Tektronix TDS 310 repair

[carl_lab]

Still, crap design IMHO... the buttons on the front panel hold up just fine, so they could just have put a "proper" PCB for the soft buttons, instead of a Flex, and then used a flex only to link the bezel board to the front panel board.

That's not a Tektronix specific problem, HP also made similar designs in 8647A, 8648A-D signal generators for example.

[Vince]

Good to know... I do plan on buying an RF sig gen (for repair/cheap, probably... money is tight) at some point, and these particular units are in my short list, I am quite fond of them. Well, I mean purely on a look & feel perspective... don't know how they perform technically, not competent enough to judge, but since they are fairly "modern" HP stuff, and I am just doing this as a hobby with no particular tight specs to meet, I guess any top brand model from this era will do just fine !

OK, Just bought a new, yet another, repair kit/conductive "varnish"/compound, might be able to make it in time for this week-end. Today I also bought some double-sided tape, to stick the flex PCB back into place, and today again, I received the brand new cooling fan for the cabinet. So looks like I am all set.. with some luck the scope might be fixed for good, completely, during the week-end... crossing fingers. 

At any rate, I have already a new repair project lined up as soon as this TDS 310 is done, so stay tuned !     

I expect to receive this mystery unit in the next few days...

[carl_lab]

Good to know... I do plan on buying an RF sig gen.

HP 8647/8648 are not bad, I like them very much.
A few years ago I had some with partial dissolved silver tracks on the keyboard flex. I assume the silver reacted with sulfurous evaporation from some rubber parts to silver sulfide (black), that bunk off and left some local disconnections. I bypassed them by conductive silver varnish and the generators worked without problems since that time.

[shakalnokturn]

Congrats on the fix! You'll be saving TDS scopes by the dozen soon.

When fixing broken tacks/vias a single strand of multi-stranded wire can be handy, they come in different diameters too.

For your next attempt on a flex print, try either a needle point or toothpick to apply the paint. The blue insulation coating can sometimes be scraped/chipped off with a good blade but you need to get the pressure and angle right to not remove the track with it.

The other option (it would probably be mine at this point) is to scan the remains of the flex print and rebuild the as a stiff version. Of course that gives you an excuse to invest in some etching equipment for your lab.

I see a Hameg scope made its way to your bench, is that in need of attention too?

[Vince]

Hi there, le Chacal Nocturne, welcome aboard EEVBlog ! 

Congrats on the fix! You'll be saving TDS scopes by the dozen soon.  

I have to say that these repairs are quickly becoming an addiction ! 
No more TDS to fix at the moment, but I DO have another Tek scope repair lined up ! ... found a lovely piece of Tek gear, cheap, just could not resist...

Hint #1 : model number is "TDS 300 like" .. but it's not a TDS.
Hint #2 : fixing this one will most likely require getting some help and advice from the gray beards at TekScope !

When fixing broken tacks/vias a single strand of multi-stranded wire can be handy, they come in different diameters too.

Hmm thanks, will look into that...

For your next attempt on a flex print, try either a needle point or toothpick to apply the paint.

Yea, that's my intention too ! No more brushes for me, no matter how tiny they are... they are still way too big for the job !   

The blue insulation coating can sometimes be scraped/chipped off with a good blade but you need to get the pressure and angle right to not remove the track with it.

Yep, that's what I have started doing a couple weeks ago... see pictures below. Looks like it's working, the top plastic layer of the flex is quite thin in fact, so it's not that hard to get rid of so as to expose the underlying traces. It's a long and tedious though.. so I still need some motivation to resume work on it, and scrape some more, scrape, scrape, scrape....... with naked eyes... it's a form of torture I say ! .... I think I will teat myself with a microscope for X-mas....

The other option (it would probably be mine at this point) is to scan the remains of the flex print and rebuild the as a stiff version. Of course that gives you an excuse to invest in some etching equipment for your lab.

