Wanted - Insight Into Forbes Scope Clock Blanking Circuit

[Tom Katt]

I continue to do thought experiments with the easiest to accomplish crt blanking circuit for a vector display project.  The blocking capacitor design used on many scopes is insufficient for the blanking I want to do, as they only work well for regularly repeating short pulses and my needs are more random.

Anyway, looking at how Mr. Forbes approached this for his Scope Clock it seems to me that what's going on is a low voltage (maybe 35 - 40 volts) rectifier being applied to a high voltage line, basically summing the voltages together to raise the -1200V supply to about -1240V, which is enough to provide cutoff for the crt grid.

Is there anything special going on here that I am missing, or could I take virtually any low voltage and impose that on a high voltage?  It looks like caps C7-C8 in the schematic would block the high voltage source from the transformer windings.  So even if I did not have a special multicore transformer, I could potentially add a separate small 40 VCT (or 24 VCT?) transformer to my existing supply and end up with the same idea?  Or even maybe a small DC-DC board that provides about 40VDC with a series capacitor and sum that with the higher voltage?

Suggestions welcome!  Thanks!

 

[gregebert]

C7 & C8 are for a voltage-doubler. The isolation is thru U3 (TLP2361 opto-isolator) and from what I can tell it's a blanking circuit that goes all the way down to DC, unlike the Z-axis inputs on scopes that use a capacitor and have a lower-frequency limit.

[Tom Katt]

I get the voltage doubler - but that is creating the additional 40 volts that gets summed to the -1200 line and switched by the opto.

My question concerns creating the -1240 for the opto to switch.  The transformer windings on pins 4, 5 and 6 aren't creating the HV, but only the 40 volts.  I'm trying to come up with a way to add that additional voltage needed for cutoff to my existing -1100 volt supply, which will be switched in a similar manner.

[gregebert]

You just need to be careful about the isolation, because it's 1100 volts rather than the typical 120V for a typical transformer. Good transformer manufacturers such as Triad will provide isolation info in their datasheet.

[David]

Tom,

This circuit is rather complicated for a blanking circuit, because I wanted high performance without needing an expensive 1600V film capacitor as is used in high end oscilloscopes. 

It may be possible to achieve decent performance with a voltage multiplier driven from the CRT heater supply to drive the grid switch circuit. I recently did a test of lowering the voltage from 40V to 20V with decent results. I haven’t explored all the possibilities yet, but I do want to find a way to simplify it while keeping good performance. 

David Forbes, Tucson AZ

On Aug 23, 2026, at 9:37 AM, Tom Katt <tomk...@gmail.com> wrote:

I continue to do thought experiments with the easiest to accomplish crt blanking circuit for a vector display project.  The blocking capacitor design used on many scopes is insufficient for the blanking I want to do, as they only work well for regularly repeating short pulses and my needs are more random.

Anyway, looking at how Mr. Forbes approached this for his Scope Clock it seems to me that what's going on is a low voltage (maybe 35 - 40 volts) rectifier being applied to a high voltage line, basically summing the voltages together to raise the -1200V supply to about -1240V, which is enough to provide cutoff for the crt grid.

Is there anything special going on here that I am missing, or could I take virtually any low voltage and impose that on a high voltage?  It looks like caps C7-C8 in the schematic would block the high voltage source from the transformer windings.  So even if I did not have a special multicore transformer, I could potentially add a separate small 40 VCT (or 24 VCT?) transformer to my existing supply and end up with the same idea?  Or even maybe a small DC-DC board that provides about 40VDC with a series capacitor and sum that with the higher voltage?

Suggestions welcome!  Thanks!

<sctv-schem-1.png>

 

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To view this discussion, visit https://groups.google.com/d/msgid/neonixie-l/f5dbeada-8178-4cc0-a23a-e894f93df86cn%40googlegroups.com.

<sctv-schem-1.png>

[Tom Katt]

Thanks so much for weighing in here!  I'd say 'complicated is in the eye of the beholder - personally, I see it as very clever.  The issue I am struggling with is that I have a fixed neg HV supply for the CRT and need to find some way to increase the grid sufficient for cutoff on a logic level as you have done.  I'm working towards converting an old Leader 310A scope to have a proper Z axis.  

I do have access to a rather unique HV optoisolator that would be great if I could figure a way out to apply it - I have considered using it to switch the entire neg HV rail on and off for blanking (which would be cathode, grid and A2?), but some reading suggests that fiddling around with the cathode HV supply may affect the beam focus as well.  So if I switch the entire negative supply as my blanking that might result in some visible distortion.  But maybe worth a try.

The opto I have is an MSS-OPTO-150-RevB.  Too expensive for mass production, but as I have it seems like a possible solution.

I thought about the heater supply as well since it is already there and tied to the HV section - that would be a fantastic solution!  I may play around with that as well ;-)

Thanks again!

Tom

[Tom Katt]

Just for future reference, I ran an experiment and achieved logic level control of crt blanking with a single component.  Granted that component retails for $90 and the company might be going under as their website is no longer active, but it IS an option ;-)

I simply inserted the opto in series with the crt cathode supply so that logic level signals control whether the crt cathode is connected to HV or not.  I left the other crt leads connected as normal, so the grid and focus elements remained connected to the HV resistor divider chain.  Sending 3.3 volts to the opto led (it has a built in current limit resistor that consumes 100ma) allows the cathode to see HV and thus emit electrons to the phosphor.  Removing the 3.3 volt opto signal blocks the cathode HV, instantly cutting off the beam.  The beam remains off until the opto gets the enable signal again.

