My first push-pull vacuum amplifier

[Herald1360]

CV428 is 5B/254M- a rather nice compact 807 equivalent.

[bikerhifinut]

A couple of things Miguel.

Lay your amp out with the valves vertical, its easier to work on and better for your valves.

Investigate alternative valves. it's surprising what equivalents are out there especially if you are sufficiently agnostic about valve bases.

Stick to a circuit that has been proven to work for a first attempt. Mullards amplifiers are a starting point, and the Leak circuits are well worth a look. All are known to work and by the EBay activity and my experience they are reliable and serviceable.

Good luck Miguel.

A.

[joebog1]

Those valves are free to Miguel.
It may take a couple of weeks to reach Cuba by mail.

Regards
Joe

[bikerhifinut]

I stand by my comments.

research the equivalents. That way you know what you have got and what will work in its place.

as far as you can use a circuit that someone else has perfected, then all you have to do is worry about layout.

lots of russian types about that equate to western valves. EG 6SJ7 = 6J8 for instance. thats 6"ZH"8 if you dont read russian. ( I do).

take your time. plan and then plan again.

The internet is a good source of info, and its free.

good luck.

[joebog1]

A couple of pics for Miguel.
I have also included the Australian documentation that is packed with every valve.



Regards
Joe

[joebog1]

CV428



Joe

[Diabolical Artificer]

Looks like your sorted for spares Miguel, postage from Oz to Cuba won't be cheap Joe.

I'll have to get my finger out, we were going to build an 807 amp at the same time, but Miguel is at the chassis stage while I'm still in the "research" stage.

Andy.

[Miguel Lopez]

Some strange and unknown phenomenon has happened while testing the output stage for the PP amplifier. I was testing a PL36 valve with 220V on anode, cathode to HT-, and grid to -30V (fixed bias). It was pentode connected, with the screen at 2/3HT+.

I was trying the suitable time constants for both the HT and bias circuits in order to avoid the situation of having HT without bias, resulting in a current surge. Well, that situation ocurred for a brief period (1 or 2 seconds), and the Selena and the TV set lost their signal while the valve was on the surge condition. Once the valve got off, they resumed their normal operation.

The radio was very near (60 cm from the valve at best), but the TV set was on other room at about 8 m from my bench, and behind a wall.

What phenomenon ocurred there for both devices lose the tunning?

[bikerhifinut]

HF Oscillation?

A few thousand ohms in series with grid one usually cures parasitic oscillations.

[G6Tanuki]

HF oscillation is indeed something you need to address.

Any amplifier is - to be honest - an oscillator that's just not yet managed to achieve its goal.

100-Ohm *wirewound* resistors in the anode-to-output-transformer, and 10KOhm resistors fitted directly between the drive-source and the output-valve grids are standard fitment when you're using HF/VHF-capable PA valves in tame audio applications.

A fistful of Ferrite-beads on the grid- and anode-leads is also a good way of stopping any audio-amp being distracted and diverting its energies into unrestrained 10-250MHz RF activity.

[Miguel Lopez]

I think that I did not expressed myself well.

There was no signal present, as I was just adjusting DC (HT and bias), trying to find the optimus value of capacitors and resistors on those circuits to avoid the anode current surge in the event of a temporary blackout, which are very frequent in Cuba.

So no grid stopper were used as there was no signal, just DC. Anode was connected directly to HT+ as there is no OPT yet.

In one of the test, the situation that I was trying to prevent, ocurred, and then the radio and the TV set lost their tunning. Why?

[G6Tanuki]

Your amplifier ia becoming an oscillator. And a radio-transmitter!

You need to fit "Parasitic suppressors".

[joebog1]

It doesnt matter if there is no signal Miguel!! When the "gain" of the circuit is increased above a certain level, when you are changing the bias, the circuit will oscillate.

What then happened is that the AGC of the radio ( Automatic Gain Control) and the AFC of the TV (Automatic Frequency Control) took over when a large signal was received from your "transmitter" and tried to correct the various circuits in the radio and TV to handle such a big signal. The EL36 or 6CM5 will not last long when it is an uncontrolled oscillator!!!

This phenomena was probably caused by your layout, and the output signal or wiring was too close to the grid input wiring.
It could also have been caused by insufficient decoupling within your power supply circuit, and the required energy to oscillate was transmitted to the screen grid.
I cannot say exactly, without looking at the layout of your breadboard, or the circuit as you made it.
Best regards
Joe

[Miguel Lopez]

OK, that will not be a problem as precisely I was testing the circuit to avoid such condition, and it only ocurred during the surge. I will avoid the surge by all means.

On the other hand, once I connect the OPT to the anode of the PL36, the inductance of the winding will block the surge for a short period of time too.

I've been testing a relay controlled by the bias voltage, so when there will be no bias, the relay will discharge the HT capacitors very quickly via a 330 ohms bleeder resistor.

But my original question remains. OK, it oscillates, but at what frequency?

The radio was tuned at 630kHz AM, and the TV set was tuned at channel 6, which is several MHz, and bot of them lost their signal. Can a single oscillation

BTW, the layout for the test circuit is really messy, but this is only my high tension breadboard (HTBB), and not the final layout. See the attached picture

[Herald1360]

If it goes off at lowish RF, there'll be harmonics from dc to light, more or less, with plenty of grunt and from the look of the lashup, lots of UHFish antennae to radiate nicely. It could also be squegging and delivering HF bursts of VHF oscillation.

How's your 'scope situation? A few turns of wire on the end of a probe makes a useful search coil for RF sniffing.

[Diabolical Artificer]

That's a good tip Chris, is the coil connected to the scopes input or connects by induction?

From what I've read amps can oscillate audibly or way up in the Mhz range with destruction possible if not tamed. With the 807 the datasheet recommends using a 100r on the top cap. But as the oscillation occured only during the overvoltage conditions this is all academic.

BTW Miguel, could it be that your mains is pretty mucky anyway? I doubt many devices in Cuba have proper mains protection to stop crap going onto the mains. The generating station is possibly put together with typical Cuban ingenuity and therefore not clean like ours is most of the time.

What I'm trying to say is could the oscillation have come off the mains ? That or the extra voltage caused your amp to have more gain and nudged it into self oscillation?

Andy.

[Radio Wrangler]

You use a small coil (say 30mm diameter two or three turns) connected to the end of a length of coax. The coax is connected to the scope.

You need an oscilloscope with enough bandwidth to be able to see if there is anything there at the sorts of frequencies at which there may be spurious oscillation. The coil on cable "sniffer" works even better on a spectrum analyser because of the greater frequency range and sensitivity compared to most oscilloscopes. For progressively higher frequencies you use smaller coils. Having a bit of plastic tube to stiffen the coil end of the co-axgives you a handle.

If you don't have a high frequency scope, then adding a series diode to the coil can allow the probe to show DC in the presence of a strong, constant, RF field,and to show the fluctuations if the field is not constant.

David

[Restoration73]

Note that as RF valves without metallised glass or metal bases are designed to be
mounted with the valve socket recessed, i.e. below the mounting panel. RF parasitics
are a well known problem, even experienced with MOSFET power amplifier.

[Miguel Lopez]

I think again that I have not explained myself well. Oscillation is not a problem, at least not until now. Normal operation it does not oscillate, but work fine. The interference occurs only when there is HT, but no bias, and the valve suffers a current surge for a very brief period of time.

Originally Posted by Andy

could the oscillation have come off the mains?

Negative. As I said, it occurs when I test the response of the circuit to a
blackout, so no mains was connected, only the energy stored in the capacitors.

I'm not interested to know the causes. I'm just interested to understand the phenomena.