Yes this TDS 310 repair is truning more and more like an excuse to get this and get that, learn this and learn that...
Not sure I want to get back into etching my own boards.... I sure did 25/30 years ago, when I was a kid... quite messy !
These days I would just outsource the job to some PCB service !  Price is affordable as long as the surface area of the board is small, which would be the case here. I would have to find a company that offers thin 0.8mm PCBs though, as I don't think a standard 1.6mm board would work : I fear it would raise the flex too much, ie have the same effect as pressing the buttons ! Not fully of course, but half-way... that's why I am giving up on double-sided tape to secure the flex on the frame work : it comes on a pretty thick foam substrate... Glue is really the only thing that would work I think...

Anyway, I really need to move my butt and finish the repair on this flex, so I can put this scope back together : I need/want to move to move on to the next repair ! 

I see a Hameg scope made its way to your bench,

"made it's way" ?! It's one of my very first/early scopes !    My old man gave it to me 20 years ago or so. So not quite my very first scope but almost.
All the other scopes in the lab were acquired very recently, since I built my lab last fall.

is that in need of attention too?

It is now !!    Worked just fine when I used it a couple months ago when I used it to help me fix my Tek 2232, but yesterday I noticed taht the vertical position control knob on the Channel one, is borken, spins round and round, doesn't work any more ! 
See pics below. I had this same problem on this scope, a few years back, on some other button, delayed trigger fine adjust I think. There is a plastic coupling that links the long rod to whatever is on the board (a pot, a trimmer, a switch...), and these develop a crack along one of the tiny screw that secures the couping to the shaft. The screw therefore becomes ineffective and fails to grab the shaft. Anyway, this first coupling was easy to fix because it was very easy to get to, as you can see. However this second broken coupling, CH2 vertical position control... appears to be buried deep inside the guts  of the scope, between two riveted metal plates forming part of the framework/chassis. I don't even know how I can get to it to attempt a repair of some kind ! Looks like the scope was literally built around this coupling !     I will have to take this bloody scope completely apart just so I can get access !   

[tautech]

Vince, get yourself a head mounted magnifying visor for the small work, I've had one for years and wouldn't be without it. 

For the Hameg shaft couplers, member Specmaster solved that one in this thread.
https://www.eevblog.com/forum/testgear/hameg-408-1a/

The shafts might have retainers just inside the panel and once loosened will allow the shaft to be removed....check for them.

[Vince]

Vince, get yourself a head mounted magnifying visor for the small work, I've had one for years and wouldn't be without it. 

Yeah I think you might be right ! Anything is better than nothing at all... and this kind of things I guess should be pretty affordable, will check that out !

For the Hameg shaft couplers, member Specmaster solved that one in this thread.

Oh, looks like a common problem, some one has been there done that !
Looking at the huge number of views, 10,000 in 4 months.... I take it it's not just me and Specmaster having shaft coupler problems on these scopes ! 

[tautech]

Vince, get yourself a head mounted magnifying visor for the small work, I've had one for years and wouldn't be without it. 

Yeah I think you might be right ! Anything is better than nothing at all... and this kind of things I guess should be pretty affordable, will check that out !

You don't need anything very flash depending on your preexisting eyesight.
I only got a set when I started SMD work but my eyes are worse now and I now use them on minimum magnification for most work. 
2nd from the left in the images at the top of the page but I didn't pay that price. 
https://www.google.co.nz/search?q=head+magnifier+visor&rlz=1C1ASUM_enNZ497NZ497&oq=head+magnifier&aqs=chrome.2.69i57j0l5.11825j0j8&sourceid=chrome&ie=UTF-8
Some prefer the ones with LED's.....but they are heavier and a little bulkier.
Don't be tempted by anything other than those with one piece acrylic lenses....they have have all sorts of field of view and alignment issues.

For the Hameg shaft couplers, member Specmaster solved that one in this thread.

Oh, looks like a common problem, some one has been there done that !
Looking at the huge number of views, 10,000 in 4 months.... I take it it's not just me and Specmaster having shaft coupler problems on these scopes ! 

I don't think it's at all common as I haven't seen mention of it before but it was a interesting problem to solve and gathered a following of sorts. Even your threads will gather huge numbers of views over months/years as viewers just browse to gather info from others experiences.
It's always good to share info. 

