you made it... i calculate the inductance of the choke in two steps, i chose a Z value for it that is sensibly higher than the impedance of the tubes and coil, like 3X more (or even more), then calculate the inductance form the F and Z so... like 2kohms of the tubes times 3 = 6k .... 6000/(6.28 * 27'000'000) = 35 microhenries .... that (according to online calculators) is like a pipe of 3cm of diam. with 50 turns of wire (organized in like 3 groups) for a length of like 5cm, you can add more turns to be even better, DC resistance is irrelevant
that's what i suggest you, apart this it may be a busted tube or the layout that is "secretly" not good for some reason, this happens at hi f.
I didnt think the choke could be that critical, but yes i see, it does make sense.
I should wind one instead on something non inductive, you're right.
Thanks a lot for commenting, i will also try making that Tesla coil on it, because i havent seen something like this myself and i love the idea of it.
Hah yoh wouldnt belive it!
I found out the reason with someone who was uneducated on tubes, he checked the datasheet and found the example schematic that had g2 decoupling, but had a 100ohm resistor between the grid 2 and the bypass.
I out such resistors in, they smoked, i out higher power ones in, flawless operation together with the new RF choke, i can light uo CFLs but strangely only part of it lights up.
Remmeber when you had g2 issues, this couldve been your fix
@@janno288 these things happen, honestly i saw those resistors there and also G1 on audio amps, but original schematics never had those, only amateur ones, so i guessed that the cap should go tight on G2 and G1 can be without, never had problems myself also
i said it could be the choke cos maybe 27MHz isn't too high for a ferrite but who knows, i'd not trust random ones
@@PosthumanKindergarten if you look at some tube datasheets they sometimes show a resistor in series with g2 from the bypass capacior to the tubes g2.
(PL519 datasheet)
@@janno288yeah, i see... first time i see that datasheet, i don't have any 519 or 6P45S, didn't need to see that... honestly it's incredible that you solved such a big problem with this simple thing, MAYBE this is mostly for the PL519.... in my case i doubt this would solve a thing, i strongly believe that putting an SCR on a pentode VTTC is prone to fuck ups, or it's just the GU50 or i dunno, maybe some day i'll work on it again