Test Equipment Anonymous (TEA) group therapy thread

[mnementh]

So... this is how MY morning has been going.     How about YOURS? 

mnem
 That is my LAST custom-sleeved SATA cable.   

Ham fisted dwagon ? 

Today looks alright as I take in some caffeine but yesterday was a real bitch.........our main water pump broke, really broke. 
Castings smashed and gunna be a quite involved welding job that I really don't need right now. 
Much smaller backup pump installed.
Damn Murphy !

A little bit that and a little bit "I know I need 12" but all I can get is 10" on the SATA cables. Went to plug it into the MB, the other end broke.  So now i have some 18" cables in there until my 12" ones arrive and I can sleeve 'em up.

I remember in my youth, we had a jet pump in the barn freeze up; cracked right up the threading for the inlet nipple. After working the casting shut using a rosebud and the bench vise, I welded it up using 6013 rod on DC reverse. Had to do it about 3/8" at a time; weld a little, heat everything around the weld to relieve the stress, let it cool to "tsssst" hot and then weld another 3/8" or so. Got right up to the edge of the threads with my weld, then stitched back over itself for 1"; again, 3/8" at a time... but when it was done, I assembled with silicone RTV and it held water and drew up against a 22 foot head. For another 15 years. 

Amazing what you'll do on a Sunday afternoon when you're faced with schlepping water for a barnfull of animals via 5 gallon buckets.

mnem
"That's what SHE said..." *Ba-Dump-BUMMP-TSSHHT!!!*

[0culus]

He's talking about the balun on his 20m rig; it's effed up, meaning he's TXing all kinds of RF interference when he keys up.

mnem
Murphy just keeps punching me in the dick today...

OHHH. I totally misread that. 

[GregDunn]

So, while discussing another piece of HP equipment on the HPAK group, who should drop in but the engineer who designed the HP 3468/3478 meters and worked on the 3455 before moving on.  One thing led to another, and... well... my HP 3468A just arrived a few minutes ago.  Listed as "fair condition" it actually looks good except for the expected 40 years of mild scuffing on the case.  The front panel is fine.
I haven't run it through the full test suite yet, but DCV and Ω seem well within spec.  I'll open it up and check the battery later.

[med6753]

More playing around with the graphics on the Siglent SDM3055. This is the old 723 based 1.000VDC reference. I knew this reference drifted downward quite a bit for about a half hour then would settle down. I now use this reference just as quick check rather than a precision calibration source.

 

[Wolfgang]

this is the *first* 723 I have ever seen drifting downwards (after the settling period, of course). When you look at your 5 1/2 Siglent and the digits that move, maybe the meter is drifting upwards instead ?

https://electronicprojectsforfun.wordpress.com/silly-circuits/silly-circuits-a-heated-lm723-reference/lm723-long-term-stability-results/

[Neomys Sapiens]

My ADSL router was dropping out if I keyed down on 20m. This turned out to be that damn BNC connector in my balun I mentioned earlier. No shield continuity. Check this:
.....

A friend of mine bought a phantom-fed preamp for a scanner from the notorious Kuhne company.
Then he called me to tell that he still has mediocre reception without the preamp, and NO reception with the preamp, whether he powers it or not. Only available t&m gear: a multimeter. At least he was not averse to open the blasted thing. Turned out that the RF out/phantom power in N-connector had no continuity on THE CENTER CONDUCTOR. 
Cheapest piece of excrement I ever saw posing as a RF connector!

[bd139]

That really doesn't surprise me. I worked on a production line once. The connectors of choice on a module I was working on were MIL-STD-348 (TNC but more expensive). Assembly didn't work. Turned out that the supplier hadn't actually inserted the receptacle into the bulkhead mount connector. What did they do on the production line? Just not solder that because it wasn't there and sling it over to QA (me) to unfuck.