[Vince]

Thanks... looks like the ones on the first picture, left of yours then, is only 12 bucks shipped from China, and not so bad... dixit this 90 year old half blind gentlemen ! 

Even comes with different lenses for different magnifications factors.



Worth a try at such low price... not sure I wanna wait a month for them to come, though... don't want to wait this long to fix my flex PCB on the TDS. Will be handy for all future work though, obviously...

[Vince]

OK I worked  on the thing today... not looking good !

My second attempt at fixing the flex was a failure... it's too much of a mess, I officially give up and am now searching for a replacement flex... if anyone finds one before I do (Sphere, Qservice, Ebay, wherever... ) give me a shout ! 

In the meantime I thought I would at least try and fit the new cooling fan I bought for this scope... should be straight forward I thought.... well not quite.

Fan uses a 3 wire connector, luckily it can be used as a simple 2 wire fan without any drawbacks.

However there was one apparent problem already... even though I bought the slow/quiet version of this fan.. I tried it on the bench supply and at its nominal 12V it's very loud and air flow is wayyyy more than this scope needs... and the scope's power supply actually output more like 13/14Volts not 12... not going to make things better/quieter ! So already some hack would be needed here. Either brute force/adding a beefy series resistor (at 12V the fan draws about 500mA), or the elegant way : modify the power supply so as to adjust the voltage to a lower setting...  but that might not be possible because this "12V" is also I fear, also going to the CRT board... plus, for all I know it might also be used as a starting point to created one or more lower voltage rails.  I wouldn't know because it's a third-party power supply so the schematics are nowhere to be found, as usual....
So I thinks the series resistor is the sager way. I guess I were motivated, I could use a compact adjustable DC-DC converter to waste less energy and let me easily adjust the fan speed. Yes, that would be a better approach... not going to spend any money on this ultra low end scope though, so would have to be either a home made converter, or a cheap Chinese power brick from ebay... I guess for a bloody fan this would be perfectly acceptable...


Also had to remove the connector from the power supply board so as to fit a connector that would match the one that came with the fan. Could salvage one from my pile of scrap boards, great.

From this point.. things would start to g downhill !

1) PSU powered off, I plug the fan and it SPINS !!!  Eh Very slowly mind you, but sill, how could the fan spin AT ALL,  given that the PSU is not powered up ?!       I unplugged the PSU then plugged it back, fan would not spin anymore, phew. Still, a mystery...

2) PSU does not work any more !!!!   

NOOOOOO !!!! This scope is driving me crazy, I did the hard job at figuring out and fixing the real/difficult problem, and then I got stopped byt this stupid flex keypad and now by a bloody PSU ?! This thing is getting on my nerves now... clearly the educational value of this thing, through the repairs that it forces me to investigate/learn from, is way higher than the technical/performance worth of the machine !!!
So I try to remain positive about it : I see it as an educational tool, not much more. This means that, though, there is no way I will spend any significant amount of money on this scope, it's just not worth it. If I can't fix it cheaply, then I will keep it for spares  because at some point I would like to geta higher spec TDS 3000, a 380 or something, with 2GS/s and 400MHz of BW, it's much more appealing.

So now I am having to trouble shoot this HV SMPS, with no schematics...  a good exercise I guess, so will try to see if I can manage that. "Luckily" my Tek 2232  blew its SMPS recently so I learned the basics of HV SMPS while fixing it, thanks to the good souls on this forum. I do  have some basic knowledge of these things now... at least the topology. And indeed when I look at this  third-party SMPS, I am not totally in the dark, phew.

I am not sure if I feel like working on this SMPS at the moment, will depend on my mood (and I have other gear that need fixing as well, more on that in future topics ! ). A few minutes reveal the following, as a starting point !

-  All power rails read 0V.

- The standby voltage is gone : read ground, whereas it should be around volt whether the board is powered up or down.

- Even though the standby voltage is missing, somehow when I ground this pin (as would happen it I pressed the power button on the instrument), teh board DOES react : I can hear the relay click as it should.. but still, all power rails read zero volt.

- There is a smell from the "hot"/primary side of the SMPS : the heat sink carrying the power FET and another, progressively gets warm... then slightly hot, but one can still maintain permanent finger contact. Still, why would it get "that" hot if all power rails are down ?

