Author Topic: Agilent ESA E4401B Memory (B72) and Firmware upgrade (Read 7446 times)

Nitrousoxide · « **on:** January 30, 2019, 03:39:16 am »

Hi all,

I recently acquired an old E4401B. It is a fairly old model as it has an earlier serial number (less than 4144), and thus does not come with the expanded memory upgrade. I would like to install this as it allows the latest firmware to run on the device and allows for features such as 1 Hz RBW (It also has the high accuracy source upgrade, however, they only added 1 and 3 Hz RBW in a later firmware upgrade).

My attempts so far have been somewhat successful. There are two components to upgrade, the DRAM SIMM and the flash SIMM. The DRAM simm was trivial as any old stick that has comparable timings would work (and in fact did so), the devices detects that it now has 32 MB of RAM.

The flash SIMM was not as trivial. I had managed to source some identical chips off ebay from China (they are obsolete, that was the only source I could find). Another source on the internet stated that the last two jumpers (out of the four) on the flash SIMM should be opened, whilst the top two should be shorted.

Before attempting that, I wanted to check if the device could boot into the firmware updater. Alas, it could not. This is where my problem is.

I followed the instructions given by the firmware update guide. I tried formatting the disks both from windows and from the spectrum analyser, and then dropping the files onto the floppy. But once I insert the disk and reboot (as said by the service guide), it just boots straight into the operating system.

The spectrum analyser is running possibly the earliest firmware version of A.01.03

Does anyone have any tips, suggestions or experience?

EDIT:
When I insert the firmware update disc, I can hear the floppy drive seek twice and then it seems to just give up.
The machine just hangs on a white screen afterwards.

EDIT 2:
I tried formatting the floppy disk in the ESA, then proceeded to copy the bootloader onto the disk in linux. The machine attempts to boot from the floppy disk, It seeks a few times and then loads. However, the screen remains white and the machine sits in a locked up state.

Nitrousoxide · « **Reply #1 on:** February 11, 2019, 04:18:28 am »

Update:

I assumed that the internal battery being dead may cause operational issues. So I ordered a replacement from Arrow (reputable source) replaced it with the higher mAh version in hopes of prolonging the need to replace the battery again.

However, it seems to be even worse now. Before the time would be held since the last power-up, i.e. It was set to 10th of January, but the time would only increment once the device was on, but it would still maintain the time. Now, when I enter the correct time and power cycle the machine, the date resets to some bogus value. (i.e. Date feild just has 122, 44 instead of 11th, Feb 2019)

Also, of note, is that every time it boots I receive a "Copywright Hewlett- Packard Company 1999". I assume this is standard in earlier firmware versions before boot screens.

I feel like a better place to ask this would be the keysight forum.. if it wasnt for their god awful user interface....

Any ideas why the memory seems to be not holding values? Should I just contact their official support?

Ordinaryman1971 · « **Reply #2 on:** February 14, 2019, 03:07:22 pm »

So, the battery you've replaced is the Yellow TimeKeeper battery that's in the socket on the board?
Do you still have the old one, did you measure the voltage on it, is it still acceptable or it's dead?
How did you replace the battery, did you take the board out to do it? I don't remember if it's possible to replace it when it's in the chassis.
Sorry for asking trivial questions, I really apologize but I just don't know your setup.
Do you have ESD protection in place, everything nicely grounded?

Nitrousoxide · « **Reply #3 on:** February 15, 2019, 04:04:08 am »

Quote from: Ordinaryman1971 on February 14, 2019, 03:07:22 pm

So, the battery you've replaced is the Yellow TimeKeeper battery that's in the socket on the board?
Do you still have the old one, did you measure the voltage on it, is it still acceptable or it's dead?
How did you replace the battery, did you take the board out to do it? I don't remember if it's possible to replace it when it's in the chassis.
Sorry for asking trivial questions, I really apologize but I just don't know your setup.
Do you have ESD protection in place, everything nicely grounded?

Yes, I made sure to get the exact same timekeeper snaphat. Only difference is that it is a higher capacity.

I have the old one, measured the voltage. It was way below acceptable, around 300 mV.

