|
I have an interesting phenomenon on my Agilent 54641D MSO. Upon cold powerup the trace of channel 1 is "all wiggly". The wiggliness is highly sensitive to mechanical movement of the BNC socket of channel 1. See photo below. After (say) 10 minutes the problem goes away, stressing the BNC no longer has a visible effect. The scope works just fine once warm. Channel 2 has none of this nonsense :-) Any ideas? I am toying with the assumption that the socket contacts of the channel 1 hybrid are dirty/temperature sensitive. But that is more like a wild guess. Wilko 
|
Does it do that if you try moving the input with a non conductive tool? Or just your fingers??
toggle quoted message
Show quoted text
On Feb 19, 2023, at 1:26 PM, Wilko Bulte <wkb@...> wrote:
?I have an interesting phenomenon on my Agilent 54641D MSO. Upon cold powerup the trace of channel 1 is "all wiggly". The wiggliness is highly sensitive to mechanical movement of the BNC socket of channel 1. See photo below. After (say) 10 minutes the problem goes away, stressing the BNC no longer has a visible effect. The scope works just fine once warm. Channel 2 has none of this nonsense :-) Any ideas? I am toying with the assumption that the socket contacts of the channel 1 hybrid are dirty/temperature sensitive. But that is more like a wild guess. Wilko <0EF0777B-AE21-4A52-B69E-1F1378F81A6C.jpeg>
|
It also does it when I use the plastic-covered BNC of the original HP probe to "rattle" the BNC input.
Wilko
|
Did this just start out of nowhere or were you testing something?
toggle quoted message
Show quoted text
On Feb 19, 2023, at 1:33 PM, Wilko Bulte <wkb@...> wrote:
?It also does it when I use the plastic-covered BNC of the original HP probe to "rattle" the BNC input.
Wilko
|
It has been doing it for like 2 months or so. No abuse or measurement that triggered it (at least not something I recall, it has not been used much lately)
Wilko
|
If I had to guess I would say bad contact at some point. I haven’t pulled the schematic to see what the front end looks like but I’m thinking something loose. It only does it on channel 1 correct, channel 2 functions properly??
Andrew
|
It only occurs on channel 1, not on channel 2.
And once warmed up, channel 1 is also OK. Which points to something "reconnecting itself once warm". I guess I have to once again pull the mainboard, I have stared at it before but did not find anything out of the ordinary.
Wilko
|
If you have another scope once opened you may want to inject a small signal into channel 1 and even channel 2 for comparison of signal at certain points along the circuit path. I can’t seem to find a schematic but that would definitely help follow the signal.
Andrew
toggle quoted message
Show quoted text
On Feb 20, 2023, at 4:00 AM, Wilko Bulte <wkb@...> wrote:
?It only occurs on channel 1, not on channel 2.
And once warmed up, channel 1 is also OK. Which points to something "reconnecting itself once warm". I guess I have to once again pull the mainboard, I have stared at it before but did not find anything out of the ordinary.
Wilko
|
On Mon, 20 Feb 2023 at 12:54, Andrew <kb2lmn@...> wrote: If you have another scope once opened you may want to inject a small signal into channel 1 and even channel 2 for comparison of signal at certain points along the circuit path. I can’t seem to find a schematic but that would definitely help follow the signal.
if you don't have another scope, a 1Hz signal will be visible on a voltmeter.
My first action would be to carefully inspect everything connected to ?the BNC input, especially since it sounds like the mainboard has been removed/replaced. Solder joints and SMD devices might be cracked.
|
Thanks for everyone's suggestions. I decided to crack it open for the 2nd time and go for a mechanical checkout on the BNC connectors and "everything connected to them in the close vicinity. Although I initially suspected the non-optimal soldering of the central BNC pin that proved to be a red herring. Getting the old stereo microscope gave, literally, a whole new perspective. I managed to take some pictures via the microscope and let them speak for themselves.. Interestingly enough only K700 and K800 were affected, the other 2 relays soldering appeared OK. In any case I resoldered all relay pins and that seems to have fixed it. Using real solder, the Pb/Sn version I mean, not that newfangled RoHS rubbish ?? I'll keep the patient under observation of course but for now it seems to have been cured. A user cal has been completed successfully. Wilko      
|
Looks like the most relevant pictures never made it.
Retry below.
|
Although much better now that the relays have been resoldered, the problem is not completely gone yet.
It now takes approx 2 to 3 minutes from a cold start before the problem disappears, that was more like 10-15 minutes before the rework was performed on the relays.
So, I suspect there are some more, and likely similar, soldering issues to be resolved. The input hybrid runs very hot, its heatsink one can barely touch. Maybe the hybrid's socket has become temperature sensitive? I am a bit "chicken" about removing the hybrid, given its unobtainium status..
Wilko
|
Sounds like you’re heading in the correct direction. I would look for more solder issues also since the relay joints helped.
toggle quoted message
Show quoted text
On Feb 22, 2023, at 9:30 AM, Wilko Bulte <wkb@...> wrote:
?Although much better now that the relays have been resoldered, the problem is not completely gone yet.
It now takes approx 2 to 3 minutes from a cold start before the problem disappears, that was more like 10-15 minutes before the rework was performed on the relays.
So, I suspect there are some more, and likely similar, soldering issues to be resolved. The input hybrid runs very hot, its heatsink one can barely touch. Maybe the hybrid's socket has become temperature sensitive? I am a bit "chicken" about removing the hybrid, given its unobtainium status..
