6P43P Strange Plate Graphs

[TheGimp]

I"ve been playing with plots for various tubes since I got my "VACUUM TUBE ANALYZER" and find that the 6P43P has a strange plot compared to other similar tubes.

I've got screen captures of 6P14P, 6P15P, 6BQ5 and 6P43P in the word ".doc" file attachment.

Thinking it might be an oscillaton, I have installed a 0.001uf Cap from the grid to cathode, but it has little effect. Grid drive is via coax to shield it from surrounding circuitry.

I've attached the plot of the 6P43P-E and a zip of a word file with all plots.

[Yvesm]

Hey,

Your vacuum tube analyser names itself as a "Lampemètre" ! That sounds French

BTW, a common curiosity in all your captures is that the plate current MUST BE null when the plate voltage is.

What is this lampemètre ?

A simple test is to measure a simple resistor that must produce a tilted straight line crossing the origin: 0 Volts = 0 mA, and for, says a 10K one: 10ma at 100V, and so on as Mr Ohm told us

Yves.

[smoking-amp]

What voltage was the screen grid at? 120V?

[TheGimp]

Good point on the resistor calibration Yvesm, and yes the test set is french.

The Lampemètre came from :

MODULE LAMPEMETRE ANALYSEUR - VACUUM TUBE ANALYZER - eBay.ch (endet 28.04.11 14:33:56 MESZ)

It is a nice little test set for the price, but the plate current starting above 0 is an issue with it, along with the fact that the cathode current, instead of anode current, is sensed via a 10R resistor which produces regenerative feedback and includes the G2 current as well.

I'm trying to figure out a way around those limitations, but for now I'm using the Lampemètre for initial testing.

[artosalo]

It is a nice little test set for the price, but the plate current starting above 0 is an issue with it, along with the fact that the cathode current, instead of anode current, is sensed via a 10R resistor which produces regenerative feedback and includes the G2 current as well.

I'm trying to figure out a way around those limitations, but for now I'm using the Lampemètre for initial testing.

I discussed about this problem with the seller Alain, who is possibly the designer as well. The conclusion is that I try to replace the 10 ohms current sensing resistor with 1 ohms and add the gain of the DC-amplifier by 10. Let's see if it works.

[Yvesm]

Oh yes, the screen current increases dramatically at low anode voltage and, since you measure in the cathode, here is the problem !
<edit to answer artosalo's message>
You are measuring the sum of the plate and the screen current, what could be the value of the sense resistor.
</edit>
If possible, try to return the negative terminal of the screen PSU to the cathode rather than to the ground so the screen current will no longer flow thru the 10R resistor.

Yves.

[TheGimp]

I will need to run a second ps for the G2 to be able to isolate it's supply return. I'll see what I can do about that. With low gG2 current requirements, it may be possible to install a filament transformer backwards to and run a voltage doubler to generate enough voltage for a grid supply.

I think some error is introduced as well by the zero Cross Detect used to start sampling data since AC is used for the anode supply and it is sampled to find the start point.

Artosalo, let me know if that works and I'll modify mine as well.

[Yvesm]

Originally Posted by TheGimp

I will need to run a second ps for the G2 to be able to isolate it's supply return. I'll see what I can do about that. With low gG2 current requirements, it may be possible to install a filament transformer backwards to and run a voltage doubler to generate enough voltage for a grid supply.

Even the screen of a large (KT88) tube draws less than 10/20mA as long as the plate voltage is above 50 to 100 volts, below it sky rockets and may even blow the tube.
It's not a bad idea to use a weak PSU with a cap large enough to maintain voltage on peaks.

[TheGimp]

Originally Posted by smoking-amp

What voltage was the screen grid at? 120V?

G2 was 175V for the 6P43P-E and 250 for the others.

I used one leg of the B+ winding for the anode and the other for the G2 supply. Transformer is 250-0-250, and the G2 supply is regulated by a simple FET pass regulator with zener ref. It won't regulate above 350V because of sweep loading of the anode side.

[smoking-amp]

"G2 was 175V for the 6P43P-E"

The screen current usually picks up strongly with plate V somewhere below the g2 voltage. Seeing as the g2 current is going thru the cathode sense resistor too, this looks to be the cause of the kinks (although the reduction of the plate V momentarily is odd, unless the g2 loading is feeding back thru the plate V supply). As Yvesm suggested, returning the g2 supply to the cathode directly should eliminate the plate current sense problem. Likely will have to use a separate xfmr for the g2 supply anyway, so the plate V feedthru will be fixed as well.

How many bits of resolution does the Lampemètre A/D converter have? Or maybe using a sound card interface?