HomeBrewerDude
Lifer
- Jan 18, 2001
- 14,465
- 1
- 0
I think its time to consult with the customer, review specs standards, and figure out whether the QC trumps Eng or vice versa
I think its time to consult with the customer, review specs standards, and figure out whether the QC trumps Eng or vice versa
Well if the requirement says -01.0 then it depends what the allowance is for a variance. For instance on a computer power supply variance may be around 5% which is according to the specifications. So if the -1.0 has a value to the first decimal point then that reading is only accurate to 1/2 of the next decimal point or ".05" so it may be within specifications. It depends if the specs allow for variance and how much they allow for variance. You must realize that with electrical current there will be some variance. The Ideal reading should be above the minimum and below the maximum or around "-1.1".
So it really is a matter of how sensitive this piece of equipment is to the specification. Engineers tend to write specs that are sometimes giving allowance to around 200% of what is needed. They tend to expect fluctuations in requirements. That is why equipment can often be run beyond the redline of a maximum. The figure is only .03 off from the quoted amount.
How accurate was your testing equipment? The more specific the requirement the more sensitive the testing equipment has to be. You have to use some of the current to get a reading for the test. Even a volt meter uses electricity. We are talking about a miniscule amount.
So it really is a matter of how sensitive this piece of equipment is to the specification. Engineers tend to write specs that are sometimes giving allowance to around 200% of what is needed. They tend to expect fluctuations in requirements. That is why equipment can often be run beyond the redline of a maximum. The figure is only .03 off from the quoted amount.
How accurate was your testing equipment? The more specific the requirement the more sensitive the testing equipment has to be. You have to use some of the current to get a reading for the test. Even a volt meter uses electricity. We are talking about a miniscule amount.
The QC is absolutely correct. You have the measure 10 times more accurately.
I'm a test engineer, and have three times rejected things because the readings were in spec, but not "right". Once, after the unit was opened, there was a loose cable that would have failed during vibration test. In another case, there was a flaky regulator in a subcontractor's part, that would have failed shortly after it was fielded. The third was because the DMM had a low battery, and so the reading it was giving was not correct. In all cases, the company saved money because I said something.
I'm a test engineer, and have three times rejected things because the readings were in spec, but not "right". Once, after the unit was opened, there was a loose cable that would have failed during vibration test. In another case, there was a flaky regulator in a subcontractor's part, that would have failed shortly after it was fielded. The third was because the DMM had a low battery, and so the reading it was giving was not correct. In all cases, the company saved money because I said something.
-0.97 is pretty damn close to -1.0 (within +/- 5%)
And, the specs say -1.0, -1.1, -1.2.. That does NOT mean a range of -1.0 to -1.2.. Otherwise, you could say -1.15 would pass while -0.97 would fail, which seems to me would be against the intention of the original specification.
IMO, most companies would want the lowest voltage possible for a working part that is within their spec. So, I would say -0.97 is almost spot-on.
PASS
And, the specs say -1.0, -1.1, -1.2.. That does NOT mean a range of -1.0 to -1.2.. Otherwise, you could say -1.15 would pass while -0.97 would fail, which seems to me would be against the intention of the original specification.
IMO, most companies would want the lowest voltage possible for a working part that is within their spec. So, I would say -0.97 is almost spot-on.
PASS
Based on the info you've provided, out of spec and QC is right IMO. If I were the customer, I would reject the parts if testing showed the same results you're getting.
Since I don't know the details, I'm not seeing why this is a debate when they could just ask the customer like I originally said.
But, what I am curious about is, IF they do decide to pass the product, what is the head engineer's position on disclosing this QC information to the customer?
To me, it sounds like he wants to sweep it under the rug and just give it a "pass" rating. Whatever your opinion is on the question of pass/fail, I don't see how anyone could agree with not disclosing this information to the customer.
But, what I am curious about is, IF they do decide to pass the product, what is the head engineer's position on disclosing this QC information to the customer?
