Using 1r to measure plate current

[loverocker]

So, it's a fairly common mod to add small 1r resistors to cathodes in fixed-bias amps to measure cathode current (via the voltage dropped across them).

This got me thinking - why not add 1r resistors to the plates to work out the plate-only current? I don't think I've see it done, and I'm scratching my head trying to work out what the reason is?

I've got a layout that has the room for them... what do you guys think?

[Paul P]

Originally Posted by loverocker

This got me thinking - why not add 1r resistors to the plates to work out the plate-only current? I don't think I've see it done, and I'm scratching my head trying to work out what the reason is?

I've just been through this idea myself. See the thread :

Bias measurement

While putting the resistors at the plates might be a bit more accurate
the test equipment used to measure the voltage drop will be at 400+
volts whereas measuring a resistor at the cathode things are close to
zero volts. Long wires with high voltages on them are to be avoided
both for safety and interference reasons.

(By stretching my imagination I can see the case where, while measuring
the voltage at the plate you knock your multimeter off the bench. The
probe disconnects and the end of the end of the wire is at 400 volts,
until you decided to remove the other end from the resistor...)

I like the idea of putting a probe between the tube and the socket. I
plan to make myself a biasing tool with a couple of meters on it. You
can buy them ready made.

[loverocker]

Hi Paul - thanks for the link. And also for being a proper gent and not telling me to do a Search (like I should have!).

In this particular amp, the 1r plate resistors wouldn't create a big diversion in OT lead dress (the relevant turrets are on the way between the OT wire grommet and the output valve sockets). Safety's a potential issue, but this would improve safety as it'd avoid the need to probe for plate voltage at the valve sockets - I can do that at these turrets on the board (easier access and more space).

Think I'll give it a go.

[Wild Bill]

Originally Posted by loverocker

Hi Paul - thanks for the link. And also for being a proper gent and not telling me to do a Search (like I should have!).

In this particular amp, the 1r plate resistors wouldn't create a big diversion in OT lead dress (the relevant turrets are on the way between the OT wire grommet and the output valve sockets). Safety's a potential issue, but this would improve safety as it'd avoid the need to probe for plate voltage at the valve sockets - I can do that at these turrets on the board (easier access and more space).

Think I'll give it a go.

If you're careful it should work fine. The advantage with a cathode resistor is that the voltage on the probes will be millivolts. If something slips and shorts to ground - who cares?

With a resistor in the plate lead you'll have virtually the entire B+ on the probes with respect to ground. Easy to get exciting...

---Wild Bill

[Enzo]

And the ONLY reason I can see to do it that way is to eliminate the contribution of screen current, and that is so easily determined that it hardly seems worth the effort. To me. If you want to make measurements, you can check the screen current across the screen resistor in a few seconds and subtract it, or you can check a few and get an idea of what the typical screen current might be in general and just make a rule of thumb. It is only a couple milliamps.

ANd keep in mind also that if the bias is off a couple ma, it won't much matter.

[Don Symes]

One word Loverocker - clipleads.

If you do this thing, use ONLY meter leads with clips on the ends. 'gators are a bit big and a huge opportunity for 'excitement'. I'm thinking of those j-lead hooked clip things Pomona makes - http://www.pomonaelectronics.com/cgi...RABCLIP&page=3

Hook them on, then power up. Power down, then remove.

[loverocker]

Originally Posted by Enzo

And the ONLY reason I can see to do it that way is to eliminate the contribution of screen current, and that is so easily determined that it hardly seems worth the effort. To me. If you want to make measurements, you can check the screen current across the screen resistor in a few seconds and subtract it, or you can check a few and get an idea of what the typical screen current might be in general and just make a rule of thumb. It is only a couple milliamps.

ANd keep in mind also that if the bias is off a couple ma, it won't much matter.

Yes, it removes screen current from the figures. But that's sort of the point: the 1r resistors on the cathode are a make-do compromise. For the same amount of effort as adding 1r resistors at the cathodes (and some layout planning from the get-go), 1r resistors on the plate supply give a true figure for plate current.

By the way, in all of my builds, I measure screen current directly and calculate and check screen dissipation as well as plate dissipation. According to KOC and others, exceeding screen dissipation is as bad as exceeding plate dissipation, and I can't know if I don't measure.