I replaced the battery step by step as per the service guide. Removed the support bar, unplugged the front panel cable and then lifted the board out. When removing the snaphat, I used an IC extractor tool, pulled the thing off in one go.

Yes, I had ESD protection. I was grounded.

No need to apologize, any feedback at this point is excellent as I am stuck against a brick wall

xrunner · « **Reply #4 on:** May 14, 2021, 01:22:12 pm »

Quote from: Nitrousoxide on January 30, 2019, 03:39:16 am

Does anyone have any tips, suggestions or experience?

EDIT:
When I insert the firmware update disc, I can hear the floppy drive seek twice and then it seems to just give up.
The machine just hangs on a white screen afterwards.

EDIT 2:
I tried formatting the floppy disk in the ESA, then proceeded to copy the bootloader onto the disk in linux. The machine attempts to boot from the floppy disk, It seeks a few times and then loads. However, the screen remains white and the machine sits in a locked up state.

Hi. I found the solution for this as I was trying to do the same thing. I'll go ahead and post it here so as to help it be more available to others. It turned out to be quite a rabbit hole! I will PM you to let you know about this.

Quote from: xrunner on May 13, 2021, 11:01:28 am

Agilent E4411B Firmware Upgrade

OK let's review this mess.

Trying to install the firmware for an Agilent E4411B that does not have the memory expansion. Current old version was A.05.01 (Sep 2000). The latest firmware that can be installed without the memory expansion was A.07.06 (Nov 2001). With the memory expansion I can install A.14.06.

I created the floppy set using my PC exactly as instructed by Agilent's directions (see attachment). I did a clean format for each disk. It did nothing because the 1st floppy that had the loading program (ESALOADR) was not being read by the E4411B.

After several more attempts I realized I had to go down the rabbit hole. What was wrong? Was the USB floppy bad? Was Windoze 10 making bad copies to the floppy? What phase is the Moon in? I copied some small files to the floppy and it performed flawlessly reading and writing. The files could also be copied onto the C: of the SA and renamed (more on that issue ...).

Doing more research on this I ran across this thread from Keysight -

Keysight help thread

What? Format the disks on the SA and it fixes the issue? I reformatted the disks on the SA but still had to get the files on them at the PC. So I did all that and it still wouldn't load and run ESALOADR !

What I wanted to try next was to keep all the file actions in the "ecosystem" of the SA. To do that, I had to get the ESALOADR onto the SA drive C: (so as to have the SA copy it onto the floppy). When I tried to copy that file - nothing happened. I find out that the SA file menu choices do not operate at all with files that do not have an extension. It simply ignores them! OK, so back to the PC. I took the disk back and added an extension to the file name like this (so it would recognize it) -

ESALOADR.TXT

After that, the SA copied the file to a TEMP Dir I made on it's C: drive. So I had ESALOADR.TXT loaded onto the system. However, it won't do squat that way. Has to be on a floppy disk. Next I re-formatted the floppy on the SA, and copied the file back to the floppy. Now to re-name the file ... oh nope ... it won't re-name a file to anything that doesn't have an extension. In order to do that I'd have to go back out of the SA system. I didn't want to do that, so I thought, why not try it as-is and see what happens? Well I'll be damned if it didn't start loading/running ESALOADR.TXT!

So it seems, for some reason, writing ESALOADR to the floppy from the PC, even if the floppy is formatted on the SA, doesn't sit well. I do not know why. I have not found any other reference to this exact solution on the internet. I ordered the flash chips and memory stick and I plan to do the memory upgrade and get this beast all the way up to A.14.06 firmware. I'm putting that first floppy disk with ESALOADR.TXT in a safe place too!

Nitrousoxide · « **Reply #5 on:** May 15, 2021, 05:24:52 pm »

Thanks to xrunner, I have made some significant progress (all be it 2 years later).

Copying his instructions (almost verbatim). I got the firmware loader to work. The only difference is that when copying the files on to the local storage, I had to rename it as a .csv rather than a .txt file.

