Bias Oscillator black hole

[kalee20]

Thinking further, the 2V p-p seems very low, but it depends where you are measuring it (if it's on TR2 secondary then it could be credible for a low-impedance erase head).

I'd expect well over 100V p-p on the 6V6 anode, for instance.

I'd expect maybe a fifth of this on the 6V6 grid.

It would be good to know how much current the 6V6 is taking when oscillating - I'd be surprised if it was more than 10mA. If you can measure that - great!

And - now it's working - you can have some fun with C4 value. Try tacking 2200pF in parallel, then 4700pF, and find at what value the oscillation breaks up. It's worth noting that C4 and the 5nF capacitor on the grid form a capacitive divider, so larger values of C4 both increase the amplitude of oscillation on the grid (which makes squegging more likely) and increases the grid time constant with the 250k (which also makes squegging more likely). You've made a huge step forwards, so now it's time to try to get things as the original designer intended (which we'll have to guess at).

And - most importantly - you're having fun and learning!

Originally Posted by Clay1905

Umm,? I don't know what you are alluding to with "a quick snip". I can export drawings to pdf, or jpg, but the jpegs are pretty awful. I get that you are trying to steer me towards a smaller file to upload, but what is a quick snip?

I'd do it by viewing on-screen (the pdf or whatever!), hit PrintScreen - this copies the screen view to clipboard. Then open Paint, paste the clipboard in, trim it as necessary, save as a .png file AND as a .jpg - and try uploading to the Forum.

The jpg will probably be smaller size - and it will probably do - if not, use the .png file.

[Clay1905]

Here's a picture of the oscillator output.

[Radio Wrangler]

When posting a picture of a waveform from a scope, you need to state what the voltage and time scales you used are, otherwise people looking at it have no idea of the voltage or frequency. On most modern scopes, voltage and time scaling are set by switches and have a 1-2-5-10 sequence, but these values are only accurate when the variable control knobs (There should you want to fit a waveform to some specific size) at in their 'cal' position. Witht them out of cal, the real scaling will be larger than the switches say, but you won't know by how much.

David

[kalee20]

It's quite a clean sinewave! Which is encouraging.

Reference David's comments, yes it would be helpful to know the scaling (though you have already told us it's about 50kHz), and also exactly where you have picked it off from (TR2 secondary to the erase head maybe, which is the 'output.'
Looking forward to more info!

[Clay1905]

Ok, that was a bit daft wasn’t it?
Here are three pictures. First is of the output, at the head terminals. Second is what’s on the anode, and the third is the grid.
What thinkest thou?

[kalee20]

I'm thinking, are you using a x10 probe? (I hope so - it's good practice!).

If you're not, then the valve is only oscillating weakly, which would be rather surprising. But how many volts is the HT feed to it?

The waveforms do look good and clean, nice and sinuous.

[Clay1905]

Yes, it's X 10 probe.
B+ when running is at about 270VDC.
The waveform is nice. I'm pleasantly surprised by how little residual hum there is on the B+. and no cross-talk between the oscillator and the amplifier.

[kalee20]

OK, so we have the anode swinging about 100V each side of 270V, which is a healthy amount. Good to know - thanks!

Can you measure the current taken by the oscillator?

[Clay1905]

The oscillator is passing 49.1mA at the anode.
Still no action at the head. I had a look and found the erase coil was not grounded well. Now it's too well grounded. Seems I've made a short. Drat! I'll revisit the thing later on.

[kalee20]

49mA at 270V seems a lot (13.2W! Something is getting hot!). I'd have expected 10 - 20 mA.

Erase heads can run warm - but this seems excessive. However, right now you're still fault-finding - it might all 'come right' together.

