Test Equipment Anonymous (TEA) group therapy thread

[mnementh]

WELCOME TO 5000!           

From the Agilent 54654N diagrams, that were linked to recently.
The monitor cable has three connections for +15V and ground, six total.
Three connections for the intensity pot.
Last four connections are half bright, full bright, HS (Hsync?) and VS (Vsync?).
Interesting the HS goes through a 74393 to provide the probe compensation signal.
HS & VS come from the big Jackal custom IC and the half bright, full bright from a PAL IC.

Other relevant pages have been cropped, rotated & added below.

This forum post suggest the Hsync is 19.72kHz and the Vsync is 60Hz.
https://www-elektronik-si.translate.goog/phpBB2/viewtopic.php?t=30373&postdays=0&postorder=asc&start=30&sid=c8ec2c01726d6cc1ee123fe5ebbe8c3d&_x_tr_sl=auto&_x_tr_tl=en&_x_tr_hl=en-US&_x_tr_pto=wapp

Guess I need to locate those on my 54615B, to try & narrow down where my intermittent horizontal roll problem is, but it's not the same board.

David

Okay, I've traced out the 54621D and the VCC, GND are in the same locations, plus Pin 10 dssig_grd is also tied directly to GND. However the BRIGHTNESS pot is not brought out to this header on the 54621D.

   
                        VSync                                                    HSync 
   
My guess is that 13/14 are still the sync signals, and they have repurposed 7, 8, 9, 11, 12 for their greyscale scheme. 54645A agrees; trace 1 is the 54621D, trace 2 is the 54600A.


   

However, pin 8 is a different story; all I'm getting from the 54600A is noise.   

Then the light dawns; the pot isn't connected to the monitor PCB, duh. Then it occurs to me the voltages usually present on a brightness pot, so I connect the cable back in the 54600A and probe the pot. Yup... (-58V) sent out pin 9 to the pot. Without knowing how the 54621D does its greyscale, I'm not too sanguine about connecting that (-58V) back to its mainboard.

I think this is probably a good point to call it a night.

mnem
 

Stay tuned.   

mnem
       
*yawwwwwnnnn* "This will probably be at least a 3-coffees move."

[mnementh]

... being their biggest focus rather than staying in their lane and sticking to their knitting ...

Would Sir like his metaphors diced, or scrambled?    (Only Mickey Spillane is allowed to have his metaphors hard boiled.)

"Please Hammer, don't hurt 'em!!!"

mnem
*fulminating at the mind again*

[mnementh]

       
*yawwwwwnnnn* "This will probably be at least a 3-coffees move."

Allright... maybe slightly more than 3 coffees... The mouse on my workshop PC has developed a hinky left button, resulting in very random double-clicks and dropouts while highlighting.

mnem

[BU508A]

Repair Shimano EC-E6000 Battery Charger

A friend of mine brought me the other day a battery charger from Shimano.
Failure description: Doesn't work anymore.

Opening was easy, just four screws.
First thing I've measured was the fuse. Jep. fuse open.

Next step: measuring the conductivitiy between the + and GND rail after the rectifier. Jep, 0.02 Ohms.

A closer visual inspection revealed two exploded LCS703H. Ordered a bunch of five at Mouser, they'll arrive next week.

And here are some pictures:

Plastic case, easy to open:


Top and bottom view of the PCB:


One of the exploded LCS703:


Parts overview and detailed view. Interesting, those LCS703H had a separate plastic case on top.
An annoying thing was: they've putted some Loctite on the screws. Had to be very careful to get them out.

Edit:
Forgot to add the details of the parts:
- bridge rectifier: D15XB 60 B1362
- 2x LCS703H
- fuse: T6.3AH 250V, 5x20mm
Anybody an Idea where to get those fancy caps for the fuse with those legs?

I've attached the datasheet of the LCS703H below.

[bd139]

Ooh explodey. Hope it works afterwards. I love how the fuse goes about 200ms after every semiconductor goes pop in SMPS designs. Fuses are literally just fire prevention these days.