To me, it sounds like he wants to sweep it under the rug and just give it a "pass" rating. Whatever your opinion is on the question of pass/fail, I don't see how anyone could agree with not disclosing this information to the customer.
As a civil, that is fine and dandy since none of us can get anything that precise anyways
.
Guessing that's not ok for electric...
.Guessing that's not ok for electric...
Obviously depending how critical getting the exact voltage. For the most part a .97V and 1V is the same, most electronic parts tolerate +-.03 V without any performance differences.
I would say .03 difference is acceptable if the product functions as it should. I can't see it being dead-on all the time, but I'm not an electrical engineer (but engineering related).
I certainly hope you guys don't make precision components, then. 10x precision of the spec can get really expensive, really fast, and for no good reason.
The engineer is right that this is within the spec. Head of QC is right that the customer is probably a moron who doesn't realize that 1.0 = 0.95 to 1.04.
Solution is obviously to call the customer and ask for tolerances. Is the actual tolerance +-.1 (in which case .97 gets rounded and is OK), or is it +-.10 (In which case they really should have specified)?
No kidding? So if someone asks you to machine something to 1.00 inches (+-5 thousands is a very standard tolerance for machining), you'll do it to 1.000 (+-5 ten thousands is doable, but not nearly as much of a piece of cake)?
Originally posted by: Specop 007
Ouch. Ship an out of spec part on a .mil contract?
Give it a shot, see what happens.
*cough* Osprey *cough*
Uhh, he said MEASURE to 10x, which is very standard, not produce the part to 10x the spec tolerances.
Well they DID measure to 10x the standard. Standard is +-.1V, they measured to .01V. Why bother measuring to 10x the standard if you have to round anyway?
Zenmervolt
Elite member
- Oct 22, 2000
- 24,512
- 21
- 81
That's not the problem. It's still bad practice to aim for the edge of a range.
Let's say you have to hit a 12" diameter target from 15 yards. Do you aim for the very edge or do you aim for the center? You aim for the center, that gives you much more margin for error.
Your design team aimed for the edge of the spec and got burned. If they had only aimed for -1.1 volts initially there would not be a disagreement now.
The part fails.
ZV
I understood your question as asking if they would produce the part 10x the spec precision. I can't answer and have no idea why they would measure to 10x and then round. It sounds dumb to me too.
Originally posted by: Specop 007
Ouch. Ship an out of spec part on a .mil contract?
Give it a shot, see what happens.
Funny, I was just thinking ask the engineer if he would feel confident putting that "device" in his son or daughter's hands during a life-threatening situation...
So a bit of an update, I guess. I inquired further about this issue, since as I mentioned before it is not my project and anything I've said to this point has been hearsay, really.
The spec was "-1.0V, -1.1V,-1.2V." That is to say, the test point can see any of those 3 voltages at a given time for a given function. So the -1.0V was not "aiming for the edge," it was aiming for 1 of 3 centers.
As far as the tolerances go, I have yet to get a straight answer, and I haven't seen the exact spec myself, so I don't know. I'll ask about it more Monday. Typically there is a +/- 5% or whatever range on the spec, but it could very well be that the spec didn't state any acceptable range, just a set voltage. Regardless of the tolerance, there will always be rounding - so I see it more as a company policy problem than a customer spec problem.
I didn't work today (have a flex day every other Friday), so I don't know the final result of whether or not it shipped or got delayed, but they were still arguing over it yesterday.
Specop 007
Diamond Member
- Jan 31, 2005
- 9,454
- 0
- 0
Originally posted by: ITJunkie
Originally posted by: Specop 007
Ouch. Ship an out of spec part on a .mil contract?
Give it a shot, see what happens.
Funny, I was just thinking ask the engineer if he would feel confident putting that "device" in his son or daughter's hands during a life-threatening situation...
See my first post on page 2
TRENDING THREADS
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 10K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 7K
-
-
Facebook
Twitter