Matching EL34s in triode vs pentode modes

[RobLM]

Hi everyone, been lurking for a while, first post!

I can see, from EL34 datasheet graphs, the performance differences in pentode vs triode modes. I am wondering, in terms of matching tubes on a curve tracer, can one match a pair in, say, triode mode, and assume that pair will be a similarly good match in pentode mode? Or could the differences between modes be such that well-matched pair in one mode is not a well-matched pair in the other?

The question came up as I get ready to put together a tube adapter for a 576 curve tracer (thereby connecting the plate to the "collector" port, the grid to the "base" port, and the cathode to the "emitter" port, with an external supply for the heaters). I am wondering if I can match el34 type tubes well enough by tying the screen grid to the plate, or if I'd need to have an external power supply for the screen and a switch to toggle triode and pentode modes. I started looking at the prices of external supplies that can provide a high screen voltage, and am hoping that simply tying the screen to the plate will do the trick for identifying tubes that would be well-matched with each other in either triode or pentode mode.

[loudthud]

The biggest problem I see is the step generator can only go to about -40V. Ten two Volt steps plus 20V of offset dialed in. I read about a guy in Yugoslavia that build an X10 amplifier to drive grids. It needs to be DC coupled with something like a -100V rail.

Pentode mode presents it's own problems. Most curves you see specify a DC Voltage for the screens. When the Plate Voltage collapses to zero each half cycle, the screen current can go to more than 100mA. You need a good adjustable current limit to avoid burning up the screens. You should seriously consider a series resistor on the screens. Occasionally you see UL mode curves, in that case the screen supply will be rectified AC of some percentage of the Plate supply.

As to the question about how close EL34's might match between triode and pentode mode, that's anybody's guess. You'll need a variac and a true RMS meter on the heater supply.

FYI, I own two 576's and a 570.

[RobLM]

Thank you for the info! I think i’ll start by getting set up to test in triode mode first, make sure I’ve got that down, and come back to the idea of setting up for pentode mode/comparing the two down the road a bit.

[RobLM]

Would summing, with either an opamp or maybe a passive summing circuit, the 567's step generator output with an external power supply capable of down to something like -100v, and then returning the summed output to the 567's base connector, be one way to do it?

[loudthud]

On the front of the 576, the part that sticks out is called the "Front Porch". The right portion is removable, the left portion contains the Collector supply Variac and circuit breaker. On the right portion there are three banana jacks down near the bottom, the Step Generator Output, an External Input and a Ground. You would use the step generator output, and then rout the signal back into the external input and set the switch on the front porch to External Input.

Behind the front porch there are four connectors that allow access to various internal circuits within the 576. The standard plug-in only uses one of these connectors. One of the other connectors has 120VAC (using the primary of the 576's power trans as an auto-transformer) so it would be possible to build a small power supply to power an opamp.

An opamp circuit is really the only option because you need gain to get 5 and 10V steps.

[RobLM]

Thanks again for the tips, loudthud. I was a little too intimidated to try and figure out how pull power from inside the curve tracer itself, but picked up a Heathkit Ps-4, which gave me sources for screen voltage, V- for the opamp, and filament voltage as well, and used a Danelectro variable 9v guitar pedal supply to give the opamp a V+ of 5v. I put together an opamp circuit from this paper by Dennis Tillman that describes how to adapt an old tube tester to work with a curve tracer (I just lifted the opamp section and took his suggestion to try an OPA454). http://www.ke5fx.com/A_VTCT_Adapter_for ... s_W7PF.pdf.

I was able to confirm with a triode that my opamp circuit was amplifying the step generator's voltage x10, getting identical curves with step generator amplitude=-2v/external amplifier=x1, and step generator amplitude=-.2v/external amplifier x10, etc.

Then I killed the opamp, I think by giving it too much V- and putting it outside of its maximum voltage swing . New OPA454s on the way, and then I'll get it up and running again and test the external amp circuit with an EL34, supplying the tube with screen voltage from the Heathkit. Appreciate the help!