-
greetings EEVbloggers,
I'm new here, but, this being the internet, there is little stopping me from diving right in and making a fool of myself.
I have recently developed an affliction (not a typo) for Tek 5*** equipment. These units can be thought of as the little siblings of the popular 7000 and TM500/5000 frame systems. They also share a great deal of technology with the awesome 4** portable scopes. Produced for two decades, ending in 1991, the 5000 series occupied Tek's lowest price point in the market. Indeed, the 5A24N plugin, at $25 1971 USD, has the distinction of being the least expensive Tek product ever.
The low bandwidth (2MHz) 5100 parts were targeted to dedicated industrial applications, and their D1* displays were sold to various third party OEMs for integration into their systems. The 5400 parts (50/60MHz) were more general purpose low end instruments.
There were nearly 30 different plugins produced for the series including a curve tracer, an audio spectrum analyzer, and some quite sensitive differential vertical amps, as well as some fairly nice dual/delayed timebases.
I first became interested in these systems because I already had a network analyzer that used the D10 display, and I was looking for a new display unit for my home-built W7ZOI spectrum analyzer. I had been viewing the spectrum analyzer on an Hitachi V-209 (legendary sharp trace, but dinky 3" display). The non-storage variants of the D1* and D4* displays are nice 6.5" units, so I figured a 5000 series scope in XY would make a great display for the spectrum analyzer with the added benefit of sharing parts and "know how" with something I already had.
One charming aspect of these systems is that they can be configured either as "benchtop" or rackmount if you have, or can fabricate, the appropriate mechanical parts. This appealed to me as my lab is largely built around the RS310C rackmount form factor. Another advantage is that the series has been mostly, so far, free of the ravenous vintage Tek mania that has driven up the prices of the similar systems.
There are not a lot of resources specific to the 5000 series on the web. One very nice overview has been put together by Christian Weagle at http://www.i9t.net/5000_scopes.html . I have attempted to contact Mr Weagle to no avail, and his page was last updated in '07 so I fear he may have gone SK, or at least SKbd. I have saved that page locally against the day it goes 404.
There is also a useful writeup by Peter at tekwiki http://w140.com/tekwiki/wiki/Introduction_to_the_5000-Series_Oscilloscopes .
Repair and restoration information specific to the 5000 series is rather scarce, so it is well that they share so many parts and architecture with the other Tek products of the era. Much of use can be gleaned from 465, 485, 7603 etc. threads.
I am creating this thread as a place to seek assistance from the community in my personal repairs, with an eye toward leaving something of a resource to those who come after. Several of your threads and handles (I'm not going to name names for fear of leaving someone out) have become very familiar territory in recent months as I have been lurking about, and I hope to collaborate in leaving something of use. -
So, to bring us up to date;
The first one I got in the shack was a 5440, so, in the early parlance, a 5403 acquisition frame with a D40 display, in the rackmount configuration. It was an Ebay offering and came loaded with a 5A18N dual trace vertical amp, a 5A21N single trace differential vertical amp and a 5B12N dual/delayed timebase.
A couple quick notes on bandwidth and nomenclature; 5 is obviously the series, the letter indicates the plugin type. The suffixes are somewhat random but there is some convention. Suffixes starting in 1 or 2 are low bandwidth plugins, 2MHz or less, while suffixes beginning with a 3 or 4 are intended for 5400 frames and are the 50/60 MHz units. System bandwidth is, of course, limited by the weakest link, so in this case I had a high bandwidth frame loaded with 2MHz plugins. Though I find the 5B12N quite nice, more on that later.
The Ebay listing classified it as the all too familiar "for parts not working" (translate as "whatever it is you're stuck with it") but the pictures showed it displaying its 400mV calibrator square wave with a decent looking trace. The details elaborated "may have noisy or unresponsive controls". Well, that sounded like precisely what I was looking for from a project perspective and the price was, I thought, reasonable so on it came.
It arrived safe and I can not fault the seller. The controls were noisy indeed, and the entire center plugin compartment was unresponsive...fair enough considering. All LV supply rails were dammed close to bang on, with acceptable ripple, at least for now. The better part of an enjoyable day with wee paper strips and IPA, both the isopropyl and India pale varieties, and all three plugins were working nicely, albeit only in the outer two slots...on to the frame.