Wilko
|
Today I spent some more time on it. First attempt was to remove the hybrid, its pogo pin socket and clean all conact surfaces. So the hybrid gold plated contacts, the gold plated PCB contacts and both sides of the pogo pings. Once removed, everything looked very clean, so it was no surprise to find that the cleaning had no visible effect on the problem. Hmm.. Second I decided to touch-up all soldering on the white ceramic component (not sure what exactly is inside, input amp along with attenuator?) as well as resolder every solder joint that looked suspicious in the channel 1 input section. Good old Sn60Pb40 0.6mm. With the original RoHS soldering .. (never mind, you get my drift). And yes, it seems th    at this has finally cured the issue: on cold powerup a stable timebase line is produced. Also, mechanical stressing of BNC #1 has no longer any effect. Will again keep the patient under observation but this looks promising to me. Wilko
|
Hi.
If you need to resolder anything with a ceramic substrate (even SMD parts these days.)? Use one of the "eutectic" silver containing solders. ? That will help prevent later failure, and also has no "plastic" region in it's transition phase, so creating "dry" or "cold" solder joints is almost impossible.
Sn-Pb-Ag (62% tin, 36% lead, 2% silver; melts and solidifies at 179 degrees Celsius/354.2 Fahrenheit)
Sn-Ag (96.5% tin, 3.5% silver; becomes eutectic at 221 Celsius/429.8 Fahrenheit)
Sn-Cu-Ni-Ge (99.3% tin, 0.7% copper, 0.6% nickel, 0.005% germanium; 227 Celsius/440.6 Fahrenheit)
Perhaps you just disturbed some "Tin Whiskers", a known issue with many lead free soldered products.
Probably the definitive article on that subject!..
No wonder Lead Free solder is prohibited for Aerospace, Military, Medical and other "safety of life" systems.
Regards.
??? Dave G8KBV.
--
Created on and sent from a Unix like PC running and using free and open source software:
|
No wonder Lead Free solder is prohibited for Aerospace, Military, Medical and other "safety of life" systems. But strangely (AFAIK) not for automotive systems which I would have though should be covered by the "safety of life" issue. Or did I get that wrong? D.
|
?As for whiskers: I also cleaned everything that look "mèh" under the stereo microscope. I once had this whole whiskering mess explained to me by a HP colleague. Albeit the context was zinc whiskers, growing on zinc plated floortiles in a UK data center. Imagine microscopic whiskers being circulated by a powerful airco ?
A closer look at the solder I used reveals Stannol as the brand. HS10 as the product.?
Mine is Sn60Pb38Cu2. ?Which is not in the datasheet?
So whatever the problem was exactly, it is gone now. Quite chuffed with that outcome, I must admit.. ?
Wilko
|
Hi.
Re the Leaded/lead-free solder issue.
I didn't mention "Civilian Automotive" use, as I've heard conflicting information about that aspect.
But yes, I would have thought the regulations would have erred on the side of safety, especially with "driving control" systems. (Engine Management, Throttle controls, Electronically controlled Power Steering, ABS etc...)? But I don't know for sure.
Some of the info on that NASA site is truly scary!?? Components that were procured with a "No Pure Tin" clause in the purchase spec/agreement, but supplied WITH Tin, and problems ensued.? Not only, but also including Nuclear Power Station systems affected, causing unscheduled "incidents" and such.? (Reactor Scrams, and shutdowns!)
I also vividly recall the Mullard/Philips AF11x series of PNP transistors that would fail with a short to the (grounded) metallic can.? It was a common thing to cut the "Shield" wire, to restore normal service...
A quick and profitable fix, for a 13 year old at a school where many carried portable pocket AM radios (against all the rules!)
73.
??? Dave G8KBV.
--
Created on and sent from a Unix like PC running and using free and open source software:
|
Who needs tin whiskers in a car when you can have a Tesla with a loose steering wheel? Not that you need a steering wheel with Autopilot. Oh... hang on.. never mind ?
Wilko
|
The really sad thing about the "science" behind the move to lead free solder is that it only takes 1% lead to mitigate the tin whisker problem. I believe that Dr. Howard Johnson wrote about?this years ago. Silver can have the same issues with whiskers but I think that is only in pure form. I think there was a NASA report on that as well I read probably 15 to 20 years ago.
Think about how much lead is in a car battery. I would bet that the number of those in landfills far exceed the amount?of lead used in solder. It has been proven that the lead in the ground does not leach out anyway. It seems that the scientists in Europe that pushed the lead free solder did not really do the proper research and now we are stuck with this problem.
Personally I really don't like the lead free solutions very much. If I am building something for myself I still prefer 63/37.
IIRC there was a formulation by Nihon that had nickel in it that was also said to abate the whisker problem.
I found Dr. Johnson's writings on Lead Free Solder:
Sam
Sam Reaves
ARS W3OHM
Owner / Moderator of: LeCroy Owners Group on Groups.io
Sencore Owners Group on Groups.io Sprint Layout Group on Groups.io?? Pulsonix EDA Group on Groups.io
LPKF Owners Group on Groups.io
Electronics and Mechanical Hardware Design Engineering Manager
Staff Scientist Andritz Rolls Global Research Center (RETIRED)
|
KB2LMN
David C. Partridge
Sam Reaves