- more blatant, still on the primary side : there are two beefy resistors, which are literally BOILING hot !!!!  It's a 100+ kohm resistor, maybe 2Watt. Reminded me of the 150K resistor in the 2232 SMPS that makes the mains rectified voltage trickle to the PWM chip.  The other resistor is a big 5Watt ceramic 3,9k resistor.

Obviously these resistors aren't at fault, they are just the poor victims of the real cause of the problem, somewhere downstream.... but what problem is it... that is the question ! 


I don't have the schematics so I will have to reverse-engineer the thing to some extent I guess, armed with the little knowledge I gleaned during my 2232 repair

The main IC's on the primary side are (datasheets attached below)

- PWM chip : 8 pin CS3842A

- Power FET : a BUZ 355 ... not sure what all the "buz " is about though, because it's a pretty horrific unit  : 1.5ohm Rds(on)  ! 

- TIP 50 : 1A BJT , mounted on the same H/S as the FET. Maybe its drives the gate of the FET, the 2232 had a similar arrangement, though using a tiny T0-92 packaged BJT, not a big TO220 like this TIP 50 ! So maybe it's doing something else entirely, who knows.

- LM393 comparators... 4 of them on the board ! They must love them I guess.  2 of them on the primary side, 2 on the secondary. No idea what they do. I guess one of them must be tied to the standby line in some way.

- MOC 8103 optocouplers for primary/secondary feedback isolation, 2 of them.


So I guess I should turn my attention to semi-conductors first, power ones first (diode bridge, FET, TIP 50), then low power stuff (PWM chip and comparators), the passives last..

At least the good thing is that I can easily work on the thing : unlike the 2232, the SMPS in the TDS is stand-alone so I can have it on the bench on its own, and super easy access to all the components. Working on it will be much easier then.


So here I am, second stage of this scope repair.. front panel at first, and now the PSU.. you don't get to choose...

I will try to repair it with my limited knowledge, and if that fails I might create a new topic for this particular issue, so as to get a better exposure/larger audience.

Anyway, topic far from closed then....

[tautech]

With two PSU transformers is this not a pre-regulator configuration ?

[Vince]

Yes it looks like a similar topology as the 2232 PSU :  common mode chocke, then mains is full wave rectified, then a big filter cap, then a PWM chip with a big FET on a H/S... just not sure yet where the coil is, though !
Then the "big" transformer in the middle must be that of the inverter, whose outputs are associated with beefy diode pairs mounted on a H/S, just like the 2232. Then the last/second big transformer is on the secondary side of things, and delivers all the low voltages then regulated with simple 3 terminal regulators for most of them at least. The 12V for the fan and CRT board being less sensitive, I think it may be possible that they are regulated using simple discrete components rather than  a "proper" regulator.

So I am using the schematics from the 2232 PSU as a rough guide to help me figure this third party unit....

At least we have all the datasheets for all the actives devices on the board, a good starting point to probe around...

[Armadillo]

What exactly is cut the flex cable? Maybe a closeup of the cut would be easy to understand.

[Vince]

Hi,

I explained and illustrated the problem in detail in my previous messages ;-)

I am not wasting any more time and money and effort on this thing. I ruined it despite having the best of intentions for it, my bad, will know better next time I have to work on these things...

[Armadillo]

Do you mean that single carbon key you partially rub the carbon away?

[Vince]

No.. this was the ORIGINAL problem... at this point it would have been easy enough to fix with the repair kit I bought... but if you read the following messages you will see that I made a complete mess of ALL the keys, basically cutting pretty much each and every track that runs on this flex ! I tried to fix all that, I did... but no more, I give up. You must know when to stop, sometimes !   

