开云体育

The R5030 is a very unusual scope, ti has the large CRT and fiber-optic display like the 576 curve tracer, and
the lighted function switches from the 7K plug in series, along with amazing sensitivity down to 10uV/division. yes, microvolts. Kind of a 7A22 merged with a 576 and a 5K rackmount.

I got three repair candidates from Dennis (to fulfill a long time desire), and was able to get one fully working, the others have too many fatal flaws to be anything other than organ donors. I found many strange problems, and was amazed to discover that on all 3 units, the blue covers had been installed incorrectly, (differently!) blocking the scope's required ventilation. In the one I repaired, the output stage got so hot, the power resistors actually de-soldered from the pcb.

Once re-organized, the covers work correctly, and the scope runs OK, although I have to say, like all 5K systems, it just gets TOO hot. The tricky part was getting the fiber-optic display up and running again. In my case, every display was bad, with so many pixels gone, nothing was readable.

The manual pretty much ignores the display in every way, there's no info on it at all other than a mention of the lamp types in the parts list (150-0048-01 / 683 5V T1, flying lead types). The big question was, how the heck do I get to the display to fix it? It is buried behind the panel, with no obvious way to remove it.

Thanks to some list members who have worked on the 576, the answer was to remove the CRT bezel and all the overlays (there are three, remember the order!) to get at the front of the display itself. The display then pulls forward (no hardware), and can be extracted if all the mating harmonica cables at the rear are removed. WATCH OUT, some are smaller than the headers they plug into, be sure you know how to align them when replacing the cables! All should start at the right side (viewed from the front), there will be a few exposed pins to the left on same middle cables.

Once out, the display module has three covered assemblies, remove the aluminum covers (two screws) to get at the lamps and fibers below. I REALLY wish they had used a based T1 lamp, but every lamp has to be hand soldered in place in a tiny board covered with interconnect wires. not the best bit of production design at Tek, that's for sure. The black assemblies pull to the rear to expose the lamps.

Be careful of the brittle fiber optic light pipes, they run through the lamp assemblies and could be easily damaged if you get impatient. gently pull the black light assembly free of the metal mousing, and expose the rear board. you will have to cut some cable lacing to do this, it is too tight otherwise. My technique was to use a pair of surface mount test tweezers hooked to a bench power supply (set to 4.9V, and c/l of 0.12A) to probe each set of lamp contacts, there was no other easy way to make both connections at once. Once side of the lamps is common on the boards, the other goes to a wire that heads back to the connectors. Polarity is not important in probing the lamps.

As you hit each lamp, you should see it light (even from the rear), I found one lamp installed that was the wrong type, it drew way too much current, and was dim (good lamps will draw about 50-60mA)., I also removed any marginal lamps. it took me about a dozen lamps to restore my assembly. You have to carefully de-solder each lamp (don't burn the wires), clean the holes, then tread in a new lamp, being careful not to twist the leads so they short. Once fully in and centered, re-solder the lamp. Test each lamp when done, you don't want to have to go back in this thing again.

Once all the lamps in any section are all good gently re-seat the black base into the metal housing, be careful not to damage any fibers or pinch any wires. Once a full display is good, re-install the metal cover (2 screws), being careful not to trap any wires between the cover and the chassis (this can happen very easily). Once all three displays are done, the entire display board can be re-slid into the fitted slot next to the CRT. The last step is to re-connect all the harmonica connectors in sequence, being careful to align them properly (un-connected pins will be to the left).

Adjusting the display dimmer (below the two intensity controls) will adjust both the CRT graticule light and the displays. with luck, all the digits and symbols will now illuminate correctly!

If anybody wants to see the pics of my completed unit let me know I can email them to you. I was able to transfer the Option 04 desktop carry kit to my unit, and remove the rack parts to make mine into a proper bench scope. The final result is very nice looking, and operation down to 50uV/Div works very well, below that noise can be a real issue. Operation at 10uV would have to be into a very low impedance via a short shielded cable. They were maybe optimistic on that range.

All the best,
walter (walter2 -at- sphere.bc.ca)
sphere research corp.

Hallo Walter,

I have been looking for an R5030 or 5031, just to get the 10uV sensitivity. I got the taste of this, from my HP130A oscilloscope, which can do 200uV differential, but absolutely humfree and noisefree. It has only differential inputs.

