Agilent E8357A

[tv84]

Having seen the old and new HostID from Joe, I guess the conversion rule may be very simple:

Old PNAID=01ABCDEF
New PNAID=FEABCDEF

So, probably, the conversion rule is

FF - 01 = FE

or

FF XOR 01 = FE

Others may confirm.

[caesarv]

It sort of warms my heart to see all my licensing code being examined/reverse-engineered.
FYI...in case you are curious, the 95403 number was picked because it is the zip code for the HP/Agilent/Keysight factory in Santa Rosa, CA.

[tv84]

It sort of warms my heart to see all my licensing code being examined/reverse-engineered.
FYI...in case you are curious, the 95403 number was picked because it is the zip code for the HP/Agilent/Keysight factory in Santa Rosa, CA.

I love to know the story behind these small nuggets... Kudos. 

[joeqsmith]

Gave it a spin.   Shown with the home made 70mm Beatty, uncalibrated.  Using the 9GHz range for this. 

Also shown is the $120 LiteVNA vs the 20 year old PNA sweeping a 7GHz HP filter.   The Lite's IF was set to 4kHz, 401 data points.  The PNA was 20k, 801 data points.  PNA supports unknown thru but using a simple SOLT cal on both. 

Good to see it all start to come together. 

It sort of warms my heart to see all my licensing code being examined/reverse-engineered.
FYI...in case you are curious, the 95403 number was picked because it is the zip code for the HP/Agilent/Keysight factory in Santa Rosa, CA.

Best post in a very long time!!! 

[TheSteve]

This is all good stuff! I love free bandwidth from heaven.

[joeqsmith]

I felt I owed it to everyone involved to make one last action shot to show off the PNA's higher range settings.   

Both the LiteVNA and PNA were set to 1601 data points, 1kHz IF, 300k-9.0GHz.  No averaging was used.  Both ports were terminated during the measurement.

The LiteVNA was calibrated using the standards supplied with the V2Plus4 which includes the second best sorted load from Mini-circuits using the ideal model.  The leakage error term is being corrected. 

The PNA was calibrated using my homemade standards that I characterized using Mario's MatLab scripts against a set of Agilent standards.  The best of the eight sorted loads is used (not a bad $80 investment).   These were characterized to 6GHz.  Coefficients are being used and the unknown thru model is selected. 

For fun I set the PNA's IFBW to 100Hz as well.  This is about as fast as the LiteVNA at 1kHz. 

[Forzaman]

Hello. I have a newly acquired E8357A that has option 015 (configurable test set) already installed. At the time I purchased my
VNA, Keysight had a store on eBay that was offering the time domain option (010). After I received the VNA I realized the Keysight eBay store was no longer selling the options. I asked Keysight Used for a quote on the time domain option, but was told they do not deal with consumers only commercial clients. I was pissed, so now I find myself here looking to hack the time domain option, and to throw in the 9GHz upgrade for good measure.
Reading the posts here, I see the procedure to get the new HostID, but how do I calculate the FLEXlm licenses afterwards? I'm a newbie to this option hacking process. Any help would be very much appreciated. Thanks!

[PA0PBZ]

Reading the posts here, I see the procedure to get the new HostID, but how do I calculate the FLEXlm licenses afterwards? I'm a newbie to this option hacking process. Any help would be very much appreciated. Thanks!

Since I missed all the fun with the .exe files (I guess that was a PM to TV84), I will jump in here  Post your new HOSTID and the wanted options and some magic will happen.

[tv84]

It seems this quest had a bit of history in recent weeks...

For reference only.

[Forzaman]

Reading the posts here, I see the procedure to get the new HostID, but how do I calculate the FLEXlm licenses afterwards? I'm a newbie to this option hacking process. Any help would be very much appreciated. Thanks!

Since I missed all the fun with the .exe files (I guess that was a PM to TV84), I will jump in here  Post your new HOSTID and the wanted options and some magic will happen.

Thanks PA0PBZ! I will clone my HDD before messing around with the files since the unit came with a 6-month warranty. I ordered a HDD that is scheduled to arrive mid-week. Once that is done I will post my HostID. Very much appreciated!

[joeqsmith]

It seems this quest had a bit of history in recent weeks...

For reference only.

I thought maybe we would get lucky and Keysight would provide support, or at a minimum they would provide CV with approval to support them.   CV's comment about the end of his career certainly paints a grim but all too common, picture of today's company culture in the USA.   I'm thankful we have member's not only capable but willing to help.   

Another thing that came out of that thread was CV's response about the lower frequency limit and there was no option to bring it below 300kHz.  This is why I had told you that I don't think there is anything else worth going after.  If he had said there was some experimental, unreleased option,  I think it would have been worth spending more time with. 

