IFR 1600S Processor Bd Repair

[jms6612]

I recently repaired the processor boards on two IFR 1600S service monitors. These units were purchased at an auction about 15 years ago. They work very well, especially for ham radio applications, however the processor boards needed some repairs and upgrading. The processor sockets (100 pin PQFP) needed to be replaced due to heat-related warping, so I replaced them with AMP 822064-4 sockets and 821939-1 covers. Unsoldering the old sockets was the hard part. Installing the new sockets and IC's was fairly easy. I also added some heat sinks to the processors to provide some additional cooling. Space for the heat sinks was not a problem since there was a vacant extra memory board socket next to the top side of the processor board. I replaced the processors with new 80376's while I was in there. I also replaced the real time clocks (Dallas DS1287) since their internal lithium batteries had fully discharged and were no longer providing accurate time.

After reinstalling the processor boards, everything works well (passes all self-tests) except the service monitors have forgotten their serial number and option info (using *IDN? over RS232), and I am no longer able to see the user programs contained in flash memory. The cal data is intact (memory board is unchanged), and everything else seems to function normally. I used to run the Easy Sweep program to graph my ham radio antenna's SWR versus frequency, which I can't access any more. It should still reside on the memory board, but I can't get to it. I can run some Aeroflex evaluation software on a PC to essentially do the same thing using external TMAC commands, however I was wondering if anyone knows of an easy way to restore the original functionality of the IFR 1600S units, without sending the boards in? I know ... most people would probably have ordered a new processor board, but I wanted to see if I could meet the challenge of repairing one. So far, I was 99% successful.

If this helps, I get the following info for the CPU boards using the configuration screen:

Unit #1): Main CPU: 3.02 3/21/1994
Unit #2): Main CPU: 4.00 9/12/1995

Any help would be greatly appreciated!

I have enjoyed using these service monitors greatly, especially with the color displays. Besides the issues with the processor boards, they have not had any problems at all. They are wonderful to have on the test bench, especially for servicing ham radio equipment and checking antenna SWR.

Thanks, John

[kcbooboo]

This is not necessarily what's happening, but some of the Dallas parts have non-volatile RAM in them which manufacturers love to use to hold things like serial numbers and other "gotcha" stuff. Perhaps when you replaced that, you lost some very important information that only the factory can put back in. I'm not saying that IFR did this, but I've seen it in other company's products. Perhaps the factory can give you a procedure for updating that info.

Did the CPUs happen to have any EEPROM within them? That would be another place where they'd store stuff like that. It might be worth putting the old parts back in to see if the missing functionality is restored.

Bob M.

[jms6612]

The datasheet indicates that the Dallas RTC has nonvolatile RAM in it, which would likely be erased when the internal lithium battery dies. If the designers placed the SN and options in there, then it would be erased as well. Actually, I first noticed the problem right after replacing the RTC, so that's probably what caused it. I was hoping there might be a 'secret procedure' for restoring the info. At least I can hope ...

[Wowbagger]

Yes, the Dallas part stores that information. It's not a big deal to reset the serial # and options, EXCEPT: the information on how to do that is company proprietary (read: no, I will NOT tell you how to do it). The best thing if you can afford it would be to send it back to a cal center and have them set it back up. Worst case: how close to Wichita are you?

[jms6612]

That's what I was afraid of ... Thanks for the honest answer. I think I will use the 1600S 'as is', since I can still use external Aeroflex PC software to run the SWR program. The other programs that were in flash memory weren't of much real value to me anyway, since they were for the CSA option. I've disconnected the CSA option hardware since I'm not interested in cell phone testing. The 1600S is still a very useful service for what I need, even without the user programs.

[kcbooboo]

I've had Dallas chips in older PCs where they'd lose some of the CMOS settings - but not all of them - at power up. I'd go into the SETUP screens and reset the date/clock but many of the other settings seemed to still be OK. I don't think the NVRAM relies on the battery for non-volatility. If you can put it back in and the S/N comes back, then you'd know for sure.

Bob M.

[jms6612]

I reviewed the datasheet for the DS1287, which describes 50 bytes of nonvolatile static RAM. It also describes the real time clock as being nonvolatile for at least 10 years, which implies nonvolatility as long as the internal lithium battery is adequate. But as you say, the real test will be to put the original RTC IC back in. I put in an IC socket in anticipation of future replacement, so the task will not be difficult. I'll let you know the results. One other thing that I should mention is that the user programs were still available when the time clock initially became corrupted. That could be explained by a partially discharged battery, with the voltage threshold for losing RAM contents being lower than the threshold for the running clock. I wish the chip had an external pin for supplying backup voltage, but unfortunately it does not appear to have one. I suppose I could have added a couple of diodes in an OR configuration to the Vcc pin so that it would receive power from either the instrument or an external battery, but it might be too late for that.

[Wowbagger]

That's the failure mode for those devices: the battery gets low enough the clock won't run, but the memory will be held up.

As for being able to externally supply battery: that was rather the idea of an integrated solution - to avoid needing that. Later devices had a removable battery module on top. Now-a-days the RTC/NVRAM function is usually built into the southbridge chip or a super-IO device.

[jms6612]

Update: I reinstalled the original DS1287 (the only one that survived the removal process) in the IFR 1600S and the user programs are still unavailable. The configuration screen indicates options: 0 whereas before it listed the options correctly. Using the *IDN? query over RS232 produces a SN of -1. I'm thinking that perhaps the soldering iron heat reduced the low battery voltage to below the minimum threshold for the RAM. The manufacturer indicates on the datasheet that when power is switched off to the device, the backup battery supplies both the real time clock and the RAM, indicating that the RAM is (seemingly) only nonvolatile while the battery is good. It's just a matter of time before they all lose their options memory, unless the power is left on continuously. Perhaps some day after the instrument support life is terminated, someone can share the information about resetting it? Until then, I can live with it ... Thanks for the helpful suggestions.

-John

[Wowbagger]

The problem is the AN-GRM4317 - the military version of the 1600. That is still being supported (and likely will continue to be supported for slightly less than the heat death of the universe, the way the Marines et. al. operate). So as long as that is true, then you won't see any information like that get out.

[sreaves]

Just for the record if anyone needs or wants to replace one of the Dallas DS1287's or any other of this series I do have a way to read the data (assuming that it is good) and write it to a new DS1287 but the part does have to be removed from the target board and must have all of the pins intact. I have done this for other instrument brands that use these dreaded parts. But I would like to know if the DS1287 is the only Dallas part of this type that is used in the 1600/1900. I could also provide as a service to the group the removal and replacement on the CPU PCB but obviously your machine will be down the whole time the board or part is out of your machine. I would not suggest replacing this part with an UN-programed one as that may what have led to the loss of the options but this is just a guess on my part.

Once I get the CRT/Display issue fixed on mine refreshing this part is the first thing on the list.

Sam
W3OHM

[jry]

Is there anyone still selling the Intel 80376 processor in the QFP package ? either 16 or 20Mhz would work

Think I need one or two to fix some dead processor boards that had the usual socket failure and re-seating the chip did not help.

Not sure IFR ever broke out a separate part number for the individual parts on the later version products

[jry]

Is there anyone still selling the Intel 80376 processor in the QFP package ? either 16 or 20Mhz would work

Think I need one or two to fix some dead processor boards that had the usual socket failure and re-seating the chip did not help.

Not sure IFR ever broke out a separate part number for the individual parts on the later version products

[Alexspiter]

Hello.

It will be possible the useful, here article how to make RTC Dallas DS1287 external power supply.
http://mcamafia.de/mcapage0/dsrework.htm

Yours faithfully, Alexander.
Callsign – UA1ACV
Russia, St. Petersburg

[jry]

The DS12887 is still available at very decent prices and is compatible with the DS1287 ...just has more memory ,
I have used these on the 1600S processor boards ...used a socket as well and a dab of RTV to make sure the RTC does not pop out.

[jmsb6612]

I'm the one who started this thread. My username has changed due to login difficulties. The 80376's in QFP form are getting harder to find. I obtained two of these several years ago since the original processors were heat damaged in both of my IFR 1600's. You could see obvious discoloration on the center of the external package, so I replaced both IC's, along with the sockets. I don't know where you can get them nowadays. I would start by internet searching for the part number, and check eBay occasionally. Also, it might be helpful to check with someone who works at Aeroflex, since they have to support the AN-GRM4317 military version until the end of the universe, according to Wowbagger. I wonder if Aeroflex might have already replaced the 80376's with a newer processor? Compatibility with older boards might be an issue though, if this is the case.

Regarding the Dallas DS1287 external battery modification, I did that to one RTC that had an expired battery, so I'll put that in when the current one dies. Unfortunately I have lost the options configuration since its internal RAM contents have been erased, so I can no longer access the user programs. No big deal though, I just use the Aeroflex evaluation software and external TMAC commands to run the SWR program, which is the only one I'm interested in for ham radio.

[Wowbagger]

I know we had done some work in the GRM to support the 386sx processor, but I don't know if that work made it back to the commercial version. Also, trading a 376 for the 386sx is like trading a red unicorn for a white unicorn - more common relative the original, but still damn rare. I know we had to do a last time buy to support the GRM.

As for replacing the whole CPU assembly for something newer - nope, that was never done (and I'd suggested it as a project several times).

[ablitz]

Does anyone know how to restore the options on a 1900 after replacing the DS1287 with DS12887? Any way to copy the old data to a new chip?
I have 2 1900s I would like to save