Testing SM's Valve Heater Phase

[Peterae]

Time to move over to an appropriate thread for testing results.

The plan now with this is first to build a heater drive circuit, i thought about using all sorts of fet/transistor class A drivers for this but have ended up going the good old 2n3055 transistor route but will modify the below circuit, i will use 2 x 2n3055 in parallel and 1 2n3055 to drive the base, tap the sine from a small LV mains transformer and feed the phase board then the output sinewave to the base of the 2n3055 via a preset to adjust it's amplitude.

I have an old computer AM2 heatsink and fan assembly i will mount the 2n3055's to.

http://www.rcepi.com/make_an_adjustable_voltage_motor.htm

I still need to get hold of a Valve HT transformer, not an easy task for 500-0-500 which is why i was going to go lower 250-0-250 and use a different valve, if i wait for a reasonably priced 500-0-500 240mA transformer i could be waiting for the rest of my life, these things sell for over £100 in the UK  

[Peterae]

At last i am nearly there.
The above driver circuit worked but i could only get about 2V pk-pk across my heater.

Here's a summary of what i am doing.

Mains supply feeds a 6V transformer, this goes through a vol control to adjust amplitude, this then goes through a 0-360 Degree phase shifter, this then feeds a 15 Watt(upgraded from 10Watt) Class A amp which feeds one secondary winding of a 9V large toroidal transformer, the other 9V secondary winding is connected across the heater of the 5U4G.

Tonight i finally got to the point of powering up, and the 15Watt Class A easily gives me my heater amplitude and a lot more if needed, it glows  

I still have a couple of issues that should be no problem, i am using a 0-6 0-6 transformer, 1 winding provides my mains frequency and phased 50Hz reference, i used the other 0-6 winding to rectify and supply my phase adjust board, it looks like i am getting some distortion because of this, so i now need to move the phase shifter to a new supply and just use the 0-6 transformer for my reference.

I also have a 450-0-450 transformer hooked up via a variac - not had the bottle to fire this up yet.

Something strange happens also, with the variac turned to minimun which is giving me about 3-4 volts ac on the anodes, i connected 1 scope probe to my driver circuitry and then connected the second channel across the heater which is isolated from the driver circuit and suddenly i got lots of hash if the scope gnd is connected to only one side of the heater, i find this strange everything is isolated and connecting the scope ground to the heater should not matter anyway, so the ground probe is shorting gnd from the driver side to the heater side of the valve, surely it should not matter if 1 side of the heater is grounded?

Not got my good camera here but took a snap on the notebook.
I have also attached a shot of the sine across my heater, you can see some distortion on the top of the waveform.

Also i have uploaded the 2 pdf's which show how to build the Class A, i had to modify the design, it was originally a 1969 design, this pdf gives value changes to drive a 3Ohm load, so i added this mod to the 2000 version, i used 2 TIP3055 with 0.1 Ohm emitter resistors in place of each 2N3055 and a BD139-16 in place of the 2N1711

[Peterae]

First Test with HT to valve, i explain the setup.

[Peterae]

Varying Heater Phase versus HT Phase, Scoping 0V Ht Tap to each side of the Heater, i am just checking to see if there is any difference between each side of the heater during phase adjustment.

[Peterae]

OK so now i need a plan 

Quotes from SM

Well, I measured the output from the tube and the result is 500 volts DC at 250 mV.

I don't have this right now, i need to add a filter cap and load

The point I wish to make here is that also along with the 500 volt DC is, yes, you guessed it, the 5 volts three amp AC current!

Indeed there will be, so we need to forget about the high DC potential and zoom into the 5V AC across the heater filament by switching scope to AC and turning the V/Div down

They are both completely independent of each other except for some very interesting things I will mention to you some other time..

Hopefully we will be able to see this, we are concentrating on low mV noise and relationship between heater noise and plate noise, both will be present.

First of all, obviously you can have several different output components in the power output signal.
You can have DC and AC together without any problem.

Indeed we can but we must scope from the other side of the heater as opposed to the HV side to include the Heater AC waveform.

I was able to analyses everything coming out of this simple two transformer AC high voltage circuit.

I don't think i am looking for xrays as Lindsay says, the noise is very low voltage, i am more interested in harmonic currents.

In most power supplies there is lots of hash coming out and designers use a .05 or so to short out as much as possible before it gets to the smoothing capacitors.

And here is the confirmation we are looking at low volts hash and noise, we don't want a 0.05 across the smoothing cap otherwise we will remove what we are looking to analyse.

This hash comes from the mains supply and especially from the transformers themselves.

I can quiet believe this, i have previously seen mains noise get coupled to the secondary of a transformer and remain the same amplitude.

Then the smoothing capacitors take out the rest of the multiple frequency hash along with the gigantic 60 Hz ac left in the B+.