Edit: trying not to buy that 3478A I don’t need at the moment 

[BU508A]

I'm having here some NOS LT1185CT (negative LDO from Linear Technology).

From the datasheet (attached below):
"It can be used as a positive output regulator with floating input or as a standard negative regulator with grounded input."

It can do 2.5V to 25V output and current up to 3A.

Example:


I'm having 5 lots of 4 LT1185CT per lot and would ask 16 EUR plus shipping per lot.
If you are interested, please PM me.
Side note: Mouser is currently out of stock and is asking ca. 10 Euro per LT1185CT.

Overview:


Detail:


Edit:
Added this offer to the buy/sell/wanted section:
https://www.eevblog.com/forum/buysellwanted/fs-five-lots-of-four-lt1185ct-per-lot-voltage-regulators/msg4290862/#msg4290862

[Specmaster]

       
*yawwwwwnnnn* "This will probably be at least a 3-coffees move."

Allright... maybe slightly more than 3 coffees... The mouse on my workshop PC has developed a hinky left button, resulting in very random double-clicks and dropouts while highlighting.

mnem

And that ladies and gentlemen is why Dragons are now extinct, they sat way too close to the computers they were trying to delete with far more than 50gm of explosives   

[AVGresponding]

Repair Shimano EC-E6000 Battery Charger

A friend of mine brought me the other day a battery charger from Shimano.
Failure description: Doesn't work anymore.

 

Parts overview and detailed view. Interesting, those LCS703H had a separate plastic case on top.

Obviously they realised they needed additional blast protection...

Anybody an Idea where to get those fancy caps for the fuse with those legs?

All the usual suspects, I should think?  https://uk.rs-online.com/web/p/fuse-holders/0379274

[BU508A]

All the usual suspects, I should think?  https://uk.rs-online.com/web/p/fuse-holders/0379274

Thanks! This gave me some hints where to look and found some caps from ESKA:

https://www.ebay.de/itm/194817117353

[xrunner]

Ready to put in my Ryzen 5 5600G 

[mnementh]

TBH if you can be bothered to cobble something together use the oz2cpu/Defpom method.
Why, because of mwyatt's contribution simulating my measurements to come up with the values of the passives in the networks of the OEM probe.
These things OEM whom as yet I've not found who makes them are damn expensive but don't forget a 20 lot of decent  quality grabbers alone will set you back a C note !

Or if you not in a hurry, keep an eye out for one that looks identical but made for LeCroy, Tek or Rigol and a little sleuthing on each of their websites will get you the Pt. # to look for.......then post them in the DIY thread for the benefit of others. 

I'm 100% they all have the same OEM and the twin thin silicone 8 way ribbon coax (that's a mouthful ! ) is a thing of beauty and near impossible to emulate unless you got mates in the company where it's made to get offcuts !
That's why Defpom went the route he did so to get something nice for his efforts rather than some cobbled together twisted pair mess.

Last year Siglent were specialing SPL2016 + MSO/LA and AWG/FG licensing with a new SDS2000X Plus for just $ 219 that was one hell of a deal compared to shelling out retail for each.

Actually, I was thinking to make a header so I could plug the HP pod and whip leads directly into the oz2cpu scope PCB. As I said, those will probably be more than long enough for anything I'd be doing, unless I change my mind and put the scope up higher on my bench, which is pretty unlikely ATM.

I was asking about MK's project since you linked to it at the beginning of the SDS2000X thread. Not knowing what the difference is between the Siglent frontend and the HP braided cables, I'm leery of just plugging random lengths of either braided cable or micro-coax into the front of the scope without any compensation, or the wrong compensation. I saw the issues oz2cpu was having there in the beginning.

mnem
 

OK so if you already have the HP pod den go widdat !
Be aware you only need get something 1/2 clean into the LA input terminals as what's shown on the display in Digital mode bares little resemblance to the actual signal. IIRC there's discussion about this in the DIY thread.
IMO coax LA probe set compensation is likely to be pretty much an industry standard however the truth of the matter is suck it and see.
But the fact that we don't see cries that 'it don't work properly' for any of these DIY creations which points to them being fairly forgiving.