As for probing around high voltages and the risk of slipping - it would be true if this wasn't a safer approach to measuring ALL the high voltages. Because we have to measure plate voltage and ordinarily that's only available at the actual valve socket - where there's far more risk of slipping and shorting the probe to ground (or another socket pin).

My style of turret board (see below) is far easier to probe safely than a wired-up valve socket. There's always a big chunk of phenolic between probe and chassis, whether I'm measuring a few volts on preamp cathodes or full B+ on the filter caps.



Don - those clip probes look great, thanks for the pointer I use croc clip leads for hands-free occasions, but those look really useful for tightly packed chassis.

I appreciate the comments - it's always useful to get a reality check on these ideas. I'd just assumed that there were other reasons for not putting the 1r resistors where they'd be most relevant.

[Matt T.]

My opinion is that one of the nicest things about the 1R cathode resistors is that you clip your negative lead to the chassis and make ALL measurements without ever removing it. You can check all your voltages AND your idle current just by probing with the positive lead. With the 1R plate resistors, you'd have to disconnect/reconnect the negative lead for the idle current measurement and then change it back again to check your plate voltage wrt ground.

Also, nobody mentioned it but I was curious: the dc resistance between the plates (through the OT windings) is pretty low, is the series resistance (the 2 each series 1R resistors) added to it low enough to be insignificant?

[SlidePicker]

Originally Posted by loverocker

According to KOC and others, exceeding screen dissipation is as bad as exceeding plate dissipation...

It's a lot worse in terms of frying tubes. They'll tolerate running the plates too hot a lot better than overheating those grid wires.

[aletheian-alex]

Why not make those 1r plate current resistors 100r? Then you have both a place to check the true bias current, AND a plate snubber as a talisman against oscillation--which is usually not an issue except in the case of a higher-slope tube like an EL84, in which case it might be of some good (sometimes when I look at the specs for el84's, I swear that they were preamp tubes in a former life...)

[Tom Phillips]

I would not recommend 100 Ohm resistors added as plate current sensing resistors. The resistance is a significant % of the primary OT winding resistance. That along with the power dissipated in the resistor when the amp is played, you will be loosing power output.
Regards,
Tom

[aletheian-alex]

Originally Posted by Tom Phillips

I would not recommend 100 Ohm resistors added as plate current sensing resistors. The resistance is a significant % of the primary OT winding resistance. That along with the power dissipated in the resistor when the amp is played, you will be loosing power output.
Regards,
Tom

Hmmm... I have seen several data sheets that recommend 100+ ohm plate snubbers for EL84's and a few other small pentodes. Maybe it was just a hair-brained idea of the 1950's...

[Tom Phillips]

That's interesting. I'll have to look into that. Let me know if you run across a specific reference or schematic.
Tom

[Enzo]

The transformer runs on turns ratios, not internal resistance. What the tube sees is the reflected impedance of the speaker in proportion to the turns ratio. If the primary impedance is maybe 4000 ohms on each side, I don't think the extra 100 ohms will matter that much.

If the tube is conducting 20ma of current, the voltage drop across the resistor is 2 volts. Not much out of several hundred volts. 2 volts across a 100 ohm resustor dissipates .04 watts, 40 milliwatts. I think a 20 watt amp can afford a 40mW loss without much trouble.

[Tom Phillips]

Right. Thanks - I spaced out on the AC vs DC circuit model.
Tom

[Amp Kat]

You could get a little advantage from a 100 ohmer if your meter is not very accurate as it increases it 100 times I think but I'm not sure it would be enough cause for concern. The current sense resistors are great for point to point amps but are useless in PCB amps where the traces are underneath the board and I sure as hell ain't taking off a PCB just to install a plate or Cathode current sense resistor and cutting traces. The current from Cathode to plate will be identical with repect to the resolution of your meter. Now if you could go back to say .0000000001 you may see a slight movement but for the risk of slipping and zapping your ass I can't see a benefit of a plate current sense resistor unless you just like to experiment. I don't think you'll get a more accurate answer taking the reading off the plate. I've matched Cathode current sense to the shunt methode several times to no avail as they were dead on to the nearest tenth !

[kg]

fwiw i use 10r cathode resistors. never had an issue with 'em.

also, slidepicker is right.. i'd rather overdissipate plates than screen grids, any day.