T.C. measurements on precision resistors

[hwj-d]

May I ask here, what the cost is for two kx sets of Mr. Edwin G. Pettis resistors to germany incl. tax, customs fees, sending, packing, approximately to germany?
Sorry, have absolutely no plan for that
  Thanks.

[Andreas]

May I ask here, what the cost is for two kx sets of Mr. Edwin G. Pettis resistors to germany incl. tax, customs fees, sending, packing, approximately to germany?
Sorry, have absolutely no plan for that
  Thanks.

Hello,
 
although its somewhat off topic here:

current pricing of Edwins resistors was mentioned here:
https://www.eevblog.com/forum/metrology/resistor-set-for-ltz1000-positive-standard-7v-circuit/msg1396976/#msg1396976

Tax: you will have to pay it at your local customs office when you pick up your resistors. (19% VAT here in germany)
        note that customs uses monthly fixed exchange courses between US$ and EUR so it will not be exactly 19% of that what you payed.

Customs: currently there are no customs fees on electronic components like PWW resistors.
You should have the HTSUS/TARIC number at hand for the customs to avoid confusion or selecting the wrong custom fees.

HTSUS: 8533.21.0080
Description: Electrical fixed resistors, other than composition or film type carbon resistors, for a power handling capacity not exceeding 20 W
MFN Duty Rate: Free

Shipping cost: depends on what option you have with Edwin. I payed something between $8-$24 for USPS.
The higher fee usually does not pay out (does not speed up transport), since customs handling takes at least 1 week.

I don´t know if Edwin offers sending per courrier (which handles customs) now.
But usually courriers charge you another EUR 20 for the customs handling.

with best regards

Andreas

[Andreas]

Thanks Edwin for your valuable insight on materials for resistors.

What I take from this:
The precision resistors are all indivdual made (hand crafted) and many sources of error are possible,
depending also on experience of the actual maker.

So if you have high demands it is reasonable to test every single resistor.

with best regards

Andreas

[Andreas]

Hello,

some results of my brand new VHP101 & VHP202Z,  I'm afraid there is only 20% chance you will get an almost zero TC resistor from the batches your order or so.   

what is not clear to me:
Resistor A319 seems to have a defect on rising temperature near 10-15 deg C.
visible on the first picture over time. (even outside the diagram below -25 ppm).

I cannot see this defect on the second picture (over temperature).
has it been omitted?

25 deg C temperature rise in 2 hours is quite a lot.
Will the defect also be visible with slower ramp speeds?

with best regards

Andreas

[branadic]

Hello,

some results of my brand new VHP101 & VHP202Z,  I'm afraid there is only 20% chance you will get an almost zero TC resistor from the batches your order or so.   

Beside 10k_741 with almost flat temperature response all other resistor have significant T.C. and you can now combine resistors to get almost zero T.C.

-branadic-

[branadic]

I just measured a 9K9850 VHP202Z at -1.3ppm/K.  What a bummer.  Now I understand what they mean by "0.2 +/- 2 ppm/C".  That means it could be as bad as 2ppm/C.

If you now plot change in resistance over temperature and put a first order polynom fit on it all data will be more clear on first sight.

-branadic-

[branadic]

What about x-ray of a 12k econistor, in comparison?

So here are the other resistors of the LTZ set.

1k: date code 1704
12k: date code 1726
70k-1: date code1719
70k-2: date code1719

-branadic-

[3roomlab]

im not sure if i understood the purpose of having 2 windings, esp the 12k pic. isnt it that both winding should be equal to negate inductance?

[texaspyro]

It looks like one of the wires  loops back and touches the weld strip.  That seems like it would be a source of error if that contact was intermittent... the effective length of the winding could change.  The error would be greatest on the resistors with the shortest winding length.

