EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: energy_hunter on November 20, 2016, 08:21:48 pm
-
I have an HP8753D VNA that is intermittently working.
History
- VNA wasn't used for ~6mos. I moved it and broke off the keyboard connector. When I started the VNA it displayed a "ROM 1M FAIL" msg.
- After a little research, I found that the CPU board needed a new EEPROM (U4). I ordered a Firmware Upgrade kit from ebay for V6.14.
- When I replaced all 4 EEPROMs the machine started fine. Hurray! But then I moved it from the floor to the bench and the VNA would not start up. |O
- I read the service manual and the CPU boots up properly with all 4 red LEDs on. The front panel lights are correct and boot up properly but the display is blank.
- When I hit PRESET, I see "hp" in white letters on the display so I think the display is working but the instrument is not sending display signals possibly from the CPU board.
- I replaced the electrolytic caps on the CPU board and tried a new Graphics processor board but still have the same problem.
- Once when I tapped the machine on the display side the VNA booted up properly and showed the normal display.
Questions
1. Does anyone think I need a new display? These are $400.
2. Should I try a new CPU board? These are also $400.
3. Any advice would be appreciated. Thanks.
-
Seems like you should quit moving the instrument - it's bad luck.
Have you tried hooking up an external monitor to the video connectors on the back?
Power rails all verified???
-rastro
-
Wiggle some connectors, I had same on 8753A with CRO module. After banging unit few times it came back on :D. :box:
-
Is it possible it's a boot up problem and the dark display is just the symptom? It may hang-up after the system displays and clears the "hp" message - right??
Some quick checks:
If you have an s-parameter test set like an 85046/47 then you should be able to toggle between "Trans FWD" and "Trans REV"; the LED's on the test set should toggle.
After the 8753 has had time to boot try switching between S11 to S22 measurements.
-- Press [MEAS] then top monitor menu key for S11
-- Press [MEAS] then 4th from the top monitor menu key for S22
If this does not work then it may be getting stuck somewhere in the boot cycle.
You should also see responses on the RS232 and GPIB interface if you have the interface capability.
-rastro
-
Thanks everyone for their ideas. I will try to address them all in one reply. Your help and ideas are truly appreciated. :-+
- The power test points on the Post Regulator have all been verified and checked w/ an O-scope. I also opened up the power supply (pre-regulator) and didn't see any bad electrolytic caps.
- It could be a loose connector that is needed for the boot up sequence. I am able to get the VNA display to turn on if I pull out all the cards and put them back in without securing them with the cross bar. This is not a sure thing but it seems like moving the cards, cables or mother board connectors around gets the unit to boot up properly. At least I know the general area of the problem.
- It looks like the instrument is booting up properly but the display signals are not turning on. While the display wasn't working I put the Green BNC to my o-scope and saw a regular 25.5kHz pattern but no signal riding on it. No signal on the Red or Blue. Then I pulled the IF and PHASE LOCK cards out and the display started working on the VNA and I saw the signals on the o-scope. Also when the monitor was on I ran a display self test and it returned the correct results.
- I will try sending GPIB commands when the display is off to see what I can measure.
Does anyone know what the motherboard looks like? Could putting the cross-bar on the PCBs break a cold solder joint on the motherboard connectors? Or does the crossbar move some of the connectors on the first three PCBs and it prevents the system from doing a measurement?
Overall I agree with rastro that it is a boot up problem. It's hard to diagnose without a display. So far after removing the IF and PHASE LOCK boards and turning on the instrument I can get the display to turn on. When cards were in and it was working I have left it on for an hour and it measured fine on both channels with a load on each so I believe the instrument is worth trying to save.
Could be a cable problem. I think that's my next step is to take off the cables and test their integrity with my VNA...aw man!
-
You find dead caps through measuring, not by just looking. It is not consumer junk ;) Just repaired my Marconi SA, had to replace over 100 caps. Not a single one popped but DF varied between 1 and 5. And Only the 33 and 10 uF.
It sounds to me like a bad solder joint or dirty connector. I often use a stereo microscope to check solderjoints. Then suck the old solder away and resolder the component.
-
I think PA4TIM has some very sound advice regarding caps and solder joints. A 7X to 10X stereo microscope can really aid identifying cracked and loose parts that you will not notice even with a magnifying glass. I am really glad I invested in one.
And YES it sound like your system is worth fixing.
It really looks like an intermittent associated with re-seating a few PCB's. It can be tedious tracking down an intermittent like this but probably means you won't have a major assembly to repair/replace.
Also if it hangs while booting the GPIB/RS232 will probably not work either. This route would be useful if the problem was after successful boot-up.
Best course of action is to poke around looking for the unstable connection.
-rastro
-
You all are awesome for offering your advice and it is sincerely appreciated. I did some more troubleshooting. It may be possible that by pushing the cards on the motherboard side that I'm shorting something on the CPU board as the aluminum platform beneath the CPU board flexes. When I pull out the IF and PHASE LOCK cards I can get the display on with the terrible looking sweep you see when there is no IF. This may be pulling the aluminum plane away from the CPU board. So my plan now is to focus on the CPU board and look for any contact points that may cause failure. It could be a cold solder joint somewhere or a cap but I think I'll try adding insulating standoffs on the CPU board first.
-
Hello,
I found this post after much searching. It describes the exact issue that I am seeing on my 8753D. The screen is blank, system boots fine, front panel leds ok, four illuminated cpu board leds, and when you press reset the small white "hp" in the upper left displays. I'm just wondering if you ever solved this issue.
thanks
-
Sorry for the late reply on this to close the loop. I finally managed to fix the problem by changing a capacitor on the pre regulator power supply. Deep within the firmware upgrade instructions are also instructions that require reducing a capacitor on the power supply by a factor of 2x. Once I did this the problem was fixed! I can now cycle the power on and off and the display does not go blank. There must be some power-down error that sticks in memory and protects the display and when you reduce the value of this capacitor then it works.
Reference: pg 28 of the Installation Note
8753D and 8753D Option 011
Network Analyzers
Firmware Upgrade Kit Revision 6.14
Kit Number 08753-60185
08753-90393.pdf
"This modification applies to serial numbers 0000A00000 to 3410A06400".
"Remove pre reg board and replace C7 with a 220uF, 50V electrolytic capacitor"