[Vince]

OK making progress on that PSU.  I Googled the manufacturer and part number (Zytec 22917300) and got 4 matches. No, not 4 thousands pages of results like you always get with Google.. not 4 pages, no.. 4 matches... of which 3 are useless because they all point to the same site, which is no help whatsoever.    So basically there is only one, onnnne single match on this PSU, which gives useful technical information. So I thought it's worth sharing !

https://www.electronicspoint.com/threads/tektronix-tds350-psu.25635/

Basically a forum where there is a poor soul like me in need of assistance with this PSU, then comes along a lad who fixed a LOT of these TDS 300 series scopes, and no less than 18 times, this bloody PSU was the issue ! Fixed them all as I understand it, and has some useful information... I will summarize :

1) The big 5 Watt resistor that's boiling hot on the primary side... it's a common fault apparently. Eventually dies and goes open circuit after a while. Mine is not dead yet, but I might as well measure it to make sure it hasn't suffered/drifted significantly.

2) Said resistor is part of the start-up circuitry... so that would be consistent with the standby voltage having disappeared on my board !

3) Topology : the 5V rail is used by the feedback loop to adjust/correct the primary voltage. So if the 5V is not working, all other rails will be down as well even though they work just fine.  So the 5V must be checked first... as long as it's not working, no point wasting time trouble-shooting any of the other rails.

4) The two beefy double diodes packages (FEP16DT) mounted on the secondary side heat sink... are used to form the full wave bridge rectifier of the afore mentioned 5V rail. Sometimes these diodes are shorted.  If shorted,  then a  NTE 6240 works fine as a substitute.


OK, that's quite helpful, better than nothing at any rate, that's for sure...  maybe I could try to see if this Zytek company still exists and ask them nicely about the schematics... you never know your luck ! Product is 20+ years old so maybe they won't mind...

[Vince]

OK I think I sussed it... in the end it's no coincidence that this PSU decided to "fail" just when I so happened to plug the new fan on it...

Spent some time checking each and every semi-conductor on the board, quite a few of them (when you add up all the signal diodes and small trannies)... all were fine. The big 5W 3,9 resistor s fine too.
BTW the PSU is well designed, a joy to work on : the two heat sinks  are not soldered to the board but rest on stands, they just click into place. The components are not screwed on the H/S but simply held in place by a metallic "spring"/clip, neat. And there is no dirty/greasty white cthermal compound either, but rather a rubbery pad. So it was very easy and quick and "clean" to removed all these power semi-conductors and their H/S.

Also, the PSU is really safe : the big mains filter cap, which is charged at 320V DC then (in my neck of the woods anyway), must be fitted with a bleed circuitry because it takes literally only a mere 5 seconds for the voltage to drop to a single digit voltage !

So I just didn't know what could be wrong.. decided to power it up to try and probe around... go figure, the thing worked just fine ! 

So... I plugged the fan in and... again it failed to start, though you can see it's trying to. It's jerky. At the same time... I notice that all the other rails are WAY down, like 5V instead of 8.5V for example !  40% off or so, quite a hit.

So I thought, maybe the 15V output supplying the fan (and CRT board), is faulty and fails to supply enough current. So plugged the fan back in, but also measured the current it was drawing, or trying to at least !  Didn't bother hooking up the scope, just did a quick check using my DMM and its "PEAK" detect mode... first time I use this feature, quite handy in fact !  So, the short-circuit current the fan draws every time it tries to start is about 250mA tops., but the fan requires about twice that.

So, looks like we are on the right track here ! Might explain why I tossed the previous fan 4 years ago then : I probably noticed it failed to start, assumed it was faulty so just figured I would have to replace it.... which I am doing 4 years later, and the fan still doesn't start...

Looked at the parts list in the service manual... I was obsessed with the CFM/air flow rating, not wanting to overheat the scope, so bought a fan with a similar flow... and in doing so I overlooked this important piece of info : OEM fan was rated at 2 Watts, the new one I bought is 5W !  It all adds up : the board peaks at 250mA or so, enough to drive a 2 or 3 Watt fan, but definitely not my 5W fan.

Still.... the previous fan, if it were OEM... it should have started !  So maybe it was indeed faulty, or it was not the OEM fan, it got replaced in the past before me, and the guy ran across the same problem I am having right now.

