Is everyone using the 6625AF now?
We've started using ours at Nellis, but only to chart the 1060 boxes for now. I think more will come online when its resistors finally come in.
BTW, I get WTF but BOHICA?
Yeah we have been using our system for over 2 years now. We have the Resistors...got them a few weeks back. They are nice and do speed things up.
BOHICA......Bend Over Here It Comes Again
Here's one for you all.
WTF! it's the SOSDD!
Well Steve SOSDD.........Same Old Sh!t Different Day
Well on to the 6625AF. I have done alot of test and research on this item in the reversal rates, samples....etc. For charting the SR1060's a reversal rate of 10 is all that is needed....I know what the book says.....use a rate of 60, but I also know the results I've gotten along with Ken Eddy @ AFMETCAL. For samples......set the samples for 30. This will give you the desired results. Also you can chart the 1 and 10M decade boxes in either manual bridge operation or by using the Test Utility. You CANNOT use the Transfer Standard Utility for the High Ohm boxes in bridgeworks. I have some data on the reversal and sample rates. If you want it to prove to the evil QA that you can get these results by using other rates than what the manual says.....reply requesting it, shoot me an email or PM me and I will be more than happy to send the data to you.
Another thing is you need to make and attach guards from the block in the back of the unit to the UUT. This is paramount expecially with High Ohms. And remember that this is an active guard and will have voltage present on it.
And for an alternate email charles.dale AT moody.af.mil
edit: Post edited to change e-mail address making it unrecognizable to spambots. - Jesse
Also Doug at Barksdale has done alot of research. Truly if you have questions on the system ask. Doug is very knowledgeble compared to me. he is a K8 guy as I'm a K6 guy. Though don't let me being a K6 guy scare ya......I know the 6625AF.
Has everyone gotten the Scanner excerciser off of METWEB yet???
Boxes may not be charted with the Transfer Utility. Chart whatcha need in the Test Utility. The results I get are the same....though it states that the Transfer Utility is only good for the calibration.....ie 50ppm. Everyone take note.......thanks Barksdale for that one.
If anyone is ever sincerely interested in the full history behind the original specification (which was modified for acceptance) for this automated secondary resistance measurement system and any of the contracting details in regards to the 6625AF please feel free to contact me directly anytime. Do you realize the original contract award was made in July of 1999? How many systems are actually being utilized around the world today? :(
Okay, here comes the fun part. We have been trying to cal the SR1050 boxes with the 6625 with a new K-Pro. So far, all we get is that that "Servo has railed" using the method that the K-pro calls for. If you are not already familiar with it, let me explain.
The 1050 box goes on channel 32 and the reference channel of 22 or 23, depending on what box you are doing. After the hook up ,putting in the values ,channels, and references into the computer, and hitting start it will start to run. Then the readout on the 6625 goes all green and says "error 2" which in turn produces a dialog box on the computer which states the servo has railed. According to the book this can mean that Rx and Rs are not connected or that the ratio is too high. Just by using the K-pro, this should not happed. At least, that is to assume that this has been tried before the K-pro was put out. Has anyone out there tried to run it with this new procedure? Maybe you know something we don't? :?
Quote from: cobychuck on 01-30-2006 -- 13:46:23
Okay, here comes the fun part. We have been trying to cal the SR1050 boxes with the 6625 with a new K-Pro. So far, all we get is that that "Servo has railed" using the method that the K-pro calls for. If you are not already familiar with it, let me explain.
The 1050 box goes on channel 32 and the reference channel of 22 or 23, depending on what box you are doing. After the hook up ,putting in the values ,channels, and references into the computer, and hitting start it will start to run. Then the readout on the 6625 goes all green and says "error 2" which in turn produces a dialog box on the computer which states the servo has railed. According to the book this can mean that Rx and Rs are not connected or that the ratio is too high. Just by using the K-pro, this should not happed. At least, that is to assume that this has been tried before the K-pro was put out. Has anyone out there tried to run it with this new procedure? Maybe you know something we don't? :?
Do you have the 4310 resistors???????
Can you copy and paste the setup in the new procedure on this thread????????
Well, my question is now moot. It turns out that everyone I have talked to for help has not had the same level of difficulty getting their system to work. Since my original post, our system has gone nuts. We're working the issue right now but we'll see how this all turns out. As for the procedure, I don't have anywhere I can copy from, unless I sit down with the k-pro and type it out. The procedure is: 33K8-4-467-1 dated 20051030.
We found the problem! Our whole system has to go back to Guildline for repair! At least I now know it wasn't something I was doing wrong.
Okay, a little more detail on the system failure. We ran the acceptance test from the com data and found that the current for the servo was half of what it was supposed to be. The 6622 had to go to the company for repair. That's why we kept getting the servo has railed error. Like I said before, at least it wasn't something I did! :wink:
Dang!
If you had that thing running you could have saved some time on the 5700As! We did one earlier this week and the resistance ran itself (except for three readings that didn't meet the TAR) and was done overnight (took about 8 hours). We had to run the three manually, but it was still a time savings over the old way. :-)
You have the 6625 set up to cal the 5700's? How's that? Do you have an automated setup that switches the 5700?
Please forgive me as I have only limited experience with the new standard, and never had the 4310's to work with. I was basically only running the 24hr charts on the 1060 boxes when I left the lab for the schoolhouse.
I know that there are way more uses for the 6625 than meets the eye, so maybe we can use this thread to share different uses that are developed.
This was a new addition to the program by AFMETCAL (as far as I know) recently, and you have to have the 4310 to use it. My tech tells me that it works great, but doesn't meet the current TAR in the procedure for three of the values.
