Repair: Anritsu MS2651b Spectrum Analyzer with failed Display (Missing SYNC)

[jkf1000]

I bought this unit a few months back and have been looking at it between other projects but now its time to sort it out. I have spent a lot of time with the scope trying to understand the digital signals around the MPU and the video chip.

What I have found so far. No HSYNC or VSYNC timing pulses from the HD63484 chip, but I did notice that when measuring levels with the DMM the LCD did start scanning line by line as I touched the HSYNC connection on the video chip with the probe which suggests to me that the LCD is fine and just waiting for a sync pulse. All clock lines are present and healthy at around 6.4Mhz, the power on reset delay is working perfect with a 214ms delay to the processor and video chip.

My initial assumption was a failed video chip but replacement made no difference.

I have looked in more detail at the signals from D0 to D15 on the MPU and all look healthy and visually similar when viewed on the scope, the signal lines then go in two different parallel directions with one direction passing through a pair of HC245 transceivers then on to the Scan/AD board, the other direction passes through a pair of AC245 transceivers and then it gets more confusing.

The output of one the AC245 chips feeds directly to the D8 to D15 pins of the HD63484 video chip and the data viewed here is again healthy and visually the same for each data line, all good.

The output from the second AC245 connects to the D0 to D7 pins, but also connects to the M66011FP serial/parallel chip and the ADV476 Colour RAMDAC. When scoping around the D0-D7 lines at the video chip it looks as if the data is becoming corrupted/changed. I have what appears to be a good data stream on 3 of the eight pins but very unusual pulses on the other 5.

The problem I have is my experience with digital electronics is very limited, so with that in mind I may just be mis-interpreting what I am seeing. It appears to me that the video chip is not receiving the correct information for it to operate correctly. My suspicion is one of the other chips sharing that part of the data line has gone bad, but that is just a best guess based on the little I know so far.

I welcome any input from anyone that be able to put me on the right track with this one as I am getting a little lost and don't want to just start randomly replacing ICs which may or may not be at fault.
What sort of signals should I expect to see on the D0-D15 pins?

I have added a photo with the data paths marked, what isn't shown is that the two AC245 transceivers are between the MPU and the rest of the lower chips but are mounted on the underside of the board. ( As a quick check I have removed and swapped around the AC245 chips but the data issue stays the same)

Any help or suggestions as to which direction to go next are greatly appreciated.. Karl

[singapol]

I'm no expert and anritsu does'nt give out schematics. Sometimes we tend to think problem is complicated
like what you did..checking signals to and fro but overlooking the simple and obvious like:

Is the lithium battery still good at 3 Volts? the date code on the altera chip is 1995, a good 20 years old.

Have you checked the esr of the electrolytics caps especially the small one near the black wire at the bottom right corner same goes for the other scan pcb you mentioned?
                                                                                                       Good luck.

[jkf1000]

I fully understand what you are saying about overlooking the obvious sometimes, but unfortunately not the case here. I started with the simple and most obvious and worked up from there. All the electrolytics have been replaced on this board and also the power supply as a matter of course. The lithium battery is holding a solid 3v and no signs of leakage. As for the Altera CPLD that is always a possibility but as with most repairs it is a process of elimination starting with the simplest and most likely suspects first.

I know in the end it will be a simple fix, the problem is how to find the fault in the first instance especially with a lack of supporting information.. Karl

[singapol]

So I did some reading of ms2651 operation manual that has block schematics- 4 pages.
You mentioned video chip HD63484 was replaced and you mentioned measuring clock lines at 6.4Mhz. is that
the oscillator clock signal at pin 50- 2CLK? HSYNC and VSYNC originates from this chip. Do you know there are
3 speed version of video chip which is 4, 6, and 8Mhz. Since you measured 6.4Mhz. What marking is on the oscilator, I suspect that your new chip is wrong speed at 8 Mhz.? What speed is original chip?

Edit: My apologies I guess a higher speed device can be run at a lower speed but still why no H and Vsync?

[jkf1000]

Thanks for taking the time to look into this. The block diagrams are all I have and I have studied them for a long time trying to make sense of what I have with my machine.

There are actually 4 different speed alternatives for the video chip right up to 9.8Mhz but I replaced mine with the 8Mhz option as that is what was origionally fitted to my unit and it suited the available clock speed.

I still suspect that the data is being corrupted by something else on the circuit which is preventing the video chip from initialising but without removing/isolating other components in that part of the circuit I am not sure how to narrow down the search.  . Karl

[singapol]

When you say failed display,do you mean it's dead no backlight at all? If that's the case power supply is suspect but do know that the lcd backlight uses high voltage .I don't know what's the high voltage like for a 5.6in. display but for a 17in. one it's over 600Vac  so don't use your multimeter if it's only rated for 600V! Also check the power circuit for failed components like power transistors,diodes or hope it's not the stepup transformers.

If there is a screen on power up then suspect fault is somewhere else like CPU or front end rf circuits.

[jkf1000]

The LCD module itself appears to work fine but I have tried a spare one just to be sure. The fault is in the digital side somewhere. I still suspect that the video chip is not receiving the correct digital signals.. Karl

[singapol]

Are you saying there's a screen display when you boot up except that it can't show measurements?

[jkf1000]

The LCD screen powers up and so does the backlight. All that is displayed is a few vertical lines and random colours. The only time I had anything different was when I measured voltages around the video chip with the DMM which caused the LCD to scan lines randomly down the screen when I touched the HSYNC pin which again suggests that the LCD isnt the problem.

[singapol]

Now it's more obvious, there is a break or interuption between microcontroller and video interface
 ( HD63484).
Touching probe to Hsync showed line scanning...that's s good sign. If we assume the obvious could be a contact problem.

