Reading here I here people speak of this thing called "grid current". What is it? I understand its current and varies based on ohms law like any other current. It must be very small as the grid basically floats in the space charge right? It has no resistor across it to draw current. Maybe if a tube is gassy that might be like a resistor that would cause the grid to draw current?
Also on the breadboard how do I measure such a small thing as grid current without having the measuring instruments itself throw it all off? Certainly the sense resistor method wont work by measuring the voltage drop across a grid stopper, seems like it would be an impractical way to measure it, the resistor would have to have ultra-precision right? I have this 100 uA full scale analog microammeter, in bygone days would people measure grid current by simply hooking a microammeter in series with the grid? If so, should it be hooked up on the input or output side of the grid stopper? If so, what polarity to hook it up? If so, should I maybe pick up a 50 uA full scale microammeter for better accuracy? If so, Should I pick up a center needle 50 uA meter so I dont have to consider the polarity hooking it up?
If these are too-many stupid questions please set me straight on the practicalities of measuring grid current 🙂
Also on the breadboard how do I measure such a small thing as grid current without having the measuring instruments itself throw it all off? Certainly the sense resistor method wont work by measuring the voltage drop across a grid stopper, seems like it would be an impractical way to measure it, the resistor would have to have ultra-precision right? I have this 100 uA full scale analog microammeter, in bygone days would people measure grid current by simply hooking a microammeter in series with the grid? If so, should it be hooked up on the input or output side of the grid stopper? If so, what polarity to hook it up? If so, should I maybe pick up a 50 uA full scale microammeter for better accuracy? If so, Should I pick up a center needle 50 uA meter so I dont have to consider the polarity hooking it up?
If these are too-many stupid questions please set me straight on the practicalities of measuring grid current 🙂
You could use a very large grid resistor for the measurement.
Grid current is very small unless the grid is driven positive or at least not sufficiently negative, or the valve is gassy. Sometimes the grids of valves used as oscillators and those of power valves (used in class B2 or AB2) are deliberately driven positive.
Grid current is very small unless the grid is driven positive or at least not sufficiently negative, or the valve is gassy. Sometimes the grids of valves used as oscillators and those of power valves (used in class B2 or AB2) are deliberately driven positive.
Ideally there should be no grid current with negative grid potential. However, in a real tube the vacuum is not absolute, so there is some small conductance due to gas ionization. Also, as tube ages, metal vapor from hot cathode is deposited on the insulators, creating bridges of conductance.
Under normal conditions, grid current is difficult to measure directly because it is very small (nanoamperes). There are indirect methods though. Such indirect methods are used, for example, in tube testers to detect poor vacuum.
Under normal conditions, grid current is difficult to measure directly because it is very small (nanoamperes). There are indirect methods though. Such indirect methods are used, for example, in tube testers to detect poor vacuum.
Are you talking the nanoAmps of small-signal bias? Or the many milliAmps of a Power tube driven positive grid?
Any meter with a decent millivolt scale should be accurate enough. Use a 1k or even 10k series resistor.
The OP asks why grid should draw current if, apparently, there is no physical carrier for such current other than gas. This implies that the question was about conductance at negative grid potential, and not about conductance at positive grid potential, where the current carrier is obvious: flow of electrons from negative cathode to positive grid.
Radiotron Design Handbook has a good passage on grid current; I like the 3rd edition for many things.
Two more in old style in searchable text. Both have sections on triodes that talk about grid current.
https://ia601606.us.archive.org/22/...e_the_Vacuum_Tube_John_F._Rider_1945_text.pdf
https://ia801006.us.archive.org/4/i...ons_of_Electron_Tubes_-_US_Army_1952_text.pdf
Could you also measure grid current with equal valued resistors in both the cathode and plate circuits and look for difference on an oscilloscope? (Would want a high voltage or differential probe for the plate circuit in most cases.)
https://ia601606.us.archive.org/22/...e_the_Vacuum_Tube_John_F._Rider_1945_text.pdf
https://ia801006.us.archive.org/4/i...ons_of_Electron_Tubes_-_US_Army_1952_text.pdf
Could you also measure grid current with equal valued resistors in both the cathode and plate circuits and look for difference on an oscilloscope? (Would want a high voltage or differential probe for the plate circuit in most cases.)
Measuring where???
Meaning:
* where would you put your probes?
* which meter scale would you use?
or: same in different words:
* what would the galvanometer sensitivity be?
What sort of circuit/tube are you looking to measure grid current?
Most form of grid current will not be constant.
There are different forms of grid current.
Contact potential grid current may exist as a slight negative current and has been used as a biasing method in small signal circuits. The current is usually small and measuring the current can be tricky to do right. Also may vary from tube to tube.
If the grid is driven positive with respect to cathode there will be grid current. This current also depends on plate and screen voltage. There is less grid current when the plate voltage is high. Electrons might as well just go for the plate if it is a high voltage. As plate voltage drops, electrons might as well stick to the control grid.
Screen grid current is also dependent on control grid voltage and plate voltage.
Most form of grid current will not be constant.
There are different forms of grid current.
Contact potential grid current may exist as a slight negative current and has been used as a biasing method in small signal circuits. The current is usually small and measuring the current can be tricky to do right. Also may vary from tube to tube.
