Anyone repaired a Racal RA1792 receiver?

[RF Wanderer]

Hi,

I recently acquired a Racal RA1792 receiver. In fact we found it languishing in a small warehouse about to be thrown out. It had been kept in a dry clean environment and appeared in good physical condition having originally come from a coastguard station years ago.

I'm wondering if there is anyone here who has repaired one of these Rx's?
I will hold off on going into specific details until later. In short though it has almost pristine LCD display and I've gone through the tantalum capacitor replacement routine to bring all bus voltages back to normal. What we are now faced with is dealing with the all zero's/random digits resulting from the microprocessor board.

As I say more later if there is anyone here who has had experience of this?


Carly

[frsimen]

Hello Carly,

Welcome to the forum

The CPU boards can be troublesome. They often suffer from battery leakage and that can damage tracks/cause unwanted paths on the PCB. Check carefully in case your one has had that problem. Be very careful if you are probing around the CPU and its peripheral ICs, as I discovered, they don't respond well to pins being shorted together!

Check that any socketed ICs and the ribbon cables are properly seated.

It would be worth checking that the clock oscillator and reset circuit are working, as they are needed to get the CPU to start up properly. All zeroes is what you get in the absence of data on the data and address bus.

Does anything happen if you press on any of the many buttons?

In the end, I couldn't repair my CPU board, but the radio is now working again using a home designed STM32 based microcontroller board.

Paula

[RF Wanderer]

Thank you Paula for you reply!

Originally Posted by frsimen

Does anything happen if you press on any of the many buttons?

Er..No!

Let me give you a little more information of where I'm at.

So the Tantalum replacement phase involved a lot of power on and power off's. In all this time I only got "all zero's". Maybe a single letter in the right hand display where the S meter is like eg "00A"

I then moved on to the micro proc board A6A2. The NiCad battery was mostly OK with minor evidence of slight oozing around the seal. No leaking onto the PCB. Nether the less I replaced that with a new NiMH one.

Further inspection of the board revealed some slight tarnishing of a few tracks but very limited. I then tested every track I could and every VIA through the board for continuity. All tested OK.

Further detailed examination under magnifying revealed evidence of a white deposit around the major socketed IC's such as the EPROMs and uProcessor chip. Research informed me that the gases alone from the battery can cause corrosion and cause an alkaline residue to form (I found it also on the SCORE) module that connects to it)

All socketed chips were removed and reseated. I cleaned the board with a weak acetic acid solution and then thoroughly cleaned down with meths and isoprop alcohol.

At this point something interesting happened. Although mostly still all Zero's on power up very occasionally I essentially got a full normal display. Memory channels, filter selection, S meter and SSB mode etc all displayed. But still no button control. A bit like trying to communicate with a faulty unmanned space probe!!

That last result intrigued me. I reasoned that the problem was bad connections somewhere. Again IC's removed (under strict static protection) The pins were carefully and thoroughly cleaned to bright shiny metal. I cleaned the IC socket holes extensively, difficult though as not much will fit in them. I used a sewing pin, bent a bit, to try and scrape each and very hole and a old DIP IC to plug in and out many times. TBH I'm still not 100% satisfied that I did a good enough job.

Powered up yesterday and the result was this. Occasional all zero's but mostly random digits mostly in the 3 or 7 range. They change.

So my work has caused so change and I think the data is getting through partly. (the clock is working and the reset has been checked). I'd like to find a way to clean the sockets better though.

One final note. This model has a small extra modification board A6A2A1 fitted above the Xtal. Apparently this was produced to eliminate lock ups. It has very few components on it, one IC, two resistors and a 100nf capacitor. There was some corrosion so I resoldered this. In the process one of the resistors fell apart. So I replaced all four components on this tiny board. It interfaces via a ribbon cable to the processor and adjacent chip.

So thats where it stands!

[frsimen]

You’ve had quite a battle with this already. Your fault sounds a lot like the one that I had originally. There are quite a few different versions of the A6 boards, my one didn't have the additional board.

A question that comes to mind is that when the LCD has a full normal display, is it a sensible display or do you have odd combinations, like USB and FM both showing at the same time?

Is there any sign of life on IOC or the IOD buses? These are used to drive the displays and the various functions in the rest of the receiver. You may find that one of the lines is stuck high or low, which will give you a clue as to where to look for a fault.

Paula

[RF Wanderer]

Originally Posted by frsimen

A question that comes to mind is that when the LCD has a full normal display, is it a sensible display or do you have odd combinations, like USB and FM both showing at the same time?

No its not a normal display. There are numerous combinations such as filters and modes which would not normally be there together.

Carly

[frsimen]

Hello Carly.

That illegal combination of icons suggests that the latches which are used to store the data for the drives to the displays and so on are not clearing properly at start up. That is consistent with the microprocessor board not sending out any valid data. If you unplug the connector to the front panel board, you will find that on start up the display sometimes is a some random combination of icons and digits, other times it will start up only showing zeroes.

I don't know if a normally working CPU board will start up without the SCORE board plugged in, but it might be worth trying that in case that is doing something to cause trouble.