The schematic for the interface board is on pages 124 and 125 of http://w140.com/tek_5440.pdf . With a plugin inserted and enabled the beam finder revealed the missing trace hard up, normally off the screen, the position control had no effect. Reseating u620 did not improve the situation. While a fault in u620 remained a possibility attention turned to the driver pair Q610 Q614. A quick probe revealed that Q614 contained no diodes, open every which way, patented not-a-transistor. the parts list on page 101 list these as "sel from" mps6521, so either binned components or a matched pair, I'm guessing the latter. an NTE cross search fingers the 6521 as...wait for it...NTE123AP, a bog standard 2n2222, the most ubiquitous npn transistor ever conceived. I have a batch of motorola 6521s in a saved shopping cart and some day I will get them here and do some matching, but for now a quick trip to the silicon bin and the center bay came to life, oh the joy!
Here is that configuration displaying a 2MHz RF signal and its line frequency calibrator
And here it is in X/Y mode doing what I bought it for, this is the local FM broadcast band.
-
Well, even though the setup above performed its intended function admirably, as a TEA sufferer, the idea of a 60MHz frame stuffed with 2MHz plugins was simply unacceptable. Soon a new Ebay listing appeared which was quite intriguing. The frame was listed as a 5443. Now, there is a 5440 (single-beam non-storage), a 5441 (single-beam-storage) and the dual beam 5444...but nowhere that I know of in the old Tek catalogs was there ever a 5443. Indeed, at the time there were a grand total of ZERO google returns for "Tektronix 5443" save the ebay listing in question. The photos clearly showed "5443 oscilloscope" printed on the rear panel and the display was a D44, the same display as the rather uncommon 5444. further, the unit was stuffed with 2 5A48 dual trace amps and a 5B40 timebase. I simply had to have it.
In sober reflection, I was probably a bit hasty. The auction was again "for parts not working", but this time the AC cable was clearly cut...never a good sign. Also the fuse cover was missing. Further, it was in over/under "benchtop" configuration less than ideal for my shack. The unit seemed to be in good cosmetic condition, however, and was loaded with just the plugins I needed so the die was cast.
It arrived safe and I loaded the plugins in my working frame straight away. Annnnnnnnnd, my frame failed to power up. I quickly determined that one of the 5A48s would prevent the frame powering if inserted in any slot so it got set aside.
One of the things I was really eager to see was the on screen display offered by a combination of high bandwidth frames and non "N" plugins. With the frame powered up with the two working plugins the behavior was quite odd. Only a few sweep rates seemed to work. Some were just blank. Others would have the same sweep rate for several adjacent settings. The timebase produced no OSD whatever. One channel of the "working" vertical amp produced no OSD, and the other, bizarrely, was calibrated in...seconds. Touching any of the red variable knobs resulted in horrible noise or complete loss of trace, as did many buttons. A trip to the store was in order as I was not in possession of sufficient stocks of one of the IPAs for the task ahead.
Fully disassembling the attenuators and variable controls in these is fiddly work. Extreme care must be taken to not damage the delicate gold contacts or break brittle 30 year old plastic parts. A few bullet points on this;
>the markings on the buttons will come off with ANY fluid, and I mean ANY! IPA, water, sweat on your fingers, touch these with any moisture whatever and you will have a smudged or blank button. You have been warned. I don't know if they were like this from the factory or if it is somehow a factor of age, but it is certainly the case now. One member, robrenz, has resorted to engraving the tiny buttons on a CNC mill and filling the engraving with black epoxy which I have to say is just top notch.
>the gold contact pads and fingers clean up just like the 4** and 7*** stuff; use ONLY isopropyl and paper, no abrasives whatsoever. No, not your special stash of infinity grit sandpaper, just don't.
>I found the best way to remove the plastic switch extensions was with an Xacto knife and a very fine slotted screwdriver. Use the knife to gently pry up the tab while using the screwdriver to lever the extension away from the switch shaft. I have not broken one...yet...but it is pretty scary.
>the buttons themselves are best left on the shafts. They are sometimes quite tight and I marked the sides of a couple with the pliers before I figured it out.
>my method for cleaning the switches is to drip some IPA in with a needle applicator while operating the switch continuously followed by air in a can to force the IPA deeper and blow crud out, still operating the switch. I do this operation twice and then finish off with a very small squirt of MG Chemicals "super contact cleaner with poly phenyl ether" the stuff with the gold band on the can. Mr Carlson turned me on to it. The Caig might be just as good, but the MG product is impressive. This procedure has treated me really well.
> for the "uncal" variables operated by the red knobs I first carefully removed the aluminum operator shaft and coupler, made sure the pot was hard CW with the cal switch engaged, then loosened the four screws until the pot came away from the switch body. I could then perform the same cleaning operation as above on the pot before cranking it hard CW again and re engaging it with the switch. All of these were VERY noisy. They are a pain in the ass.