[Edwin G. Pettis]

Oh my, if I was QC I'd have to reject all but the bottom 70K; first, the wire sizes selected is poor, the bobbin should be much fuller, at least 50% when possible, the finer wire sizes also makes calibration difficult, the poor bobbin windings, one pi has barely any turns in it.  It is unlikely that all of these resistors were wound by the same person indicating that something is very wrong in production.  Look at the huge difference between the 70K resistors, one is wound well, the other one is a reject, the wire size is different too.  Sometimes for low ohm resistors the bobbin fill isn't going to be too full but for all of these values there is no reason for such poor wire selection and winding technique.  While the wire lay in a Pi is not strictly smooth by nature, some of the windings in these resistors is just plain sloppy.

texaspro.....the only portion of the ribbon that the wire can touch is from where it exits the molded bobbin to the welded end of the ribbon, the rest of it is embedded inside the bobbin so the wire can't touch it.

3roomlab.....please read my earlier post about resistor inductance, it cannot be cancelled by reverse winding pi, but in these resistors, the inductance is probably even higher than equally wound pi.

[branadic]

I wonder how Burster resistors made in Germany do look like. Someone here in the group willing to lend some of them for X-ray image?

-branadic-

[mimmus78]

I think run a resistors over X-ray for quality check is a little bit overkill for a resistor, especially if you plan to do for each resistor.

What are QC check we can run in house? Wondering if this tests are sufficient to find a lemon or if some bad one can sneak out.

[MisterDiodes]

You probably want your money back on that 12k and 70k...something went wrong there for sure.  To be fair: I opened up a few resistors from a recent order from GR and they are much more even and look like a PWW reverse winding resistor should look like:  Even turns count at both ends of the bobbin, and the bobbin isn't too empty or too full.  I have never seen that problem on Pettis resistors, and we've even ordered some "custom" models from Pettis that were perfectly fine.

I think G.R. had a real QC inspection problem that day - and you probably want to bring that to GR's attention and see what they say.  Rhopoint is just a middleman reseller, that is out of their control - that's why we order direct from the manufacturer if possible, just in case something like this comes up.

I should point out that Mr. Pettis' point about the dual windings not cancelling out AC fields is observable at the very local area around the resistor: The very close local area near the center of the resistor (where the two windings meet) does achieve near perfect mag field cancellation, but as you move out to the resistor ends that cancellation effect is less and less perfect if you are very close to either -end- of the resistor.  In the final analysis of the dual winding - it is still a very valuable feature in a PWW in a low freq AC application:  If you stand back from the resistor and look at the final disturbance in the magnetic field while the resistor is carrying AC current, all you'll see is two -very weak- inductive coils that are counteracting each other, and the end result to the surround area is close to zero.  As Mr. Pettis points out:  If you try to measure the Q on these at low freq AC or even up to some 10's of kHz it is very, very low and hardly measurable.  You do see more inductive effects when you try to measure Q at 100kHz or higher, but you don't use PWW for that application anyway.  Conversely, as long as the resistor is illuminated by a low freq mag field that has about equal field strength at both ends of the resistor, it is unlikely to have much effect on the circuit, thanks to the dual winding technique.  If you had a near mag field very close to one end of the resistor, you might see that inductive coupling in your circuit, depending on what the circuit is doing.

So in other words, don't mount your PWW with one end butted up against a mag field source, and it's probably OK.  And all precision circuits benefit from some sort of enclosure that offers shielding from whatever low freq EMI junk is in your local environment.

You can also order PWW resistors with Ayrton Perry or other bifilar winding designs that work well into AC ranges - but those are generally very difficult to build to a tight value tolerance compared to standard dual winding design and I guarantee you won't like the price.  Those designs will also carry additional penalty as possible lower working voltage and increased inter-winding capacitance.  It's always a tradeoff.

[branadic]

To be honest, I've X-rayed resistors from Edwin because of issues I had with them in my fist LTZ1048B. I found two resistors reading high, after they were in cicuit. I dropped all of them and replaced them by G.R. 8G16D ones I ordered from Rhopoint and had no more issues.

