---

A very compact build-it-in oscilloscope module - 20150206

       This is a very small oscilloscope unit measuring 2.5 x 2.5 x 5 inches. The screen size is a mere 1 inch (25.4mm). The entire unit is designed to occupy the panel space taken by a standard mechanical meter movement of its day. The 1950s. I have seen ads for this unit in literature dating to 1957. Possibly earlier. This unit was made in 1960.

       It operates with two power supplies. One provides 6.3 VAC for the tube's heater. The second supply can be as low as 250V and as high as 800 volts. (Pure speculation based on literature for similar products.) I am running this test at 340V. The power and deflection connections are clearly marked on the peg boards. All components are soldered to discretely placed metal posts attached to an insulating phenolic board. Old school construction for sure. As I test this unit, it is obvious that it has never been installed. The silver plating on the contact posts is tarnished and there was no solder present until I applied it to make the necessary connections. This makes it NOS, New Old Stock!

       The 90901 was designed to be built into any piece of electronics that may require local monitoring of complex waveforms. AM and side band transmitters for instance. It couldn't be easier to use. Besides the modest power requirements, it also needs X and Y inputs of about 120V to deflect the spot from the center to one edge of the screen. There is a slight difference in deflection voltage between the horizontal and vertical plates due to their relative positions behind the screen. Both of these inputs are AC coupled. Controls are provided for brightness, X position and Y position. There is no focus control because the 1CP1 CRT is self focusing.

Powering the 1960 scope with my 1960 power supply - 20150206

       Millen also made another version of this scope, model number 90911 which used the slightly larger 1EP1 CRT at 1.25 inches. The module still fit within the same form factor as the 90901 and could directly interchange in most cases. The circuits are extremely simple and similar to the 90901. The 90911 incorporated the additional, and more common, pot for the focus control.

Cossar 1CP1 / DH3-91 one inch CRT with detail of loctal base - 20150207

       The circuit is quite conventional except for the slight goofiness required by this tube having only one of the deflection plates (D1') tied common with the second anode. This is done because of the limited number of pins in the tube's glass envelope. The base of the 1CP1 is referred to as loctal. This base was devloped in the 1950s to lower cost and make a connector that did not vibrate out of the socket in car radios.

       The 1CP1 designation is based on the U.S. JAN (United States, Joint Army Navy) nomenclature developed for components during world war 2. The DH3-91 designation is European in origin. I have no historical context info for elaboration beyond that one fact.

Schematic diagram of the James Millen & Co. 90901 one inch oscilloscope - 20150207

       Here is a copy of the above [schematic in PDF] format. Feel free to copy this design and build your own version of this scope. With 3D printing, it would be reasonable to expect that a good to perfect mechanical facsimile of the Millen scope could be made. If you just want to learn more about oscilloscopes then visit the [Oscilloscope Museum]. A vast array of vintage o'scopes and their history.

       The vertical deflection plates are closest to the screen. The horizontal plates sit farther back, from the screen, than the vertical plates. So, with the slightly longer distance, horizontal deflection is a little more sensitive than the vertical. More millemeters of deflection per volt than the vertical plates. This effect is much more pronounced in these shorter cathode ray tubes. According to the Philips datasheet, the horizontal deflection (D1/D1') voltage is 9.1V/mm, or +/-115V for full screen deflection. The vertical plates have a sensitivity of 10.5V/mm, or +/-130V for full screen deflection. The reason the deflection voltages are so high in this tube is because it is so short. Got to induce a lot of "bend" in the electron beam in the very short distance from the end of the elctron gun to the phosphor screen.

       Total resistance from HV+ to the HV- terminals is 285K ohms. With this info, we can calculate the total current and power needed for any given operating voltage. In any case, it is not very much. At 340 volts, the current drain is approximately 1.2mA for a total power consumption of less than half a watt of high voltage.

       The cathode heater in the tube draws 550mA at 6.3 volts. That's an additional three and a half watts of power. This makes the total power consumption of the 90901 about 3.9 watts.

       You may have noticed that this scope module lacks any means of external beam intensity modulation. That is, it has no blanking or video input. I may develop a modification for this some time in the future. The next task will be to interface a deflection amplifier so I can display my Sparkfun scope clock on this totally cool antique module. As of 2015, this unit is 55 years old. Almost as grizzled and fossilized as an old Labguy I know...

---

Millen 90901, Sparkfun Scope Clock and Labguy X/Y deflection amplifier - 20150207

       Above is the Sparkfun scope clock board set up with my high voltage X/Y amplifer and the James Millen 90901. Not shown are the 12 volt adapter for the scope clock board and the bench power supply providing the low voltage +/-12 volts to the deflection amplifer board. The chassis with the two vacuum tubes provides the 6.3 volts CRT heater voltage, the +300 volts for the deflection amplifier outputs and +340 volts high voltage to the Millen 90901 oscilloscope.

       The Sparkfun scope clock board outputs only a 5 volt signal. The 90901's CRT needs over 200 volts (pk-pk) to drive to full screen deflection in the 1CP1. That is the function of the deflection amplifier board. I designed it to take in five volt ground referenced signals and generate two output signals that are 300Vp-p in opposite phases. This gives the effect of having up to 600Vp-p on the deflection plates at the full gain setting. Output is completely variable from zero to full scale.

Millen 90901 running the Sparkfun Scope Clock with unexplained hum in the Y channel - 20150207

       That darned black bakelite is vey hard to photograph! Sorry for the dark photos.

       Without firing up my bench oscilloscope and searching for it today, the scope clock setup has an AC hum (oscillation?) in the vertical axis channel. But, not on the horizontal. This not a failing of the 90901, it is most definitely in the amplifier board. Perhaps a bypass capacitor is open or inadequate for the job. I admit that the ground system on this prototype is lame. This is why the display looks fuzzy in the close up photos. The image is bouncing up and down at high speed. Two identical amplifiers. One stable, the other not. Hrumph! (Her what?)

       The reason I had not seen the oscillation before was because the unwanted AC signal's amplitude is constant and the scan amplitude was much greater for the three inch tubes. Looking at previous photos, I now see this vertical smearing. The oscillation is going to turn out to be around 10 Vp-p. That will be covered later when I solve that problem.

       Despite the deflection amplifier problem, it has been successfully demonstrated that the James Millen & Co. 90901 oscilloscope module works as well as the day it was manufactured over half a century ago.

       More to come. Stay tuned.

REFERENCES:

1. The James Millen Society Dedicated to the memory of Mr. James Millen, W1HRX, founder the company that bears his name. The Society provide lots of documentation for Millen products. But, seems to be lacking info on this one. Can anyone help add to the knowledge base?

2. Philips 1CP1 / DH3-91 one inch green phosphor CRT Datasheet. This datasheet also applies to the 1CP11 / DB3-91, blue phosphor tubes.

3. Electrostatic CRT Projects, Labguy's Wolrd dot com. Some more detailed work with electrostatic cathode ray tubes.