From what he said, it's plug and play, you just set it and walk away. Except for running those pesky three. Maybe they'll fix that in the K-pro. He ran it over night and then just had to compare the values in the morning (enter them in on the error display and read the difference in PPM.
Excellent! I didn't even know that there was a k-pro out there using the 6625.
FYI, we are adding a block in the Adv. Cal school for the 6625 and I am going up to get some training from the manufacturer in April. Should be some good stuff.
As Cobychuck mentioned before, the procedure for the 1050 boxes uses the 6625AF now. The 5700A procedure hasn't incorporated it yet (thus the problem with not meeting TAR on three resistors). I expect it's in the works though.
There was some reluctance to use the 6625AF initially, and I believe there are a lot of labs that have it collecting dust in a corner still. It's been a great help to our K8 area, but it's a bugger if you don't have anyone to train you and pass on the handy tips. The initial training was severely lacking, and the setup and software wasn't exactly intuitive. Hopefully a block in the advanced course will give us some experts. :-)
Here's the deal...I just spoke with our K8 tech and he said the procedure to cal the 5700A resistance is in the Bridgeworks 2.1 software upgrade we got from AFMETCAL. you need to look in the Bridgeworks help file on the desktop. There's also a Transfer standrard procedure that looks interesting.
The 6625 is an awesome tool if used correctly. You are correct Freezer about the training. it took myself and another to figure ours out, unfortunatly for that site we are no longer there. Call Doug Smith at Barksdale, he has it figured out pretty well also......heck he was calibrating the 5700's with it years ago.
We talked to Doug pretty extensively at first and continue to use him as a reference. We've been able to wean ourselves to the point that I'd stack our K8 tech favorably against most out there. He's got a good handle on it, from what I can tell. There also seems to be pretty good crossfeed as well. I've seen a few tips, tricks and things to watch for passed around. Most we were aware of already, but it's good that they're sharing the wealth.
Hey guys,
Just so you know, I'm up at Guildline for training this week. We are supposed to be adding it to the Adv Cal course. It is very interesting and I am learning all about the 6625 and its capabilities.
Anyway, I wanted to let you know so that if you have any questions or concerns, you can post them here and I will do my best to get an answer. We pretty much have access to anyone in the co. including the president.
Oh yeah, and Freezer, you can tell your K8 guy that the TAR issue for the 5700 is going to be fixed in the next release (Bridgeworks 2.2).
Hoopty,
Here's a question from our K8 tech:
In the current version of Bridgeworks (2.1a) when you are running a single resistor from "Test", there is no way to tell the system a rough target value of the unknown resistor. Apparently there was a field for this in an earlier version that gave the system a starting point.
What this creates is a situation where the system sometimes rails when it shouldn't. Sometimes when doing high Meg resistors (10Meg and above) the system can't determine the correct rough ratio, and rails. This only happens if you have previously done a low resistance and then proceed to a high resistance. To fix it, you have to let the system "sneak up" on the unknown, by getting it to figure a ratio closer to 1:1, say using a 1 Meg unknown and a 1 Meg standard (4310). Once it has figured the ratio, you can stop the test and move up to the 10 or 100 Meg unknown you were trying to test to start with!
If there isn't a fix for this, we'd like to see the rough target value field entry again.
Freezer,
It sounds to us like both resistor profiles (Rx & Rs) are not being loaded in the test. The old version (1.9a) showed both the Rx and Rs windows on the page at the same time. The new version (2.1 & eventually 2.2) shows only one channel at a time and doesn't distinguish between Rs and Rx. You should be selecting your Rs channel and loading the profile for that resistor and then selecting your Rx channel and loading a profile for that resistor as well. That is how the system knows where to start off at. Using this method, there should never be a "rough ratio" displayed. That is actually bad for the servo, as it must travel its entire length searching, as well as it will make for much slower measurements. Telling it both resistor values by loading both profiles will put it much closer to where it needs to start off.
If this doesn't solve your problem, I'll need some more specifics on the ratios you're trying to achieve. At first I thought it might be a case of a less than 1:1 ratio, which is a no-no with this system (It will work, but you just have to swap your Rs and Rx...). Anyway, after talking to the guys up here, it sounds more like the problem I described above.
Good luck.
Jesse
Thanks Hoopty, I'll pass this on to our K8 guy.
I went back and talked to the K8 guy and explained what you said. He immediately marched over to the 6625 to prove you wrong....then said "it might work that way..." He's back there testing it now :lol:.
He indicated that the documentation didn't say anything about loading the Rx (I have no idea, but he's a very literal guy, if it didn't say "Load the Rx profile here", he wouldn't do it).
Anyhoo, we're testing right now.
thanks again!
Freezer
Hoopty,
Looks like the problem was just as you said. We weren't loading the profile for the Rx. I'll probably try to crossfeed the info to our other labs, just in case they're as in the dark as we are!
That's the problem with stuff like this, you think "everyone else can't be as dumb as we were!", so you don't say anything :oops:. Then someone buys an inspection because of it!
I'm used to looking dumb, so I'll send it out.
For me, this once again validates the usefullness of this site! :-)
Thanks again!
I'm glad that it worked out for you.
It is somewhat confusing w/ the new layout vs. the old. With the old you actually saw the separate entries at the same time. Now you only see one. So I wouldn't say it was a 'dumb' mistake really.
Can you say if it made any significant difference in the amount of time it took to run the same type measurements?
Check this out:
http://www.mintl.com/documents/6242a.ds.rev.4.pdf