Revisiting the storage and memory situation- I had soldered on two extra NV memory chips bought from Aliexpress, hopefully, they're not full of sand and are functional (well... hopefully they are full of sand... technically...).

Apparently, I had bought a 32MB RAM stick before, but it wasn't in the machine so the only logical explanation is that it evaporated and I will have to get another 32 MB stick.

Now, herein lies the final hurdle. The non volatile SIMM stick has 4 jumpers to indicate the capacity (I assume they aren't the same as EDO ident). So the question remains, how to configure them. I found this on another forum and will attempt to this when the rest of the components arrive:

I restored the flash chips to their original locations, and confirmed
that everything was working as before. Then I ran the flash erase
with several of the other jumper configurations. After finding the
correct jumper configuration, it allowed me to install the latest
firmware, and everything appears to me working well.

If anyone else wants to try this, you need to have both the jumpers
closest to the SIMM's edge connector open, and the other two shorted.
You will have to run the secure flash erase to reformat the SIMM (hold
"view/trace", "ESC", and "return" while turning on the analyzer. The
screen will be blank, but you can tell it's working by checking for
the Vpp voltage of one of the chips. Give it about 10 minutes, and
then pull the plug to turn it off. After erasing the flash, it won't
work until you install the firmware.

Now, the real challenge is if I have the guts to wipe the firmware. I had previously contacted Keysight support (about a year ago) about this and the response I got was, effectively, a very polite middle finger: "No, it's broken (the motherboard), we won't help you, but our new stuff, bye." (paraphrased).

xrunner · « **Reply #6 on:** May 15, 2021, 09:48:00 pm »

Agilent E4411B flash memory secure erase and restore firmware

Quote from: Nitrousoxide on May 15, 2021, 05:24:52 pm

Now, the real challenge is if I have the guts to wipe the firmware ...

So we can both feel cozy about it (this is going to have to be done anyway when I install the two flash chips) better see how it goes now and know it will work, rather than wonder if this or something else is wrong later - I'll take one for the team.

Let's do this ...

I powered on the unit with ESC & VIEW/TRACE buttons held down and it went into flash erase mode. Absolutely no indication of progress or completion - just a white screen. Service manual sez "wait approximately 10 minutes ...".

I decided to wait a bit longer ~ 12 minutes. I powered it off, then on, and nothing but a white screen. Yep definitely nuked, but that is expected. Inserted ESALOADR floppy and re-cycle pwr again, and sure enough it automatically starts the firmware loading program. That fail safe ability resides in another chip on the main board. Firmware re-loaded and back up and running like nothing happened.

Now that wasn't so scary was it?

Well OK just a little bit.

Ordinaryman1971 · « **Reply #7 on:** May 15, 2021, 11:05:01 pm »

Very cool.

Nitrousoxide · « **Reply #8 on:** May 16, 2021, 04:10:31 am »

Quote from: xrunner on May 15, 2021, 09:48:00 pm

Firmware re-loaded and back up and running like nothing happened.

Question:

You had to reformat the ESALOADR drive on the SA. For the floppy's that contain the firmware, did you have to follow the same procedure for each disc? Or could you just format them as FAT on windows for them to work?

xrunner · « **Reply #9 on:** May 16, 2021, 11:29:03 am »

Quote from: Nitrousoxide on May 16, 2021, 04:10:31 am

Question:

You had to reformat the ESALOADR drive on the SA. For the floppy's that contain the firmware, did you have to follow the same procedure for each disc? Or could you just format them as FAT on windows for them to work?

I formatted all the disks on the SA. The only file I had to have written by the SA was the ESALOADR. All the others had the files written to them from the PC on the PC's USB connected floppy drive (Win 10).

It's possible that the floppy disks other than the one having ESALOADR on it could have been formatted by the PC, I just did them all on the SA to be sure. But if there is a mysterious reason they have to be formatted on the SA, and you formatted them on the PC, you wouldn't be able to re-format on the SA if you are trying to load firmware AND you have nuked the system. At that point you can't get back into it to get to the format menu. That would be a bad day.

