Mr.French's Dynamic Mutual Conductance Tester.

[Radio Wrangler]

34mA in the triode-connected list, 48mA in the pentode-connected list for the EL84, with appropriately different Vg1 bias values.... Enough to give a noticeable difference in Gm just from that, so the ratio of Ia to Ig2 is easily lost.

David

[ms660]

Originally Posted by Radio Wrangler

34mA in the triode-connected list, 48mA in the pentode-connected list for the EL84, with appropriately different Vg1 bias values.... Enough to give a noticeable difference in Gm just from that, so the ratio of Ia to Ig2 is easily lost.

I think there's a misunderstanding, do the calcs using the same Va-Vg voltages as the pentode, I make it not a lot of difference, compared to the KT88.

Lawrence.

[Radio Wrangler]

True, there is not a lot of difference. But we're going round the same circle again.

Strapping G2 to A to put a pentode or a tetrode into 'triode mode' for the purpose of measuring its Gm, as asked in post 33, does introduce a SMALL error in the result. Not strapping it avoids this error, so why not avoid the error and get a more accurate result.

Most people, if offered a choice between a 100 gramme and a 105 gramme bar of the same chocolate for the same price in a shop would not choose the disadvantaged deal. Even though the difference is small.

Errors are best avoided if they can be. If avoidance has a cost, then a judgement has to be made on whether it's worthwhile, but if it's easy then it's the proverbial no-brainer.

Small errors can be insidious, they can be cumulative, and systematic errors tend in this direction, so minimising errors is a good habit.

David