Solartron CD 568
Now here is something unusual from I think, the early 50s; In some respects, it is similar to the AN-USM Synchroscopes. Here is what Solartron have to say about it, quoted directly from their catalogue:
“These Solartron Oscilloscopes have been evolved from the well-known type D 300, based on the original design of the Government Radar Research Establishment. The instruments use a new post-deflection acceleration cathode-ray tube to give a marked improvement in brilliance and resolution of trace. Facilities are provided for accurate time measurement by the use of markers which can be superimposed on the trace in CD 518 and a sine wave calibrator in CD 568.
These oscilloscopes are intended for use in radar, pulse and general communication work. They are particularly suitable for display and measurement of transient and continuous phenomena in the low frequency, audio and radio frequency bands.”
It arrived in somewhat dirty but otherwise functional condition. British quality electronic equipment of that era was just that, quality. This one has no electrolytic capacitors and two hermetically sealed double C core power transformers as well as two similar chokes; all the terminals are on hermetically sealed ceramic posts and this is true for all the power supply capacitors also. (As a kid, I had a Cossor 1035 MK111, this was built similarly.) A previous owner saw fit to cut the sides of the enclosure away, also to punch lots of holes in the back; I am not sure why, the instrument is quite large for the power it dissipates, around 100W, and I would think that heat is not an issue. In a way though, it suits me because it leaves the insides visible! It has a PDA CRT (pictured below) that is round-faced. I understand from a correspondent that Cintel developed this CRT (90EG4P/CV1587) especially for this application. This surprises me since flat-faced CRTs such as the 4EPx were available when it was made; it must have been done for economy; my unit has serial number 59296 so a lot were made. Like many British instruments of that era, it is superbly constructed but was functionally somehow peculiar. Tektronix created not only superbly built instruments but ones that satisfied a broad base of user’s needs. The British, clever and often brilliant; Americans, clever, also often brilliant and practical too. Damn!
It took me a few hours to familiarise myself with the circuit and then after lubing every tube pin with Deoxit, ditto the switches, I was able to get it into calibration and it is quite good too though I did subsequently do some further work on it because I was not satisfied with the trace definition:
1/ Correct the astigmatism (there is no astigmatism control).
2/ Replace the PDA rectifier and filter capacitor.
3/ Replace the external Z modulation coupling cap*.
These three corrections have resulted in a display that is now quite a bit sharper and brighter.
4/ Correct the trace length (it was too long) by padding R68 down to 57k from 68k. This correction serves to increase the repetition rate which reduces the annoyance of low sweep-rate flicker. This is part of a potential divider network, I note that R49, the lower part of the divider is given on the schematic as 33k (as it is in the scope) while in the simplified schematic, it is given as 51k. My modification put the voltage fraction close to what is indicated on the simplified schematic.
* The coupling cap is a TCC Visconol, these are often leaky. I measured about 80µA leakage current. Disconnecting it resulted in a noticeable improvement in brightness. I actually placed a larger capacitor in series so as to avoid removing a rather obvious physical feature. This will not affect function much since any Z modulation is likely to be a pulse.
Here are some pictures of the beast. If you want a headache, since this instrument is unique in my experience, I have created a fairly intense functional description below these pictures.
Here is the front showing the meter provided to measure amplitude, it is quite accurate and makes this a form of valve-voltmeter but with a display. If you are sharp, you will spot the “crowsfoot” on the lower left of the meter which indicates that it was procured by the UK Ministry of Silly Walks, err, I mean Defense.
Here is the CRT showing the beautiful label that was revealed, industrial archeology! The label states “Max. Anode Volts, 6kV”, in this case it is approximately 3.5kV.
Here is the gun and massive deflection plates, I will say more about that when I get to describing the circuit.
Right side (Y system), the pipe is a 0.6µS delay line. The plastic cubes on the base contain the inductors for the frequency calibrator which has a most unusual design. It is simply a resonant circuit that is “pinged” by the trigger so that a damped oscillation occurs each time the timebase sweeps. The resonant frequency(s) are quite accurate and so it can be used to set the timebase sweep rate in say, mS/in.
View showing one of the two top-quality power transformers, the two main power supply rectifiers are on the right with the negative rail gas-regulator and the HV negative rectifier is on the left.
There are two B+ lines, one gas tube regulated negative line, one HV negative line and a positive PDA supply.
How it works and what it was for:
This instrument is paradoxically, perhaps the most interesting instrument in this collection while being the least usable. It has a triggered time-base that is specifically designed to trigger from fast edges. While it can trigger from sine waves, that is not the primary intention, the maximum frequency or better said, repetition rate of the wave-form that it can trigger from is 1MHz. In addition to this, there is no internal connection from the signal under examination to the trigger, the connection has to be made externally which is consistent for its intended field of application that may be summarised as systems that have trigger signals and exhibit fast events that repeat relatively – to our modern systems – slowly, up to 1MHz. In other words, it is intended to examine the triggering of external circuits together with the results of the triggering. For example, if you wanted to know when an event in an external circuit occurs after triggering, this oscilloscope is designed to do that. Furthermore, it also has a 0.6µS delay line that can be switched into the Y system and a variable time-base delay. It does anticipate the Tektronix trigger (Ropiequet, not Vollum) in that it has an auto-run feature when switched into ac mode. It can be calibrated and is capable of amplitude and time measurements, most “instruments” of the period could not do this. In addition the graticule is engraved on both sides to help avoid parallax errors. However, the way these measurements are made, when viewed from the perspective of a Tektronix-like calibrated oscilloscope, is somewhat cumbersome, yet interesting. I don’t have a date for it, I suspect the early 50’s since it has a domed face CRT (with PDA). It is fun to review this instrument in a historical perspective, I would hate to have to use it for real.
