Test Equipment Anonymous (TEA) group therapy thread

[Vince]

I played with Wedeco brand systems. They are French I think?

I just Googled it, apparently it's German, founded in 1976. They wanted to be eco-friendly and be able to treat water without chemicals. So they went the UV way.

I can't get away from these Frenchies!  It's ok though, all the French people I've met so far are always pretty cool.

You are lucky then !
As for me, for the past 20 years the society has been going south and let me tell you that with every passing year there is an ever growing chunk of the population that I would rather shove into a giant shredder so that short of their brain being able to contribute to society in a useful way, at least their dead shredded body can serve a useful purpose as compost to grow trees and plants, which are much more pleasing and useful.

UV water treatment does interest me.
Next year I plan on digging a 30m3 reservoir in my garden, to collect rain water (with my roof area, the stats for the past 30 years in my village, show that I would be able to collect between 60 to 120m3 a year, so filling up a 30m3 reservoir is plenty realistic), to cool down my house that's overheating in summer, and I don't want to go the A/C route, nor can I afford it.

Anyway, the price of water is also going up and up and more and more, sometimes they restrict the use of water to various extents, so even if you could afford to pay for the water, it's still no guarantee you would even be allowed to use it. So I thought I could also use my giant water reservoir to be autonomous, long term, and produce the potable water I need, should things go really badly in the water department. So I would need some way to make it potable. Not that many liters. Just enough for drinking water, cooking and a quick shower. So I am hoping a fairly compact and affordable unit, easy to install, does exist somewhere, but if not, the way things are going, I am sure companies will develop and specialize in designing just that kind of thing, as the market grows more and more.

In the much shorter term I intend to use that rain water to supply the toilet flush. The configuration of my house would make this fairly easy to implement "cleanly", and would be an instant significant saving on my potable water bill. Actually as a little kid I remember thinking about that very issue, and always assumed toilets used non-potable water. When I later realised that no, we were throwing lots of potable water literally down the toilets... I was shocked.

Some guy the other day told me that the water companies do not like people doing this, because that's sending untreated water into the system.... yes, surely rain water mixed with poo is extremely more harmful to nature than potable water mixed with poo.
In practice they don't bother trying to enforce their view because virtually nobody bothers implementing a "grey" water system for their toilet.... so I will just do it and nobody will ever know or care...

Anyway, these UV rays are very useful it appears, not just for blanking EPROMs ! 

[factory]

Depends on the type of UV, make sure you know the difference between UVA, UVB & UVC tubes.

I used to work in water treatment, we usually used 4 or 5 foot long tubes in banks. They'll hive you a nice sunburn without too much trouble.

I played with Wedeco brand systems. They are French I think? I can't get away from these Frenchies!  It's ok though, all the French people I've met so far are always pretty cool.

PDF warning... This is actually who I used to work for:
https://pncwa.memberclicks.net/assets/2010ConfTechPresentations/Session12/2010%20pncwa-%20session%2012-2%20-%20disinfection%20-%20matthias%20boeker.pdf

Looks like those would give a nasty sunburn, the pdf was interesting reading, we only got the units in when they failed (usually the DC electronic ballast), don't which part of the train the small units were fitted to, hopefully the catering coach and not the other waste going into the tanks, or on the track (in the bad old days). 

David

[m k]

Depends on the type of UV, make sure you know the difference between UVA, UVB & UVC tubes.

One example is that you're not tanned through windshield but side window is different.

Purification is an interesting thing.
We have the same water in every tap, USA city citizens may have different system.
Then industry can have completely different requirements and still none is talking about toxins, maybe levels are fundamentally low, though I remember one old case, MPTP.

PS,
the frog was floating in the air.

[Neomys Sapiens]

How many of you have a 11K series scope ?!  I can't remember seeing anybody posting about it since I joined TEA 3 years or so ago ?

Nah, certainly not. 10bits are nice, but not with that sample rate. There was a better digitiser that could be fitted to a 7xxx scope. 20Msps/s? Yaaaawn! My ole' 2430 did 100Msps/s and a 2440 would do 500! I would consider your 544A racy in comparison, and more useful and capable too!

[TERRA Operative]

TDS220 has 1Gsps! 

Don't worry about the bandwidth limitation... 

[Vince]

You are comparing apples and bananas people ! 