After I get the system upgraded with all the memory and latest firmware, I might try another experiment. I'll format all the disks other than the one that holds ESALOADR on the PC and see if the SA will load the firmware from them. I would keep the disk with ESALOADR (once it works) in a safe place and don't muck with it ever again.

I should have the stick of 32 MB ram in the mail Monday. I'll install it and see if it's recognized. Even though it won't do much at this point until I get the flash upgraded, it's one more step completed.

xrunner · « **Reply #10 on:** May 17, 2021, 11:47:42 am »

I got a minor issue fixed with this testing.

Several years ago I was saving some state files for fast instrument configuration. Since the file naming menu had the options for both upper case and lower case letters ... well why not use them all? Well, later I found out the file deletion won't delete any files created with lower case letters.

Since the flash erase nuked all those files, they are gone now.

Looks like it was fixed in this revision under "Defects Repaired" -

A.08.02

...

Prevented the File Rename operation from creating filenames which could not be loaded or deleted (previously it permitted lower case characters)

...

So this should be covered when I get the last available firmware loaded, provided all the memory upgrades work. It will prevent you from creating file names that can't be deleted from the file menu.

xrunner · « **Reply #11 on:** May 17, 2021, 09:59:56 pm »

Couple of things showed up today for this project.

The flash memory board, besides needing the two additional flash chips, will need the missing decoupling capacitors installed. Four more for each chip (0805 size). I can do that now while still waiting for the chips to arrive.

When I got this unit in 2016 I replaced the Timekeeper battery. It contains a 3V backup battery and common 32.768 kHz crystal for time. Figured might as well replace it while doing all this other stuff.

xrunner · « **Reply #12 on:** May 19, 2021, 10:50:03 am »

Replaced Timekeeper crystal/battery module. Old battery (installed 2016) read 2.8 V - New one 3.4 V.

Received the 32 MB SIMM stick (used part). Removed original stick and put new one in. The unit recognized it with no issues (see pics). However, it doesn't enable any further functionality (as far as I know) until the flash is upgraded ... which then allows the latest firmware to be installed. That version is A.14.06 1 SEP 2008. That's what we're shooting for.

I included pics of the old and new SIMM sticks. Nothing special about them, but if it helps or answers any questions in the years ahead for anyone reading this, so much the better.

Tracking on flash memory chips states "May 18, 2021 3:07pm Departed from Origin Country"

xrunner · « **Reply #13 on:** May 20, 2021, 07:10:38 pm »

I had the 0805 decoupling caps so I went ahead and installed them on the board. Eight total - four for each additional flash chip.

xrunner · « **Reply #14 on:** May 23, 2021, 10:12:35 pm »

So the flash memory isn't far away, but I felt like playing ...

What happens if I try to install the latest firmware without the additional flash memory?

Since it's going to be nuked again anyway, let's see! I wrote all the floppies and started the ESALOADR utility. Is it going to check if I have the B72 memory option when checking disk one - before starting?

Nope.

It actually finished the firmware upgrade and asked to be restarted. After it successfully re-started, it automatically asked to have the PowerSuite utilities disks inserted. This is where I suspected it would fail, and it did. But ... the actual system firmware is now A.14.06!

Might be interesting for anyone who wants to try this. You can have the latest firmware without the utilities if you desire. The C: drive is also a bit smaller. It only has 1.9 MB free now.

Party on ...

Nitrousoxide · « **Reply #15 on:** May 24, 2021, 07:11:51 am »

Quote from: xrunner on May 23, 2021, 10:12:35 pm

So the flash memory isn't far away, but I felt like playing ...

What happens if I try to install the latest firmware without the additional flash memory? Since it's going to be nuked again anyway, let's see! I wrote all the floppies and started the ESALOADR utility. Is it going to check if I have the B72 memory option when checking disk one - before starting?

Nope.

It actually finished the firmware upgrade and asked to be restarted. After it successfully re-started, it automatically asked to have the PowerSuite utilities disks inserted. This is where I suspected it would fail, and it did. But ... the actual system firmware is now A.14.06!

Might be interesting for anyone who wants to try this. You can have the latest firmware without the utilities if you desire. The C: drive is also a bit smaller. It only has 1.9 MB free now.

