Workbench tinkerings with PL509's.

Hi Dave,

1 Transformer, fixed and working

Well done, another transformer saved from the recycle bin. I am not sure how far the drill has gone into the laminations but it might affect core performance. I suspect for practical purposes it may not make a discernible difference. It will be interesting to compare with the other untouched Elizabethan transformer.

This transformer is not suitable to be driven by a PL509 due to high DC anode current, but biasing the grid back to -50 ~ -80 Volts may help but places operation in a very non linear region. The PL509 was not designed for audio amplification but as a switch to drive a high inductive load in TV timebase, i.e. either on or off, hence the linear transfer characteristic is less than ideal. To use this valve, it has to be operated at higher anode current with correctly designed air gapped transformer to avoid core saturation.

1) - If a valve is run at normal bias current, lets say half of it's rated anode current. Does this cause the cathode material to wear quicker than if it was biased at 10% on?
Valves are designed and specified that when operating with prescribed cathode current they will give so many thousands of hours reliable service. Operation at reduced ratings will more than likely increase service life due to reduction of heat generated failures. What is more likely to poison a cathode is operating with a large negative grid bias cutting anode current off. As long as there is some current the cathode will be OK.

2) - Is this normal operating procedure for an output valve? i.e. quiescent current approx. 1/2 of max permissible current.
No, the correct operating point is somewhere on or below the maximum anode dissipation line (orange plot in attached PL509 graph) and is directly related to anode supply voltage. Selecting the correct slope for a low distortion transfer characteristic takes a little bit of guess work. The load line as shown can swivel about the selected point 160V (anode) -30V (grid) and oops 175mA anode current which is way outside the current allowed in the Elizabethan transformer which has been designed to work with up to 50mA (green line)

3) - Does biasing a valve at a lower current reduce its gain?
Usually, but is varies greatly with valve type. Some are designed to exploit this characteristic for automatic gain control such as IF valves designated vari-mu.

4) - If gain is relatively unaffected in 3), is the resulting output waveform any less symmetrical at low output levels? (I.e. low volume, within the available 'headroom').
Usually once the operating supply voltage and maximum anode dissipation is determined from the graph, by inspection a load line can be drawn, shown in light blue which represents the AC load line presented by output transformer plus speaker impedance. The line can be swivelled to even out the distance between grid bias and anode current either side of chosen operating point.

Maximum 30W power dissipation shown in orange, Maximum cathode current shown in yellow, both must never be exceeded. Maximum estimated current allowed for Elizabethan transformer shown in green. A load line based on -30V grid and 160V anode shown in blue.

Of course the valve does not have to be operated at 30W so if it is required to operate at 25W, recalculate new maximum required power dissipation and re-trial load line. In practice it would be better to create a new graph expanding out the area of interest between -50 and -80 Volts

Graph kindly provided by Philips and jiggered by me...

Cheers
Rich

have you thought of screen grid modulation like this circuit ?
http://www.reocities.com/bobdanielak/technoteNo33.html
rob t

To use this valve, it has to be operated at higher anode current with correctly designed air gapped transformer to avoid core saturation.

Or Push Pull. Then the DC bias currents cancel and limit is copper and AC power. A mains transformer can be used.

The Anode load for AC on a transformer is the loudspeaker impedance times voltage ratio (turns ratio) squared. i.e 4,000 / 8 = 500, which is voltage ratio of about 22. Or about 220V to 10V transformer (about 115V to 6V).

You can use feedback from speaker connections to cathode and no cathode decoupling capacitor to improve linearity.

Thank you for all the replies.

Rich, cheers for taking the time to draw out the graph and answer my questions, it does make sense now.

There's some interesting things to think about there, I've got a few ideas floating about for hybrid and straight valve amplifiers using probably some unusual topologies.

I'm thinking first of a P-P amp based on a 509/519 with low biasing current. Its all in my head at the moment but I'll pen something down soon.

Next question - how far can I push the Cathode/heater insulation on one of these valves? would it withstand anywhere near the HT rail? (probably going to be 160v Max).

Thanks, Dave.

Next question - how far can I push the Cathode/heater insulation on one of these valves? would it withstand anywhere near the HT rail? (probably going to be 160v Max).

Dave, if you look at a typical circuit using a PL509 you will notice that it and the boost diode are usually at the top end of the heater chain which would be a very similar voltage difference, albeit with respect to earth, so the heater/cathode insulation has to be pretty good for that!

The data sheet - http://tubedata.tubes.se/sheets/010/p/PL509.pdf - give heater/cathode voltage as 250V max. However, in a note under 'Hum' a reference is made to "Vk/f = 220V RMS" so you can take the maximum rating as being 250VAC (354V Peak) ...

Transformer isolated (audio drive phase splitter and speaker transformers) 5+1 channel Push Pull audio Audio amp running direct off rectified mains?

Transformer isolated (audio drive phase splitter and speaker transformers) 5+1 channel Push Pull audio Audio amp running direct off rectified mains?

Nah, if I was doing a 5.1 system then I'd use loads of little SE amps.

I've got lots of ideas flying round, variable gain, low/variable bias, P-P, Totem-pole, maybe even some class-D theory thown in for good measure.

Maybe I should just make up a chassis and see where that takes me!

Dave.

There's a nice design using PL509s and a 100v line output transformer for class AB with distributed load. I'll see if I can find the circuit diagram.

I have thought of doing class D with 509s though, that could work even if it is a bit of work to achieve!