Resistor testing and directionality

[Peter Daniel]

Originally posted by MRehorst


My point was that your simple test method is NOT a way to do the test. It is like sticking the resistors in your ears to see what the "sound" difference is. When the contact is unreliable, and the device is unshielded, you don't know what you may or may not be hearing.

If you want to do such a simple test setup and use only ears to evaluate the result, then shield the thing (just put it in a grounded box), an connect the resistor or other part to be tested using some binding posts or nuts and bolts and use a torque wrench to tighten the nuts uniformly and repeatably.

Whether THAT is a valid test is still debatable, but at least it should lead to repeatable results with minimized "noise".

There is another issue to consider- do resistors sound the same when they have only ac signals across them? Does ac + dc cause different sound? Does dc polarity and level affect the sound?

If none of this matters, and the whole point of the exercise is to fool yourself into thinking that you have actually tested something and have some valid results then your test set up really is optimal.

MR

We could talk for days about it (in this manner), without having any useful results.
Did you try it? If not, you don't know what you are talking about. It's that simple.

[MRehorst]

I don't have to try it to know it is not a good method. It is OBVIOUS that it is a bad method for all the reasons I have listed.

If you want to deny the obvious, you are free to do so.

When people become irrational, there is no point in continuing the discussion.

MR

[mrfeedback]

MR,
As Peter says, it really is as simple as connecting a resistor and having a listen.
If the contacts (or solder joints) are good enough, then the test is perfectly valid.
Your point about shielding and inteference is cancelled if the Devices Under Test are exposed to these same conditions.
Experienced ears are indeed a perfectly valid measuring tool - if musicality is what you want to measure.

Eric.

[MRehorst]

Let me put it into terms that will be more clear to the parties involved: if I were to claim that using a stick of spagetti I can drill holes through aluminum faster than a drill bit, and stuck to that argument no matter what anyone said, I would be laughed off these forums. If I said you have to try it to know, again people would laugh.

That scenario is roughly equivalent to this "debate" about the test method for determining the directionality of resistors.

I am a professional engineer with 20+ years of experience designing ICs and test circuits for them. I have spent thousands of hours in test labs with real test equipment devising methods and measuring characteristics of components. Over the years, companies have invested millions dollars based on the results of my lab work.

I have pointed out OBVIOUS flaws in the test set up, yet people keep denying they are flaws. It is like I am saying an apple is an apple and people keep insisting that it is an orange. Even my five year old son is more reasonable than this.

If the test method is bad, the results are meaningless garbage. There is no rational way to argue against this.

MR

[Peter Daniel]

Originally posted by MRehorst
That scenario is roughly equivalent to this "debate" about the test method for determining the directionality of resistors.

I will agree with you that this scenario isn't a perfect set up for testing directionality of resistors. .


However, it is good enough for comparing sonic signature of resistors "on the fly". I can easily identify different brand of resistors, without knowing which one was placed in a circuit. As you noticed that post appeared originally in "Comparing resistors" thread.

If Dorkus had similar ideas about how to conduct a proper tests, I now know, why he had never had it done.

Originally posted by MRehorst
I am a professional engineer with 20+ years of experience designing ICs and test circuits for them. I have spent thousands of hours in test labs with real test equipment devising methods and measuring characteristics of components. Over the years, companies have invested millions dollars based on the results of my lab work.

That's probably why listening tests are not really meanigful to you.

[MRehorst]

Well, even among engineers there is a spectrum that runs from thorough to sloppy.

When it comes to engineering matters, I will trust my own experience, judgment, and rationality. When it comes to making nice boxes, I will defer to the expertise of others every time.

MR

[Peter Daniel]

Originally posted by MRehorst
Well, even among engineers there is a spectrum that runs from thorough to sloppy.

When it comes to engineering matters, I will trust my own experience, judgment, and rationality. When it comes to making nice boxes, I will defer to the expertise of others every time.

MR

Couldn't agree more.

[haldor]

Originally posted by mrfeedback
Phil, solder joint problems are not included here - note that I said after 'blanket resolder'.
I am well familiar with noisey contact problems too, and this is not the sound that I speak of, and besides all pot and switch contacts get treated before this initial runup.

I suspect that the resoldereing process demagnetises residual fields in component leads, and clipping level current pulses cause a remagnetising of component leads, and a consequent bias in characteristics.
The solder joints themselves may be part of this equation also, especially if a magnetised component lead is contained within that joint.
This sonic change occurs only once after a blanket resolder, and happens when the amp is pushed to momentary overload.

Eric.

Hi Eric,

If there was a blanket resoldering then the first thing I would suspect is a bad solder joint. If you don't get all of the old solder out of the joint and clean it (with zero residue solvent or alcohol) before resoldering, then there is an increased chance that contamination (burnt flux or corrosion) is going to be trapped in the solder joint and result in a bad connection.

I don't know about any effect of magnetising leads, but I do know from personal experiance that a cold solder joint will cause the exact behavior you described. Perhaps the next time this happens you should take a closer look at your soldering. I find that even when I am being careful resoldering a PCB I can still get a bad solder joint. My best results come from using a fresh unpopulated board with fresh components.

Also the comment about small signal contacts is not something that any "treatment" I am aware of will correct. If the gold flashing has been disrupted by an overcurrent condition then the only fix is to replace the contacts.

Phil

[UrSv]

Originally posted by Peter Daniel
Here's the pic of those connectors. I was looking through pages of Digi-Key catalog and couldn't find anything suitable. Those I found in my local store. Unfortunately they had limited stock and I bought them all. They expect to restock in 2 months. Anybody knows another source for them?

Actually they look like what my standard supplier refers to as connectors for CB radios.

http://www.elfa.se/elfa/produkter/se...htm?_42_243_90

[mrfeedback]

Hi Phil,
I have been resoldering too long to let crook solder joints through, and my method prevents that.
Caig Deoxit contact treatment removes oxides and leaves audibly cleaner contacts.
There is general consensus around here that magnetic component leads can affect sonics.
Like I said, the effect only occurs once after blanket resolder, and remains stable thereafter.
It takes a couple of experiments to understand this initial sound, and once you have noticed it, you will hear it repeatedly, but only once per each resoldered channel.
Also this sound is unlike that due to a dodgy contact.
This is an observation only, and my thoughts about causes are at this stage speculation only, but the effect is real ime.

Eric.