Party on ...

Wait... That's mental. Meaning that even if the flash chips are fake, then the upgrade would still work...

Ok, I will attempt the same when the SIMM stick of ram arrives.

I guess I owe you a beer

xrunner · « **Reply #16 on:** May 24, 2021, 11:40:44 am »

It's funny what you can find out if you just do something. Of course there are risks too. I've had this unit since about 2016 and I just took the word that it couldn't be upgraded without the B72 option. Like I found out, you can't install the Power Suite utilities without additional memory. But the latest base firmware can be installed.

This whole thing started for me when I was reviewing the Signal Path's video on fixing an E4407B -

https://youtu.be/gb1QMJtwumQ

I thought I'd delve into it and update my E4411B. After research I learned that there was a version of firmware that could be installed (A.07.06) without the memory upgrade (according to Agilent). That's when I had to figure out the trick to getting the damn ESALOADR to run ...

Anyway, having that in-hand, I didn't have any fear of experimenting because I could get back to the A.07.06 version. So why not try A.14.06 ?

Hey wait - I just realized something. Shahriar installed new firmware on his unit in the video from floppies and never said a thing about overcoming any problems running ESALOADR.

xrunner · « **Reply #17 on:** May 26, 2021, 11:06:29 pm »

The flash memory chips arrived (Sharp LH28F320SKTD-L70). Might not get to them until the weekend.

The jumpers need to be the same as the fully populated example picture. So I'll remove one jumper at the bottom of the four possible positions.

xrunner · « **Reply #18 on:** May 29, 2021, 09:30:26 pm »

I removed the bottom jumper and parked it to the side in preparation for the board upgrade. I soldered on the two new flash devices. I carefully inspected my work and cleaned the board. I installed the board again, and performed the flash erase (power on while holding ESC and VIEW/TRACE). I already had all the floppies ready to go with A.14.06, so proceeded to re-install the firmware. Unfortunately I got the message shown in the pic -

"Create a drive with 100 k free to continue"

Bummer

After trying several things I decided to re-inspect the flash board. I found what looked like several pins that could have remained unsoldered. Very hard to tell, but after touching them with the iron and re-installing the board, the message went away.

The firmware installed and re-booted. The system indicated I indeed now officially had Option B72 memory expansion. It then asked for the personality disks for installation of the Power Suite utilities. But it failed on the second floppy "Failed to write staging file". I turned the unit off then on again and tried it through the system menu with the same result, but it failed at a slightly different point. I could tell this by watching the "Bytes Remaining" counter. I then tried the "Verify Disks" selection, and it also failed to verify. When it fails to verify it does not give the "Failed to write to staging file" error.

I then suspected that a floppy was either bad or needed to be re-formatted and re-written. I fully re-formatted all three personality floppies and copied all the files to them again. However, this failed again at a slightly different point. I discarded the three floppies and used three from the firmware set, and again it gave the same error but never at the same point. I also downloaded the package again from Keysight, and re-wrote the floppies and it failed again.

OK this is getting really aggravating.

I went back and got the Power Suite utilities from the previous firmware A.07.06 which only resides on one floppy instead of three like this latest version. This installed OK. So right now I'm beat down and need to take a break. I have the B72 memory option installed, as shown on the system screen, the latest firmware, and the Power Suite utilities from the previous version. The unit is working fine. I will have to think about why the utilities won't install and see if there is anything else I can try. The error is always "Failed to write to staging file".

Nitrousoxide, I'd like to know how this goes for you, if you get to this point.

xrunner · « **Reply #19 on:** May 30, 2021, 02:07:04 pm »

Whew - I finally got it all installed.

I re-re-inspected the flash board this morning with a clear head and two address lines on U2 were still not soldered.

It sure looked like they were but it can be deceiving sometimes. That explains why the installation could not write a file at some point.

The first two disks of the three-disk supplemental file set were finally read and installed without a hitch. It skipped the last disk on purpose (as indicated in the display), I assume because it was already installed by the latest firmware.

