Help to repair or reuse parts of Agilent N9360A GSM Tester

[RichardFr]

Hello,

I have recently got opportunity to buy an Agilent N9360A Mobile Station Tester for a very cheap price. It's a very specific device, intended to simulate or test GSMs and related protocols.
I don't have direct usage for it, so I do not forcibly want to repair it, but at least I would be really happy if I could reuse some of its parts, like the 10MHz reference.

https://www.keysight.com/us/en/product/N9360A/mobile-station-tester.html


The device is faulty, the stand by LED comes on, but when I press the start button, the front LED color changes (from orange to green), and if I look very closely to the LCD screen with a light source, I can see some random black lines (and no backlight). But nothing else happen : HDD not spinning, Fan not spinning, no sound...
There's only another LED on inside of the device, it stays lighted only as long as I have the finger on Start button.

When I release the start button, the device goes back to Stand By.

I did some measurement on SMPS outputs.


When the device is in "stand by" I get :
- 12V on the yellow wire, so it looks like to be a permanent voltage.
- nothing on the red and the oranges ones.

When I put it "on" :
- The 12V remains steady (not a single mV of change)
- The orange is to 3.3V
- The red one is 1.18V.
I find this last one quite suspect !
Yellow / Orange / Red makes me think to AT/ATX PSUs, so I could attempt to feed in some 5V (in case the PSU would be faulty), but it could also be intended to power up some 1.2V ASICS (for example), so if I put in some 5V, we all know what would happen .

When the device is on Stand By, power consumption is 17W (that's a lot to light up an orange LED !)
When I press the Power On button, it goes to 30W. It comes back to 17W when I release the button.

I have also :
- Replaced the CR2032 on motherboard (completely dead : it was giving 0V when I measured it)
- Verified if the fan is OK (I have already seen high end devices to refuse to start in case of dead fan)

Just for those interested, here's the mainboard and its HDD :


The 10MHz ref is on a dedicated board, and plugged only with a 4 wire connector for power and control, and some coaxials for 10MHz in/out, so it should be easy to reuse if I know how to wire it. There are two things written on the PCB : CGD-230A and 7PCMD0592B. I thought maybe the same ref was used in other Agilent devices. Here's a picture :


Also, the PSU has a really interesting design, with 3 completely independant PCBs, each of them directly plugged to the mains :


Thank you for your help, of any kind (informations about the 10MHz reference, confirmation of PSU expected outputs...)

Happy christmas to everyone !

[RichardFr]

Hello,

Quick update : while dismounting the PSU, I found a sticker with no details, but with just what I was needing : the output voltages of each module.
And I found the red one is a 5V, so the "middle" PSU is faulty.
I have injected a +5V with proper grounding on this power line with a lab PSU, and the device booted fine.
It's based on a customized Linux, and the mainboard is a P3 700 with 256MB of RAM and a 40GB HDD.
I can boot it to the main screen. If I start the "GSM" operations, it fails later, but at this stage it could mean anything.

Few notes, if it can help somebody one day :
The PSU startup command is made by two wires, it's not a short circuit as in an ATX PSU, instead apparently it requires to inject a +5V, which comes from the permanent 12V. The "+" side is close to the outputs, the "-" is turned to mains input.
The startup signal is present on my device, so the issue comes from the +5V module itself. I'll have a look on it.

I have also noticed that the 10MHz is always powered up (stand by drawn power fall from 17W to 10W if I unplug it), this explains in part why the device consumes so much power even on stand by.

I'm still interested by any technical infos on this device and its parts

[RichardFr]

Nobody ? 

[RichardFr]

Meanwhile I repaired the PSU. It was a dead capacitor (C10, 50V, 39µF Nippon Chemi-Con LXV. Visually perfect, but found dead while measuring in-circuit, with very high ESR and almost no capacity).
BTW, the thermal design of the PSU is pretty bad : three boards with little heatsinks stacked in an almost closed metal box, and the fan is located on the other side of the device, so it does not help. The three PCBs, especially the 5V one, have browned close to the power components, especially transistors from the primary. At this price, this device deserves a clearly better design.

After booting and starting the "GSM" feature, it still displays an error, "00001002, unit initialization failure".
And also, I was unable to have anything in output of the 10MHz reference, even if I let it warm for almost one hour. Apparently the error is related to it, which is not a good sign, as the reference was primarily one of the components that I wanted to salvage !
The reference is connected properly, and it gets hot, but nothing at the output (except a short spike at 3V during the power up). The device is configured to use internal reference.
I will have a look to the manual for the error code and try an external reference, just in case.

[PA0PBZ]

The reference is connected properly, and it gets hot, but nothing at the output (except a short spike at 3V during the power up). The device is configured to use internal reference.
I will have a look to the manual for the error code and try an external reference, just in case.

Can you measure directly on the OCXO? Maybe there is something wrong with the INT/EXT logic.

Also a reply to let you know that you are not just talking to yourself in this thread 

[RichardFr]

Can you measure directly on the OCXO? Maybe there is something wrong with the INT/EXT logic.

Also a reply to let you know that you are not just talking to yourself in this thread 

Thank you for the answer and for letting me I'm not alone
Direct measurements on the OCXO was my next intention when I will be able to spend a little more time one it, it was already late yesterday when I finished to repair the PSU

I will also attempt to figure how the 4 wires of the module are used (obviously at least one for the permanent 12V, and one for ground).