It's a sampling scope not a real-time scope, so please compare it to similar scopes, which are contemporary to it 

[Vince]

Tek 2465 PSU RECAP & REPAIR

OK last evening I pulled the PSU and took it apart, and took note of all the caps in there. Will sit down and order all that sometime today.

Then I looked at the fan motor control board, which is part of the PSU, to try to diagnose it.

Well the good news is that the fan control is on a little individual board, that's soldered directly at the but of the motor.
you just have to removed a 3 pin cable that attached to it to power it with 15V, and that's it. The motor/board can be removed in an eye blink.



So that's fantastic, that means I can actually power the fan stuff alone on the bench, don't need the PSU never mind putting it back into the scope.
So that will male troubleshooting orders of magnitude easier/faster... so here is the result :

Initial problem was that the fan would not spin.
Well now that the fan is removed and the motor is 100% unloaded... I can see it does rotate at power up, but very, very slowly. So far so good, after all it's temperature controlled.
Problem is that it draws only 40mA or something... again no problem you say, current and torque will increase as necessary when a load is present ?
Well, no. I can stop the shaft/motor with my fingers effortlessly. This motor will not develop any torque whatsoever, and will not draw and more current than the 40mA it does when unloaded.

OK so I tried to speed it up. I shorted the thermistor in order to simulate a high scope temperature. It worked. Motor now spins much faster. However it still draws pretty much the same current, and again if I try to load it with my fingers, it fails to draw more current, and fails to develop more torque. It just slows down and stops eventually, with only moderate finger pressure.

OK, so you say the electronics on the board fail to provide enough current, right ? No.
I buzzed all the components on that board, all good. Except for resistor R1691 at the bottom of the schematic. It's supposed to be 15R but registers only 10.4R.




However I don't think it's a problem... it's not a precision resistor used for current sense/feedback. To me it looks like it's just there to limit the short-circuit current (when the motor starts initially) to something safe. In this case 15Voltsd supply, 15R... that's 1A grand maximum. Which is also the current rating of the transistor array that that drives the windings, so that adds up.
So with a 10R resistor that means the max current is more... but who case, since the freaking motor won't draw more than 50mA no matter what...

ISTR replacing the quad transistors with individual transistors mounted on a standard IC header. I don't see why they need to be thermally coupled.

Since this transistor array is prone to failure it seems, and since I got some strange in-circuit readings, I removed it and replaced it with discrete transistors.
2N2222A. current rating is only 800mA not 1A, but since the motor only draws 50mA... I though it would be at least good enough to see if it made any difference, and help me further the diagnosis.




Well, doing this... it made absolutely zero difference what-so-ever. It's no better or worse.

So as far as the components surrounding the motor, I don't see anything wrong.

Only the motor itself is left. Manual says it's got hall effect sensors in it, which are in the feeback loop to control the amount of current that the windings get.
So maybe they are tired and/or : I ntice a lot of RADIAL play in the motor shaft.. axial play is normal to a degree, but lots of radial play hmmm... not good me thinks. It maybe causes the hall effect sensors to misbehave. The gaps are not as tight and consistent as they should.

Well that's were I am at right now, not looking good !

I don't know if one can find a cheap replacement motor (preferably with the control board still attached to it...).

If not, maybe it's possible to retro-fit a "normal" fan, like I noticed, later 2465 scopes were fitted with.

Requires  some investigations....

Anyway, I am well pissed, because without a working fan to keep the hybrids cool, I just can't take the risk of using this scope any longer !

So for now I will order the caps and recap the PSU, which is the biggest/most painful job that this scope needs, and when that's out of the way I can investigate what can be done to fix the fan situation...

[TERRA Operative]

https://jestineyong.com/tektronix-2465a-dv-oscilloscope-siemens-hall-effect-sensor-controlled-fan-motor-rebuild-part-1/

Apparently you need to rewind the coils to replace the bearings.....

I'd be looking at complete replacement with a modern fan myself. 

[Vince]

Wow, thanks for that Terra ! 

What an awesome repair guide !

Unfortunately yes, it's beyond my abilities and motivation. That would a major project for me and I am not looking for yet another "project".. I have plenty enough as it is.

However it's still a very interesting read, as it confirms my diagnosis. It makes it clear that it's a common problem for these motors, both the slow speed, hall effect sensors, and mechanical play in the rotor.
So I now am not hesitant any longer in incriminating the motor itself rather than the electronics around it.