It was Andreas how asked me, if I can do X-ray of the 8G16D as well. So I did and shared them here. I don't won't debatting on which design is the better one, which is more reliable and who did something wrong. I do have my own opinion about both designs from a reliability point of few and I let it up to you to have your own.

@ MisterDiodes
If you have a contact to G.R. please let me know, I wasn't able to find one on the web.

-branadic-

[Edwin G. Pettis]

At this point, I have no knowledge of the history of these resistors with Andreas so I can't comment on why two of the resistors apparently failed, they were QC'd at the time of shipment and fully met spec.  I believe those resistors were from August, 2015, although age is not necessarily an indication of problems.

My comments on the G.R. resistors are quite valid and have nothing to do with G.R. being a 'competitor'.  The faults with the G.R. resistors are not design faults but execution of production procedures.  As I explained earlier, there are many variables in PWW manufacturing and at any time something can go wrong with even the best resistors, they could possibly even pass final QC and develop a problem much later, there is no easy way to tell.

I would strongly object to anyone drawing any conclusions about a given manufacturer based on such a small sampling.  I have manufactured thousands of resistors and have one of the lowest failure rates in the industry.

[branadic]

Well, Andreas received them and send them to me. They were bend and soldered using a gripper for keeping heat away, nothing fancy happening so far. However, two broken resistors in a set of five for the LTZ circuit is a rejection rate of 40%. This is just a simple fact. Maybe they are made on a monday or whatever. I don't complain even though the money for them is lost.

All I want to say is: Edwin, don't complain about the quality of G.R. or any other PWW resistor manufacturer, because if we take a serious look on your design - I avoided showing X-ray images of them here - and would have a member of G.R. in here I'm sure he would also throw hands in the air and scream. If your quality was much better, then my resistors from you should have worked 100%, right?

My general suggestion to all PWW resistor manufacturer is, bend the resistor to the appropriate grid and quality check them afterwards. Thus, broken wires due to bending can be indicated and the resistors are bend in the way manufacturer specifies.

"people who live in glass houses shouldn't throw stones"

-branadic-

[Andreas]

Hello,

again some measurements of 8G16 120 Ohms. DateCode 1522

#2 shows a several 10s ppm relaxation effect on the first day (10.01.2018)
which has gone on the 3rd day. (12.01.)

#3 with no surprises

but all measured 8G16 120 Ohms have large T.C. + hysteresis and large ageing drift during temperature cycling.

and again a overview: the 120 Ohms resistor table with Z201, Ultrohm Plus style 805, and 8G16.

with best regards

Andreas

[MisterDiodes]

@ MisterDiodes
If you have a contact to G.R. please let me know, I wasn't able to find one on the web.

-branadic-

http://www.primetechnology.com/content.aspx?ParentId=3&PageId=81

Either call them or use the Sales email address, and politely ask to send the 12k resistor X-Ray photo - I doubt at this point they aren't going to offer a refund on parts from 2015 (and who knows how long before that they were sent to Rhopoint) but it would be interesting to see what the reaction is - especially on 12k resistor.

It's not like the wire got up off the bobbin and unwound itself to the other side of the bobbin - that really looks like someone wasn't paying attention during the winding process.

I do know that recently (late 2016) they changed the resistor manufacturing plant to a Central America factory (Costa Rica) so maybe they were having QC issues. That's only an observation, there is really no excuse for that 12k nonsense.

The samples I'm looking at today are some 5k units from the new plant, and they are built correctly. 

[Edwin G. Pettis]

In fact, we do not know the cause of the failures, you latched onto bending as the cause, it was just one of several possibilities I mentioned.  To use such an exceedingly small sample as the basis for a claim of a high failure rate is just plain absurd.  No one has a zero failure rate and while I do claim a very low rate, it is not zero at this time.  At this time, your conclusion as to the cause of failure is unproven, you have only proven that there was a failure.