[joeqsmith]

I will clone my HDD before ...

Smart!  This is the first thing I typically do when working on old equipment.    I am using a Transcend 2.5" PATA SSD in mine.   I have used this same drive in other XP & 2k equipment with very good results.  Last time I bought them, they came direct from the factory so they may be getting harder to come by as there is little demand. 

[Forzaman]

I will clone my HDD before ...

Smart!  This is the first thing I typically do when working on old equipment.    I am using a Transcend 2.5" PATA SSD in mine...

I was thinking about trying a PATA SSD, but was reading on the Internet that it wouldn't make much difference on Win2k because it doesn't support the TRIM function among other things. A YouTube video showed the bootup time of a Win2k PC with a HDD, and after the drive was swapped with a SSD. The time was essentially the same, but you are saying there was a noticeable improvement in the VNA performance? Cheers.

[Forzaman]

It seems this quest had a bit of history in recent weeks...

For reference only.

I thought maybe we would get lucky and Keysight would provide support, or at a minimum they would provide CV with approval to support them...

I have purchased a lot of test equipment during my career, and still do so as a consultant, so I have a good relationship with the R&S, Keysight, Anritsu, etc. sales reps. Sometimes I can get them to pull some strings for my equipment that I use for my personal use, but not much luck with Keysight. I was looking at the N4416A test set for my E8357A to expand it to 4 ports, but it requires the N4425A Balanced Measurement Software. The guy at the Keysight Used website told me that he never heard of that software, and the unit is not supported anymore anyways. My Keysight sales rep said he would ask around, but told me that it was nowhere to be found, essentially saying it was lost. There are plenty of these test sets up for sale on eBay, but they are no use to anyone without the software. Such a waste of perfectly good hardware.

[joeqsmith]

I will clone my HDD before ...

Smart!  This is the first thing I typically do when working on old equipment.    I am using a Transcend 2.5" PATA SSD in mine...

I was thinking about trying a PATA SSD, but was reading on the Internet that it wouldn't make much difference on Win2k because it doesn't support the TRIM function among other things. A YouTube video showed the bootup time of a Win2k PC with a HDD, and after the drive was swapped with a SSD. The time was essentially the same, but you are saying there was a noticeable improvement in the VNA performance? Cheers.

I made no mention of any noticeable improvements in VNA performance.   That said, I think on my one LeCroy DSO, after upgrading from 2k to XP, with and without the SSD, the boot times were a wash.  The small amount I gained with the SSD, I lost with the OS.   Still the time allowed to stabilize the equipment is much longer than they take to boot, so nothing to be gained there.

The reason I chose that particular drive was it supports many of the features not supported by these old OSs internally with it's firmware.   While I was told by members on this forum about SSDs having a very short life (weeks) and lots of fine tuning of the settings required, I installed one where I imaged the drive with no changes what so ever.  Basically I was trying to force the worse possible condition according to the experts to validate their claims.  That was in 2015 and that drive is still in use today with the same setup.

PNA performance wise, I basically use it headless.   In other words, I tend to do all the heavy lifting with an external PC.   The first VNA I bought basically sent me down that path as it had no computer.  So to make it work without the aid of a grease pen, using a PC was the only way to cal it or create Touchstone files.....  The up side to that was it forced me to learn how the VNA worked and understand some of the math behind it.  It also provided me with an appreciation for the real experts in this field, of which I am not. 

Why use an SSD?  I've lost mechanical drives in test equipment before and my choice of moving to SSDs was hopes of improving long term reliability.   Verdicts still out on that one.   That said I have an old 80s DSO that uses a 68k in a VME chassis.  A relic.  It uses an MFM drive.  The Miniscribe drive was locked up, typical of that brand with their out gassing.  I bought a used Seagate ST-225 and hacked it's hardware to mimic the Miniscribe drive.   The PSOS OS they used only supported a couple of different drives.  That old scope with it's hacked up unknown history drive has been in my possession for over 20 years now and is still in operation.   So, there's that data point...    

[Forzaman]

I will clone my HDD before ...

Smart!  This is the first thing I typically do when working on old equipment.    I am using a Transcend 2.5" PATA SSD in mine...

I was thinking about trying a PATA SSD, but was reading on the Internet that it wouldn't make much difference...

I made no mention of any noticeable improvements in VNA performance.   That said...

Thanks for the clarification, and detailed background info! BTW, did you write that VI program that you posted the screen capture of with the impulse response? Very impressive!

[joeqsmith]

I made no mention of any noticeable improvements in VNA performance.   That said...