General principle: assure yourself that you have sufficient analogue signal integrity. Then flip to the digital domain with an LA ot printf() etc.

If you do that then there is a good chance that there is sufficient leeway on the analogue waveform for it to be correctly interpreted as a digital signal, even with a crappy termination. That, after all, is a main benefit of digital signals.

The exception could be with the clock signal, especially if the other signals are close to setup and hold time limits.

Pretty sure the LeSiglent doesn't have the CLOCK input that older LAs use to sync everything; tho I suppose I could break that out to the EXT Trig... might be useful in some cases. 

I have been looking for a "correct" set... but the other brands are as expensive or even more so... I even looked at the older SPL1016 and they're still $170 shipped. By the time I get all that happening and the oz2cpu scope PCB built, still looking at as much money as I have in the scope itself.

Meanwhile, the HP Pods by themselves are usually available for $25-35. Seems a no-brainer, if you don't need a braided cable long enough to reach to the back of a Pliocene-era LA. 

A good compromise for hobbyist use, no? 

mnem
 

[bd139]

Ready to put in my Ryzen 5 5600G 

Fingers crossed you updated your BIOS first 

[mnementh]

       
*yawwwwwnnnn* "This will probably be at least a 3-coffees move."

Allright... maybe slightly more than 3 coffees... The mouse on my workshop PC has developed a hinky left button, resulting in very random double-clicks and dropouts while highlighting.

mnem

And that ladies and gentlemen is why Dragons are now extinct, they sat way too close to the computers they were trying to delete with far more than 50gm of explosives   

"I'm fireproof. You're not." ~Hellboy

mnem
*nuking the mouse*

[mnementh]

Yeah I would not bother either ! Well I would not own such a piece of junk to begin with.... but as I explained I am under a moral obligation to go the extra mile to rescue this thing.


However after what you and Dwagon said, it's realllly looks like it's that cheap disposable CPU that's toast USB wise, so the "extra mile" is 99% covered already....

All I can do now to be 100.0000% sure it's the CPU, so I can look my friend in the eyes... is to make 100% sure there is zero discrete component between the CPU and the USB connector power pin.

So I need to find the datasheet for the CPU... and sadly I just spent 10 minutes on Google and NO LUCK ! This CPU is so crap apparently that you can't even find a datasheet for it....

So looks like I have reached the end of the road.... pfff....

People keep thinking/assuming somehow that electronics and simple and can be fixed easily in 2 minutes for free. I keep telling them that no, it's the other way around : there is a millions things that can prevent you from repairing a device, and that a quick cheap repair is not the rule it's the exception ! But when I tell them that they look at me suspiciously, like they are thinking I am either incompetent or trying to find excuses because I am lazy and don't work to spend  any time on their stuff...

Had that for years. I impolitely explain my time is worth more than the device is worth. That usually gets the point across. Same with computer problems.

Had a recent one with my uncle who wanted his laptop repaired. It’s about 8 years old and quite frankly disgusting. I don’t even want to touch it. I impolitely explained that it doesn’t matter what you want me to do you will have to pay me the hourly rate for what I could have been doing instead because my time isn’t worth it. He bought a new one.

And that’s the reality of a lot of repairs these days. I only play inside old test gear for fun.

Yup. Same for me and old computers. A while back I bought a "WalMart Special" Acer laptop with  N2840 CPU/4GB RAM/500GB HDD. I took on the usual warnings in here that it wasn't worth the money to fix it up, but I wanted to see what was inside of it and fix it up to keep my diag skills sharp.

Right now I'm in it for $67 and it runs half-decent but still spinning rust slow; I'm toying with the idea of popping in a old 120GB SSD I unearthed the other day to make it more attractive before I flip it.