I do not have the proper test gear to safely measure HV current, but as best I can tell there are no issues and the scope continues to run normally with nothing getting warm or appearing stressed.  Likewise, the crt display is unchanged and remains sharp. 

Obviously, there are better ways to achieve blanking, but cost aside I doubt many are easier to achieve lol.

Now on to completing the vector controller board ;-)

[Dekatron42]

There are several similar extra high voltage opto couplers available at Digikey for far less money, I've used the Vishay CNY65-serirs successfully in some applications, Optek also has some.

[Tom Katt]

I'm hardly a HV expert nor do I have vast experience with related components - but I always thought that while devices like the CNY65 series provides HV *isolation* they are not capable of actually switching or controlling HV circuits...   The Vishay example shows a Vceo limit of 80 volts - isn't that the limit of what can be switched by the phototranistor inside the opto?

I did quite a bit of searching when I stumbled on the opto I used and while I did see many that provide kilovolts of isolation, none had switching or reverse breakdown voltages much higher than perhaps 500 - 600 volts...

Now, I may be reading the datasheets wrong - I'm willing to learn (and save money lol) !

Thanks!

[Dekatron42]

I thought we were discussing "Isolation" voltage here in the thread - but after reading the datasheet you linked to I see that the device you are referring to "MSS-OPTO-150-RevB" actually mentions "15kV Reverse Voltage Rating" of the photo diodes.

[Nikolay Miroshnichenko]

Tom, you are reading them correctly — those are two different specifications and they get mixed up constantly.

The isolation number describes the barrier between input and output: how much the insulation between the LED and the detector will stand off. The Vceo number describes what the output transistor can hold on its collector. Nothing connects the two. The CNY65 is a good example: the catalogue entry says 8.2 kV isolation and, in the same table, a load voltage of 32 V (Vceo 80 V absolute max in the datasheet). A factor of 250 between them, and both are correct.

The same trap sits in the part used in the Scope Clock. The TLP2361 datasheet shows 3.75 kV rms — that is a one-minute test voltage. Its working rating, V_IORM, is 707 V peak. We hit exactly that in our own design review: the tube cathode sits at −1600 V DC continuously, so the coupler was outside its working rating even though the "5 kV" headline looked generous.

So an optocoupler can carry the command across the barrier, but the thing that actually stands off 1600 V has to be a high-voltage device on the hot side — a MOSFET or a bipolar cascode rated for it. The MSS/HVM OPTO-150 you bought is the exception that proves the rule: it costs what it costs because the photodetector itself is rated 15 kV, so it needs no help.

Two other numbers worth having before you standardise on any of them:

• Speed. The CNY65 is 2.4 µs rise / 2.7 µs fall; the OPTO-150 datasheet says 2 µs on and 2 µs off. That is fine for turning a display on and off, but if you ever want blanking per vector, a point on a vector display lives a couple of microseconds, so the coupler alone eats the whole budget.
• What actually limits your cathode-supply switch is not the coupler at all, it is the capacitance on the cathode node. Breaking the HV feed does not pull the cathode down: the multiplier column and wiring stay charged and bleed off only through beam current and leakage. Coming back is slower still — recharging a Cockcroft-Walton column is a fraction of a second in our simulations.

For what it is worth, here is the direction we went on our own CRT supply, still on the bench-less side of the project, so take it as a design in progress rather than advice. Instead of switching the high voltage we let the whole grid domain float on the cathode: heater, grid drive and its own local supplies all sit at −1600 V, and what crosses the barrier is digital, through a reinforced digital isolator (V_IORM 2828 V peak, DC-qualified, about six dollars). Blanking and brightness are then done at ordinary transistor level inside that floating domain, so the accelerating voltage never moves and focus does not drift.

Simulated, not yet measured: cutoff at −80 V with a 147 ns blanking edge, and 6-bit per-point brightness carried across as serial data into a shift register and an R-2R ladder on the floating side.

And one warning from the same work, since it cost us a week: our first version of that grid stage could not blank at all. We had a complementary emitter follower whose base sat on the isolator's 0/5 V logic output, and an emitter follower cannot drive its output beyond its own base — the −80 V rail on the collector was never reachable. Six review passes cleared it before a five-line SPICE transient showed 0.3 V of grid swing instead of 80. If anyone copies that topology, check it in simulation before trusting it.

Nikolay

[Tom Katt]

While I wait for a DigiKey order to complete the analog board for the Leader scope Z axis hack, I figured I'd begin work on a crt test setup to design a better platform from the ground up that would allow grid blanking.  Unfortunately, switching power supplies remain black magic to me so I will go a more comfortable route using a small 400 VCT toroid from the tube amp world - Antek in NJ USA makes a variety of reasonably priced transformers which should be sufficient with a few multiplier circuits.  Riad Wahby's website jfet.org has a lot of good information and his design allows blanking by pulling the grid down to the negative rail via a transistor controlled by low voltage logic - I'll throw a bunch of variable resistors in the mix to let me play and see what voltages work best.  The grid supply is obtained from the HV resistor chain so that should simplify things.  It would be nice if crt designs could be more easily simulated, but I don't think there's anything that can calculate what the beam actually looks like in real life with different voltage parameters.

Antek toroids: https://www.antekinc.com/as-05tc200-50va-200vx2-transformer/

jfet.org: https://web.jfet.org/vclk/

If nothing else, this will balance the Nixie / crt ratio in the office ;-)

Tom