Cleaning up the controls straightened those two plugins right out. I had a great time putting them through their paces and marveling at the operational OSD, but the empty bay was bothering me so attention soon turned to the misbehaving vertical unit.
Repeat after me; it is always a capacitor, well, not always, but pretty dammed often. In this case the -15V rail in the plugin was dead short to ground, and here is the bit of wisdom that many of you will already know; the tantalum drop electrolytics used in this era of Tek gear have a bad habit of failing short. It took a good bit of tracing, schematic studying and lead lifting to finally track the fault to C153 a bypass cap between the emitters of Q150 and Q155 in channel 2. A quick rustle in the cap bin and all was made right.
Here is that configuration triggering on a 100MHz signal
and here it is with the 5B12N back in displaying four signals in two timebases, note that the 5B12N is triggering on a signal over three times its rated bandwidth, verrrrrrr nice /borat
-
I make use of the 7000 series but the 5000 series has some interesting capabilities including a variety of differential plug-ins, the 4 trace plug-in, and a better low frequency spectrum analyser.
-
Thanks for stopping by David, you are one of the folks I was hoping would notice, for sure.
At this point I was feeling pretty pleased with myself. Like Midas, everything I wave my soldering iron at turns to gold. Time to pull the covers off the mystery frame. It had a sticker on the back that indicated that it had been in a medical school environment, hmmmmm, Weagle does mention that these were sometimes used in EEG setups. Perhaps the 5443 was only sold to institutions or 3rd party OEMs and this explains its absence from any retail catalog. Also on the back was the "option 1" marking, uh-oh. Option 1 was a money saving option that deleted the readout board, so no OSD in this one.
Off with the covers and, whats this? Someone had glued thin acetate sheet over nearly every perforated section of the panels, leaving only a tiny bit of the bottom front as an intake. Pretty sure Tek didn't do that. Something to do with the medical environment? Jeez, I hope I don't catch ebolaids from this thing. The fuseholder was broken. Replacing that and getting a temporary power cord hooked up was rather tight work down next to the HV board. While the HV safety cover was off to enhance that access I probed the .6A slow blow fuse there and it was open. I did not have that in stock, but snapped in a regular .5A. I also noticed some rusty corrosion around one of the mounting screws of the right side vertical board that looked like it might short to a trace, so that got pulled off and cleaned up.
Moment of truth, installed my least favorite 2 plugins and pulled the switch. A quick safety aside before I continue. My father happened to be visiting for this part, a fine troubleshooter himself but with only a passing interest in electronics, and my partner was present as well. The point being, I personally do not like, and do not recommend, working on high voltage without other folks around to yank the cord should the bad stuff happen, not that it ever will because we all have great respect for, and take great care when working with dangerous voltages...RIGHT? Sorry to belabor the point, but yeah, the CRT anode in these things is -3KV and the accelerator is no less than 12KV! so as Carlson says "if you're following along at home, just take care."
Anyway pulled the switch, and heard a raspy buzzing sound. Shut it down and had a think. "Could be the fan" dad said, well, yes it could be. Pulled the fan cover and jammed a pencil in to stall it and tried again. Still buzzed, shut down. Hmmmm, I don't much like where this is going. As I started to remove the HV safety cover my GF, who had been hovering about for this part, was like "What are you DOING? STOP. Recap that shit, you're going to break something". Yeah, she's a keeper. I am a stubborn individual, however, and removed the HV cover and procured a length of dielectric hose. The plan was; we would power up one more time and I would diagnose the display from the front while dad listened around the HV supply with the hose. He did not need to be told, but was anyway, to keep his hands WELL back from the supply.
OK, deep breath, here goes. I really did not like the buzzing, but in a few seconds there was life in the display. The beamfinder allowed me to get both traces in pretty quickly. There was some sweep, though not full width and one of the traces had a large blotch always stationary on the screen regardless of trace position or sweep. The blotch could be made flat or round with the focus control. The intensity controls had no effect. About this time the buzzing, and one of the electron guns, stopped simultaneously. I shut down and hung my head in shame. The GF rolled her eyes and withdrew, and dad informed me the buzzing had been coming from T410.
Sorry, no pics of the display, this all happened in about 20 seconds.
Several days intervened with the project sitting there on the bench as a constant, depressing, reminder of my foolishness before I mustered the motivation to do what I should have done in the first place.