1. You need a magnifing glass to see if the affected pin is deformed or touching the PLCC ic.

2. Use continuity check to see if bottom pin of pin connects with top spring tongue of H/Vsync.

3. Try reheating or reflow of solder joint or replace with new PLCC socket. A lot of work I know and can damage copper tracks.

4. Clean contacts with contact cleaner or alcohol.

Then there's the mysterious front board designation in the block diagram of Anritsu 2651 in it there is a sync separator block although it may not be directly related to this video problem.

PS: Also check if there is clock signal for mcu. I hope you are using a proper PLCC ic. extractor tool instead of small narrow tip screw driver. Hope wire harness or flexi cable and connector to LCD display is good

[jkf1000]

Unfortunately it is not a connection issue, the plcc socket is brand new which I fitted to assist in fault finding and because it is easier when fitting a new IC. Just as a sanity check after fitting the socket I used needle point probes to check continuity between the IC pin and the socket itself and all is fine.

The HSYNC and VSYNC signals leave the video chip and go to the Altera PLD chip where the signals are combined before being sent to the sync separator on the front panel board and the Sony composite generator chip on the right of the main CPU board. I have no suspicions of any issues on the front board with the way the LCD started scanning when probing HSYNC pin with the DMM probe.

I do have the correct PLCC removal tools so limited risk of damage to any chips.

The HSYNC is held at 0v and the VSYNC is at cmos high (5v) I have also measured this with the CPLD chip removed from the board in case it was faulty and creating a low impedance path to ground dragging down the HSYNC line but measured results were still the same. The impedance to ground measured at both sync lines is around 5Mohm with all chips in place so no loading by other components.

I still think that the data on lines to D0 to D7 is being corrupted and not allowing the video chip to initialise correctly or the data from the MPU is corrupted.. Karl

[singapol]

Check ADV476 ramdac and esr of 47uf 10V tantalum next to it.

[jkf1000]

Already done it, reads at 50uf (47uf) with an ESR of 0.35ohm. I have checked that the fault is still present even with the RAMDAC removed so unfortunately another dead end. the only thing left on the D0 to D7 data lines now is the M66011FP serial/parallel chip for which I don't have a spare to try so maybe time for an Aliexpress order.

[singapol]

Agreed next suspect is M66011 serial bus controller it converts parallel data from mcu to serial and serial to parallel backto mcu. There is also 74HC164 serial to parallel output and 74HC138 3 to 8 decoder possibly
glue logic. Both are near M66011.

[jkf1000]

Just had a look on Aliexpress and I can source replacements for all the chips in that area plus a new MPU (if needed) and a ADV476 for around £25 shipped which will have to wait a few days until payday. I would just be happier if I could narrow it down to one or two components that may or may not be at fault. I don't have a SOP converter for my logic IC tester so am not able to test the chips individually so the Aliexpress order may be the only option.

My main suspicion is still the M66011 IC as that is the only other IC on the D0-D7 data lines that I cannot easily remove. It is too close to the PLCC socket and the oscilator can to remove with hot air so will have to have each leg cut and removed one at a time to release the chip.

The Altera PLD device is a different story, I checked each pin last night with the scope and it appears to be handling the secondary oscillator division and buffering for the different chips on the video circuits (all working 100%) as well as the sync signal combining for the Sony composite generator and the front SYNC separator chip (on the disply board) as well as different memory addressing tasks. If the CPLD is faulty (which I don't think is the case) I dont know how I would replace it. The chip itself to replace is reasonably cheap but its the programming that would cause the big problem.

From what you have learned from looking at the block diagrams do you think that any of the information from the Scan/AD board would have any impact on the generation of the HSYNC/VSYNC signals? From the little I understand of it, all the required digital information for the video signals originate from that board which are then translated by the MPU and sent to the video chip, but the timing signals for the sync lines should be seperate from the image information and generated only on the main CPU board based on the timing requirements of the LCD in use (PAL/NTSC)

I was discussing my fault a couple of weeks ago with someone else on the forum who has a nearly identical analyzer but he is having either level or mixer problems but his display worked 100% so we were able to assist each other with swapping measurements but at the moment I think he is busy with other life related issues which is why I decided to start my own thread just about the issues I am having.

Any further thoughts or suggestions greatly appreciated.. Karl

[singapol]

I attach link for MS2601a (crt display) service manual vol.1. It gives a detailed desciption of the subcircuits and hardware interfacing. It also has a troubleshooting flowchart. It essentially gives an insight into how Anritsu designs a spectrum analyser.

20mb file:
http://exodus.poly.edu/~kurt/manuals/manuals/Other/ANRITSU%20MS2602A%20SERVICE%20VOL%201.pdf

After browsing through it there are a number of things going on in a system. You mention the scan AD board, this is part of the measuring section of the SA. What is key is the interupt control system linking the various ICs. So main MCU- common bus- specific ICs.

I would check what is connected to ADV476 this ic outputs signal/RGB to LCD or possibly the 75ohm composite video output. Also the transciever logic ICs like 74AC541 and 74HC574 (data latch?)communnication with the MCU + ROM ICs which contains the firmware + front panel,keyboard etc . There are over 20 logic ICs in the photo. HD63484 is the graphics display plotting engine. Use resistance measurements of multimeter to check logic to ground or between pins you may get lucky.

[jkf1000]

Just dug a little deeper with the DMM and it appears that the two RAM chips and 4 ROM chips share the common signal lines. If I have a failed ROM chip that would make a lot of sense with the issue I am having. I don't suspect the ADV476 or the composite generator, but I am now thinking of maybe a failed memory IC.. Karl