So I guess it will all end up nicely, though : as said before, this new fan is way too noisy and produces way too much air-flow at its nominal 12V, never mind the 14V that the board is outputting. So adding a series resistor to halve the current should be able to make the PSU happy as well as make the fan spin more slowly. I just tested it using the bench power supply : adjusted to 14V and current limited to 250mA so as to simulate the PSU. Under these circumstance, the fan is a bit lazy to start obviously, but it does start reliably, and once the speed is  stabilized, noise and airflow are at adequate levels.

Phew ! That's one problem sorted. Lesson learned : next time I replace a fan... don't assume that the PSU has ANY headroom current wise, assume it can not deliver more than what was strictly necessary for the OEM fan to work ! In this case my 5W fan 2.5 times more powerful than the OEM 2W fan, my bad.

I do have a bucket load of power resistors salvaged over the years, I am glad they an be put to use at long last !   



Oh, and as far as topology goes, I was confused about the apparent lack of transformer/coil for the FET in the pre-regulator section... in the end there is no pre-regulator ! The FET drives directly the inverter transformer apparently !  I guess it works...

Also, the TIP 50 / power BJT mounted on the FET heat-sink... has nothing to do with driving the gate of the FET (like it does in the 2232). The PWM chip has a totem-pole output capable/designed to drive the FET directly, so no need for an extra tranny to drive the gate. So, the TIP 50 what does it do ? It is connected directly to the 5W ceramic resistor of the start-up circuitry, so must be involved with this part of the PSU...

So I will add this series resistor and call it done. A bit of a hack sure, but that will do. Next time of course I will make sure I get an appropriately rated fan (if at all possible) to avoid this kind of hack ! 

[tautech]

Nice one Vince.
Did you check the LV caps...value, ESR ?

[Vince]

Nope, I don't have an ESR meter jsut yet... so I have been making progress recently on what gear brand/model might find mty fancy and fit my budget. So as soon as I can afford it, will do. By the end of the year I think.

As far as measuring values, measuring in circuit is not reliable in circuit since I don't have the schematics to see the surrounding circuitry...

overall I didn't bother because they all "looked" fine and the PSU was working just fine so it seemed odd to me that it would fail suddenly due to caps. I was right of course, retrospectively...

But yeah I can't wait to get the gear to check ESR (will get I think an LCR meter altogether actually, and avoid paying twice for this feature...)

[Vince]

OK, just worked a bit on the thing. Tried to add a series resistor to the fan , as described earlier... doing so was a total failure and turns out the repair might be more interesting than envisioned at first : might well be that the PSU is indeed faulty, and my fan replacement far being a more powerful than the OEM one may not be the actual issue ! Might still be an issue will see, but it's definitely not the only problem here.... here goes :

At first, I used the bench supply to figure out the appropriate resistor, again set to simulate the PSU open circuit voltage (14V)) and short-circuit current (around 250mA), as it tries but fails to get the fan spinning. As a safe side approximation, I simplified and took the fan winding resistance out of the equation, assuming it had to be small, and the back EMF being zero of course since the fan fails to spin. So 14V divided by around 250mA, 56ohms. Had a 47ohms, tried that, yielded to a 250mA current so perfect ! Fan would start reliably but air flow was not good enough to me. Tried a 33 ohms one, air flow was now adequate.

So... I then tried that on the actual PSU.... failed to start. Reverted to the 47ohms one... still fails to start (and again all power rails pulled low consequently). Now I am starting to worry... so I tried higher value resistors. Fan obviously would not start but I wanted to  see at what point the power rails would start NOT going down anymore. Tried 150ohms... rails still down (about 40% lower than normal, 8,5V rail would read 5V, and 5V would read 2+ V)  !

Tried 150ohm, rails still down. 330 ? Still down. 470 ohms ? YEAH they are up at last !
Measured voltage drop across the resistor, it implied a current of around 20mA !

Soooo.... this fan connector is supposed to spit out at least 165 mA in order to be able to drive the OEM fan (parts lists says 2 Watts 12 Volts), but can managed only 10+ of that ?!

Surely my fan can't be responsible for that can it... 

So looks like the PSU/ fan output is too weak... and the fan I tossed 4 years back when I first got the scope... probably was the OEM one and was probably just fine... oops.... 