Again, what I said about the G.R. resistors in your X-rays is totally valid and has absolutely nothing to do with them being a competitor, practically anybody looking at those pictures would know the construction is faulty and I pointed out that was not a design flaw but a failure to follow production procedure which an entirely different matter.

As for G.R., it appears that they have also changed their interior construction from the resistors in your X-rays, when they did this, I don't know.  It appears they are still experiencing some problems in production from Andreas' tabulation.  Due to the complexities of PWW manufacturing, there is bound to be problems now and then...guess what, Vishay does too.  You also might note that Vishay has all but divested themselves of their PWW operations...gee I wonder why.

[branadic]

In fact, we do not know the cause of the failures, you latched onto bending as the cause, it was just one of several possibilities I mentioned.

Right, I don't know the initial cause for the failure, but do know the failure, a broken weld and I have shown you the broken joint with the x-ray image of the 120R resistor. As the wire of the 70k resistor is very small we were not able to see the failure at first sight, but we will investigate on that.

To use such an exceedingly small sample as the basis for a claim of a high failure rate is just plain absurd.

I've never said that, this is just your interpretation of the rejection rate I mentioned on my resistor set for my LTZ.

Seems like you do have a serious problem with factual criticism. We could have done a factual analysis on how the failure showed on the resistor and what could be done to prevent such failures in future (one of the things we do for many companies almost every day), but instead you act like your ego was flawed.

-branadic-

[Edwin G. Pettis]

I quote: "two broken resistors in a set of five for the LTZ circuit is a rejection rate of 40%,"  that is exactly what you said, that is absurd.  That is an inaccurate statement of failure rates and no one would accept such a statement, care to try that one on General Resistance and guess their response, come on.

As to failure analysis, you have to have the items in hand in order to do a proper analysis so you can't claim I didn't make an analysis, I gave you some examples of possible causes for the failure, which one it might be remains to be resolved, since the cost of returning said resistors to me for analysis exceeds the cost of the resistors, it is ridiculous to make a point of it.  You also first said the resistors were reading high, then later changed it to open, obviously they appear to be open, the cause remains a mystery and since you're not qualified to do the analysis, you must accept the fact that the cause remains unknown.  I accept the fact that they failed for some unknown reason, there is nothing left to complain about, we both agree they are broken, the cause at this point is not really relevant.

I don't have a problem with criticism when it is accurate, you are missing the point, I know all about weld failures and how they can occur, they happen now and then for a lot of different reasons, the point is to strive for no weld failures.  In fact, I demand you produce any evidence of my denying that the resistors failed or any thing else that was inaccurate.

While I agree with you about our dimwitted pres, I object to your silly comparison, it has no place in this discussion.  I think all that needed to be said has been said, two resistors have failed, reason unknown, end of story.

Nothing personal here, just disagreeing with your position.

[dl1640]

be cool, no one is perfect

[branadic]

since you're not qualified to do the analysis, you must accept the fact that the cause remains unknown.

If you tell me that I’m not qualified to do a proper analysis you have to accept that I’m the opinion that you are not qualified to manufactur reliable resistors. And as you wrote me, since the production of this resistors you have changed your construction. So if there where no problem, why would you have changed your construction? You also had problems with the expoy in the past, I’m not sure on which post you gave that statement. So if everything was fine, why would you have changed the epoxy?
Sorry Edwin, you don’t know nothing about my daily business, so keep your statements on my skills for yourself.

-branadic-

[Echo88]

@branadic: Accusing one af not accepting factual criticism reminds me of this thread: https://www.eevblog.com/forum/projects/oshw-24bit-adc-measurement-system-for-voltage-references/ in which you remarkably sit on your high horse and continue to refuse to just attach pdfs, because everyone should just install Eagle to look at the schematics. Glasshouse and stones huh?

[dl1640]

sometimes words are not that accurate and people lose their Temper Co.