want to expand tetrode class A specs to AB2, have curves

[Opcom]

The 12A6 is a nice tube taking 250V on the plate and using 12V for the bias and for the heater. Who could ask for more?

I have the datasheet showing class A audio operation and the curves are included. I can see how to draw a line on the curve that matches the datasheet class A conditions well. I think it was properly done.

OK well the tube can take a little grid current so I want to investigate AB1 and AB2. I don't know how to use the curves for that and how to determine power input and output so my work is wrong there.

Does anyone know or is there a program that lets one enter some of the data and get an approximation?

[WD5JKO]

Patrick,

   I am not sure if you refer to single ended or push pull operation. For single ended, class A must be used to conduct 360 degrees. With triodes, the power can be boosted about 50% by increasing the grid drive such that you draw grid current on peaks. This is called class A2 whereas A1 has less grid swing, and no grid current.

  Driving the control grid(s) to +10v on the 12A6 should not be a problem with cathode follower drive such as from a 12AU7. The guy at Tubelab has a way to do this with FET's, and calls it "PowerDrive":

http://www.tubelab.com/PDcookbook.htm

  Whether SE or PP, the RL on the tube(s) depends on the placement of the loadline on the plate voltage versus plate current graph. The slope of the line is important as is the intersection points on the X & Y axis. So with your load line, the intersection points are about 330V/.15A so the RL would be 2200 ohms. That sounds kind of low to me for a 12A6.

   You might look over the tutorial from Patrick Turner of Turner Audio:

Single Ended Beam Power Tube:
http://www.turneraudio.com.au/loadmatching1sebeamtetrodes.htm

Push Pull Beam Power Tube:
http://www.turneraudio.com.au/loadmatch4-pp-beamtetrodes.htm

Jim
WD5JKO

[WA1GFZ]

I suspect the gain will dive when you reduce G2 voltage. You could always build a little test fixture with a load resistor in series with the plate and meter the voltages and find the AB2 load line.

[Opcom]

Thanks, yes it will be push pull for audio. I'm not interested much in single ended class A2, although I can see it might be useful.

I recently got a little box of these tubes new in the metal version, some dated NOV 1944.

The red part of the line in the curve is my attempt to draw the class A1 single ended audio amp ratings from the sheet but it's likely wrong.

The blue is a guess at extending it, considering more drive to the grids and less than 360 degrees conduction for AB2 push pull.

So only the red can be considered.

The sheet calls for 7500 Ohms. The bias point is 12.5V, the only spec the DS reveals.

330V/.15A is 2200 Ohms, but I'm not sure how that works out to the datasheet recommendation of 7500 Ohms unless the red line is at the wrong angle.

I'll read those references.

As for G2 volts, the max rating is 250 and that's what should be used.

[w4bfs]

hi Patrick .... you probably already know this .... the 12a6 is a kissing cousin to a 12v6 or 6v6 ... RCA Radiotron manual should have your answers ...73

[WBear2GCR]

As far as I can see, at the bias point for AB, about -20v on the grid, you'll see about 20ma resting current, that being somewhere around the 300v mark... I'd have to stop and look at a text to remember what they want for the load rating of the xfmr, but my seat of the pants says that typically a 5/6k transformer it typical for this sort of tube... afaik there is no reason you can not drive the tube into A2 or AB2 without any modifications.

Watch the screen current!

Take a look at that 1570B schematic, that's my favorite way to do the deed.

I should *know* this stuff cold, but alas my brain is like cheezo-del-swiss.

[KM1H]

I believe the Heath A7 is running them AB1

[Opcom]

This all gave me plenty to think on.

The A4, a PP class AB model, makes 6 watts into a ? Ohm p-pload.
? bias, maybe 15-20V, via 470 Ohm cathode resistor,
?V plate and screen, beam tetrode configuration.


The Mapletree S5, a PP class A model, makes 5 watts into a 10000 Ohm p-p load.
-12.8V bias via 200 Ohm cathode resistor,
262V plate and screen, beam tetrode configuration.

Either way, it looks like the best answer is in those schematics.

Let me ask this:

If a single tube in class A makes 3.4W into 7500 ohms, then would a push pull pair make 6.8W into a 30K plate to plate load?
That would be 7500 ohms for each tube. As we know, if the turns are doubles the Z quadruples.
 
Some how this seems weird compared to the Mapletree schematic.

Look at the Mapletree schematic. This is a class a amp, almost taken right from the book. It has a 10K CT transformer. That would appear to be 2500 Ohms plate load for each tube.

Look at the 6V6 datasheet.
The 6V6 data shows a 2500 Ohm load gives 1.5 Watts. Two would be 3W with huge distortion and this isn't right for the Mapletree.

What gives?

[KM1H]

Here is the A7-E Heath which expands a bit on the WW2 12V tubes.

[Opcom]

390V in ultralinear? Yikes. Someone must have had a truckload of those tubes..

[KM1H]

High voltage with low current doesnt hurt as long as the tube doesnt have voltage breakdown. At 7W its not being worked very hard.

Not much different than those on here running 813's at 3500V, 4D32's at 1000V, etc. except they are also running them much harder.

[W2VW]

.

Not much different than those on here running 813's at 3500V, 4D32's at 1000V, etc. except they are also running them much harder.

813's what?

Nobody who actually gets on the air using AM runs an 813 at 3500 volts for more than a minute.

That is strictly low buck, plenty of spares low average power mode activity.

[KM1H]

I believe a few on here have claimed 3200V + on AM a few years ago. On SSB/CW the graphite plate versions dont even get beyond a very dull red; the sheet metal ones get holes as easy as 811's in Ameritrons.

[W3RSW]

Concerning the 390 plate voltage, if you have 60 ma quiescent (class A) across the 470 ohm cathode(s) resistor then voltage across the tube is "only" 362, so not too bad.     In reality if this amp is operating class AB, quiescent current at such high plate voltages may be closer to 100 ma.   I wonder what kind of service life Heath found from its customers at the time?

The a7 looks like a pretty nice little amp.  I see Heath beefed up the first filter cap to 40uf, right up there with the table radios of the era.  

Looks like RIAA eq. in preamp circuit, double slope with step - low pass filter circuit.  ...staying up with the times.  This amp is sufficient to drive a nice 10 in. Bass reflex speaker cab. in smallish living room or den of same era.  

[KM1H]

Id be more prone to playing with that 3.3K resistor to get ~300V for the 12A6's and then 250 and 180V. I dont like changing tubes often.

Ive used that bass boost circuit on my Scott 800B in the dining room, it made a huge improvement over the original as well as loafing 6550's in place of the 6L6G's.

Carl

[Opcom]

An excel file helped to show me where the bias point might be, but despite the values in it, there's only a chance of it being right. I think it's close enough to show a 17.5V bias voltage for the plate and screen volts in the chart itself.

The chart can't apply to the A7 because:
1.) Only assumed 11% of the cathode current goes to the screen grid as suggested by zero signal class A conditions from the data sheet.
2.)There is no curve for plate current vs. G2 voltage - so the 250V curve was used.

The curves didn't show anything but 250V screen operation, so the plate current is proportionally lower than it might otherwise be in the A7 if Eg2 = Eb.