Old school Tek porn

[jh15]

Thanks for keeping the Tek Porn thread active. At one point I was going to part out my Tek sturf on ebay to help out collectors/restorers, but now I think I will clean up my 535, in perfect operating condition but icky inside, figure if I should keep the Hickok Tek clone for parts, sell it, or restore it from my 535,. Also the 535 era thing I can't make out on the top shelf in my garage.. has tunnel diodes, maybe pull them (I had to design a tunnel diode oscillator in college) and play with them.
I might have a 545 kicking around...

[GK]

More porn. I decided to re-cap and calibrate my Type 127 power supply unit today. The 127 allows a pair of 500-series plug-in units to be used independently of a mainframe oscilloscope. For example a vertical amplifier plug-in can be used as a stand-alone measurement/laboratory amplifier.

The signal path in the 127 is fully differential for both channels and the output signals from the respective plug-in amplifiers are provided via PL259 connectors on both the front and the rear panels of the unit. The design is a bit weird though in that the 12AT7 triode parallel-connected cathode-follower pair for each output does not deliver a low enough output impedance to terminate a 50 ohm coax. The 127 signal amplifier outputs were therefore specified to drive 170 ohm terminations. The manual for the 127 specified a Tek part number for the 170 ohm terminators but says absolutely nichts about the type of coax to use. My Type 1121 amplifier (showcased earlier in this tread) drives 93 ohm terminators via RG62A coax, which has a characteristic impedance of 93 ohms. 

I don't have or know of any coax type with a 170 ohm characteristic impedance, nor have any 170 ohm terminators, so for the time being as a temporary (maybe) compromise I calibrated my Type 127 to deliver the correct signal gain and optimal pulse response into a 1m length of 50 ohm coax unterminated. The 127 terminated into 170 ohms is supposed to give a rise time better than 20nS. Testing my 127 with my TU7/1M1 Pulse/Load Calibrator plug-in, which delivers a ~3nS rise-time high frequency square wave, the best I can tweak into my 50 ohm unterminated lengths without over shoot and with a flat square wave response is in the order of ~48 nS.

Anyway, the 127 comes in handy when you need more traces on your single plug-in mainframe, as shown in an attached photo. Here I have a Type CA dual-trace plug-in inserted into my Type 545B mainframe. A pair of 1A1 dual-trace plug-ins inserted into the Type 127 provide signal inputs for four independent signal traces. I used my analogue chaos computer as a signal source and all of the plug-in amplifiers are operating in the "alternate" dual trace mode.

The four independent traces on the 545B are a bit tricky to set up though. You actually have to use both timebases of the mainframe oscilloscope. On the rear panel (why not conveniently on the front panel too I don't know) the Type 127 is provided with a binding post for each plug-in amplifier which serve as the bistable trigger inputs for any dual-trace plug-in amplifier (Type CA, 1A1, etc) which has the "alternate" trace function.

In the 545B mainframe, this trigger input of the plug-in amplifier is internally wired to the "+ Gate A" signal output binding post. This signal is a pulse that occurs at the end of each sweep of Timebase A; so when you set the plug-in amplifier to operate in the "alternate" mode, the plug-in's signal output is alternately toggled between CH1 and CH2 at the termination of each horizontal sweep.
Now here is where is gets to be a bit of a pain getting 4 traces to display with a pair of external dual-trace amplifiers. The external plug-in amplifiers, for obvious reasons, must not alternate at the same rate as the internal plug-in amplifier, but at 1/2 (ideally) or less the switching rate. To achieve this you connect the external trigger input for Timebase B to the "+ GATE" output of Timebase A, and set Timebase B's sweep duration to double that of Timebase A. The "+ Gate" output binding post for Timebase B is then wired to the trigger input binding posts at the rear of the type 127 (this is the yellow wire in my set-up photo). This basically sets Timebase B to operate synchronously with Timebase A, but at half the frequency, and this half-frequency "+ Gate" pulse from Timebase B is used to toggle/alternate the external amplifiers. Timebase A is obviously set to be the displayed timebase.

Because this, in addition to accessing the trigger input binding posts at the rear of the 127, is a pain, I'm currently building a 1U 19" rack patch panel, to breakout the trigger inputs, which will mount in the rack either above or below my 127. This patch panel will also incorporate a bistable/toggle flip-flop and driver circuit to drive the 127's trigger inputs. This will provide the half-frequency alternate trace switching function, thus eliminating the need to use Timebase B at all and permitting the same trace number expansion method to be used on oscilloscopes without dual timebases (for example a 6-trace setup with my 551 mainframe). 

[GK]

Okay, a prototype flip-flop knocked up on the breadboard and I have six traces on the 551. In theory, if I had a pair of 1A4 plug-ins in the 127 and another in the upper beam slot of the 551, I could have as many as twelve traces.