So... PSU troubleshooting !  Without schematics...  a first for me.  I meant this TDS 310 repair to be a learning exercise above alll, and looks like this scope heard me ! LOL

Had a quick first go at it.  There are 4 main rails : +/- 5V and +/- 8.5V.     We know that +5V is regulated from the primary side/PWM chip, so we are left with 3 rails that need regulating.

In the bottom right corner of the board, just near the main transformer and main output connector, we see three x3 terminal regulators, good start. A positive one (LT1086) and two negative ones (LM 337), it all adds up.

Now for the two auxiliary connectors, teh fan and the CRT supply...   The fan is supposed to be 12V and the CRT is supposed to output 15V. but in practice, open circuit, they both deliver a similar 14V.  OK, so maybe Tek decided to use a single 14V, half-way between 12V and 15V, to cater for both needs with one supply only, why not.  Still, problem remains : 14V from... from what ? Higher voltage main rail is 8.5, so it can be used to derive 14V from it without a DC/DC converter. We need at least 15V to produce 14V... and I using 15V to feed 5V and 8.5 Volts linear regulators would produce way too much heat and waste way too much energy... so...... the 14V must have it's own tap on the transformer, and it's own voltage regulator.... but I can't see any. We have run out of 3 terminal regulators already, and no fancy DC/DC boost up converters of course, and no DIP package regulator either...

Hmm. Wait, in the top right part of the board (highlighted in yellow in the picture), where the fan and CRT connectors are located (in light blue), I see a tight group of 4 little T092 packages (in red) ... hmm maybe some discrete/hand made regulator ? Let's see what the part numbers say :  we have two TL431, which are adjustable Zener diodes, and two 2N4403 which are general purpose PNP transistors, capable of a continuous 600mA collector current, plenty enough for the fan... So looks like the CRT and fan use independent regulators after all, and both use an identical discrete circuit made up of a Zener and PNP pass transistor.
The CRT on this scope works just fine so let's not worry about its 15V supply, it's just fine even though it output only 14V not 15V.

So I guess I must figure out this fan supply then...  maybe the pass transistor is faulty, or maybe the Zener is faulty, or maybe some of the passive around either of these two guys...

The good thing is that the two regulators being identical in design and open-circuit voltage, I guess I can use the CRT regulator as a reference, and why not,  swap the Zener and transistors to help figure things out.

I think I most likely have a decent substitute for this 2N4403. The 2N222 counterpart, 2903 is it ? The other day I spotted these in my drawers.... must be good enough for the task, could try that.
The Zener I can always test on breadboard.

Passive/resistors around the regulators, can't see any burnt or damaged ones. 

There is a "big" electrolytic cap next to both connectors, I guess they must be in parallel right across the connectors. Maybe the one for the fan is leaking badly and sucking up most of the output current, leaving only a spare 20mA for the fan...

Who knows, but anyway I have a clear path of investigation now, so let's gooooo...

I guess the very first thing to do is probe around to figure out which tranny and zener correspond to the fan output.

[Old Don]

Vince, here in the States Harbor Freight has a Chinese made head mounted magnifier levels of magnification plus a flip into position "loupe" for extra magnification  and it sells for about $5.00. Or well under 5 Euro's. Don't know if you have a company like Harbor Freight, but even if you have to pay 10 Euro's that should be all you need for most work.

Here's a link so you can see it:  https://www.harborfreight.com/catalogsearch/result/index/?dir=asc&order=EAScore%2Cf%2CEAFeatured+Weight%2Cf%2CSale+Rank%2Cf&q=magnifer+strap

[Vince]

Hi Don,

Yep as discussed farther up this page, I will definitely give this type of thing a try, cheap and can only better than nothing, so not much to loose.... 

Made some progress on the troubleshooting last night. Another lesson learned : just because something seems obvious and all adds up... does not mean it's true !

I mean, these two Zener and tranny pairs around the fan and CRT outputs looked all too tempting and plausible for sure, but after spending some time probing around both power off then on... I just could not make any sense from it !

I Had no choice but declare that these Zeners/tranny combos had nothing to do with the fan and CRT !  No idea what they are used for, would love to have the schematics one day, to find out,  but this is never going to happen of course, sadly...

