CRT grid bias for cutoff

[Tom Katt]

Apologies for more CRT discussions, but this has me very confused.  I've got an old Leader LBO-310A cro scope that I'm repurposing into a vector display.  Like many basic old scopes it has a Z axis input that is just tied to the grid through a blocking capacitor.  The manual indicates the Z axis requires about 20V p-p for blanking.  I'm looking for a way to apply a digital blanking signal for my controller.

But one thing that has me really confused is the schematic shows the grid to have -1070V while the cathode is shown at -1100V.  Doesn't that make the grid positively biased to the cathode?  And I always thought a positive bias grid was BAD because it could damage the tube by stripping the cathode.

Am I looking at this correctly?  Manual here if helpful.

Thanks!

[Tom Katt]

What would happen if I added a bit more resistance to the cathode to provide about 40V differential between the 1100V supply so I could use some kind of switching to put the 1100V on the grid for blanking?   That would probably affect the overall bias a bit, but maybe the various adjustments for astigmatism would compensate?  Maybe I could come up with some kind of optocoupler arrangement as used in David Forbe's design...

[gregebert]

If you have a very-high impedance voltmeter, you might want to check the voltages while the CRT appears "normal". I suspect you will see notable differences between your measurements and the schematic.

Even a DMM with the standard 10Meg input impedance will overload the circuit. You need to make a 10X attenuator for your DMM, such as a 90Meg series resistor (easiest to make it with nine 10-meg resistors). The attenuator will need to be calibrated by measuring a lower with and without it, and confirming the readings differ by 10X.

Back to the blanking question, you should be able to confirm it works with an AC source such as an audio oscillator; the blocking capacitor wont allow you to do blanking down to DC.

[Tom Katt]

On Wednesday, November 5, 2025 at 5:59:08 PM UTC-5 gregebert wrote:

If you have a very-high impedance voltmeter, you might want to check the voltages while the CRT appears "normal". I suspect you will see notable differences between your measurements and the schematic.

Yeah - I learned the hard way many years ago working on tube amps that normal meters and tube voltages don't mix well.  I may try the 10X resistor method, if only to compare ratios and not worry about exact accuracy.  Thanks for the suggestion.

And yes - the goal is to achieve ttl level signal control, so the capacitor arrangement needs to go.  I've seen a few examples of using transistor switching, but the trick seems to be getting a manageable potential difference of maybe 40V to be able to sneak in an optocoupler.  

[gregebert]

I still think the capacitor coupling will work as long as you use an AC signal to modulate the Z-axis. Putting a DC level-shifter into the circuit that needs to isolate 1000 volts is going to require some very careful design. From what I can see, the time-constant for the Z-axis is pretty long (around 100msec ??), and it's very likely whatever you are displaying is much faster, so DC coupling isn't needed.

An op-amp circuit driven from TTL-level can easily translate from 0-to-5volts to -10 to +10 volts (20V p-p per above target), for example, and that can hook directly to the Z axis jack. Basically, a gain of 4, with an offset of 2.5V if your TTL-level swing the full 0-to-5, though 7400 series TTL devices typically drive out 4V, so a gain of 5 and an offset of 2.0V.  LTspice is a good way to simulate your design before building it. The nice thing about inverting OPamps is that the (-) input is already a summing point, so adding the offset is trivial.

Thinking about it more, you might not even need an offset. It functions the same as a contrast control, though I dont see such a control on the schematic. Having adjustable offset wont hurt if you dont need it, but not having it if you do need it means another round of circuit changes.

[Tom Katt]

On Thursday, November 6, 2025 at 2:16:30 AM UTC-5 gregebert wrote:

I still think the capacitor coupling will work as long as you use an AC signal to modulate the Z-axis. Putting a DC level-shifter into the circuit that needs to isolate 1000 volts is going to require some very careful design. From what I can see, the time-constant for the Z-axis is pretty long (around 100msec ??), and it's very likely whatever you are displaying is much faster, so DC coupling isn't needed.

I don't think I'll have an AC signal source for blanking, unless the pulsed signal qualifies.  As a vector display, I'll want to blank the crt every time it moves to draw a new vector not connected to the current one.  The display refresh rate is currently 60Hz, which works out to about 16.7ms for an entire screen - obliviously the blanking signal will be significantly shorter than that as there will be potentially dozens (if not hundreds) of discrete vector starting/ending points in that span.

I use Proteus VSM for SPICE simulation.  Given the small blocking capacitance, I suspect the RC curves will be influenced by the inherent capacitance within the crt tube itself.  I'd prefer to have instantaneous on/off transitions to avoid any 'smearing' on the display.  But the simplest thing is to at least test what happens when I put some signal on the blocking capacitor line to the grid...

[Tom Katt]

I've been looking at how older scopes handle Z axis and even my trusty old HP180A (my age lol) looks like it uses capacitive coupling as well...  Although it has an amplifier circuit that allows a +2V signal to blank the crt.  So perhaps I'm giving up on a solution before I've even tried working with it.