So there you have it. I hope all goes well for you Nitrousoxide, and I hope this all will help someone else in the future.

xrunner · « **Reply #20 on:** May 31, 2021, 01:23:45 pm »

One final experiment that I wanted to do. I wondered if after all the upgrading whether the first floppy with the firmware loading program (ESALOADR) would now work straight off the drive, instead of having to run through all the hoops to get it to load (if you missed all that hoop-la go back in the thread ...).

I re-formatted a floppy and loaded ESALOADR on it from the PC and stuck it in the E4411B. I turned it on and ... nope, it still will not read the program like it should.

If anyone out there has an explanation I'd love to hear it.

Nitrousoxide · « **Reply #21 on:** June 26, 2021, 06:51:15 am »

Well,

Much delay in the revival of the project. I got around to playing with it again today.

The new stick of ram I got didn't have any jumpers soldered in (not a good sign!!!). However, I soldered the appropriate jumper on the PCB and gave it a shot, and.... nothing. White screen of death.

The new stick has KM44V4004BK-6 chips on it. Seems to be almost identical to the chips on your stick yet... nothing. So either the stick was DOA or even worse, the motherboard in the SA is faulty. The funny thing is, with no jumpers it detects only 8MB of ram.

I guess I just go for a shotgun approach and just buy a few different ram sticks and hope that one... sticks.

This upgrade is probably the more headache-inducing venture I have ever done in my EE career.

EDIT: I just looked at the really fine details of the memory part numbers...

I got a 3v3 2K ref whilst you have a 5V 2k ref... I hate technology sometimes...

I guess I had some form of foreshight and ordered some of the identical flash chips from aliexpress, so I will solder them on to the existing module and give that a shot.

ua4yhz · « **Reply #22 on:** April 15, 2022, 06:27:04 pm »

My E4411B firmware loads normally. Only need to format the floppy disk on the analyzer itself.
But the firmware A.14.06 does not want to download. And knob "continue" don't work.
I ordered memory chips LH28F320SKTD-L70 for option B72.

ua4yhz · « **Reply #23 on:** April 16, 2022, 08:38:58 am »

Quote from: Nitrousoxide on June 26, 2021, 06:51:15 am

Well,

Much delay in the revival of the project. I got around to playing with it again today.

The new stick of ram I got didn't have any jumpers soldered in (not a good sign!!!). However, I soldered the appropriate jumper on the PCB and gave it a shot, and.... nothing. White screen of death.

The new stick has KM44V4004BK-6 chips on it. Seems to be almost identical to the chips on your stick yet... nothing. So either the stick was DOA or even worse, the motherboard in the SA is faulty. The funny thing is, with no jumpers it detects only 8MB of ram.

I guess I just go for a shotgun approach and just buy a few different ram sticks and hope that one... sticks.

This upgrade is probably the more headache-inducing venture I have ever done in my EE career.

EDIT: I just looked at the really fine details of the memory part numbers...

I got a 3v3 2K ref whilst you have a 5V 2k ref... I hate technology sometimes...

I guess I had some form of foreshight and ordered some of the identical flash chips from aliexpress, so I will solder them on to the existing module and give that a shot.

As an alternative to the LH28F320SKTD-L70 chip, I think it is possible to use LH28F160S5T-L70A or LH28F160S5H-L70.
But I didn't check it.

philpem · « **Reply #24 on:** April 16, 2022, 09:47:52 am »

Quote from: ua4yhz on April 16, 2022, 08:38:58 am

As an alternative to the LH28F320SKTD-L70 chip, I think it is possible to use LH28F160S5T-L70A or LH28F160S5H-L70.
But I didn't check it.

Depends if the analyser needs the full amount of memory. The 28F320 is a 32Mbit part, the 28F160 is 16Mbit -- half the size.

Two 28F160 chips would give 8Mbyte Flash
Two 28F320 chips would give 16Mbyte Flash

Edit: from the screenshots above, 8MB Flash is the unexpanded module, which won't fit the latest firmware - so the 28F320 chips are required


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: Agilent ESA E4401B Memory (B72) and Firmware upgrade (Read 7446 times)

Share me