I’ve been putting this off for a while, but after a series of easy wins, I thought maybe I was up for the challenge. I have had in my possession for a number of years, a decent collection of analog sampling plugins and accessories from Tektronix, including 2 digital readout units, which allow for automated voltage and time measurements. Getting those readout units working is the goal, but first I have to start with a reliable sampling system.
I’m doing all this on a tektronix 568 mainframe, which has all the connectivity for the readout and automation systems, but these plugin pairs would work just as well on a 561 or 564 if you just wanted basic sampling functionality.
First I’d like to catalog the state of all of these plugins. I think the majority of them don’t work, so I want to find the units that work the most, and start from there. If I have a working pair, it will help evaluate the other plugins, as sometimes it’s hard to tell if the fault lies in the timing unit or the sampling unit.
3S76 – Dual Sampling #2443
This one seems to work on the A channel. I have had it successfully send a trigger to the timing unit, and was able to position a signal on screen. The B channel is not working. After some time, I do get a sweep on screen for it, but I can’t position it. Getting the B channel working might be a good step in understanding how repair other sampling units. More on the repair below.
3T77 – Sweep #2810
This seems to work. Needs trigger sensitivity up pretty high, and it’s not what I’d call rock solid, but so far it seems like the best timing unit of the lot.
3T6 – Programmable Sweep #110543
This also seems to work well. External trigger provided on BNC J123 on the rear of the 568, via the gremar connector on the plugin. According to the Wiki Entry, it’s a Mod 651C based on the 100/1000 samples / sweep switch, although it’s not marked as such. It needs a few indicator lamps, Tek part 150-0057, manufacturer part # CM8-725.
A entry on Archive.org would have me believe it’s a 5V lamp. It’s being driven via an SCR by the 6.3V filament supply, so it’s only on 50%. They’re ~ 2.25mm diameter x 4.5mm long. I couldn’t find the exact match, but I’ve ordered a few variants and we’ll see if any of them work.
3T77 – Sweep #2250
3T77 – Sweep #0881
3T77 – Sweep #2246
3S5 – Programmable Dual Sampling #40199
This seems to work on the A channel with the 1 known good S2 sampling head I have. The B channel is not working
Side note: the channels are lettered on these early sampling scopes, and numbered on traditional scopes. No idea why, but it holds true on the 661, and all of the 560 series plugins. On the 7S14, it’s back to numbers.
3S5 – Programmable Dual Sampling #30151
A channel works, with internal triggering. The B channel is not working.
3T2 – Random Sweep #30320
3S76 – Dual Sampling #1209
Doesn’t work. With normal/invert switches set in the middle (a troubleshooting step in the manual). I get a sweep that can be moved vertically on Channel A, but not on Channel B, suggesting there are a few things wrong with this unit.
The sample diodes seemed to pass a multi-meter diode test, for what that’s worth.
Update: I stole a Nuvistor V1073 & transistor Q2163 from this to repair #2443
3S2 – Dual Sampling #40267
Both channels work, although I wasn’t able to get internal triggering working. Unfortunately I only seem to have one working sampling head at the moment.
I set out to get Channel B working again. Worked through the troubleshooting outline in the manual, which starts out checking the sampling gate pulses coming from the timing unit – obviously this wasn’t the problem, but I figured I’d play along.
Getting some of the measurements in this plug-in requires patience; in some cases, I’d have to turn off the scope so I could get probes in place without fear of shorting anything. In a few cases, I had to use some helping hands to hold probes in place so they weren’t placing any excessive stress on the component leads. This thing is a complicated, layered, 3 dimensional puzzle, with some fragile assemblies.
Triggering was fine, so I went on to measure the gate voltages. Channel A was fine, B was not.
This led me to suspect the sampling diodes, and the B channel had one diode that didn’t make my Fluke 87V beep in the diode check. The numbers were good, in terms of voltage drop, forward and negative resistance, but it just didn’t beep. Turns out those diodes were fine, but I have no idea why that one diode didn’t trigger the beep. If I was more ambitious, I would have lugged out the curve tracer.
I swapped diode sets from A to B, and B still didn’t work (and A still did). Reading a little deeper into the troubleshooting steps, it talked about the Nuvistor immediately following the sampling gate as a next possible trouble. V2073 was indeed the issue, and swapping it with one from what’s quickly becoming a parts unit gave me a trace on both channels.
In going through calibration, I discovered another issue – B channel would only show a trace on the screen when in inverted mode. Checking the output signal on the front banana jack showed me that it was pinned up around 20v, regardless of position or DC offset controls. Looking at the schematic, I saw that there’s actually an inverter that gets switched in when in ‘Normal’ mode – so the bulk of the signal path is inverted, and this just flips it back near the end. One of the germanium transistors in that inverter was bad, so I grabbed one from the parts unit, and I’m getting a trace in both Normal & Inverted settings.
This one always worked
This one was marked as ‘bad, shorting’ by my friend who gifted it to me, and I remember experiencing the same back when I fired up this rig a few years back. I opened it up, took it apart and put it back together, and it’s working now. I think one of the boards was sticking up a little too much and it’s pins were making contact with the body.
Marked as Bad, doesn’t work. I tried swapping the strobe & preamp boards from #93650 to no avail, which leads me to believe it’s the sampling diodes – there’s really not much else left on that central board.
S-2 # 40345
Marked Bad, doesn’t work.