Wire wound resistors failing

[Grandchuck]

Hi and happy Thanksgiving to those who celebrate it.

I have 3 copies of Dr. Frank's LTZ1000 reference design and used resistors from Edwin Pettis.  Several resistors have failed.  They became intermittent and were sensitive to physical pressure; it was easy to find the problem.  The ones that have failed have not gone open but moved around in value when pressure was applied.

Mr. Pettis no longer seems to be a member here.  I am posting this to alert other users of these resistors.

[floobydust]

Which model is it? I have 802, 803's and they are fine.
Wild guess- it's the soldering at the resistance-wire ends to the leads. That would mean your intermittent readings can never be lower, only go higher- than the rated resistance.

[Grandchuck]

I realize that and that is why I wrote that they did not go open. Indeed, as you surmised, they went higher in value with pressure.  No doubt a soldering issue.

I did not know there were various models offered by Edwin?

[floobydust]

It looks like the resistance-wire has special needs for soldering that may have not worked out. If possible, check under a microscope for flux or corrosion etc.
Is there Green Crud? You could try a PM to Mr. Pettis for help.

I would just do surgery and resolder the ends (with special flux and cleaning). But then that looks like the resistors are newly upset and need mechanical stress relief and I have no idea what the secret sauce is for that.
I don't know what wire type is being used, example Manganin.
"Notes on Treatment:
MANGANIN® can be worked easily. Though the alloy can be soldered, it develops in air a thin oxide film; this must be removed before working. With an appropriate flux MANGANIN is also suitable for dip-tinning. Furthermore, MANGANIN can be brazed and welded. Resistors made of MANGANIN  must be aged in order to remove mechanical stress. For further details see Technical Information."


At first I'd asked for HC-49 packaged parts and got this reply:
"I do not make hermetic resistors at this time; it is difficult and expensive to acquire the hermetic cases in smaller quantities.  The 802 is an epoxy resistor, 0.250” D x 0.375” L with 22 AWG leads.  There is no real advantage to using hermetic resistors in the LTZ circuit.  I manufacture resistors in several different bobbin sizes and most any value from about 10 ohms to over 1 Meg (high value resistors are more expensive due to the cost of wire).  The higher value resistors are used where low TCR and very good stability is required that justifies the cost such as transfer standards."

[Grandchuck]

OK, mine are the 802s.

I have been replacing them with Vishay metal foil resistors. 

Thanks!

[Edwin G. Pettis]

Yes I'm still a member, I just don't post too often but I do look in fairly often.  Sorry to hear about the resistors, when did they fail (all at once?), on the same board.  You should have contacted me first but no matter, did you throw them away or did you keep them?  I have an exceedingly low failure rate but it isn't zero, nobody does (not even mil resistors).  After nearly two years I don't have any conclusive ideas as to why they failed without doing an examination.

Floobydust is incorrect, they are not soldered and cannot be soldered they must be welded.  To hazard a guess; there may have been some mechanical stress on the leads which eventually lead to failure.  The weld joints are very small and any flexing of the lead can lead to the weld breaking and becoming an unstable connection but that can only be determined by a proper examination.  The act of taking the resistor apart can also break it so it has to be done very carefully, I do not recommend users trying to do it.

Regards,

Edwin
Ultrohm Plus

[Grandchuck]

Sorry Edwin.  I did try to contact you first but couldn't find you ... my mistake.

I have two that I can send you for your inspection, but one I tried to disassemble and you are right about breaking it.

[floobydust]

[...]
Floobydust is incorrect, they are not soldered and cannot be soldered they must be welded.  To hazard a guess; there may have been some mechanical stress on the leads which eventually lead to failure.  The weld joints are very small and any flexing of the lead can lead to the weld breaking and becoming an unstable connection but that can only be determined by a proper examination.  The act of taking the resistor apart can also break it so it has to be done very carefully, I do not recommend users trying to do it.

I see these solder remnants on the leads and pretty shiny so I assume they have been soldered.

[branadic]

I see these solder remnants on the leads and pretty shiny so I assume they have been soldered.

What you see isn't a solder joint, but a weld joint between the copper lead and an Evanohm rod to which the Evanohm wire is welded to.

-branadic-

[floobydust]

I have no idea what the welding process is or how it looks- stared at a few more parts and it's totally possible it's actually a weld that ends up being very clean and shiny.
In my head the wrong assumption is it's like inductor and transformer terminations where the solder did not adhere due to the insulating varnish or oxide, that caused OP's troubles.

edit: a bit more discussion was here https://www.eevblog.com/forum/metrology/manganin-zeranin-enamelled-wire-availability/