The main power supplies comprise two full-wave tube-rectified C-L-C filtered B+ rails and a gas-tube regulated negative rail. The stability and accuracy of the instrument is dependent on the negative rail. The B+ rails feed the Y and X systems separately so as to isolate the Y system from the varying demands of the X system.
The negative and positive CRT HV supplies are derived from the same HV winding on the power transformer, the negative side employing a tube rectifier while the positive (PDA) supply employs a selenium “stick” rectifier.
The power switch is designed to turn the heater (LV) supply on before the HV supply.
The Y System and amplitude measurement (The schematic is below):
This comprises a 6 position capacity-corrected switched attenuator, the ranges are 500, 100, 30, 10, 3 and 1 V peak. The two highest ranges are attenuated at the input to limit the signal to the following cathode follower to 30V peak. The cathode follower drives the main (capacity corrected) attenuator for the other 4 positions. Since the system is DC-coupled, provision for adjusting the DC balance is provided so that the trace does not shift up or down as the attenuator is operated. The second attenuator is followed by a most interesting push-pull deflection amplifier:
The Y2 amplifier is what is often called (I think incorrectly) a SRPP. (I am not going to state the words for SRPP because I have never understood the SR part, for me it fails to describe the circuit in any useful way, even to EEs as far as I can tell.) The way Solartron describes the circuit is the most useful and intelligent description I have come across, it completely de-mystifies this circuit that audiophiles have become so enamored / contemptuous of mostly because almost nobody including me, until now, understood it.
Quoting from the Solartron circuit description: “The amplifiers differ from the conventional anode follower by using a valve as the anode load instead of the usual fixed resistor. This modification has the advantage, when the amplifier has a capacity load on the output (such as the Y-plates) of producing an improvement in the frequency response and a greater maximum rate of change* of output voltage for a given power consumption. Briefly, the operation of the amplifier is analogous to that of a Class AB amplifier. If the input signal demands that the output voltage shall fall rapidly, the bottom valve (V23) discharges the load capacity on the output, while for a rapidly rising output voltage, the load capacity is charged by the top valve (V22). In the static condition, the standing current is 9mA, but the current load on the output is approximately 25mA – i.e. the maximum current V22 or V23 will pass when operating near zero grid bias.”
* I italicized this phrase for emphasis.
And so the circuit provides single ended push-pull action from a single ended input. I have seen this circuit described as Single-Ended Push-Pull or SEPP, this description is consistent with the Solartron explanation of the circuit action (unlike the SRPP designation).
The Y1 plate is driven similarly by a paraphase SEPP stage driven by a capacity corrected network from the first stage that results in a balanced output to the deflection plates.
The ability to rapidly charge and discharge the deflection plates is entirely consistent with my surmise, that the instrument is intended for examining fast changes. To further facilitate such usage, a 0.6µS distributed delay line can be switched into the circuit between the cathode follower and the input impedance of the main amplifier. Since the delay line is matched at both ends, it introduces a 6dB attenuation of the main amplifier gain.
If that wasn’t interesting enough, now we shall look at how the Y-system can be used to accurately measure signals or DC level changes. A DC voltage (that is derived from the regulated negative rail) is injected, which can be measured by the Y-shift meter that is of the centre-zero type i.e., it can measure in both the up and down directions. Since the injection takes place after the attenuator ranges, the meter FSD is equal to the range i.e. 1, 3, 30, 100V. It is used this way: Line up the top (or bottom) of the signal with the centre-line of the graticule and the using the zero control, zero the meter. Then using the Y-shift, shift the level to be measured to the centre-line of the graticule, read the meter and correct the reading according to the FSD value that is given by the attenuator range in use. It is a form of valve-voltmeter where you can observe the signal being measured.
Here it is showing the top side of a waveform with the meter zeroed:
And here it is with the bottom side of the waveform on the graticule centreline and the meter showing the (correct) amplitude of 20V p-p:
The X System is quite sophisticated as we shall see:
This comprises a trigger signal amplifier followed by a gate valve and differentiation transformer. The trigger signal is fed either directly to the time-base or via a delay circuit; the time-base drives the X2 plate directly while the X1 plate is driven by a paraphrase amplifier:
The trigger signal amplifier is a long-tailed-pair that enables a negative pulse to be fed to the gate valve regardless of the sign of the trigger input pulse. (It may be noted that the terminology used by Solartron differs from that used by Tektronix in that Tektronix refer to a multivibrator that initiates and shuts the off the ramp as the gate.) The gate valve has a transformer in the anode circuit that inverts and differentiates the trigger signal resulting in a negative-going pulse from the transformer secondary; a square leading edge is required for differentiation and so the circuit will not trigger from sine waves. The gate valve is a short-suppressor-base pentode; the negative-going time-base ramp is fed to the suppressor grid to cut the gate valve off while the time-base sweeps so as to prevent re-triggering during the sweep, the flyback releases the gate valve to trigger again. The next circuit the trigger signal encounters is either the time-base control multivibrator or a delay circuit; the schematic below shows the trigger amplifier, the gate and the delay circuit, the delay is described further down.