Most people sometimes find differential inputs annoying, but they're the only way to measure free of hum at such low level. To the contrary, that works perfect and very convenient. I can use a normal probe cable from a multi meter, and measure absolutely hum free at 200uV/Div, provided I connect the negative differential input to ground of the device under test. I find that easier to use than a scope probe. This made me realise, that with a non-differential measurement, using BNC cables, the hum you see on a scope at levels of 1...5mV per division is nothing but the capacitive current between the scope chassis and device under test chassis.

With that in mind, I have been checking what the Tektronix 5A21N module can do, because it's best sentivity is 50uV. Which is quite good in theory, and 4x more sensitive than the HP130. Of course I can ground one channel with the push button, and it becomes a single channel BNC input. But then it hums so much, I can't use it at 50uV. When using it differential mode, hum is almost gone. Only I get stuck with almost half a division high frequency noise. Even with the inputs shorted. They must have noticed that at TEK, because there is a bandwidth limiter at 10kHz, and then the noise is almost gone. But all together that does not beat the HP130. Which does have 4x less sensitivity, but it will do so up to 250...500kHz and fully noise free. The trace is sharp and no hum. Whereas the 50uV of the 5A21N only makes sense when I measure at full scale. For small signal it makes not much sense to me. But you see.. that's why I was trying to use it.

So I wonder if the 10uV of the 5030 is just only for the catalog, or if it is of any use at all. You say it doesn't work well. Was perhaps in single channel mode, with a BNC cable attached? Because with 5A21N that doesn't work either. It does work well however in differential mode and no ground connected .

Have you looked at the 7A22 amplifier? It is a single-channel differential
amplifier that goes down to 10uV/division. 1MHz bandwidth with switchable
LF (down to DC) and HF limits. Have a look at the TekWiki page:



Sorry if you know all this or have tried this in a 7000-series mainframe
and it does not suit your application.

Steve H.

I know 7A22 exists, but I have at this moment the 5000 series mainframe. And I do have to say the 50uV range of the 5A21N with 5000 Series is dissapointing, because it doesn't work well unless the bandwidth is limited to 10kHz. And even then it is not very silent. So going to a 10uV system even, I am a bit reluctant to hunt for a 5030 or 5031, or change to the 7000 series just to find out the same.

The HP130A goes down to only 200uV, but at 250kHz without frequency loss, and really DEAD silent. If there is no hum or noise on the DUT, and at zero signal, the scope produces a hair sharp horizontal line. The 5A21N with 5000 Series at 50uV, without bandwidth limiting, you can not use it. With bandwidth limiting, there is still some small noise and hum. It's not much, but when measuring very small signal, very small hum and noise is not what you want. At zero signal there is still hum and noise visible, which comes from the 5A21 itself. So even when sensitivity is 4x lower, the HP130A works effectively much better. Because it is dead silent by itself, even when I use it with a normal multi meter probe cable, there is no visible hum at 200uV range. I need no probe, and there is no problem where to connect the shield clip to. Which is very convenient. If you ever shorted a high voltage circuit with the shield clip, you know what I mean. But it works also better. From this I know it is possible to measure like this, and I really like that method. It works actually with the 5A21 also, but not as good as with the HP130A.

I am afraid with R5030/5031 or 7A22, there will be no significant improvement over the 5A21. I quote this from Walter Shawlee's initial post about the 5030. "...operation down to 50uV/Div works very well, below that noise can be a real issue. Operation at 10uV would have to be into a very low impedance via a short shielded cable. They were maybe optimistic on that range".

Though I see the issue here perhaps just resulting from the way this was measured. I think with a normal probe and BNC cable, you can not use it at 10uV. Even when measuring against ground, at 10uV you have to connect the ground of the DUT to the negative input of the differential amplifier, and not to the scope ground. When I try that with the 5A21N, it fills the whole screen with hum. But when I connect ground to the negative input, the hum falls back to just a few uV. Perhaps Walter can say something about this.

Sorry, I called it HP130A, but it HP 130C. .

Isn't there a 5A22N also?? ?HTH.? ? ? ? ?Jim Ford?Sent from my Verizon, Samsung Galaxy smartphone

Jim, that sounds worth a try. If by any chance someone has a 5A22N for sale, please let me know.