Hopefully he is referring to the mains supply components, 50 & 2ND component 100Hz, we are not interested in these so this is good.
We will be riding on our +B voltage so we need to get rid of this DC offset so we can zoom down into the low volts phased noise.Simply by switching to AC on the scope

I became interested in the interaction between the two AC transformers.

So shall i, by adjusting the phase between the heater and the HT supply

The interaction can be very reveling, trust me.

I do hope so 

[Grumpy]

Nice job Peter.   Hopefully, you'll encounter something interesting.

The pulsed charging effect is an easy one to try, and indicates that strange things are afoot.

I always thought that SM's coments that you reference above were analogies to the operation of the TPU rather than actual "how to" instructions.

We know that his output was DC with some "hash".  Two transformers out of phase is synonymous with his two pulses out of phase.

[Peterae]

Thanks G

I always thought that SM's comments that you reference above were analogies to the operation of the TPU rather than actual "how to" instructions.

I am hoping to find the magic ingredient for combining these kicks into larger kicks, i am working on the principle that varying the phase allows the conditions to be adjusted to allow the hash to combine and find what makes a large kick, i am convinced there's a recipe for making a really big kick.

[Peterae]

I have added 2 x 470uF 470V caps in series from 1 pin of the heater and from the 0V tap of the transformer.
I am scoping both sides of the heater and clearly show the heater 50Hz 5V on one side co existing on 500V rectified DC, both seem to coexist as SM stated.

No Load added yet, i will put a resistor chain across and aim at a mere 10Watt load, this equates to 25K resistor, 20mA

[Grumpy]

Thanks G
I am hoping to find the magic ingredient for combining these kicks into larger kicks, i am working on the principle that varying the phase allows the conditions to be adjusted to allow the hash to combine and find what makes a large kick, i am convinced there's a recipe for making a really big kick.

I always thought that the "big kick" was an analogy to the little kicks (^_^_^_^) inducing DC (-------------).

[Peterae]

Hi G
My theory just got torn to pieces, i have 2 problems 

1 my heater waveform has replicated 50Hz pretty well bar a small amount of distortion, but it's devoid of mains noise.

2 the mains noise on top of the rectified HT is so clean i cannot believe it, there's no mains noise that i can see at all and that's on the 100mV AC range of my scope, this scope is noisy anyway so if there is noise it's below the scope's noise, must bring home my other scope and try that.

Adjusting the phase seems to have no effect at all - scratch head 

I will say that there was a glitch at one point where the rectified HT voltage jumped a little, this may have been a large mains born spike, but it did not show as one just a burst of DC, in the end i had no lowish ohm resistors to build a 10 watt load, i settled for a 220K 2Watt which gave me about a 1Watt load, so if the mains born spike was responsible for a DC rise in my HT then it had power, it lifted my rail enough to notice an offset while on AC scope probe setting.

I need to get my head around all this.

If SM did indeed adjust the phase of his heater then he also would be devoid of mains noise, so this is not the answer.

I tell you what i think i have just worked out how to build the Hungarian device 

[Grumpy]

Here are the statements from SM:

I was working at a laboratory at the time with much more sophisticated equipment then is available to even most manufacturing companies.
I was able to analyses everything coming out of this simple two transformer AC high voltage circuit.
In most power supplies there is lots of hash coming out and designers use a .05 or so to short out as much as possible before it gets to the smoothing capacitors.
This hash comes from the mains supply and especially from the transformers themselves.
Then the smoothing capacitors take out the rest of the multiple frequency hash along with the gigantic 60 Hz ac left in the B+.
I became interested in the interaction between the two AC transformers.
The interaction can be very reveling, trust me.
Also, there is another interesting analogy.
We seem to overlook so many things in our society.
They are right in our faces but we just look around them without interest at all.
When I began to study the effects of multiple frequencies combined together I found out that when you deliberately strive to create the worst case scenario of frequencies you start to get some very measurable kicks.
In themselves they are not much.
But if you make enough of them fast sendoff, you get a collectible power spike that is more then the power available to begin with.
The destructive heating caused by the eddy currents become the problem we face when we make a really large powerful coil.
Now you understand more about the heating problem and why using a fan does not work.

[stprue]

I have added 2 x 470uF 470V caps in series from 1 pin of the heater and from the 0V tap of the transformer.
I am scoping both sides of the heater and clearly show the heater 50Hz 5V on one side co existing on 500V rectified DC, both seem to coexist as SM stated.

No Load added yet, i will put a resistor chain across and aim at a mere 10Watt load, this equates to 25K resistor, 20mA

Very interesting results your getting there Peter.  Keep up the good work.

[Peterae]

Thx St

In most power supplies there is lots of hash coming out and designers use a .05 or so to short out as much as possible before it gets to the smoothing capacitors.

Most power supplies show hash, but not all? i was not seeing any hash last night, maybe a quiet time to test.

