EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: kada on September 07, 2015, 04:27:34 am
-
I bought a nonworking 8595E analyzer. The analyzer did not turn on. I was able to repair the power supply. When I inserted the repaired PSU to analyzer the -15V LED turns off and -15V disappear.
The problem was in the burned shorted tantalum capacitor located on the Narrow Band IF Board which was replaced:
(http://s10.postimg.org/pk1wgbqt1/IMG_2048.jpg) (http://postimg.org/image/pk1wgbqt1/)
After it I can power on the analyzer but other problems were revealed:
1. Very slow reaction to controls and display refresh - need to wait 10-15 seconds after each key presses. And spectrum refreshes once every ~30 seconds. Empirically I found out that if disconnect the A25 Counter Lock board the problem is resolved. The board is defective?
2. No response of analyzer to 300 MHz calibration signal. The calibration signal level and frequency was checked using frequency meter and millivoltmeter. It corresponds to 300 MHz -20 dBm. If I try to do FREQ CAL the error will appear: CAL SOURCE NOT FOUND.
I measured the signal with millivoltmeter at the point A (after Attenuator -> YTF -> LPF) in the band 0 and it seems good. So I guess the problem may be in the Dual mixer or Second converter. But I have not instrument to check the high frequency signal after mixer.
How I can to localize the defective part? I have not RF generator or other analyzer. I have only 15 MHz generator and 100 MHz DSO.
Analyzer screen after power on (A25 Counter Lock board connected):
(http://s3.postimg.org/xkoz0xv0f/IMG_2050.jpg) (http://postimg.org/image/xkoz0xv0f/)
After 10 seconds:
(http://s24.postimg.org/68d9qqiap/IMG_2052.jpg) (http://postimg.org/image/68d9qqiap/)
Thank you all who responded.
-
Does the oven manage to stabilise after the unit has been powered on for a significant amount of time, i.e., does the 'OVEN COLD' message disappear?
-
Does the oven manage to stabilise after the unit has been powered on for a significant amount of time, i.e., does the 'OVEN COLD' message disappear?
'OVEN COLD' message disappears after 5 minutes.
And I got the same error after FREQ CAL attempt:
"CAL SIGNAL NOT FOUND"
*LOCK OFF
SRQ 110
(http://s13.postimg.org/sgxpcpwtv/IMG_2057.jpg) (http://postimg.org/image/sgxpcpwtv/)
I also tried to do FREQ CAL with "-37Hz" passcode. The process starts in this case and I got the following message at the end of process: "CAL: SPAN SENS FAIL"
(http://s2.postimg.org/w5kghnejp/IMG_2060.jpg) (http://postimg.org/image/w5kghnejp/)
The same test with A25 Counter Lock board inserted I got the "CAL: ZERO FAIL" error:
(http://s29.postimg.org/40kll5m8z/IMG_2069.jpg) (http://postimg.org/image/40kll5m8z/)
-
Hope this will help.
Options:
With A25 Counter Lock board inserted:
(http://s4.postimg.org/yyc560nnt/IMG_2068.jpg) (http://postimg.org/image/yyc560nnt/)
Without A25 Counter Lock board:
(http://s30.postimg.org/wajwahmp9/IMG_2061.jpg) (http://postimg.org/image/wajwahmp9/)
CAL DATA:
(http://s23.postimg.org/gpfhaku9j/IMG_2062.jpg) (http://postimg.org/image/gpfhaku9j/) (http://s15.postimg.org/67dzsh613/IMG_2063.jpg) (http://postimg.org/image/67dzsh613/) (http://s3.postimg.org/mi0bjk6kf/IMG_2064.jpg) (http://postimg.org/image/mi0bjk6kf/)
-
Your best bet is to plow through the service manual flowchart to see where that takes you.
The only thing of note from your pictures is that it is showing cal date for 10Hz up to 5MHz, and I would have expected it to have started from 1kHz, but check the service manual to see what it should be giving.
-
Your best bet is to plow through the service manual flowchart to see where that takes you.
The only thing of note from your pictures is that it is showing cal date for 10Hz up to 5MHz, and I would have expected it to have started from 1kHz, but check the service manual to see what it should be giving.
It have OPTION 130
-
Hi,
do you have any WLAN equipment? One example of testing would be to configure a Wireless Access point to channel 36 and connect that to the Spectrum Analyzer. You should then see a Signal at 5.18Ghz if the SA is working.
That would also allow to see whether your Problem is only present on one Band. IIRC the 8595 uses two Bands: one from 0 - 2.75GHz, and one from 2.7 to 6.5Ghz.
If you connect the Input to and Select 300Mhz as CF, the 300MHz signal is first upconverted to 3.921Ghz, and than downconverted again. If you set your CF to 5.18Ghz, the signal is directly downconverted to 321.4Mhz. So the result of this test allows to isolate the Problem to a certain area.
I don't know whether the Counter Lock board is needed for tuning the YIG - if it is, i would try to fix that board first to make sure that it isn't actually causing the problem you're seeing.
Regards
dxl
-
Did you check ripple on the power supply rails with your scope?
When you say the A25 Counter Lock board is "disconnected", what is disconnected? Everything? This board has a bunch of connections on it.
Could be an LO problem. Can you set zero span and check the output of the YTO with the equipment you have?
-
This is just a random possibility, I haven't looked at the manual etc, but from my experience doing the cb repair work a pll lock issue could easily be due to a bad reference oscillator/crystal, might pay to check that too.
-
Thank you colleagues for help. :-+
Did you check ripple on the power supply rails with your scope?
I did not find any info related to ripple in the service manual but looks like the ripple much more than it should be.
