Test Equipment Anonymous (TEA) group therapy thread

[bd139]

In the interest of not cocking up this 3478A cal like the first one, I have built a better quality "heart bypass" for it. Probably overkill but I'm avoiding making the same mistake again. 78L05 + caps + series diode in this case. But there's some careful design thoughts going into this one. The voltage is higher than the battery backup which means you can measure and validate whether it's connected or not before removing the original battery. Also it is shrink wrapped meaning less short possibility (what I did wrong last time I think).





So the plan of action is:

1. Check new battery voltage before installation in case it is faulty.
2. Disconnect from the mains.
3. Mask the area with kapton before performing surgery to avoid shorts.
4. Measure the voltage across C510 and note as VOLD.
5. Solder the bypass unit across CR764 cathode and ground and then install the battery
6. Measure the voltage across C510 and note as VNEW.
7. Validate VNEW > VOLD
8. Desolder old battery.
9. Install new battery.
10. Measure voltage across CR764 anode and ground and confirm >0 and <VNEW
11. Remove battery from bypass unit then remove bypass unit.
12. Measure voltage across C510 and validate >0.
13. Power up and hopefully not cry myself to sleep

Any flaws anyone can see?

schematic (compiled) showing the OR'ed together +5v and backup supply. The "bypass supply" has another diode meaning 3 input OR.

[mnementh]

Your BOM needs to include one shot of single-malt to steady your hand. 

mnem
 

[tautech]

Your BOM needs to include one shot of single-malt to steady your hand. 

mnem
 

Just one ? 

[Specmaster]

In the interest of not cocking up this 3478A cal like the first one, I have built a better quality "heart bypass" for it. Probably overkill but I'm avoiding making the same mistake again. 78L05 + caps + series diode in this case. But there's some careful design thoughts going into this one. The voltage is higher than the battery backup which means you can measure and validate whether it's connected or not before removing the original battery. Also it is shrink wrapped meaning less short possibility (what I did wrong last time I think).





So the plan of action is:

1. Check new battery voltage before installation in case it is faulty.
2. Disconnect from the mains.
3. Mask the area with kapton before performing surgery to avoid shorts.
4. Measure the voltage across C510 and note as VOLD.
5. Solder the bypass unit across CR764 cathode and ground and then install the battery
6. Measure the voltage across C510 and note as VNEW.
7. Validate VNEW > VOLD
8. Desolder old battery.
9. Install new battery.
10. Measure voltage across CR764 anode and ground and confirm >0 and <VNEW
11. Remove battery from bypass unit then remove bypass unit.
12. Measure voltage across C510 and validate >0.
13. Power up and hopefully not cry myself to sleep

Any flaws anyone can see?

schematic (compiled) showing the OR'ed together +5v and backup supply. The "bypass supply" has another diode meaning 3 input OR.

Just watched this video of a 3478A having the battery swapped out. Idiot is doing this with the power switched on, and is using solder wick and a brush to clean up the solder pad with almost zero bristles in it, I was waiting for the sparks to come, but they didn't, he got lucky.

[bd139]

Yeah i watched the “powered up” replacements and wondered if they actually thought about it first 

I will have a drink or two first for sure 

Think UPS have managed to lose another fucking package. They’re fired.

Also farnell screwed up. Three separate consignments have been generated from 16 line items. Two dispatched one still processing  . Same day dispatch my fucking arse.

Also bought a ream of paper from Amazon. Turns out it’s shipping from Koln. Landed at castle donnington earlier today. Surely there’s an easier way for a £6 lump of paper?!?!?

Everything so damn stressful at the moment.

[vk6zgo]

Just a point on EMC and OEM PC builds. The general assumption is if all the components are independently tested then the final assembly is as well. Under EMC Directive at least if all components are CE marked then job done. If your device is found not to be compliant then push the problem upstream.

No refering to you but the assumption- it's total  bullshit.
A. Components should not be CE marked (there are exceptions like safety switches, but not relevant here)
B. Unless you use the exact configuration used for testing you can't ensure compliance.
C. It is the manufacturer or importer who is responsible for compliance. If you/they get fined good luck trying to get the board suplier to cover your cost. You certainly can't pass the prosecution on to them, they will soon point out the components do not have to comply, see A.
D. How do you CE mark a case for EMC just sit it in a chamber with no power  (the case is vital to PC EMC)

Even if an OEM has correctly desigined manfactured and test a product and it is compliant they have to show onging production is compliant. This included ensuring any changes in component specifications do not havee an adverse effect.

The big issue is there is virtually no enforcement and in th UK the agency charged with enforcement has no inhouse expertise or funding to get it.   

There is a story that HP had trouble with EMC on some VME systems, reportd by he users. Long story short due to local regulations the grounding finger strip on front panels of cards made in europe had different plating to those from the USA. All EU or all USA card stups were fine but if you mixed cards there wew poor bonding and passive intermodulation problems! Even when you try hard it can still fail.