[med6753]

this is the *first* 723 I have ever seen drifting downwards (after the settling period, of course). When you look at your 5 1/2 Siglent and the digits that move, maybe the meter is drifting upwards instead ?

https://electronicprojectsforfun.wordpress.com/silly-circuits/silly-circuits-a-heated-lm723-reference/lm723-long-term-stability-results/

Nope, I have confirmed that with other DMM's. It does indeed drift downwards then settles.

[nixiefreqq]

That really doesn't surprise me. I worked on a production line once. The connectors of choice on a module I was working on were MIL-STD-348 (TNC but more expensive). Assembly didn't work. Turned out that the supplier hadn't actually inserted the receptacle into the bulkhead mount connector. What did they do on the production line? Just not solder that because it wasn't there and sling it over to QA (me) to unfuck.

wasn't trying to bust your chops.  am really curious what the cause of failure was in that bnc on your balun.

was something crimped that came lose, rather than the more common ring with a solder lug?


or was it something else?

[bd139]

It's my bad this one so I did post mortem. Basically as it's a balun it's spending its life 8m in the air up a pole in the pissing english rain. Figured I'd waterproof it. So the connector was assembled as these layers:

1. BNC shell
2. plastic hammond box
3. solder lug
4. lock washer
5. nut

When I then sealed this with epoxy, the epoxy worked its way between the lock washer and the solder lug and the ground connection went.

I just built another one and I did two changes to make sure this doesn't happen. Firstly this order:

1. BNC shell
2. plastic hammond box
3. lock washer
4. solder lug
5. nut

I also skipped using the solder lug entirely and just used it as pressure plate for the chassis/lock washer. I got the whole assembly tight then fluxed up the BNC shell and the nut and soldered the balun's internal coax straight to the BNC shell thus avoiding any joints which are pressure only.

Works nicely. Flat 1.0:1 SWR up to 10MHz. Then 1.1:1 up to 30MHz. No funky SWR on pole shorts either. Hopefully this will solve my HF problems.

Edit: My portable antenna is slightly less shit in this respect. The balun is made by just wrapping the feed round a FT43 and cable tying it to the center piece. Don't think that would last 2 minutes in a permanent install

Edit 2: I've also used liquid tape all over the outside of this only as it doesn't get sucked under anything via capillary action.

[Wolfgang]

this is the *first* 723 I have ever seen drifting downwards (after the settling period, of course). When you look at your 5 1/2 Siglent and the digits that move, maybe the meter is drifting upwards instead ?

https://electronicprojectsforfun.wordpress.com/silly-circuits/silly-circuits-a-heated-lm723-reference/lm723-long-term-stability-results/

Nope, I have confirmed that with other DMM's. It does indeed drift downwards then settles.

Spock would say: fascinating.  The next question is: who made your 723, and how long did you allow it to settle ?

[Neomys Sapiens]

That really doesn't surprise me. I worked on a production line once. The connectors of choice on a module I was working on were MIL-STD-348 (TNC but more expensive). Assembly didn't work. Turned out that the supplier hadn't actually inserted the receptacle into the bulkhead mount connector. What did they do on the production line? Just not solder that because it wasn't there and sling it over to QA (me) to unfuck.

I would rather define it the other way round, as this is where TNC was codified: so the others are like MIL-STD-348, but cheaper.

[Specmaster]

More playing around with the graphics on the Siglent SDM3055. This is the old 723 based 1.000VDC reference. I knew this reference drifted downward quite a bit for about a half hour then would settle down. I now use this reference just as quick check rather than a precision calibration source.

 

Med, I'm not sure if you have this right or I'm misunderstanding here, but looking at your photo, the reading at 0 seconds is lower than it is at -1 seconds, that to me is a rising reading rather than a falling one, or you infact saying that after the 30 minutes or so, the reading still than that shown at the 0 line?

[bd139]

That really doesn't surprise me. I worked on a production line once. The connectors of choice on a module I was working on were MIL-STD-348 (TNC but more expensive). Assembly didn't work. Turned out that the supplier hadn't actually inserted the receptacle into the bulkhead mount connector. What did they do on the production line? Just not solder that because it wasn't there and sling it over to QA (me) to unfuck.