Goal is to fix the issue as quickly and easily as possible.
so first I will try to see if I can find a replacement motor, as it seem its's an off the shelf Siemens motor, which therefore may have been used in other kind of equipment, hell maybe small domestic appliances for all we know.

If I can find one, maybe at least find specs, and try to find an equivalent.

If that fails then looking retrofitting a "normal" fan, somehow, hoping that's even technically possible. I mean does the chassis / air management in that scope even lend itself to such a conversion, not sure.

I will post on the 2465 EEVblog thread that Tautech give the link of the other day, to see  what people say about this route, and if people have already done it and with what success.

But for now, the goal today is to sit down and spend probably an hour or two scouring Farnell's website to order all the caps I need for the PSU.

Then I will switch to other things, then when the caps arrive and the PSU is finally recapped and the scope proven to still work hopefully... then I will move on to investigating  the fan motor issue. One step at a time....


When recapped the 2467B PSU I remember there was trap... some small electrolytic was vicious and was... NON polarized. I remember the pain it was to find a replacement. Only axial audio and ultra expensive ones were to be found. Hopefully the cap shortage is less severe today than it was when I recapped the 2467B.

Anyway, I paid attention to that cap on the 2465 PSU and ... I could not find it ? So non-polarized electrolytic in this scope, anyone can confirm ?! Do'nt want to do a tragic mistake.
I will check the manual, schematic and parts list to make extra sure... don't want the scope to catch fire...

I also found ONE dipped tantalum in there. Not sure I should replace it, given it's fine ? However I see it's rated at only 10V which is not much, and from people say, too low a voltage rating severyl reduces their life time, and Tek here and there was used to doing just that ?! Will check the schematic to locate that tantalum and try to figure out what voltage it actually sees in operation, hoping it will be 5V or so tops... not 10V.

All that will keep me busy all afternoon I guess....it's already 14H30 here.

[Vince]

OK had a quick go at the fan motor thign. It's made by Siemen indeed, P/N = 1AD3001-0A

Google does know about it.... it found me two of them on Ebay... used, from old HP gear not Tek, and at 80 and 135 dollars a pop, it's not  for me. 

[TERRA Operative]

I got me a new toy.

A HP E4418B power meter, not working of course so I get to have some fun, and after 5 minutes I found the problem! A faulty 74HC14 in the soft power button circuit. I thought this was likely, so I bought some in Akihabara today, along with the 74HC74 flip flop that the 74HC14 is feeding into as they were the mst likely culprits looking at the schematics. Having the schematics was a BIG help!
So, I'll stick a new 74HC14 in tonight, and replace the backup battery too (I got one from Akihabara today too), and then update the firmware.
Then I'll source some rubber baby buggy bumpers and a handle for it too and it's all good.
I already have a 437B but it sometimes doesn't boot and is a bit old now, so I'll fix it later to sell it off.

Also, I got some replacement flat-flex PCB's for the front panel controls in the THS700 series scopes. one of mine had fractured traces, so off to the PCBWay for a sponsored video (in the making) and I got some PCB's 
I also am working on a replacement backlight driver to convert from CCFL to LED, but it'll need a second revision as I messed up the layout in v1.0.
See here for the thread if you have any input......
https://www.eevblog.com/forum/projects/sx1308-boost-converter-strange-behaviour-on-feedback-pin

[TERRA Operative]

Annnddd I just won an Advantest R6145 60V 1A Voltage/Current Source. Will be a nice upgrade from my Advantest R6144 for calibrating stuff.

[Vince]

wow you got some luck in the TE dept. it seems, good on you...

OK so I just searched my stock of new electrolytic caps to see if I could save some money.
I did. Found 14 caps in all, that I won't have to order... which is pretty much 50% of the total number of caps. 
Plus, the ones I got make up for most of the big/expensive ones in the list, so the remainder of the caps should not add up to too much money, great. 

I will even get to replace that lone 4.7uF 10V tantalum, because I have ONE single tant in stock, and it's the same value, and  better yet : much higher voltage rating, 35V not 10V, so it should make it reliable hopefully. Strangely its size is the same as the old one, so I guess it's not just aluminium ones that got smaller in the recent years, the dipped tants too..

Non-polarized cap : had a closer look at the board... I am changing my mind. I did find 2 of them, both 50V 1uF. There is no polarity marking on the silkscreen, nor do that caps themselves show an obvious polarity mark. So I searched for them in the manual. Schematic says they are not polarized, both of the are used in the feed-back look of a couple lovely CERDIP LM324 op-amps.