It's on my wishlist, although I already have a (mostly) working 7A22.? Would be cool to have a 5A22N for my 5103N D10 mainframe, though.? Good luck finding a 5A22N.? Might see one on the 'bay or even a spare mainframe with one in it if you're lucky.? ? ? ? ? ? ?JimSent from my Verizon, Samsung Galaxy smartphone

Good Day Jac et al,

ok, so the conversation related to the 5A22N's inherent noise made me curious. As I do have one, I ran a few, very simple tests...

I have created a new photo album for a few photos of my crude testing: /g/TekScopes2/album?id=267951
Note that I have terminated both inputs of the diff. amplifier with 50 Ohm termination resistors. And the test results are basically a tad disappointing, as discussed.

DUT: Tektronix 5A22N serial number 101 592 in a D12 / 5112 Dual Beam mainframe.
With 50 Ohm termination resistors connected, the 5A22N's inherent noise was found as follows:
VOLTS/DIV setting: 20?V/DIV, -3 dB bandwidth: 1 MHz, noise was approx. 2.5 DIV pk-pk
VOLTS/DIV setting: 10?V/DIV, -3 dB bandwidth: 1 MHz, noise was approx. 4.5 DIV pk-pk
VOLTS/DIV setting: 10?V/DIV, -3 dB bandwidth: 300 kHz, noise was approx. 3 DIV pk-pk
VOLTS/DIV setting: 10?V/DIV, -3 dB bandwidth: 100 kHz, noise was approx. 2.2 DIV pk-pk

The photos are showing the settings first, followed by a photo taken from the screen.
Please excuse the low quality of the photos, this was intended as a very simple test for illustration purposes only.

Cheers,

Magnus

P.S. I have two 5A22N plug-ins for sale - just contact me off list if you are interested magnustoelle /@) yahoo.com
I am located near Munich, Germany, so the high costs for shipment outside of Europe would make this inattractive.

Back in early Feb 2021, I had one of my R5031 scopes in my temporary shop at CenterStaging, LLC in Burbank, CA, needing the storage mode for working on a Vox AC30 with some Tremolo speed issues. I hadn't had this R5031 powered up since early 2001. Got it up and running, and found it too had some of the lamps burnt out. Tek listed those as selected type #683. I found Chicago Minature CM6833 lamps at Mouser for $1.99 ea. 25,000 hr lamps 5V 50-60mA. Ordered ten pieces, installed them a week later. I took photos and posted the maintenance session on that scope for restoring the display on www.music-electronics-forum.com, link is here:



Shortly after that was done, the scope died on my. I now have it open on the bench and in the process of replacing some of the Sprague Cap-Can's on their Rectifier/Capacitor board...retrofitting today's much smaller radial lead Electrolytics in place of those parts that were used in the 7000 series and this unique 5030/5031 scope.

I found a couple odd parts on the middle wall separating the Input Diff Amp circuit boards and the CRT/Power Supply area. Not listed in the manual, but are on that separating wall are a pair of Germanium T0-3 PNP Power Xstrs Tek # 151-0165. They measured normal for a Germanium, but oddly, I found while checking emitter-collector, I got a reading in one direction (DMM in Diode Test 1mA current source Fluke 8060A). I haven't found them on the schematic nor in either the R5030 nor 5031 manuals.

Steven
nevetslab

Ge power transistors have a very high C-B leakage, so it’s not at all uncommon to see conduction even withe the base floating—the base is getting current from the collector leakage.

Try shorting base to emitter. I’m willing to bet that the “diode” current you were seeing goes way down.

Cheers
Tom

Sent from my iThing; please forgive the typos and brevity

Hi Tom,

That was it! That reading, looking between emitter and collector went away when I shorted the base to emitter. But, surprisingly, when I reversed the DMM leads (to Emitter and Collector), now that direction was giving that reading...until I disconnected the B-E short.

Thanks,

Steven

Germanium devices certainly are fun. :)

There are a lot of hobby circuits from the 50s and 60s that ac-couple to the base and have no apparent way to get bias current into the base. Whether or not the designers knew it, these circuits relied on self-biasing through collector-base leakage. If you substitute a modern silicon transistor, those circuits just won't work unless you do something like add a collector-base resistor.

Silicon has spoiled us. And that's a good thing.

--Cheers,
Tom

--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070