So I stopped worrying about this section of the board, I stopped making assumptions, and just probed and followed traces.

From that, it appears that the fan and CRT are running from the same supply as I originally thought, with the only subtlety being that the CRT runs straight from the supply whereas the fan is fed through a series 16ohm 1W resistor, presumably to protect from shorts, or at least limit inrush current at start-up (to under an amp then), something along those lines...

Power comes from a two taps from the first (not the main one strangely) transformer, each tap going to one of the legs of one of the beefy/T0220 dual diode packages which are mounted on the secondary side big heat sink. So I guess it's the classical arrangement where middle tap is used as ground and then each tap of opposite phase has a single diode on it, this way with only two diodes you get a full wave rectifier. Then the output / common cathode of the diode package goes to big 1000µF filter cap, then straight to the CRT output then to the fan output via that series 16ohm resistor.

Well I still need to investigates further, with the scope this time, to make sure I am not talking BS again, but while probing with DMM and looking at the actual traces on the PCB, that's what it looks like to me.

I guess it makes some sense : the digital 5V rail is regulated direct from the primary side PWM chip, so if that regulation is good enough for digital chips, sure enough it's plenty good enough to supply a bloody fan...

Now having that all this, I am even more at a loss as to why the fan output can't supply more than 20mA !

1) the 16ohms resistor... looks like new, and measured spot on at 16ohms, so it's not this guy limiting the current

2) CRT output is upwards of the fan, and has no problem supplying current, judging by the fact  the CRT/scope at large, works just fine !

So on the one hand I am happy because I now understand better how it's put together... but on the other hand I am very unhappy because I just don't see how something this simple could go wrong, especially given the afore mentioned preliminary test result !? Grrr... not sure what to think of this... I witness a problem...  can't explain it, frustrating...  any suggestion welcome at this point ! 

[tautech]

Ask yourself, why use a TO220 dual diode capable of supplying AMPS for a measly load ?
What's the specs for it ?
Let's say one diode is OC, the CRT inverter circuit won't mind but a DC fan most likely will.
Put a scope on the 14V rail to check for full rectification.
Is the 1k uF sick and leaking badly and/or not to nominal value ?

[Vince]

Hi Tautech,

Ask yourself, why use a TO220 dual diode capable of supplying AMPS for a measly load ?

Yes sure.... I tried to make sense of that by the fact that the CRT uses this supply too. I don't know how much juice a small CRT display like this requires, but it's not minuscule I would think.
Other thing is that, though I didn't mention it in my post, these two taps from the first transformer don't just stop at the diodes... they keep going and eventually go to pins on the second transformer... so I thought maybe this is the primary winding of that transformer wand they use this 14V to produce lower voltages for the 5V and 8.5 rails, so some more juice needed there as well.

What's the specs for it ?]What's the specs for it ?

Plenty enough of course     enough to power half a dozen of these scopes I would think ?!   It's an FEP16DT,    8 amps per diode no less ...

Let's say one diode is OC,]Let's say one diode is OC,

Sure, I checked all power semi-conductors a few messages ago, they all checked OK. I guess I could have gone unlucky and one of these diodes went bad during this repair, you never know your (lack thereof) luck !

Put a scope on the 14V rail to check for full rectification.]Put a scope on the 14V rail to check for full rectification.

Yup, will do shortly.

Is the 1k uF sick and leaking badly and/or not to nominal value ?

Visual inspection revealed nothing, everything on the board looks sparking new.

Just checked the cap and in-circuit, the DMM test is inconclusive, can't get a reading. At first the DMM acts as if external circuitry was drawing current, so I thought OK maybe it's shorted, but then the display briefly indicated "open circuit", then back to trying to measure the cap. Maybe I was dreaming or maybe I just moved the probed inadvertently slightly... or maybe the cap is indeed dodgy, no idea, but sure will find out ! 

OK, so first : 1) check (again) the dual diode to be sure, then 2)  power it up and check the waveform on the scope to make sure it's full wave rectified then 3) remove the cap so I can test it properly/reliably.

Stay tuned...