Jones' first law thou shall check voltages. I got everything except the interface board, which is soldered, disconnected from the LV supply and got a fused mains jumper on to it. The rails were all over the place. the 200V unreg was more like 245. -30 I could only get down to about -34. +30 was hard against its unregulated source rail, which itself was high at +42, and the adjustment pot did nothing. By moving the mains jumper to the high input position I was able to get -30 in spec, just, but +30 was stuck at 38 and wouldn't budge. All the components in the +30 regulator circuit checked out good.
These are my weapons of choice for this sort of decapping work. Sure, a real desoldering station would be better but I don't have that. The short 14ga copper wire as a tip in the Weller is another Carlson trick. I tell you, that gentleman has it going on! The iron gets super hot, like RIGHT NOW!
So I got the caps off it without any real drama, cleaned up with some RA flux, solder braid and a more civilized iron, not as beautiful as the work of some here but usable.
And let's see here 3117, that's a little low, wait, what? nano?? and ummm .14 uhhh Kilo?? oh jeez, I think we have a winner folks
-
Good write up and enjoyable read. WT.
Following along at home............
QuoteAs I started to remove the HV safety cover my GF, who had been hovering about for this part, was like "What are you DOING? STOP. Recap that shit, you're going to break something". Yeah, she's a keeper.
You lucky guy, don't let her slip away.
-
Darn new age young 'uns... afraid of few measly kV. Your index finger is the best HV probe ever invented.
Boldly probe where no man has probed before. And I bet they don't even know the subtle joys of soldering in the nude. 
Or have a collection of suicide cords (wall plug on one end and alligator clips on the other). I ain't afraid of no stinkin' smoke... as long as it's not me smokin'.
-
Well golly tautech, thank you and greetings.
This pretty well brings us up to date. That cap was C932 which is between the unregulated -38V source rail and ground. The -30V rail has a lot to do with the regulation of all the other LV supplies. The rest of the caps measured, well, not great but OK so some mysteries remain, like why the 200V was so high, and why the -30V rail with the bad cap would adjust but the +30V would not.
For the moment I am stalled waiting for my dave_k capacitor boards from OSH Park. I think those little adapter boards are just a brilliant idea. -
Great thread - good write-up. I'll be following this.
-
So, I have one of these... A Tek 5440 with 4 channels.
It was my personal main scope until about 2010. I used it professionally to find faults on my digital designs up to 100 MHz. Yes, it's only a 60MHz scope but the waveforms don't magically disappear -- they just get smaller and rounder. Sometimes I'd slow down the clock rate to check bus timing.
I used to take pictures of waveforms with a DSLR and email them to colleagues, who would laugh at my
outdated tech. But it was surprisingly useful for outdated technology.
The coolest thing about mine is the NASA calibartion sticker.
Sadly it's in really bad shape now. I was just about to throw it out. It's sitting in the living room in the "take to the dump" pile.
The problem with this one is that the old plastic is extremely brittle. Early on I snapped off one of the long shafts that run from the front panel knob down into the guts, and repaired it clumsily. Some of the knob skirts have cracked off and are useless. One of the FPC-style cables complete disintegrated and someone handwired a replacement, badly.
And a bunch of internal connections are noisy -- the trace and readouts jump all over the place. But it still technically works!
-
I was just about to throw it out. It's sitting in the living room in the "take to the dump" pile.
uhhhhhhh...what's your zip? -
5A22N
-
Thanks for stopping by David, you are one of the folks I was hoping would notice, for sure.
You are too kind.QuoteAt this point I was feeling pretty pleased with myself. Like Midas, everything I wave my soldering iron at turns to gold.
If you mess around with and study electronics for long enough, then it seems like you can reach the point where every project works or at least recognizably works on the first try.QuotePerhaps the 5443 was only sold to institutions or 3rd party OEMs and this explains its absence from any retail catalog. Also on the back was the "option 1" marking, uh-oh. Option 1 was a money saving option that deleted the readout board, so no OSD in this one.
Off with the covers and, whats this? Someone had glued thin acetate sheet over nearly every perforated section of the panels, leaving only a tiny bit of the bottom front as an intake. Pretty sure Tek didn't do that. Something to do with the medical environment?
Someone over on the TekScopes@yahoogroups.com or TekScopes2@groups.io email lists may know about the model 5443 and the plastic sheet. I have no idea about the model 5443 and you already covered the two links where I would look; did you search every Tektronix catalog where it would likely be listed? I *have* seen a plastic sheet applied to the inside of a perforated panel to redirect airflow for better cooling in a stock instrument before so it would not surprise me if Tektronix did this.QuoteThese are my weapons of choice for this sort of decapping work. Sure, a real desoldering station would be better but I don't have that. The short 14ga copper wire as a tip in the Weller is another Carlson trick. I tell you, that gentleman has it going on! The iron gets super hot, like RIGHT NOW!