Measured with short probe ground lead at the motherboard test points:
+5V DGND:
(http://s24.postimg.org/8a85jqy6p/5_V.jpg) (http://postimg.org/image/8a85jqy6p/)
+12V DGND:
(http://s18.postimg.org/po2nn0t05/12_V.jpg) (http://postimg.org/image/po2nn0t05/)
+15V AGND:
(http://s10.postimg.org/gyg1inw9h/15_V.jpg) (http://postimg.org/image/gyg1inw9h/)
-15V AGND:
(http://s24.postimg.org/b9f26v7o1/15_V.jpg) (http://postimg.org/image/b9f26v7o1/)
And first of all need to reduce the ripple. I checked caps inside the PSU with RLC-meter and they all looks good. Maybe the reason in the caps on other boards?
-
I think the ripple looks ok. It's never going to be perfect. The +12V could be a little quieter, but it's not the cause of your problem. I'd move on. It's just good to check for anything that's wildly out of spec or oscillating.
And I'd recommend going through the procedure to backup your calibration constants (chapter 3 of the Assembly Level Repair manual), so you don't accidentally lose them while servicing. It takes 10 minutes to write all the numbers in a chart.
-
Thanks all for suggestions.
I built a simple RF diode detector circuit and connect it to oscilloscope to detect RF-signals up to 10 GHz. It works perfect, I tested it with 2.4 GHz radio modules.
I decided to remove the YTO and test it RF output. And I did not see any signal. I tried to change SPAN to 0 Hz, FREQUENCY, but I did not see any signal. Also I checked +5 and +15V rail going to YTO and it is OK. Also I checked current flowing through Main coil - it was 100 mA at zero SPAN and 3.25 GHz center frequency. The FM coil current is zero.
So I guess YTO is dead. Model: HP 5086-7903. Is it any chance to repair it?
-
I rather suspect that you'll need wire bonding kit to repair the YIG, and possibly more.
-
Good luck, they turn up on E-Bay on ocasion.
-
I agree - it sounds like your YTO is in trouble.
Besides +5V and +15V, my schematic shows there should also be -15V on pin J3-1. Check that too. Check all the supply voltages at the header connector right at the YTO. It's rare, but I've seen broken wires and bad IDC crimps on ribbon cables.
Before declaring a dead YIG, inside the can there should be a small circuit board with some individual components you could check. Unfortunately I couldn't track down a schematic for it, but perhaps someone else has one. I've attached a picture of it from a current ebay listing.
The 1826-1048 chip is an opamp. You could reverse engineer the circuit a little to figure out the inputs and output enough to determine if it's working. (Based on a very loose cross-reference search, it's possibly an OP-07).
There's also a 4-pin MOSFET and some diodes, caps, etc. to check.
-
I was looking at one of these on eBay the other day - am I right in saying that the tracking generator option only goes up to 2.9 GHz? Any other models I should consider?
-
I was looking at one of these on eBay the other day - am I right in saying that the tracking generator option only goes up to 2.9 GHz? Any other models I should consider?
You are correct.
I see you've already started a separate thread for other model suggestions.
-
Thank you all for helping. After the YTO replacing the analyzer works fine. FREQ/AMPLITUDE calibration passes.
I also have another 8595E with defective YTF :palm: But it works fine at the band B0 only (9 kHz - 2.9 GHz).
-
Glad you got it working!
I also have two 8595E and one has the exact same problem that the YTF is dead, leaving band 0 still working.
Thinking I couldn't make matters worse, I opened the YTF and one of the YIG sphere's rolled away. Doh! It had come off its mount. That's the end of that.
"Fixed" it with a sticker on the front panel reminding me band 1 is dead.
-
Thank you all for helping. After the YTO replacing the analyzer works fine. FREQ/AMPLITUDE calibration passes.
I also have another 8595E with defective YTF :palm: But it works fine at the band B0 only (9 kHz - 2.9 GHz).
Glad to hear you got it working Great. :) :-+
-
Hello Kaka and MarkL,
Just been reading this thread with great interest
I believe I may have a similar problem on my 8595E, but also with a temperature related issue on Band 0
Problem 1:
Band 0 works for a short period, but after 10 minutes warm up and connected to 300 MHz Cal, the Cal signal gradually drops by 15db, then finally at -36dBm it drifts off frequency and the FREQ UNCAL error appears.
I checked and proved the actual cal signal is good an 300MHz -20 dBm.
Problem 2:
Band 1 does not work at all, no signals present. If I switch to a Band 1 range like 4 GHz , 500MHz span the FREQ UNCAL appears and the unit becomes unresponsive.
Would you agree that a bad YTO could produce both of these problems?
and...
How on earth to you get access to the YTO ?
I can't see any easy access path, does the side PCB need to be removed along with the front panel?
Thanks
T.
-
It's possible the YTO could be at fault, but I wouldn't jump to that conclusion.
I would try to characterize the problems more, and one problem at a time. Start with Band 0 since it's at least somewhat functional. The first thing I would try is to look at the cal signal in Band 0 in zero-span mode. Does the signal still drop in amplitude after 10 minutes?
If so, does the amplitude also drop on the video out? Can you verify the YTO frequency and output level when it happens? It may be good to verify the test frequencies and levels shown on the block diagram in the Assembly Level Service manual.
For access, I've found it easiest to remove the screws from the front panel, pull it forward, and remove the entire RF section and set it to the side with it still connected. Be sure to remove the hard line SMA from the back of the input N connector. It's easy to mangle the hard line when removing the front panel.
I don't recall if I needed to add a couple of SMA jumpers to work on the RF section when it's out of the chassis, or if the existing cables were long enough. It's a big spaghetti mess of wires, but I don't know how else it could be done. The fun part is getting it back in, so take lots of photos of how cables were routed.