Then I looked them up in the parts-list and.... oh no, they make NO mention of them being non-polarized ! What to think now ! 
However they do give a manufacturer part number : " ECE-A50N1 ", Googled it and lucky me, somehow found specs for that very cap :

https://www.csgparts.com/part/ECEA50N1.html

Somewhere in there, it says it's non polarized... woohoo.

So non-polarized it must be.
But I wonder, given how expensive and hard to find these are, wouldn't it be possible to replace it with a humble film cap  instead ?! 
I mean, on the same board, in similar situations (caps in the feedback loop of those LM324), but for values less than 1uF, they were happy to use (small axial) ceramic caps. OK not film but ceramic, still I could buy that, should be much cheaper and easier to find than an electrolytic I would think.

Thoughts ?! 

[Vince]

Good lord Terra, that schematic you just posted !

I still can't believe my eyes : dimensions 16,199 x 10,497 pixels ?!   

Only 500KB which is a tour de force, I applaud you for that. However despite the file size being perfect, having so many pixels is just.... too much for my 15 year old computer.  It can't cope with it.
I tried loading the file locally so I can zoom on it and pan... but it's so slow, the computer almost freezes at every mouse movement I try to make. It just doesn't work...
Maybe crop the schematic to just the interesting portion of it...

EDIT : just checked. Once loaded into RAM, the image takes up 500MB of memory ! 

[TERRA Operative]

Hahaha, I crunched the colour space down and PNG does good things with compression on schematics.

Here you go. U33 is the faulty part.



Also, 1uF 50V NP electrolytics shouldn't be expensive. I reckon I could get some for less than a dollar each in Akihabara for United Chemicons, if I don't have them on the shelf here.

[Vince]

Thanks, works much better now ! 

Could wokr better still : 10,000 pixles wide is still a lot. I only needed 50% zoom to see the details of it comfortably. Which means you could make it only 5,000 pixels wide and the file would have 4 times less pixels in it, imagine how much faster my computer would be able to process that !
I would also require less energy to transfer, and less disk space on the server hence less energy and materials for the hard drive. It's  win-win.

Having an old PC is great, it forces me to be mindful of file size and image resolution. Just like in the old time when computers were 8 bts and a few KB of RAM and ROM... you just could not be careless of or else you would get nothing done 

Caps : just looked on Farnell, my go-to / default supplier, they have 19 different 1uF 50 alu caps, not a single NP one  :'(
I will see what TME or other suppliers I have access to, have in stock, but I am not ordering caps from 2 different suppliers, no way I am paying shipping twice for a few caps ! 
Maybe I could just put two 2.2uF caps in series, front to back, instead ?... looks like there is enough room on the board for such a montage..

[Vince]

EDIT : on second look, changing some search parameters, it appears Farnell DO have them.... they have ONE NP 50V 1uF, just one :

https://fr.farnell.com/nichicon/uvp1h010mdd1td/condensateur-elec-alu-1uf-50v/dp/1823721

.... and it's a Nicicon, and it's super cheap, a few cents, and it's RADIAL, yeah !  It's all truly marvelous except for two facts :

- It says this product is no longer manufactured
- Even if it were still available, it's in stock only in the USA and they charge 18 Euros flat fee to bring any US component to Europe. So that would have made for a 20 Euro cap, brrrr....

The search continues...

[TERRA Operative]

1uF you say?....

Throw a film cap in and call it done.

[Vince]

Yes I asked about that in my original message.... waiting for people(s feedback on the feasibility of doing that, based on the what they are used in the schematic.
I mean, even back then, Tek could have used a film cap if it wanted to, but they chose to put an alu instead. There must be a good reason for it and I would rather be able to make an informed decision.

Hence why I asked about rather making my own non-polarized out of two 2.2uF in series back to front. This way at least I keep the same cap technology/specs, hence take minimum risk.

I need smart people to look at the schematic and tell me what's the best course of action.

[wkb]

Recapping the 2465 PSU is not nearly as problematic as you think. I did it myself, not nearly as difficult/awkward as a lot of other T&M.

As for the 2465 fan: awkward and unnecessarily complex.

[Vince]

Well it's not problematic given that I have solutions for this NP cap. Just need smart people to tell me how best to approach the replacement of this freaking cap. That's all.