I would rather define it the other way round, as this is where TNC was codified: so the others are like MIL-STD-348, but cheaper.

That's a fair assertion 

[med6753]

Med, I'm not sure if you have this right or I'm misunderstanding here, but looking at your photo, the reading at 0 seconds is lower than it is at -1 seconds, that to me is a rising reading rather than a falling one, or you infact saying that after the 30 minutes or so, the reading still than that shown at the 0 line?

The graph moves right to left. 0 seconds is "now", then 30 seconds ago, then 1 minute ago, etc. Therefore, the voltage is decreasing over time.

[Inverted18650]

beanflying: I received the package you sent em today, MANY THANKS BROTHA!

[med6753]

this is the *first* 723 I have ever seen drifting downwards (after the settling period, of course). When you look at your 5 1/2 Siglent and the digits that move, maybe the meter is drifting upwards instead ?

https://electronicprojectsforfun.wordpress.com/silly-circuits/silly-circuits-a-heated-lm723-reference/lm723-long-term-stability-results/

Nope, I have confirmed that with other DMM's. It does indeed drift downwards then settles.

Spock would say: fascinating.  The next question is: who made your 723, and how long did you allow it to settle ?

I don't recall who made the 723. I'll have to take it apart and look. It takes a good 30 - 45 minutes to settle down. Here's a copy of the schematic.

[nixiefreqq]

It's my bad this one so I did post mortem. Basically as it's a balun it's spending its life 8m in the air up a pole in the pissing english rain. Figured I'd waterproof it. So the connector was assembled as these layers:

1. BNC shell
2. plastic hammond box
3. solder lug
4. lock washer
5. nut

When I then sealed this with epoxy, the epoxy worked its way between the lock washer and the solder lug and the ground connection went.

I just built another one and I did two changes to make sure this doesn't happen. Firstly this order:

1. BNC shell
2. plastic hammond box
3. lock washer
4. solder lug
5. nut

I also skipped using the solder lug entirely and just used it as pressure plate for the chassis/lock washer. I got the whole assembly tight then fluxed up the BNC shell and the nut and soldered the balun's internal coax straight to the BNC shell thus avoiding any joints which are pressure only.

Works nicely. Flat 1.0:1 SWR up to 10MHz. Then 1.1:1 up to 30MHz. No funky SWR on pole shorts either. Hopefully this will solve my HF problems.

Edit: My portable antenna is slightly less shit in this respect. The balun is made by just wrapping the feed round a FT43 and cable tying it to the center piece. Don't think that would last 2 minutes in a permanent install

Edit 2: I've also used liquid tape all over the outside of this only as it doesn't get sucked under anything via capillary action.

oh......got it.   yep.  focus on solving one problem (water infiltration) and something else bites your ass as a result.

thanks for the details.  your pic really had me puzzled,  but now it makes sense. 

[Wolfgang]

this is the *first* 723 I have ever seen drifting downwards (after the settling period, of course). When you look at your 5 1/2 Siglent and the digits that move, maybe the meter is drifting upwards instead ?

https://electronicprojectsforfun.wordpress.com/silly-circuits/silly-circuits-a-heated-lm723-reference/lm723-long-term-stability-results/

Nope, I have confirmed that with other DMM's. It does indeed drift downwards then settles.

Spock would say: fascinating.  The next question is: who made your 723, and how long did you allow it to settle ?

I don't recall who made the 723. I'll have to take it apart and look. It takes a good 30 - 45 minutes to settle down. Here's a copy of the schematic.

Hi Med,

some points to rant about (dont take it personal):

- the bias current of the 723s difference amp is not balanced. One one side, you just connected it directly to the reference, the other input comes from the divider.
- The reference should be filtered.
- Yes, you are using low tolerance resistors (0.1%, 20ppm ?). But the trimmer resistor is too large in value and probably has a much worse tempco.
- Maybe a precision trim (parallel trimmer from VOut to ground, tap connected to 723 neg. input via a high-value resistor) is better here.
  Trim range is smaller and more accurate, DC resistance can be balanced with the input resistor at the reference side.
- some prestab because of limited 723 PSRR is also in order.

