Keyboard Shortcuts
ctrl + shift + ? :
Show all keyboard shortcuts
ctrl + g :
Navigate to a group
ctrl + shift + f :
Find
ctrl + / :
Quick actions
esc to dismiss
Likes
Search
"ADDITIONAL STANDARDS ARE NEEDED" when using E-Cal
|
Hey Group -
Just got a new (old and used) 85062 ECal kit.? Consists of 85062-60002 (1-26.5G) and 85062-60001 (45M - 2G). Ran a cal with the 60002, 1-10G, perfect. Never seen such a good through connection. But when I run cals with the -60001 within its freq range, I get this message when I try to save the cal: "SETTINGS CONFLICT: MORE STANDARDS ARE NEEDED" I tried this over 100MHz - 2GHz, then 50MHz to 1GHz, same result. Both times I put the 1-26.5 standard in and hit the button for it, and it ran some measurements, but when I try to save, I get the same message above. Does this mean the 45M - 2G standard might be bad? Maybe it's not reading the proper cal coefficients? Thanks - Dan |
|
Maybe a bit more information.
toggle quoted message
Show quoted text
Which type of cal did you do? Which network analyzer were you using and what were the relevant settings (Sweep type, number of points, start / stop frequencies, sweep speed). Which signal source? Just for kicks, try doing 2 - 20 and see what results you get. Cheer! Bruce Quoting dan.meeks222@...: Hey Group - |
|
Sorry - answered the 1st e-mail B4 reading the second.
toggle quoted message
Show quoted text
Wat NA are you using? If it is an 8510, are you using a PC or an 85060 contoller? It does sound as though the E-cal unit might be defective - but I'm not sure. Cheers! Bruce Quoting Bruce <bruce@...>: Maybe a bit more information. |
|
Thanks Bruce -
I have an 8720C, and yes I have the dedicated ECal controller 85060. The 1-10 cal looked perfect so (right now) I am not worried about that standard. I just tried a 1-port cal. S11 failed (got the EXECUTION ERROR: Device is not in the RF path error). Then did S22 and that worked fine (verified with a good load). So yes it appears to me that port 1 of the ECal unit is no good. Crap. This is the second set that I purchased, first set also had the low freq unit defective (but different issue). And I looked on Ebay and the only 85062-60001 on there is "for parts". Apparently these things are easy to break. Dan |
|
Dan -
toggle quoted message
Show quoted text
I am unfamiliar with Ecal on the 87XX NAs. I thought they (87XX) had internal software to deal with the Ecal. Another "test" would to try S11 using port 2 of the Ecal - I'm not sure that the controller would recognize the reverse connection, but it is worth a try. What happens if you try doing S12 and S21 with the LF Ecal unit? Also, if one of the Ecal units was bad, I'd pick the LF one. You can cal to 2Gs with just about any reasonable set of standards. Above 2 GHz, it gets a bit more critical. Cheer! Bruce Quoting dan.meeks222@...: Thanks Bruce - |
|
Dan-
toggle quoted message
Show quoted text
It might also be interesting to take a look at the confidence check data for boty Ecal units. It will give you two options: 1) Display ratio of measured to "known"; 2) Display measured and "known" separately. Chose ratio - the problem with the raw data display is that the plots overly each other (assuming the unit is good). Cheers! Bruce Quoting dan.meeks222@...: Thanks Bruce - |
|
Bruce -
yes good point, the low freq is easy to cal with a relatively low cost kit. The problem is if I need a cal that spans the range of the low freq and the high freq. I can't do a single cal for that. With ECal I believe it will prompt for each standard, then knit the calibrations together to cover the band. So with the 85062 set I should be able to cal over the entire range of the VNA (50M to 20G). I love this particular cal standard set. F-M connectors, covers my VNA range. I may ask the seller to do a partial refund, then either wait for another 86001 to come on the market, or investigate paying for repair / calibration of this one. |
|
Dan -
toggle quoted message
Show quoted text
I've never tried this, but you can design a custom cal kit using the cal kit editor. Conceivably you could use standards below 2G and Ecal above. It might get messy and maybe other members have experience with the editor. Refend night be reasonable - how much do you ave in the lF Ecal - possibly could buy some standards or as you say another ECAL. Cheers ! Bruce Which connectors on tour Ecal units (N.SMA,3.5, etc) Quoting dan.meeks222@...: Bruce - |
Thanks - would be nice (even if only temporary) if I could do a hybrid cal: E-Cal 1GHz and above, and normal cal standards below, and have it stitch it together to get a clean sweep across that boundary. That's how the ECal units work. I have used an old 3.5 mm ECal, and found it inaccurate. I believe I read somewhere that the modern ECals are more stable than the old ones. Clearly a good ECal is very convenient, but personally I would not trust an old one. I don't believe you can get them calibrated either. It's probably better to just invest in a mechanical cal kit. Dave Dave |
|
Wayne ZL2BKC
Hi Dan,
You have a similar setup as mine (8720C/85060C/85064-6000x). ? I can safely say there is no way to stitch together 2 sets of calibrations with manual and ECal because one is driven by the 85060C and the other from the 8720C and whenever you start a new cal process the old one is overridden.??? Only way I can conceive this would be to download the Cal matrices to a PC and merge them together (and interpolate to get back to 201/401/801 points) but given they are a internal binary format that's undocumented, it will be very challenging, but technically possible. However I have had some issues with mine which I have managed to fix, and it may help you on your way to fixing yours. 1) My High Band module which failed with similar fault as yours "Device not in RF Path" after a few sweeps ran on the network analyzer.?? I resorted to using the scatter gun approach and replaced most of the SMD chips without success.??? When Steffan posted some of his reverse engineering notes we came together (his findings are included in the files section of this group) we both managed to get our units working.? Mine turned out to be a broken via causing one of the PIN switches to not activate. 