1) Replace with film cap
2) Use a crap audio cap that axial and way too large to fit and very expensive.
3) Use two polarized  2.2uF in series, meaning available anywhere, for cheap, and able to choose decent brands / quality
4) Fork out whatever cash is needed to get the proper cap from abroad, and have it cost me 10 or 20 Euros in the end.

Until I get some more feedback, for now the favored solution is #3.

[factory]

EDIT : on second look, changing some search parameters, it appears Farnell DO have them.... they have ONE NP 50V 1uF, just one :

https://fr.farnell.com/nichicon/uvp1h010mdd1td/condensateur-elec-alu-1uf-50v/dp/1823721

.... and it's a Nicicon, and it's super cheap, a few cents, and it's RADIAL, yeah !  It's all truly marvelous except for two facts :

- It says this product is no longer manufactured
- Even if it were still available, it's in stock only in the USA and they charge 18 Euros flat fee to bring any US component to Europe. So that would have made for a 20 Euro cap, brrrr....

The search continues...

Did I not add some to one of the boxes of parts I sent your way? The only way I could get them was to buy a few extra I didn't need.... 


I seem to remember they didn't like my suggestion of using a film cap on the 2465 thread. 

David

[Neomys Sapiens]

Well it's not problematic given that I have solutions for this NP cap. Just need smart people to tell me how best to approach the replacement of this freaking cap. That's all.

1) Replace with film cap
2) Use a crap audio cap that axial and way too large to fit and very expensive.
3) Use two polarized  2.2uF in series, meaning available anywhere, for cheap, and able to choose decent brands / quality
4) Fork out whatever cash is needed to get the proper cap from abroad, and have it cost me 10 or 20 Euros in the end.

Until I get some more feedback, for now the favored solution is #3.

I can't see why you shouldn't use a film cap. Except if the circuit dependet somehow on the normally disadvantageous properties of a electrolytic cap, like higher ESR, TC,...
Further, if you select the 2 caps for 3.) to be close in value, that won't be nice, but will work.
But if the scooe is a keeper, I would fork out the 10.-€.

[Vince]

Neomys yes this scope is a keeper indeed ! I know I was not too sure about it the very first day when I got it, but I am now already fond of it. It works just fine and is modern enough that it should not cause too much problems in the long term. So I think I will make it part of my stable of daily drivers so to speak. It will get a permanent spot on the bench.


Cap problem fixed then, thanks David !

Yes I had somehow forgotten about those... to my defense you sent me so much cool stuff that my brain failed to keep track of everything ! 
They were still there on the bench buried among other stuff.... took me only a few seconds to find them, here they are, all 16 of them !!! 

Thanks so much David ! 




So I need two caps, remain 14, so that means I should start hunting for 7 more of these scopes !!! 

OK so that's cool. The problematic cap is not a problem any more, even have the lone tant, half of the caps... I think I fit everything I have got here right now, this way it will be less work when the remainder of the caps arrive... will order them this evening hopefully.

As for the burnt 68 ohms resistor in series with the exploded RIFA, I don't think it's a fusible / special resistor like some suggested here, for :

1) Parts list does say any special about this resistor, other than it's a "COMPSN", I guess it means "composition" resistor ? could not find the glossary of terms in the manual. It's not a film resistor at least.

2) While checking my stock of X rated caps, found a brand new RIFA lookalike Kermet cap. It was marked "100nF + 33R ".... I bet that means it comes with a built-in 33 ohms series resistor. So this must be a thing I guess, hence why Tek did. So if they integrated the resistor inside the cap, it's hard to argue that it could be a fusible... that would force you to replace the entire cap rather than just the resistor inside.

[tggzzz]

Well it's not problematic given that I have solutions for this NP cap. Just need smart people to tell me how best to approach the replacement of this freaking cap. That's all.

1) Replace with film cap
2) Use a crap audio cap that axial and way too large to fit and very expensive.
3) Use two polarized  2.2uF in series, meaning available anywhere, for cheap, and able to choose decent brands / quality
4) Fork out whatever cash is needed to get the proper cap from abroad, and have it cost me 10 or 20 Euros in the end.

Until I get some more feedback, for now the favored solution is #3.

I doubt the value of C1291 is critical; I'd just use a 2µ2 capacitor. Check Vreg doesn't oscillate

The other feedback circuit may be optimised to avoid oscillation while having suitable transient response. What is the tolerance on the capacitor that Tek specfied; that might give clues to allowable values.