I really like the old style Weller soldering guns compared to the new cheaper ones which use screws instead of hollow bolts to hold the element in place. I made a big rectangular loop out of 10 gauge solid wire for mine for cutting polystyrene foam cleanly and it works great. Thinner copper wire or stronger brass or bronze had too much resistance to get enough heating.QuoteAnd let's see here 3117, that's a little low, wait, what? nano?? and ummm .14 uhhh Kilo?? oh jeez, I think we have a winner folks
This is a common failure for old can capacitors. They have a lot of extra electrolyte and usually operate significantly below their rated ripple current so instead of drying out, when their rubber seal starts leaking, oxygen or water gets in and corrodes the aluminum lead strips between the pins and the rolled up element until one breaks.
-
5A22N
The only problem with the 5A22N compared to the 7A22 is that the 5000 series mainframes have no vertical output so their plug-ins cannot be used to feed a second oscilloscope like a modern DSO.
-
5A22N
The only problem with the 5A22N compared to the 7A22 is that the 5000 series mainframes have no vertical output so their plug-ins cannot be used to feed a second oscilloscope like a modern DSO.
that's interesting, what are the extra 4 BNCs that I see on the back of some of the acquisition frames?
this one has them http://www.ebay.com/itm/TEKTRONIX-5440-OSCILLOSCOPE-W-3-PLUG-IN-5B40-TIME-BASE-5A18N-DUAL-TRACE-AMF-/272072117916?epid=1901516118&hash=item3f58c30e9c:g:KMYAAOSwNphWaG11
-
that's interesting, what are the extra 4 BNCs that I see on the back of some of the acquisition frames?
Mine has those... And they aren't shown in the picture in the manual. Maybe the labeling will help you?
Oddly the labels seem to be user-applied. I'm wondering if it's a semi-official mod.
Also, I'm sorry, but your fun and interesting thread as made me want to keep it now...
-
that's interesting, what are the extra 4 BNCs that I see on the back of some of the acquisition frames?
Mine has those... And they aren't shown in the picture in the manual. Maybe the labeling will help you?
Oddly the labels seem to be user-applied. I'm wondering if it's a semi-official mod.
Also, I'm sorry, but your fun and interesting thread as made me want to keep it now...
as long as its not going in the dumpster man
if you get a chance, pop the cover and document where exactly those BNCs attach hey? -
OK, pictures attached. Looks like there are too many so I will have to do multiple posts.
Do you have those pin blocks in yours?
Oh, and sorry for the horrible quality... My phone...
-
More pics...
-
One more thing.
The 5A48 has that two 4 pin connector in the module. The wires continue to the front
of the module and disappear behind some shielding up near the front panel.
The 5A15N, 5B10N, and 5B42 do not have that connector.
Hope this helps.
EDIT: Only one of my 2 5A48's has that connector. the other one doesn't. -
that is really interesting
no, I don't have the pin blocks, nor do either of my 5A48s have the mate, fascinating
would you mind photographing the interior of the 5A48 at some point? -
that's interesting, what are the extra 4 BNCs that I see on the back of some of the acquisition frames?
I think I remember reading a discussion about that modification over on TekScopes@yahoogroups.com.
The holes in the plug-in's plastic frame were originally for the fiber optic readout on the 7000 series which was replaced by the electronic readout developed by Barry Gilbert.
-
would you mind photographing the interior of the 5A48 at some point?
I'm a bit busy today, but last night I was reading the sites that you linked too previously about this scope.
http://www.i9t.net/5A48.html mentions a special version of the 5A48 with an option for back panel inputs. That would make a lot of sense on a rack mount scope. The picture at the bottom of that page is what the back of my 5a48 looks like.
Also, that site has a PDF describing the mod: http://www.i9t.net/5A48_images/5A48-mod-818H.pdf
That PDF talks about it being a coax connection. Initially I thought it was just a wire, but looking at it more carefully I can see that it is coax. -
That PDF talks about it being a coax connection. Initially I thought it was just a wire, but looking at it more carefully I can see that it is coax.
I immediately identified it as coaxial cable because it resembles the RG-174 cable Tektronix used with their Peltola connectors in the 7000 series and other test instruments during that era.
http://w140.com/tekwiki/wiki/Peltola_connector
Incidentally, Peltola sockets work for some (4mm?) probe tips. I wish this inexpensive standard had stayed around.
-
mail just came, this should help