| Electronics > Repair |
| Agilent 34461A corrupted flash |
| << < (35/41) > >> |
| analogRF:
if you are not connected to LAN why is the meter thinking it is connected? still it could be a hardware issue like not properly soldered BGA ball or a RAM issue. you can test the RAM by 'mtest' command at p510> prompt without any address, it takes a while to finish because it will check the whole RAM. as for BGA ball suspicion, I'd recommend use heat gun with a narrow nozzle on the CPU (low temp like 200C or even less) just for a few seconds while you power it on and see if the behavior changes. you can also do that same thing for the RAM and see if the behavior changes if you rule these out then most likely it is just a corrupted flash (nk.nb0 kernel is bad) and needs to be loaded by ymodem (or tfpt from LAN) to boot the device and then perform fw upgrade. your FW is 2.17 right now so you will need nk.nb0 extracted from FW2.17 I assume you have already tried the obvious first thing which is just putting the .UPD file (FW update file) on the USB flash inserted in the device and power it on. that's the first thing I would have done before anything else. I think your normal boot log shows it has fired up the USB so it may look for update file on it. Most like this will not solve the issue but it wont do any harm . I have the bin file extracted from the FW. if you do all the above tests and still inconclusive. just post message here and I will tell you how to upload and boot the device using the external kernel |
| Eng_tech:
Uboot and Wince are new to me as far as troubleshooting goes. I write plenty of programming code for work and home projects for many different microcontrollers and FPGAs, but I have never worked on an operating system like Wince or on someone else's equipment design when it comes to embedded OS. I'll follow your advice and see what happens. I have not tried putting the .UPD file on a flash drive and booting. I thought the front panel USB would not be working until WinCE fully was loaded and running. I'm pretty sure the spear is seated good and all balls are connecting. I made a small fixture to make sure the spear was aligned exactly over the pads when installed. I did redo (hot air on top and bottom of pcb) for chip yesterday just to make sure. Still, you never know as I don't have Xray machine to check it. Well, I do at work but never used it. I don't think they would be too fond of me using it for personal use but wouldn't hurt to ask I guess. It has been sitting ever since we got it. No one has ever used it yet. As for LAN why do you think that the meter thinks it is connected to a LAN? I did do a mtest yesterday but for some reason it aborted and I didn't run it again. I'll try again today after I get out of work. This is just a strange thing, The spear microprocessor goes bad and shorted out the 3.3 supply to it. Put new spear in and 3.3v supply source looks good now. It appears to be a common problem with these spear processors failing and when they do NAND or RAM gets corrupted also. But anyway I'll run some more tests. I have read through the U-Boot reference manual to get better aquainted with it and it's commands. Oh, you thought LAN was hooked up because of this. Mac address .......... (80:09:02:01:BF:42) Ip address ........... (192.168.1.166) Subnet Mask address .. (255.255.255.0) DHCP ................. (Enabled) Only did this after loading image 2. It wasn't hooked up though. Thanks for your assistance! |
| analogRF:
--- Code: ---WaitForLink Start (ticks=3751) Link Detected (ticks=3753) GMAC Init : 100 Mbit/s FULL DUPLEX (MII) Flushed Transmit Buffer phyCfg->dwSpeed 0x64 <--EDeviceInitialize --- End code --- although it is strange that it also says --- Code: ---Phy not found --- End code --- still this may not point to anything though.... if the CPU is fine then the only other thing I can think of is the kernel being corrupted which needs to be loaded by ymodem (very slow) or tftp (very fast) still booting with a USB with FW file is good idea I checked my logs and it seems the mtest doesnt work on these meters I think it is not implemented correctly or somthing it works on keysight scopes though. a pointed/localized heat gun for a few seconds on the memory chip at power on usually will detect if the RAM is bad. if those will get you nowhere I will upload the kernel and you can boot the device either by ymodem or tftp I highly recommend tftp because the loading process may encounter errors and in tftp you can repeat it very quickly many times with ymodm is very slow and if error occurs the you wasted like half an hour for nothing... |
| Eng_tech:
Well, booting with USB stick with firmware didn't work loading either image 1 or 2. Guess I'll try what you recommend using tftp and nk.nb0. I tried hooking the LAN connection to a network switch as well as my laptop but I couldn't establish a connection. Maybe I need to make up a cross over cable and try that. What are your instructions for doing this? What program are you using to run tftp on your computer? MobaXterm? |
| coromonadalix:
mobaxterm putty no gui, bitvises ssh client, voc terminal free for 30 days normally the cross over era cable are supposed to be over, any good router hub should have auto mdx negotiation, all our scopes, dmm, gens etc are straight cables, hub will help if the meter is in dhcp wireshark free maybe could be useful, when you start the meter you can see its ip address or any requests it will make on the network, once you find the right lan connection in wireshark be sure your main system is not consuming data, video audio updates ... |
| Navigation |
| Message Index |
| Next page |
| Previous page |