2) Cannot remember if it was specific to the 86050C or 85097 (for my 8753E) but one ECal didn't work.? Turns out there is a device detection loopback between pins 5 to 18 of the DB25.? This is a soldered wire jumper and was missing on the module in question.? Add the wire and module fixed During my efforts with Steffan, I strung together an Arduino sketch to manually control the ECal module to dump my data and troubleshoot my fault.? From that I was able to extract the S1P data, activate each mode manually and compare against the reference.?? I was able to get a near-perfect match with the data extracted.?? From that it was determined the calibration uses a hybrid of TRM and TRL.?? I didn't ponder any deeper than that only to determine it wasn't a simple "OSL" using a small set of numbers.? That's why there is a 256K EEPROM on board! Given your cry for help, I have started to tidy up my previous work.? Give me a week and I plan on uploading the sketch, connection details and basic operation guide to the files folder. During my efforts I managed to pencil together a block diagram which was helpful with my troubleshooting, and I also captured the basics of the low band PIN driver which uses discrete components. Hopefully you are comfortable enough in the micro-controller world, or know someone who can help if we need to test the internals. -- Wayne |
|
Wayne ZL2BKC
Hi Dan, |
|
Wayne -
toggle quoted message
Show quoted text
Very nicely done - you have been busy :-) Cheers ! Bruce Quoting Wayne ZL2BKC <w.knowles@...>: Hi Dan, |
|
Wayne - holy cow. Dude. Very nice!
Here is what I have been up to. The Ebay seller of the modules refunded me an amount good enough so that I could purchase another of the low band modules, and another group member was nice enough to sell me his. Which he checked out on his 8510 system so I am 100% sure that this module is working. However on my setup with the 85060, it is not recognized.? This new module has a much higher SN than the ones I had that were broke (but the VNA did recognize them).? Does anyone know if there is a firmware update for the 85060 controller? Mine has version 1.00. Keysite site does not list anything for firmware for this instrument. I am wondering if the newer ECals may have some difference that the 85060 doesn't like, and hopefully they fixed with firmware. |
|
Wayne ZL2BKC
Hi Dan, Wayne |
|
Wayne thanks - yes I found your firmware, saw that it was the same as mine. And I have verified that the jumper does not affect my setup. The old module SN 141 did not have it so I added it but it didn't affect anything (that module was properly identified by the VNA but had an RF problem). The new module SN 898 already has that jumper installed. There must be some other configuration issue I guess.
Thanks |
|
Wayne ZL2BKC
Dan,
Given the link is in place I would next check the voltages are present on the connector.? Specifically the -12v (which actually is 11.8 from my controller) is specific to the low band module as outlined in the table in my document. ?? Not having the negative rail may explain the symptoms you see for the module that is identified but returns "not in RF path" error.?? Given the potential to plug the wrong things into the port I can see there being a protection fuse on these rails so worthy of checking. *???? GND?????? 2??????????????????? Digital Ground ?*???? GND????? 15??????????????????? Digital Ground ?*???? +5V?????? 3?? ?????????????????VCC Digital (< 50mA - OK to power from Arduino)?????????? ?*???? -12V???? 17??????????????????? Analog -12v ?*???? +5V????? 16??????????????????? Analog 5v ?*???? +24V???? ?1??????????????????? Analog +24V (20-24V @ 750mA Startup) ?*???? GND? ????14??????????????????? Analog Ground This doesn't explain the later S/N '898 not being detected at all, but worth checking it is receiving the +5V to power the digital logic -- Wayne |
|
Thanks Wayne. Voltages look good. 5V is 4.56V, -12V is -12.6V, hopefully those are ok.
To clarify - the unit that I had previously with the "not in RF path" error, that one actually worked on one port if I did a 1-port cal. The B side works, the A side gives that error. So I am going to assume that the controls are working, but the Port A side has been ruined, and the analyzer must be telling me that it does not see the standard that it expects. The fact that this new unit 898 is not even recognized makes me think the FLASH chip must be bad, as that must store the ID of the module. What doesn't make sense is that the seller tested this with his 85097 + 8510 setup and it was fine. I would suspect that I plugged in the module while the 85060 was ON... seems like a bad idea. But I previously read the 85060 manual and it specifically states that "modules may be connected or disconnected during control unit operation,..." so I have not been turning the 85060 off when making connections. I may swap the FLASH chip between my old and new modules. I know the data will be invalid, but this will tell me if the FLASH chip has failed. Which would be bad. Maybe before that I'll ohm out the vias on the board... Any other suggestions appreciated! |
|
I checked vias and didn't find any issues.
So I bit the bullet and swapped the FLASH between the units, and mixed results: 1. the module is now identified by the VNA. Seems to indicate that the FLASH chip was bad. 2. I get the old "not in RF path" error when I try to run a cal. Reminder - this unit worked with the old owner's setup 8510 + 85097 + PC. Also remember that my1-26.5GHz module works perfectly. I love the idea of ECal but I am close to giving up and cutting my losses. Sell the working 85060 + 85062-60002 (1-26.5G) and hopefully have enough $$ to buy a standard cal kit. Thanks Dan |
Dr. David Kirkby, Kirkby Microwave Ltd
to navigate to use esc to dismiss