blanker circuit that was used to remove interference from digital
information encoded in the first few scan lines of modern TV signals
and DVDs. Old TVs, which didn't have to contend with this encoded information
in the 1940s and 50s, often display various retrace lines and blips in
the upper half of the screen, which would be nice to remove. The
blanker circuit, using an added tube, solved the problem beautifully on
the sets shown at CFP Radio.
I wrote to CFP Radio via the link provided at their website, and requested more information about the blanker circuit. Pierre, one of the people responsible for the website, answered right away, and translated the information about the blanker into English for me, as well as sending 3 schematics showing the circuit itself and how it is
You can send an email to receive the complete information, or contact Pierre directly through his website ( www.cfp-radio.com ) for the same information.
The circuit is a one-shot triggered by the vertical sync pulse. It uses a dual triode (he suggests a 12AU7), with filament wired as appropriate for the particular TV you want to use it in. Here's the circuit:
The line labeled 'noir' is ground; the one labeled 'sync' is coupled to the plate of the sync separator tube; the one labeled 'blank' goes to the first grid of the CRT through a capacitor (value not given, so experimentation may be needed), and a resistor is added between the CRT grid and the original connection to it (again, some experimentation with the value may be needed).
This circuit is a 1.2 msec one-shot, which is the correct blanking time needed to blank all the offending signals. It is not possible to build a strictly passive circuit with a long enough blanking interval to get rid of all the interference, so this kind of solution is what is needed.
Nice to see that someone still knows how to design this kind of circuit using a tube! I probably would have used a 74LS123, which would not seem very authentic in an antique TV.
built and tested the blanking circuit with my Philco Predicta Holiday.
I had to make a few minor adjustments to the circuit to get it to work
for me. First, the 1.2 msec blanking pulse was a little too long. For
me, 1.0 msec worked perfectly. Perhaps this is simply the difference
between European 50 fps vs American 60 fps frame rates. Adjusting R405
in Pierre's circuit to 270 K changed the blanking duration from 1.2
msec to 1.0 msec.
Also, for whatever reason, I needed more than 10 pF for C400. 100 pF worked well for me.
I also used a 6CG7 dual triode as suggested by Dennis, since the 600 mA heater will match perfectly with the series heater string in the Predicta.
Here is the circuit with my modifications:
The connections to the Predicta are explained in the following schematic diagram excerpts. First, the big picture, including the video output stage and CRT connections (in red oval), and the sync separator and vertical stages (in green oval):
Zooming in to the sync and vertical, you can see where I chose to grab the vertical sync signal from:
Zooming in to the CRT connections, you can see that I disconnected the first grid of the CRT from the existing circuitry (which includes a pulse conditioning network feeding the old blanking pulse from the vertical output transformer) and connected it through a 0.2 uF capacitor to the output of the new blanking circuit, along with a 100 K resistor to ground.
I would normally take pictures for you to show the circuit connected to the set's chassis, and also the "BEFORE" and "AFTER" pictures of the set's screen, showing the effectiveness of the new circuit. Unfortunately, the rest of the family left for a camping trip (I'll join tomorrow) and took the digital camera with them!
So you'll just have to trust me that it in fact works very nicely. Completely eliminates the interference lines. If you leave the blanking interval at 1.2 msec, it will also take away a number of scan lines from the top of the image .
Connections for other sets with slightly different CRT and sync connections may need a bit of experimentation. I'll try this on a couple more sets in the near future and report on what modifications were needed, if any.
are a few followup pictures (now that I've got the digital camera
back). Here is a picture of the added circuitry on the Predicta
chassis. The whole circuit is built around a single tube socket with a
single terminal terminal strip mounted with it. The whole works
attaches with a U-bracket to the chassis with a single screw (an
existing screw that attaches a strut from the tuner to the main
Here you can see the blanking pulse and the video output on the scope. The 1.0 msec blanking pulse is just long enough to obscure all of the videotext signal that creates the interference:
Note that in the above image, what looks like an extended pedestal (black level) on the video signal is in fact the real video signal. I happened to be tuned into a station that was showing letterbox video in a 4:3 format, with black bars at the top and bottom. You're seeing the black bar. The actual image starts just a tad later than 1.0 msec after the start of the retrace.
Here is the "BEFORE" picture with the usual interference:
And here is the "AFTER" picture with the blanking circuit working:
Now you have a use for a few of those 6FQ7/6CG7 tubes that we all seem to have a boatload of.
far as load on the set's power supply, only the filament would be of
concern (consistent with your comments) -- the B+ load is negligible.
For a color set, it's not obvious to me that you need three circuits. Since there is a single common brightness control for all three guns, I would study the brightness control circuitry and see if you can introduce a single blanking circuit there to "add" to the voltage level established by the brightness control. If there is no filtering in the intervening circuitry that would prevent a fast blanking pulse from getting through, you should be in business.
Even if you do need to apply three separate blanking signals (signal to each gun well decoupled from the other two), I would first examine whether feeding via three resistors would be sufficient. If, in the end, you need three separate low impedance feeds, I would just build one blanking circuit, and use a triple triode (like 6T8) to separately buffer (using each triode as a cathode follower) the signal going to each gun. But I would be very surprised if you really need to go that route.
received an inquiry by email to provide a little more detail on exactly
how the new blanking circuit is wired into the Predicta chassis. I'll
show pictures here in case anyone else is interested in the same
The blanking circuit is built onto a single tube socket, which is mounted by a U-shaped metal standoff on one of the existing screws attaching the back end of the tuner to the main chassis. The left mounting screw for the new socket also serves as the grounding point for the circuit (notice the solder lug just under the socket on the left).
This location is very close to the socket for the cable to the CRT, which is the small 5-pin chassis you see right in front of the added blanking circuit. The large yellow cap you see under the 6CG7 tube in the new circuit is the 0.2 uF output capacitor that feeds the new blanking signal to the grid of the CRT. One end connects to the 6CG7, and the other goes through a hole in the chassis to the bottom side, as do the two wires from the filament of the 6CG7.
Now here's the view from the bottom side:
The filament of the 6CG7 is introduced into the series heater string right at the CRT cable socket. From the schematic posted earlier in this thread, you can see that pins 4 and 5 go to the CRT filament. I disconnected the brown wire that was connected to pin 5 of the CRT cable socket, and connected it to the new yellow wire going to the 6CG7 filament. The green wire is the other 6CG7 filament wire, which is now connected to pin 5 of the CRT cable socket.
The original blanking circuit wire has been disconnected from pin 3 of the CRT cable socket, and pin 3 is now connected to the 0.2 uF capacitor wire (feeding down from the top of the chassis through a hole), and also to a 100K resistor, which is grounded to a nearby terminal strip ground.
The two remaining connections for the new blanking circuit connect to the main printed circuit board, shown below:
The red wire is the B+ line feeding the new blanking circuit, which is connected to an existing 150K ohm resistor on the board (see schematic posted earlier in this thread). The vertical sync signal is taken from one lead of an existing 0.1 uF capacitor as shown.
And that's all there is to it!
by Tom Albrecht
(which you can read about in French here: http://www.cfp-radio.com/resta ... anking.php and here: http://www.cfp-radio.com/resta ... 66-04.html ).