When I began to study the effects of multiple frequencies combined together

Isn't this valve just a mixer, what happens if i feed 35kHz into one anode and 70kHz into the other anode.

The plate has a high voltage potential with lots of useable power available.
You cant get to it or use it for anything without applying a heating voltage to the cathode or what is the cathode potential of the tube.

I am puzzled by the cathode potential, am i meant to be using a specific side of the heater as my +B volts or Cathode or can you use either side, the valve manual does not say, if either side can be used how is that possible.

500 v-0-500 v 300 mV plate transformer run it through a full wave silicon circuit then run it through a 5U4 electron tube rectifier.

This looks to me that he is using rectified AC and feeding the resulting DC to the plate in order to make life easier in testing the heater current.

I think i understand what he requires of me now, he wants me to measure the current in the heater circuit and measure the current in the cathode circuit.

[BEP]

What voltage is measured between a plate and the heater when only heater power is applied?

I think this measured value is a measurement of the 'kicks'.

[Peterae]

What voltage is measured between a plate and the heater when only heater power is applied?

I think this measured value is a measurement of the 'kicks'.

I can certainly disconnect the Anode's and see, when i get home 

[Peterae]

Here's the result of scoping anodes disconnected.

Pic Heater1anode is scoping across the heater(Yellow Chan) and green channel is on 1 anode which is disconnected from anything else

Pic2Anodes is scoping with scope earth connected to one side of heater and both channels are on 1 each of the anodes.

Very Interesting, not sure what to make of it, adjusting the heater phase alters the shape of the flat area on the yellow trace of 2 anodes picture.

I realized i am using a surge protector main extension for all my equipment, although i have now see mains noise and it seems to be additive to the rectified 500V DC so a sharp spike will appear positive of the DC 500V, i have tried to capture this but it only happens once in a blue moon, i had the camera recording for half an hour and only could see a very little on, before i switched the camera to record i had seen 3 or 4 big ones in about half hour 

[Grumpy]

like switching the hv over the 500v bias

[Peterae]

OK for this test i have placed a UF4007 diode in each feed of the anodes and then scope each side of the heater with respect to the 0V tap on the HT transformer.

The scope snap shot is what i have on the anodes with diodes in line.

Don't forget i have a 220K resistor which is drawing about 1 Watt from the 2 series 470uF 450V caps

This clearly shows hash over 470V, surely this must be energetic particles causing this, i need to try this without diodes.

[Peterae]

Does anyone know the difference between the 5U4G & 5U4GB i seem to be having trouble finding out.

I am also unable to find out how the 5U4 works, are the electrons directly emitted from the heater or does the heater directly heat a plate which then emits.

Should i be looking at this full wave rectifier valve as 2 separate diode half valves, each side of the heater has it's own plate and cathode?

[Grumpy]

I can't find much on the 5U4

[Peterae]

Same Here G although that pdf is a help, it uses pin 8 for both diode cathodes  

[AllPhase]

The G  means Glass.
The GT means Glass Tube.

Both electrically the same.

[gyula]

Does anyone know the difference between the 5U4G & 5U4GB i seem to be having trouble finding out.

I am also unable to find out how the 5U4 works, are the electrons directly emitted from the heater or does the heater directly heat a plate which then emits.

Should i be looking at this full wave rectifier valve as 2 separate diode half valves, each side of the heater has it's own plate and cathode?

Hi Peter,

From the link below it turns out the GB suffix designates 275mA anode current versus G suffix's 225mA current (while the GA suffix means 250mA):

https://www.tubeworld.com/5u4g.htm 

Regarding your heater question, it is surely a direct heating type, the filament emiits the electrons, no any Cathode material involved which would emit for the heat from the filament.  How do I know this?  There is no any separate cathode pin...  like in case for instance the EZ80 type, here the indirect heating is used.

Here is a data sheet for the 5U4GB: http://www.r-type.org/pdfs/5u4gb.pdf

rgds,  Gyula

[ion]

Peterae:

How does filament phase affect the mV of hash you are seeing with the pre anode diodes?

I'm guessing no effect.

Also SM suggested the use of a bridge rectifier in front of the anodes, in which case you could try two ways:

1) Both anodes tied together to plus of BR. Neg of BR to gnd.

2) Plus of BR on one anode, neg of BR to other anode.

[Peterae]

Hi ION
I did try this morning the mv hash experiment with no diodes, and i do still get the effect, so it's not the diodes.

You are also correct in saying the heater phase being adjusted has no effect on the hash.

Personally i feel this noise is from within the tube, at such high HT levels i am exceeding the specs of the 450Va limit, it's strange SM mentions 500V as this is over 50V above the specified max working voltage for a 5U4G, he is also exceeding the anode cirrent by 25mA.

I will try the bridge, i am going to be interested to see what happens with a FWB to one anode especially how the second anode will behave.