Well aware it’s bullshit but that’s the line you get fed.

Ofcom won’t enforce anything as they sold all their kit off a few years back 

An ACMA luminary some years back was quoted as saying something along these lines:-
"Our job is to facilitate business in their use of the spectrum."

Importer:-
"We want to import this non-complying crap."

ACMA (Rolling over):-
 "Rub my tummy!"

[Specmaster]

Actually, Amazon have been good for me of late, I even got those Basic USB-A to Lightning leads the same day on Sunday, call me impressed.

[cyclin_al]

It does have option 001 (8th digit), but maybe not factory installed, or the sticker fell off. 



David

Good point, it does have the 8th digit.  I was only looking for the sticker on the back...

[cyclin_al]

In the interest of not cocking up this 3478A cal like the first one, I have built a better quality "heart bypass" for it. Probably overkill but I'm avoiding making the same mistake again. 78L05 + caps + series diode in this case. But there's some careful design thoughts going into this one. The voltage is higher than the battery backup which means you can measure and validate whether it's connected or not before removing the original battery. Also it is shrink wrapped meaning less short possibility (what I did wrong last time I think).

Note to self:  I still need to check out the 3468A I picked up at the end of last summer    Seven months already without looking at it.

Which battery did you choose for it?

[bd139]

Varta 3v lithium manganese dioxide axial leaded 1/2 A sized (25mm len by 14.5mm dia). Closest to the original.

[Mr. Scram]

Back 20 years ago, people did not mount their computers in 19" racks as much as we do today. There was this concept of a "tower server". and some machines could have a rack mount kit attached to it -- I've converted a bunch of yellow pre-hp Compaqs from tower to rack, once.

I do not miss those days.

20 years ago is three centuries in IT.  Though I must admit I just remembered seeing a tower "server" comparatively recently. The idiots got their orders confused and bought a Dell workstation. It's probably still taking up nearly half a rack today.

[Specmaster]

Varta 3v lithium manganese dioxide axial leaded 1/2 A sized (25mm len by 14.5mm dia). Closest to the original.

That won't be a problem, you can make it fit if the leads are long enough.

[mnementh]

Your BOM needs to include one shot of single-malt to steady your hand. 

mnem
 

Just one ? 

No more, no less. 

Additional shots should be part of the post-op In-Service Prep BOM; look under the Parts Cleaner line item.   

mnem
It's 4:20 somewhere...

[mnementh]

In Other News...

I just survived an hour out front in the driveway, missing baskets with my son. I think I'm gonna curl up in bed with a nice warm fistful of pain meds...



mnem
somebody call me a ambulance

[vk6zgo]

Your BOM needs to include one shot of single-malt to steady your hand. 

mnem
 

Just one ? 

I'd rather go with several b--i--g  glasses of Bundy Rum & Coke!
On second thoughts,-----hold the Coke!

[beanflying]

In the interest of not cocking up this 3478A cal like the first one, I have built a better quality "heart bypass" for it. Probably overkill but I'm avoiding making the same mistake again. 78L05 + caps + series diode in this case. But there's some careful design thoughts going into this one. The voltage is higher than the battery backup which means you can measure and validate whether it's connected or not before removing the original battery. Also it is shrink wrapped meaning less short possibility (what I did wrong last time I think).


So the plan of action is:

1. Check new battery voltage before installation in case it is faulty.
2. Disconnect from the mains.
3. Mask the area with kapton before performing surgery to avoid shorts.
4. Measure the voltage across C510 and note as VOLD.
5. Solder the bypass unit across CR764 cathode and ground and then install the battery
6. Measure the voltage across C510 and note as VNEW.
7. Validate VNEW > VOLD
8. Desolder old battery.
9. Install new battery.
10. Measure voltage across CR764 anode and ground and confirm >0 and <VNEW
11. Remove battery from bypass unit then remove bypass unit.
12. Measure voltage across C510 and validate >0.
13. Power up and hopefully not cry myself to sleep

Any flaws anyone can see?

schematic (compiled) showing the OR'ed together +5v and backup supply. The "bypass supply" has another diode meaning 3 input OR.

Can the Cal data be read off the unit via GPIB or RS232? Haven't looked but I know this is possible on some of the HP gear.

[vk6zgo]

In Other News...

I just survived an hour out front in the driveway, missing baskets with my son. I think I'm gonna curl up in bed with a nice warm fistful of pain meds...



mnem
somebody call me a ambulance

When my son was young, I put up a "regulation height" basket & backboard on my carport.
What I neglected to consider is that our driveway slopes fairly sharply downwards away from that point.

Everybody wondered later why he could leap higher than most of his taller team mates.

Being up on the ladder, juggling the massive bits of the basket & backboard was more than enough exercise
for me---I only tried shooting baskets a couple of times before I decided it wasn't my thing.