If the grid is driven positive with respect to cathode there will be grid current. This current also depends on plate and screen voltage. There is less grid current when the plate voltage is high. Electrons might as well just go for the plate if it is a high voltage. As plate voltage drops, electrons might as well stick to the control grid.
Screen grid current is also dependent on control grid voltage and plate voltage.
I think I got this... Putting aside gassyness grid current... I see grid current concerns are mostly for when the grid is biased positive letting current intentionally flow along the cathode / grid path. Some classes of operation use this way to get more electrons ushered to the plate (I guess) by giving them a "springy" start accelerating through a positive grid much closer to the cathode. So the techniques for measuring grid current would be good to know when using that class of operation. But for most uses when the tube is biased negative it repels the cathode so no current. But to prevent grid current in this normal operation as well, it must be negatively biased enough so that the largest expected positive signal peaks still leave the grid negative. With insufficient negative bias the musical peaks can generate grid current spurts. So basically having a good way to measure grid current is handy when breadboarding a positive grid Amp, I really don't need to do it otherwise like on driver or input stages. Now, when making a positive grid biased power stage, the trick is knowing how much grid current is "just right"? Everything takes measuring, positive biased power is probably a subject for another thread. I think I understand what people mean when they use the term grid current now, thanks. On power stages like this measuring the drop across the grid stopper is what most people do? (not having spice).
Last edited:
In a nutshell:
there are TWO kinds of grid current, completely different, you can´t put them in the same bag and treat them as equal.
* "parasitic/lossy" grid current, caused by electrons trapped by the grid.
Very very low value, practically impossible to measure with standard bench instruments.
Typically used to self bias tubes by using a very large value grid resistor (think 3M3 to 10M or so) , any meter you use there must have >>10M impedance (say 1000M for a 1% measuring error), not easy; so normal means to measure it are indirect, say measuring plate current and comparing it to some fixed bias which causes the same effect.
It also requires an input capacitor, to avoid sound source (microphone/guitar/etc.) shunting it to ground and murdering bias.
* GK diode forward current caused by positively biasing it.
It is "a diode", literally, can easily pass a few mA and it is measurable using a mA meter and an adjustable positive power supply.
As you see, both "grid currents" are completely different.
there are TWO kinds of grid current, completely different, you can´t put them in the same bag and treat them as equal.
* "parasitic/lossy" grid current, caused by electrons trapped by the grid.
Very very low value, practically impossible to measure with standard bench instruments.
Typically used to self bias tubes by using a very large value grid resistor (think 3M3 to 10M or so) , any meter you use there must have >>10M impedance (say 1000M for a 1% measuring error), not easy; so normal means to measure it are indirect, say measuring plate current and comparing it to some fixed bias which causes the same effect.
It also requires an input capacitor, to avoid sound source (microphone/guitar/etc.) shunting it to ground and murdering bias.
* GK diode forward current caused by positively biasing it.
It is "a diode", literally, can easily pass a few mA and it is measurable using a mA meter and an adjustable positive power supply.
As you see, both "grid currents" are completely different.
So if talking about an input stage tube... These days DAC's have a healthy 2v signal so its good to bias modern input tubes minimally no higher than -2.5 to
-3 volts when choosing an operating point, so the grid remains negative, right. In bygone days you could bias hotter because your input sources were only in the mv range? Of course the selected plate voltage matters too, for hifi you want to set the operating reference to the linear area of the curves in combination with a negative bias selection point.
One more question, when I look at the curves on a tube data sheet as plotted. Those curves were all taken with an unloaded plate correct? Just by applying a range of plate voltages and negative grid voltages from cutoff bias (bottom of curve) to zero v bias (top of the curve). There is no plate resistor involved yet with that graph right?
-3 volts when choosing an operating point, so the grid remains negative, right. In bygone days you could bias hotter because your input sources were only in the mv range? Of course the selected plate voltage matters too, for hifi you want to set the operating reference to the linear area of the curves in combination with a negative bias selection point.
One more question, when I look at the curves on a tube data sheet as plotted. Those curves were all taken with an unloaded plate correct? Just by applying a range of plate voltages and negative grid voltages from cutoff bias (bottom of curve) to zero v bias (top of the curve). There is no plate resistor involved yet with that graph right?
Not necessarily.
Dacs can put up to 2V RMS out, good, that does not mean they do that 24/7/365 or "all the time"
There´s a reason volume controls exist.
A DAC is not a Lab Bench signal generator, and even so those have adjustable level out as needed.
I BET your preamp needs WAY less than 2V RS input to drive power amp to clipping if you wish, so set DAC volume/output to a sensible level, which will not tax your preamp input grid at all.
Depends on whether there is a volume control in between, like JMFahey wrote, and whether the input valve is inside a feedback loop.
Assuming that you really want the valve to be able to handle 2 V without any kind of feedback, you need to take into account that 2 V usually means 2 V RMS when playing a sine wave, so 2 sqrt(2) V ~= 2.828 V peak, and that the peak voltage at the grid still has to be somewhat negative. Some valve datasheets specify the required negative voltage to keep grid current below 300 nA, it's usually something between -0.5 V and -1.3 V. All in all, you would need a grid DC bias voltage of -3.328 V...-4.128 V or more negative.
Of course you would also need to make sure that the anode current doesn't get cut off in the negative input signal peaks and that the anode voltage doesn't get too low in the positive input signal peaks, as you already mentioned.
Hundreds of millivolts usually, and feedback was often used.
Indeed.