Thanks for the clarification, and detailed background info! BTW, did you write that VI program that you posted the screen capture of with the impulse response? Very impressive!

Yes, I wrote it.   You can follow the link in my signature to my Github account and select the Solver32 link to see a few screen shots, or for more recent efforts, you can check the following link where I post about using it with the low cost VNAs:
 
https://www.eevblog.com/forum/rf-microwave/nanovna-custom-software/2150/

Anymore, I only develop for the LiteVNA.   $120 gets you square wave drive, 2-port 1-path, 9.3GHz using harmonics and a bit of noise (see post above comparing the noise floor with the PNA).  Still, it's been a lot of fun to play with and I highly recommend it to anyone wanting to learn more about VNAs.   

[caesarv]

Regarding hard drives.  When the PNA first shipped, we used 10GB Hitachi drives.  They were pretty bad and we had to replaced a lot of them.  Then we switched to 40GB Fujitsu/Toshiba drives.  They were much better, but would still fail occasionally.  Eventually we switched to Intel SSDs.  We had virtually no failures with the Intel SSDs.  We may have switched to Micron SSDs recently. 

[joeqsmith]

Good info Caesar.  Mine was supplied with a Fujitsu MHV2040AH.   

I had started a link for this VNA that has my notes including some of your responses to my questions.   Had you stayed on at Keysight a little longer,  I would have hit you up for an computer board upgrade and alignment.  Are you aware of any companies who offer alignment for these?   I should be clear, I'm not asking about calibration. 

After upgrading to the last firmware that supported it in order to enable some of the more advanced diagnostics, it then barked about the receiver cal which appears to have never been done.   When I dug into it,  I suspect the system was old enough that that the MXCAL files were not yet being used.   Seems straight forward enough to go through.     

https://www.eevblog.com/forum/testgear/agilent-e835xa-pna-vna-notes/

[Forzaman]

I made no mention of any noticeable improvements in VNA performance.   That said...

Thanks for the clarification, and detailed background info! BTW, did you write that VI program that you posted the screen capture of with the impulse response? Very impressive!

Yes, I wrote it.   You can follow...

Very cool... both software and Lite VNA. If it had Direct Receiver Access I would be all in. Need it for High-Power S-Parameter measurements; hence the E8357A acquisition with opt 015. Cheers.

[tv84]

Result of sending DIAG:POKE:LONG?

Joe,

What is the result of?

DIAG:PEEK:LONG? 57742, 1

and/or after

DIAG:POKE:LONG 57742, 1

[joeqsmith]

All combinations cause the same error.  Sending DIAG by itself also causes the same error.  I looked at the included help built into the PNA and they also make no mention of this command.  I have attached a copy of the manual I use to program it. 

[caesarv]

That address space (57742) contains the bits to program the functionality of some of the earlier PNAs. Most of the below do NOT apply to the E835xA models.

Test Set Mother Board EEPROM (0xE18E):
      Bit 0: #014 - loops only
      Bit 2: #1E1 - add source attenuation
      Bit 5: #016 –  receiver attenuators
      Bit 6: #081 –  reference channel transfer switch

You can set or clear them as desired, but obviously they should reflect your actual hardware. You can set them via the Option Enable program; no keyword is needed for these particular options.

[joeqsmith]

I don't have any of the equipment required to perform the alignment but that didn't stop me from trying out the software.   I was surprised it created two new mxcal files without the anything attached to the GPIB. 
 
Files are named mxcalfile_r4p12 & mxcalfile_r4p21.   Maybe r4 is receiver #4 with S12 & 21.   I have attached one of the files as a reference.  Sadly, they did not make the file format available or it may have been possible to align it using equipment beyond what was originally called for. 

[tv84]

All combinations cause the same error.  Sending DIAG by itself also causes the same error.  I looked at the included help built into the PNA and they also make no mention of this command.  I have attached a copy of the manual I use to program it. 

My inferences were based on the file attached and what I saw when I looked at the code.

I guess that the way one can deal with the EEPROM to enable/disable the options that don't use via flexlm licenses is by manipulating the EEPROM bits directly.

Of course, my command syntax may be wrong, and better yet, as Caesar indicated the HW may not allow the combinations. But these commands are used for something...

No messing with Caesar! 

Test Set Mother Board EEPROM (0xE18E):
      Bit 0: #014 - Configurable Test Set
      Bit 1: #H11 - IF Access
      Bit 2: #1E1 - Source Step Attenuator
      Bit 3: #UNL - Bias Tees w/Atten
      Bit 4: #075 - 75-Ohm Impedance
      Bit 5: #016 – Receiver attenuators
      Bit 6: #081 – Reference channel transfer switch