[mnementh]

In the interest of not cocking up this 3478A cal like the first one, I have built a better quality "heart bypass" for it. Probably overkill but I'm avoiding making the same mistake again. 78L05 + caps + series diode in this case. But there's some careful design thoughts going into this one. The voltage is higher than the battery backup which means you can measure and validate whether it's connected or not before removing the original battery. Also it is shrink wrapped meaning less short possibility (what I did wrong last time I think)...

Can the Cal data be read off the unit via GPIB or RS232? Haven't looked but I know this is possible on some of the HP gear.

https://www.eevblog.com/forum/repair/hp-3478a-how-to-readwrite-cal-sram/

mnem
 

[Neomys Sapiens]

woe is me, I am out a fairly high number of pesos for an AIM I-Prober 520 ...

One really enticing device! Had a salesrep in last year who did show it live. Bandwidth is a tad low, but it is on my list of gadgets to aquire. If you do not have a hall probe yet, it makes even more sense.

[mnementh]

In Other News...

I just survived an hour out front in the driveway, missing baskets with my son. I think I'm gonna curl up in bed with a nice warm fistful of pain meds...

mnem
somebody call me a ambulance

When my son was young, I put up a "regulation height" basket & backboard on my carport. What I neglected to consider is that our driveway slopes fairly sharply downwards away from that point. Everybody wondered later why he could leap higher than most of his taller team mates. Being up on the ladder, juggling the massive bits of the basket & backboard was more than enough exercise for me---I only tried shooting baskets a couple of times before I decided it wasn't my thing.

   
https://www.eevblog.com/forum/testgear/test-equipment-anonymous-(tea)-group-therapy-thread/msg3098667/?topicseen#msg3098667

Oh, this was a premeditated plan to get us both much needed exercise. We rescued and rehabilitated a curb-score basketball goal; this was in the worst of the first round of COVID lockdowns, when nobody had anything in stock. I had to order lumber and hardware online, wait a day, then go pick it up curbside, etc... a total dicksore to make it happen. The actual repair work went well; good father-son time and we both love the result.

mnem
 

[mnementh]

Your BOM needs to include one shot of single-malt to steady your hand. 

mnem
 

Just one ? 

I'd rather go with several b--i--g  glasses of Bundy Rum & Coke! On second thoughts,-----hold the Coke!

That definitely belongs on the Parts Cleaner line item... 

mnem
 

[Neomys Sapiens]

Anybody out there willing to sponsor me with 10 millos ?
https://www.bernard-gallay.com/buy-yacht-for-sale/hugo-boss

*Sigh* *dream*

Oh come on - what's up with you? It does not have a single MEKO slot!

[tautech]

Your BOM needs to include one shot of single-malt to steady your hand. 

mnem
 

Just one ? 

I'd rather go with several b--i--g  glasses of Bundy Rum & Coke! On second thoughts,-----hold the Coke!

That definitely belongs on the Parts Cleaner line item... 

mnem
 

Yeah not a huge fan of Bundy rum either however......as Bambi season is fast approaching a procession of shooters are arriving to use our range for last minute checks to ensure they can hit Bambi at whatever range and turn up with expressions of Koha (gratitude) mostly in glass bottles. 

An all time fav materialized this morning ... a 40oz of Devils Cut which Larry Senior introduced me to on a visit to see him in Perth. Devils and Dry is something else and after decades of passing on wisdom to Larry Snr it's nice to get some back ! 
Wee sample coming later tonight after all jobbies are done. 

[Neomys Sapiens]

A LOT cheaper  It's basically a 1Mb/s transmission loss tester for 78R balanced screened network cables. It a current model and a new one from Ultra Electronis will cost about £20,000  well worth 99p 
https://www.ultra-pcs.com/air/ground-test-equipment/databus-network-analyser/

Jammy git!

Double stuffed jammy
 - UPS delivered a package today and I thought what the heck, I didn't order anything that big and heavy.
Opened it up and inside are two test sets. OK one does not have any leads or remote unit who's complaining when it's a freebe on a 99p purchase  Just the "RAE" instrumen case is worth that to me. In addition there is a nice toroid based PSU, battery pack  two 10T pots (500R & 10k), switches etc. There are some useful bits on the board including a AD636 RMS /DC converter.
Oh and the working one had a couple of spare leads with 4 three lug triaxial BNC plugs.

Grrrr....those cables alone!

[beanflying]

In the interest of not cocking up this 3478A cal like the first one, I have built a better quality "heart bypass" for it. Probably overkill but I'm avoiding making the same mistake again. 78L05 + caps + series diode in this case. But there's some careful design thoughts going into this one. The voltage is higher than the battery backup which means you can measure and validate whether it's connected or not before removing the original battery. Also it is shrink wrapped meaning less short possibility (what I did wrong last time I think)...

Can the Cal data be read off the unit via GPIB or RS232? Haven't looked but I know this is possible on some of the HP gear.

https://www.eevblog.com/forum/repair/hp-3478a-how-to-readwrite-cal-sram/

mnem
 

Hadn't seen that thread but looks like a good thing to do just in case of 'human error'