To permit observation of sine waves and other waveforms that do not have a square leading edge, the trigger amplifier can also by configured into an astable multivibrator by switching in a capacitor that couples a tap on the input side anode load to the output side grid. With no trigger, this astable multi oscillates at approximately 20Hz. This circuit behaves as a regenerative squaring circuit; receipt of a signal of any form on the input grid of peak amplitude greater than approximately 7V causes the multivibrator to change state at the same rate as the input signal resulting in a squarewave that is differentiated into a pulse in the gate transformer. Here is the schematic showing the astable configuration:
The time-base is controlled by a bi-stable (Eccles Jordan) multivibrator (that in Tektronix terminology is the time-base gate). This is because in one state of the output plate of the multi (here the low state) the time-base ramps (sweeps) and in the second state (high) the time-base discharges (flyback). The resting state of the multi is with the output anode high. When the negative-going trigger pulse arrives at the input grid of the multi, it turns the input valve off and the output anode goes to the low state. The low state is fed to the grid of the time-base switch valve cutting it off and the timebase capacitor (that is in parallel with the switch valve) starts to charge from the cathode of a bootstrap valve via a current limiting resistor. The bootstrap valve is so-named because the charge ramp is fed (bootstrapped) to its grid causing the cathode of the bootstrap valve to rise with the ramp thereby holding the potential across the current limiting resistor constant; this process results in a constant charging current thus a linear positive going ramp; this positive going ramp is also fed to the input grid of the multi (which is cut-off during the ramp period), causing it to cut-on at a pre-determined point on the ramp, the output valve now cuts-off changing the state to high that in turn turns the switch valve on, discharging the timebase capacitor (flyback) leaving the time-base ready for the next trigger signal. The positive-going ramp drives the X2 plate directly, thus the level on the ramp at which the multi changes back to the high state also sets the sweep length. The X1 plate is driven by a paraphase amplifier that creates a negative going ramp of equal amplitude to the X2 positive going ramp.
Delay (See trigger schematic above):
The trigger pulse can be delayed by switching a delay circuit into the gate circuit. The delay consists of a short-suppressor base pentode (6F33) Phantastron which is anode triggered by a negative pulse from the differentiation transformer. The Phantastron commences to discharge a capacitor in a linear fashion. The anode voltage at which the discharge commences from is determined by the fine delay control. The level at which the discharge ceases is determined when the potential between the anode and the suppressor grid becomes so low that the anode cannot draw through the suppressor grid the electrons that have passed the screen grid and so the remaining electrons return to the screen grid causing the potential of the screen grid to suddenly drop, the sudden drop is sharpened by a RC circuit to provide a negative-going trigger pulse that is used to trigger the time-base. The delay time is therefore determined by the discharge time of the capacitor; the larger the potential difference between the start and end points of the discharge, the longer the delay time. Thus the starting anode potential controls the delay period. Coarse delay control is affected by switching in different values of capacitor. There are 3 ranges, 100µS, 1mS and 10mS.
This allows the time-base to be swept very fast in relation to the signal under examination and by varying the delay between the signal and sweep initiation, any point on the signal may be slid into view for greatly magnified examination.
(Tektronix introduced a calibrated continuously variable delay so that the end of an event that is in effect many screen diameters away from the start can be slid into view and the time between the end points measured by the difference in starting delay and ending delay. They did this by using a second time-base (delaying time-base) to delay the primary (delayed) timebase. An X magnification system such as HP introduced with the 150A, cannot provide event time measurement unless both ends of the event are visible on the screen together.)
The time calibrator is equally unusual: Instead of providing a marker generator as was common before the calibrated time-base was developed (by Dick Ropiequet working for Tektronix), Solartron here provide a simple resonant circuit. An L-C tank of known frequency is placed in the cathode circuit of a triode that is repeatedly stimulated by a “ping” from the trigger. Thus when the time-base starts, a damped oscillation of accurately known frequency appears on the screen. The available frequencies are 1Mhz, 0.1MHz and 0.01 MHz corresponding to periods of 1µS, 10µS and 100µS. This is used to set the sweep rate of the time-base. For example if the time-base is adjusted such that two cycles of the 100µS oscillation occupy the 2″ graticule then the sweep rate is 100µS/inch. The system can now be used to measure event periods are 200µS duration down to perhaps 1/10 of that.
Here is the 100µS oscillation occupying 2″ on the graticule:
Finally, two pictures showing the damped oscillation: