Tek CSA7404/TDS7000 repair project (Upgraded memory/CPU)

[Smith]

Progress again: switched the Copermine Celeron 850 for a Tualatin Celeron 1200 (256kb, 100MHz fsb) with the same mods as tin posted on page 1. Runs fine at 1200MHz at 1,51Vcore.

I was looking at the mainboard, after seeing moreproblems with bad caps. I replaced the two caps (1500 and 2200uF 6,3V. The last one was completely dried out and was bulging up and down. While I was there I also placed a dip socket for the DS1245 NVRAM.

[andy2000]

I just bought a TDS7404B which boots to Windows 2000, but Tekscope freezes on the splash screen.  I've tried reinstalling Tekscope without change.  So far, I haven't seen any obvious problems.

The PPC board's diagnostic LED only shows "8." which never changes at any point.  The PPC board does show up in the device manager.  Could this be an NVRAM problem?

[TiN]

Sounds very like it Try to see if your NVRAM settings correct, if they are not - you know what to do. 

How you all guys solving probe situation on these scopes (ones with proprietary TekConnect ports, not BNC) ?
BNC adapters cost relatively a lot, and they don't allow high-impedance inputs either.

[Smith]

Yeah, that's a bummer. The tca-1meg (or something like that) easily go over a grand on ebay. With 4 channels that's a lot! I personally am thinking about building some simple few MHz preamplifiers with buf634 or something like that for basic every day use. Personally I don't need that GHz speed, and  if I do, it will take some good active probes, so that problem is solved.

Still looking at my scope. Yesterday I looked at the acquisition board with a thermal camera. Three op-amps where I think the trigger circuit is idle at 40C, and then one of them ramps up over 100C in a few seconds when the first waveforms appear. Probably it's oscillating. There are 3 of them in a row, and the other 2 just keep working at ~40C. Maybe that's the reason the trigger on channel 3 and 4 are not working. I'll replace it and see what happens.

[nctnico]

Also make sure to check the supply voltages on the opamps.

[andy2000]

One other issue is that the Windows 2000 OS is in pretty bad shape.  It takes about 15 minutes to shut down and various tasks that have to be forced to end.  That's even after upgrading the RAM to 2GB and CPU to a 3.06 GHz P4. 

Does anyone have an image of the 2000 (or better yet, XP) restore disc they could share?  It has an Intel D845GVAD2 motherboard. 

[Smith]

I dont know about the 7000B series, but this is my experience with my 7000 series: I had no recovery CD, just the old Win98 HDD and some drivers from the internet. I placed a new (bigger and faster) HDD, and just installed WinXP from a XP Prof SP3 CD-Rom, installed the drivers, and installed the latest firmware (The tools package, this includes the scope program) from the Tek website. Did windows update, and now everything runs fine! XP runs that much better than Win 98/2K. You might even be able to run Windows 7. Your HW should be able run Win 7 without problems, but be sure to check if your SW supports Win 7.

I did get some CMOS tools from the old hdd, as I was unable to find it anywhere. This was it.

[andy2000]

I dont know about the 7000B series, but this is my experience with my 7000 series: I had no recovery CD, just the old Win98 HDD and some drivers from the internet. I placed a new (bigger and faster) HDD, and just installed WinXP from a XP Prof SP3 CD-Rom, installed the drivers, and installed the latest firmware (The tools package, this includes the scope program) from the Tek website. Did windows update, and now everything runs fine! XP runs that much better than Win 98/2K. You might even be able to run Windows 7. Your HW should be able run Win 7 without problems, but be sure to check if your SW supports Win 7.

I did get some CMOS tools from the old hdd, as I was unable to find it anywhere. This was it.

I'll give that a try.  I was never able to get a generic Windows installation to work properly on my TDS5054 (problems installing the drivers for the acquisition board).  Since this is newer, it's possible the drivers are better. 

[Smith]

Windows did not include the drivers for the ppc board and front panel. I downloaded the drivers mentioned earlier on this tread.

[andy2000]

Windows did not include the drivers for the ppc board and front panel. I downloaded the drivers mentioned earlier on this tread.

When I tried to to a clean Windows installation on my 5054, I used the drivers contained in the "Tektronix" dir on an official hard drive image.  I was never able to get the acquisition board drivers to install properly.  It doesn't seem to be fully plug & play, and I was getting various errors in the device manager.  Even after finding an error free combination, Tekscope still didn't detect the acquisition board.   

Back to the 7404B... I pulled the Dallas NVRAM, and it still has valid looking data, so I don't think the battery is dead.  While I had the PPC board out, I did a little probing around with an ohm meter, and I noticed that the PPC reset button measures 775 ohms across it.  That seems low, so I will start by checking the reset line when I put it back together.  My flexible PCI extender cable will come in handy here since the ATX board sits directly on top of the PPC board.

[TiN]

How did you determine that NVRAM have correct data? I'd suggest use serial console to check all parameters.

[andy2000]

How did you determine that NVRAM have correct data? I'd suggest use serial console to check all parameters.

It's not blank as I would expect if the battery was dead, and I can see the location of the vxworks boot files in plain text, so it's not full of garbage.

The reset button seems to be working fine (at least there's 3.3v on the reset button).  There's absolutely no activity on the console port.  Both TX lines are flat from power on, all the way until the scope application hangs. 

I've checked all the power supply voltages that are listed in the manual, and all the DC voltages on the PPC board look reasonable with none that are obviously missing. 

I did a little probing around the PPC board with an oscilloscope, and I can't find any activity.  Even the DS1245 has nothing on any of the address, or data lines (all high).  Even the chip select and output enable pins are high.

[Smith]

The console port  should be outputting data from the second it starts. A standard mainboard com port "breakout cable" works great. You could try probing the  TTL pins of the max232. Maybe its defective.

What do you see on the seven segment display next to the  console port, any progress on that?

You might want to reset your CMOS from the mainboard too, cant hurt anything.There are 3 eprom/flash chips on my ppc board probably on yours too. Maybe you can check these for activity?  Small chance, but could the sdram memory be defective?

[andy2000]

There's nothing going into the MAX232.

The 7 segment shows 8. constantly.

I have probed the socketed  29LV040, but I see no activity.  I'll take a look at the others later.  I guess clearing CMOS can't hurt.

Swapping the 512MB DIMM on the PPC board was the first thing I tried.

I've scoped out dozens of random test points around the PPC board, and I've only seen DC voltages apart from the various crystal oscillators.

I think I need to focus on the PPC chip itself and see if I can verify the basics such as power supply, clock, and reset based on the datasheet.

How long after power on should it take fro the PPC board to show signs of life on the 7 segment, or console port?  Does it boot from the hard drive before, or after Windows boots?

[nctnico]

I think I need to focus on the PPC chip itself and see if I can verify the basics such as power supply, clock, and reset based on the datasheet.

I would follow that strategy also... Check whether the basic conditions for the CPU to run are there.

[Smith]

How long after power on should it take fro the PPC board to show signs of life on the 7 segment, or console port?  Does it boot from the hard drive before, or after Windows boots?

As far as i can remember the 7 segment display runs from the moment it starts booting. The last stage (running the software and viewing the waveforms appear) it shows a blinking "-".

The console port runs from booting, about until the waveforms appear. They also show info when calibrating, doing SPC, self check etc.

I'm still working on mine, it still has issues. Last week all inputs stopped working until I turned it on again. I'm re-soldering the press fit connectors for the data bus, just in case. The case had a mayor impact before all these issues started (long before I got it). Took me ages to get the frame kind of straight. It was hit on the corner besides the connector, and some connections weren't looking very nice, it looked like the solder around these pins was dented and cracked on some places. They are a pain to solder. Already done the PCB interfacing the PPC and acquisition board, just have to do the PPC and acquisition board.

At least I got the main board caps fixed and the NVRAM will be replaced soon.  Switched from Win98 to XP pro, new CPU and HDD, the computer runs fine now. Voltages look OK to, just found some distortion on the +/- 15V supplies, probably due to oscillation in one of the opamp (wich runs at a comfortable 110C).

[klaus11]

The Tek TDS5052B (2004) does not lead NVRAM uses CMOS Flash PLCC32 welded directly to the board.
The Boot is in a CMOS eprom and options trading in a CmosFlash.
AMD uses 32pin memories, pa 3M optio would need to replace the U1320 16MB ?? but never fabricate memories PLLC32pin AMD 16MB?

U2.....unlocated on board???....... IC, Memory; CMOS EEPROM; 2K, 256X8, 2.5V.......Stores touch screen calibration alignment data

U720...PCI board.........Xilinx XC18V01...... IC, Memory, CMOS PROM 1 M.....Program PAI FPGA at boot up

U1320...PCI board...... AMD AMLV010b.......IC, Memory, CMOS, FLASH,
128 X 8 bit.......Option Key, Event Log, IO Processor Code, Fan Cal Constants

U1800...Acquisition board.....?......IC, Memory, CMOS EEPROM, 2 k X 8...Calibration Constants, Probe Data

U400.... PCI board....AMD AM29LVo10B....IC, Memory, CMOS, FLASH 128 X 8 bit..... ??

U1501.... back PCI board....AMD AM29F010B....am29F010B
1 Megabit (128 K x 8-bit)CMOS Flash.....

[robert_]

Just in case anyone needs drivers for TDS7000/TDS7000B/CSA7000/CSA7000B series scopes:

Display driver from TDS7000B/CSA7000B
USB Front Panel driver for TDS7000/7000B/CSA7000/7000B
PowerPC board driver for TDS7000/7000B/CSA7000/7000B

Sadly all of these links are now dead.
Anyone still got the drivers?

[TiN]

Fixed, server had upgraded storage array, forgot symlink to /test URI.
All these links back now.

[takamichih]

Hi,
I recently picked up a TDS 7054 without the HDD, and this had almost the same issues as the one posted here earlier(the battery on PPC board was dead). Thanks to the post, I was able to replace the battery and set boot parameters through the serial port.
But when the PPC board tries to boot, it hangs when the message from serial port shows "Backplane anchor at 0x1e80000... Attaching network interface sm0... done" and the 7seg shows "L".

I installed Windows 2000 Professional, TekScope, and tried to install the drivers posted here and successfully installed two of them but when the PowerPC board driver was installed the OS displayed BSoD on dblib.sys and it is disabled now. Does this mean that the PPC board is broken?

I would like to hear any ideas..., and please tell me if I did something wrong because this is my first post.
Thanks in advance.

[TiN]

Downloaded current BIOS From CSA7404:

Tektronix CSA7404 Intel SF810 motherboard BIOS dump - CRC32 Checksum: C357A520

[andy2000]

I think I need to focus on the PPC chip itself and see if I can verify the basics such as power supply, clock, and reset based on the datasheet.

I would follow that strategy also... Check whether the basic conditions for the CPU to run are there.

Here's an update on my dead PPC board for my TDS7404B.  I'm powering the board with a standard PC ATX power supply connected to the socket that normally provides power to the PC motherboard.  This allows much better access to the PPC board for troubleshooting, and provides all the needed voltages.

The problem seems to be with the reset circuit.  Almost all of the reset pins that I can identify are low.  This includes the PPC CPU, memory controller, Intel PCI bridge, Intel LAN chip, and NI GBIB chip.  Also, the OE# and CE# pins for the PPC BIOS are high.  I've verified clock and power to all the major parts.  Apart from that, there is absolutely no activity anywhere on the PPC board. 

I'm gradually verifying everything I can trace, but it's not easy with all the hidden traces, and some difficult to identify parts around what I think is the reset circuit.  I've tried manually changing the reset lines I have identified to the high state, but so far there has been no sign of life.  I don't think anything is dragging the rest line down because each chip seems to have its own independent reset signal.  I could probably find the problem in 10 minutes if I had a schematic...

Does anyone know anything about the reset circuit?  I'm assuming it's generated on the PPC board, but someone correct me if it might come from the acquisition board, or some other location.  Does Tektronix have any favorite reset generator ICs that I should look for?

[TiN]

Have you looked into powergood from onboard DC-DCs? Usually on PC hardware like motherboards and videocards powergood from onboard bucks driving resets to ensure correct ramped voltages and proper power sequence. Might be same case here.

[andy2000]

Have you looked into powergood from onboard DC-DCs? Usually on PC hardware like motherboards and videocards powergood from onboard bucks driving resets to ensure correct ramped voltages and proper power sequence. Might be same case here.

That's a good point which I hadn't considered.  The ATX PC board must be getting a valid powergood signal since it boots, but I don't know about the PPC.  It's a new starting point that might lead somewhere.

[Daxxin]

there so many nlx bords models with different chipsets , the firmware of x86 board is standard or custom? tried to backup it?