[Specmaster]

Med, I'm not sure if you have this right or I'm misunderstanding here, but looking at your photo, the reading at 0 seconds is lower than it is at -1 seconds, that to me is a rising reading rather than a falling one, or you infact saying that after the 30 minutes or so, the reading still than that shown at the 0 line?

The graph moves left to right. 0 seconds is "now", then 30 seconds ago, then 1 minute ago, etc. Therefore, the voltage is decreasing over time.

Oh, I got it wrong  I must admit that I thought after I pressed send. Having said that its drifting slightly, if that is the lowest it's gotten then it's pretty good and is certainly capable of calibrating a 5.5 digit meter as can be seen by your meters reading of 0.99997V, it's only the 6th and 7th digits are dropping down, that to my mind is a win.

[beanflying]

@med you are well into the voltnuttery hole watching trend plots of VRefs about seals the deal 

Next step LM399 Vref for more stability then realise you cant measure its drift without 6 1/2 digits .........

[Specmaster]

@med you are well into the voltnuttery hole watching trend plots of VRefs about seals the deal 

Next step LM399 Vref for more stability then realise you cant measure its drift without 6 1/2 digits .........

So in essence you're agreeing that it is OK for 5.5 digit meters

[beanflying]

@med you are well into the voltnuttery hole watching trend plots of VRefs about seals the deal 

Next step LM399 Vref for more stability then realise you cant measure its drift without 6 1/2 digits .........

So in essence you're agreeing that it is OK for 5.5 digit meters

Depending on how you want to use that reference. If you want to use it as a transfer standard then it needs to me more stable than the Meter you are going to so 3.5 and 4.5 no problems generally even for Chinese Cheapies.

If you are trying to look at the drift of the Ref at 5 1/2 places and beyond you need to make sure both the Vref and Meter are stable and at a fairly fixed Temp. The problem then becomes which Ref is drifting and by how much is it the Meter or the Vref or both? Initial Aging drift is an issue and as the link Wolfgang put above the plastic cases 'can be' an issue on some packages. So you need to look at the Ref over days and then repeat later what you found to see how the Ref is settling. It becomes more of a probability and sadistics issue more than just a simple flick it on and test issue.

[med6753]

@med you are well into the voltnuttery hole watching trend plots of VRefs about seals the deal 

Next step LM399 Vref for more stability then realise you cant measure its drift without 6 1/2 digits .........

So in essence you're agreeing that it is OK for 5.5 digit meters

Depending on how you want to use that reference. If you want to use it as a transfer standard then it needs to me more stable than the Meter you are going to so 3.5 and 4.5 no problems generally even for Chinese Cheapies.

If you are trying to look at the drift of the Ref at 5 1/2 places and beyond you need to make sure both the Vref and Meter are stable and at a fairly fixed Temp. The problem then becomes which Ref is drifting and by how much is it the Meter or the Vref or both? Initial Aging drift is an issue and as the link Wolfgang put above the plastic cases 'can be' an issue on some packages. So you need to look at the Ref over days and then repeat later what you found to see how the Ref is settling. It becomes more of a probability and sadistics issue more than just a simple flick it on and test issue.

I built this reference about 30 years ago and at that time all I had was a Fluke 77 so it was more than adequate to check it's calibration. Fast forward to now. It no longer cuts the mustard. As a DC reference it's obsolete but it does have an 1.00VAC reference that I still use (because it's all I have at the moment).

[Specmaster]

Unless the drift is sufficient to alter the last digit on a 5.5 meter, so if at switch on it reads for the 1V range it reads 0.99997V and say an hour later it still reads the same 0.99997V, why is that not good enough for meters upto 5.5? Cleary on 6.5 and higher then the error would be visible as that digit on the photo shown has drifted from a 9 to 2