But this was all for the fun of fixing and keeping my skills sharp; I know when I do flip it I'll get $100-125 for it, so not even breaking even on my time.    But the difference is it's my toy, my time and my choice whether to spend it that way or not.

mnem
 

After reducing it twice, sold for $90. With the 120GB SSD in it. Plus I kept the 500GB WD Black spinning rust disk out of it to store movies on.

mnem
"Still beats a sharp stick in the eye." ~gramma Eddurds

[xrunner]

Ready to put in my Ryzen 5 5600G 

Fingers crossed you updated your BIOS first 

Yea I did, I got burned on that mistake once. Once bitten twice shy. 

I'm using it right as we speak. After a few restarts for typical driver updates it's running fine.

We're cooking with gas now baby!

[bd139]

Ready to put in my Ryzen 5 5600G 

Fingers crossed you updated your BIOS first 

Yea I did, I got burned on that mistake once. Once bitten twice shy. 

I'm using it right as we speak. After a few restarts for typical driver updates it's running fine.

We're cooking with gas now baby!

Hahaha yeah me too. Glad to hear it’s up and running. Enjoy 

[ch_scr]

Ready to put in my Ryzen 5 5600G 

Fingers crossed you updated your BIOS first 

Yea I did, I got burned on that mistake once. Once bitten twice shy. 

I'm using it right as we speak. After a few restarts for typical driver updates it's running fine.

We're cooking with gas now baby!

Good on you! Maybe undervolt it ever-so-slightly I was real shocked how high the stock core voltage is on these in "normal everyday single core boost".
Seems you are lucky enough to have a recent enough AGESA version in BIOS to profit from the "Curve optimizer" undervolting. If these settings are not shown in bios, you still can access them from the Ryzen master tool which will adjust the BIOS accordingly. The undervolt can be set up to 30 "points" which equal about 100 to 150mV - an undervolt of about 10 points should be a conservative starting point and well good enough to cap the spikes.

[bd139]

I’d just plug it in and use it 

[ch_scr]

I’d just plug it in and use it 

With a mini PC you have to fight for every bit of performance per watt, but it's OK - I know the drill from my Intel laptop
...CPU "nominal" power consumption numbers are almost lies, just almost...
Edit: In other news, my long-awaited favorite mint tea was back in stock - now I can finally have again mint tea so strong, there is almost a layer of oil on top 

[mnementh]

                    *yawwwwwnnnn*  "This will probably be at least a 3-coffees move."

Allright... maybe slightly more than 3 coffees... The mouse on my workshop PC has developed a hinky left button, resulting in very random double-clicks and dropouts while highlighting...

On the Bench This Morning: Nuking the Mouse

So old faithful crusty old Dell optical mouse started giving me fits this morning; or at least the fits it was having got consistent enough that I actually noticed. 

Since I'm at the bench already, I decided to minimize wasted time searching for a replacement I'm not even sure I have ATM and just service the thing using this process I learned and developed while doing fleet maintenance at the ISD. Once you get to where you know what you're doing and you have "the usual stuff" right out on your bench, it literally takes 10 minutes. Provided you aren't stopping to snap pics and do a write-up.

This mini-DIY applies to pretty much any situation where you might have one of these micro-switches and don't have a replacement handy, or are stuck without means to desolder the switch; they actually usually can be made pretty much 100% good again at least once or twice following this procedure.

I'm going to start right at the point of actually "reconditioning the switch"; I'm reasonably sure everyone in here already knows how to dismantle a mouse and desolder a switch. I had to take my mouse apart again to get some of the high-detail pics (you're welcome ); this makes continuity seem weird. You can do this without desoldering the switch in some cases where things like the scroll decoder aren't in the way.



Pop the top off by slipping a knife-point under these tabs on both sides of the switch.




Remove the leaf-spring contact by unhooking from the contact hook in switch body. Pull gently and upward slightly using dental pick or similar instrument.




There will be a little silver contact on the leaf-spring; the most common failure mode is this gets a little divot worn in it compounded by tarnish build-up. I didn't get a good shot for the before pic, I'm afraid. 

The cure is to buff it down with very fine sandpaper (like 800-2000 grit); I use this fingernail buffing board. Yeah, I know it's pretty horrible looking; it's prolly older than my daughter.