[Clay1905]

Today I finished off rebuilding the erase/bias coils. The erase coil had no insulation under it and the wire's lacquer had chaffed on the burrs along the edges of the laminates making a short to ground. Now I think the head unit would still work, even after a nasty experience in a war zone.
However, the evil spirit that dwells inside the recorder has decided that the line from the voice coil to the grid of V2 shall become open. Either a broken wire (I doubt) or something's gone awry in the rotary switch.
These are not related, or shouldn't be, but a number of posts back I mentioned that R14 was getting mighty hot. I doesn't feed the oscillator, but for the moment it's about all I know of.
I'm thinking I need to give this machine a rest for a few days. Owing to a range of circumstances I'm prone to getting unhealthily obsessed with projects. As I'm feeling a bit bummed out with the open circuit happeninging like this, I'll take the hint.

[Clay1905]

Hi there,
I have been thinking about Kalee's comment about very high current. Looking at the schematic, I see no impedance to DC through the oscillator. Also, there's not a lot to actually get hot. Maybe the primary of Tr2?
In any case, the only way current seems to be able to be limited is the control grid. Does this sound reasonable? I will measure the current again to be sure of the number. Also will measure g1 voltage compared to the cathode. The characteristic graph for 6V6's shows that the valve should be able to run with a relatively negative voltage on the control grid.
Am I barking up the wrong tree?

[Clay1905]

Here is the transfer chart for a 6V6. Am I interpreting it correctly?

I re-measured the anode current, and the number is correct. I also measured the P.D. from g1 to ground. This is -5VDC.
Looking at the chart for anode current, I see that the 250V curve intersects the -5V line at just below 90mA. If this is right, Then would it not be also right to think that the oscillator current is actually lower than should be expected?

[kalee20]

You're not wrong, and yes about 90mA for -5V bias with 250V on screen and anode is about right for a 6V6.

But a strongly-oscillating circuit should develop quite a lot of grid bias in this sort of circuit, biasing the valve well back and keeping current low. The peak currents may be several tens of milliamps (which is why a power valve is used) but the average should be much smaller.

I'm only guessing here, i admit - but it may need revisiting after you have sorted the coil and head anomaly.

[Clay1905]

Ah! Good to have my interpretation of the chart confirmed. I will need to do some reading, or thinking or somthing. I don't see why the valve should develop "a lot of grid bias" yet.
The physical fault to do with the head is a bit mysterious. There appears to be something loose or undone. Sometimes when the machine is upside down, there's continuity to the preamp grid. Other times not. Same when right side up. Wiggling and jiggling things only produces a bit of a crackle, sometimes. But nothing definitive. Curses!
I haven't found any fatigue breaks in wires. About the only place I still need to look at is the rotary switch.
A mystery:
When any device is being assembled, some things have to be the first that are fitted. Why is it that whatever the device or part, the earlier it went in, and so the harder to access, the more likely it will fail?

[Clay1905]

Yahoo! Finally a recording is made. After more than enough tedium and frustration the record function is alive.
Sound quality is remarkable. It's got a clarity, warmth and depth of tone that makes it a close rival to an Edison wax cylinder player.
A big part of the poor sound is that the erase coil doesn't seem to be having much effect.

In reality, the erase and bias coils are the very unequal windings of a horseshoe magnet. The voice coil is on the bias leg, furthest from the recording wire.
Does the actual polarity of these two coils matter? Owing to a bad blind spot of mine, I have wound these two coils in reverse to what they were. They are still wound so that the erase and bias zones on the head face are opposite poles.As these coils present relatively high alternating fields, I would have thought that the coil handedness was not an issue.

Otherwise, darned if I know.

[Clay1905]

I had a look at what's going on with the bias oscillator. Here are the results.

In order to establish if the erase and bias coils are working, I passed 1.5VDC through them. The erase coil reversed a compass needle. The bias coil caused compass deflection, but not nearly so strongly. Not surprising considering the size difference between the coils.
So far, so good.
Connecting a 'scope to the bias terminals I get a flat line. Disconnecting the 'hot' wire from the bias coil allows the oscillator to run. As I'm out of my depth here, I can't say why. I think it has to do with the erase and bias coils not presenting enough impedance. I took an awful lot of care to be sure there are no shorts in the head unit. As the bias/ erase coils are looking like close to a short to the valve, it's little wonder the current is so high.
What is it I don't know or understand here?