Next step is to retension the recurve spring by squeezing gently in flat-jawed pliers. About 1/2 to 3/4 compressed is all it takes to get good tension again; if you flatten it too much the hardened brass leaf will crack.




Next is to clean the lower contact; I do this by scrubbing with a fine-bristle wire brush. You'll need to angle the brush to get under the keeper hook and scrub the lower contact with the side of the wires.




After that, I clean everything; leaf-spring and switch base, and inside the cap, with a paintbrush and IPA. Here you can see the lower contact and keeper hook a little more clearly. When working on the bench, I'll accelerate the drying with very conservative application of a hot air gun.




Now it's time to reassemble the micro-switch; this can be done by hand (once you get to where you know what you're doing) or with tweezers. First you get the contact under the keeper hook, then you catch the end of the recurve spring under the hook in the middle. Then you pull the end of the leaf spring down over the shoulder of the outer contact hook into the first notch.




You can now test the switch before reassembling; by pressing down here between the two contact hooks in the switch base (there will usually be a visible wear/polish line from the switch plunger), you can make the switch actuate as in normal use. It should have a crisp, positive detent action without feeling crunchy or snaggy. If you've assembled it correctly, you do not have to worry about the leaf-spring flying off.




Now it's time to reassemble the cap onto the switch. Always reassemble with the cap upside-down so you don't lose this little gawddammitt. It will escape and be lost forever in the middens under the bench, desk, bedroom etc if given any chance to do so.




And last step, A little PM. Yeah, okay... the USB cable hasn't broken a wire yet and left you mouse-less. By relocating the captive portion of the wire so that the old "strain-point" is now kept inside the mouse, you can make it so the strain is now primarily concentrated in a new area of the cable, thereby increasing its future life expectancy. All it takes is to pull the cable back into the mouse 20-30mm or so, then tuck the excess away like so.

After that, finish up by reassembling the shell and screws, then the usual wipe it all down ith IPA, especially the Teflon slider pads to clean off the body-oil grunge that makes it not slide like it should.

Then you can test it like I did: by editing a mess of pics, and spending 3 hours to chronicle a 10-minute fix. 

mnem

   *mouse-ily again*   

[ch_scr]

Many thanks for the detailed writeup, mnem! Wasn't aware refurbishing the micro-switches was a thing or how they looked inside, I just stole the right buttons from dead mice to have a bag of replacements so far. 

[Zoli]

The cure is to buff it down with very fine sandpaper (like 800-2000 grit); I use this fingernail buffing board. Yeah, I know it's pretty horrible looking; it's prolly older than my daughter.

For silver tarnish removal an eraser does wonders, without removing any of the underlying silver.

[mnementh]

The cure is to buff it down with very fine sandpaper (like 800-2000 grit); I use this fingernail buffing board. Yeah, I know it's pretty horrible looking; it's prolly older than my daughter.

For silver tarnish removal an eraser does wonders, without removing any of the underlying silver.

You want to remove some of the silver, to remove the divot (wear crater) and make the contact flat again. As you can see in the pic, there's more than enough material to do this several times. This leaf spring is made of thin hardened brass, so pretty fragile; trying to rub it with a pencil eraser hard enough to remove anything is much more likely to mangle it than do any good.

mnem
 

[mnementh]

Many thanks for the detailed writeup, mnem! Wasn't aware refurbishing the micro-switches was a thing or how they looked inside, I just stole the right buttons from dead mice to have a bag of replacements so far. 

Yeah, I usedta keep a bin of deaders in my office at the ISD to rob cords, switches, scroll wheels, etc off of. But I found that A) I could often service the switch while on the ticket using just what was in my Little Bag of Dirty Tricks™ in less time than it took to walk down there for a replacement and 2) Once I knew how to do it, was literally less hassle than desoldering the switch most of the time, and C) I still had that as a back-up plan.   

mnem

[Specmaster]

@mnementh
Cracking job done with that write up, well detailed thank you.