If you are happy that there is no damage to the tracks and you have checked the continuity where you found damage, you are heading towards a faulty IC or a poor connection. The CPU board has its own internal data bus and address bus, the external IOC and IOD buses are fed via bus transceiver ICs. Activity on the internal buses but nothing on IOC/IOD will point towards a problem with the bus transceivers or the decoding logic that addresses them. Looking for activity on the internal buses is the best next step.

Paula

[RF Wanderer]

Well that's given me a lot to think about Paula. First though I was advised over the weekend that i'ts all but impossible to clean inside the socket pins properly when hit by corrosion and that I really need to replace them. And TBH I'm not totally confident I did a good enough job. So I've started to remove the sockets. After these have been replaced I'll have a better idea where things stand. Also the short ribbon cable where it connects to the A6A2A1 mod board showed open circuit on one of the wires. Poking around and it fell out of the connector. So the corrosive fumes got in there. I'll replace that as well.

It'll be a few days before I know anything further.

Carly

[frsimen]

Hello Carly,

You're braver than I am. Whoever advised changing all the sockets can't have tried it themselves. Watch out for the through hole plating coming out with the pins when you remove the sockets. If that happens you will have even more problems to sort out.

Paula

[RF Wanderer]

Not much braver I think! I removed one of the 24 pin EPROM sockets and replaced it. That took some effort given that I had to be careful not to damage the PCB tracks. After that I reconsidered if it was really worth it. Instead I opted to do one more final clean of the sockets which took some time. I was then satisfied beyond doubt that any connection issues were resolved.

So on to the re-test. Same result unfortunately. All zeros and sometimes a string of random numbers (and occasion alpha characters in the numerical display). Obviously there's something much more sinister going on here. So its time to call it a day. A shame really but it is what it is. I'll probably dismantle the modules and sell it.

Hey one positive. I found this site that previously didn't know about and all the interesting reading to keep me busy. So that's good!

Carly

[Sparky67]

Presume the A9 board is behaving itself and there’s nothing obvious that end, and it’s definitely the uP which is faulty Carly? I have know instances of EPROM data becoming corrupt with time, although not in a 1792, so that’s a possibility I guess. I have worked on several 1792s, most of which have been resurrected by tant replacement, I have seen a couple with exactly the same symptoms as yours. If it hasn’t already I wonder if leaving it switched on for 24 hours is worth trying? Give the battery a full charge…?

And while I think about it, we had a processor in a system which stopped working correctly if the DC drifted by a very small amount. Maybe run the A6 off a separate supply, see if that makes any difference?

Martin
G4NCE

[GMB]

I presume it has a ROM or ROMs for program memory? These mostly have a timed life (~15 years quoted) so you would be very lucky if the data is not corrupt, and if the corruption derails the MPU then nothing useful will happen.

I am wary of trying to mess about with IC sockets. It is usually best to leave alone unless you can measure a problem. I seem to remember that the cheap kind had a very low maximum insertion life.

And never poke anything in them that isn't an IC. I witnessed a brand new CPU built using wire-wrap totally wrecked by systematic testing of the wiring. The fatal flaw was that they used sowing pins as probes. These were too big and the result was that all the IC sockets made bad connections thereafter so nothing worked. And this was with brand new sockets!

[RF Wanderer]

Originally Posted by Sparky67

Presume the A9 board is behaving itself and there’s nothing obvious that end, and it’s definitely the uP which is faulty Carly? Martin
G4NCE

Well I dont know now? It occurred to me that all my tests had been with the uP ribbon cable plugged in. So .. I unplugged it. Exactly the same result, all zeros and sometimes random digits. So that points to somewhere else like the A9 board perhaps?

Carly
M0BTS

[frsimen]

Hello Carly,

That is normal behaviour for the A9 board if it isn't connected to anything.

In earlier posts I suggested a number of tests which might help in the diagnosis of a problem on the A6 board. Have you tried these and did you see any activity on the internal buses and the IOC and IOD buses?

I have assumed that you have access to either an oscilloscope or a logic probe. If you don't have either, you could use a diode probe with a DMM instead to see if anything is happening.

Bearing in mind GMB's comments above, it would be prudent to check that the plug in ICs are still making contact in their sockets, being mindful of static precautions while doing so!

Paula

[RF Wanderer]

Originally Posted by frsimen

Hello Carly,

That is normal behaviour for the A9 board if it isn't connected to anything.

Yes Paula I get that in the case of just all zeros. But what about the roughly 50% of boot ups that have a string of random numbers/characters? As I said the behaviour is exactly the same as when the A6A2 board is connected. Thats whats got me wondering about the A9 front panel boards. Yes I've had a scope on them and IOC/IOD seem fine

Carly

[frsimen]

The digit displays are driven by 4056B latched 7 segment decoder drivers and the icons by 4054 4 bit latches. I don't think there is any guarantee as to what state they will be in when the power is applied, so you will get random numbers and icons appearing sometimes. Normally, when the CPU writes to the displays, everything is sorted out.

Your CPU card isn't completely dead, which is a good sign. Have you checked the switches on the front panel memory board, S1A, B and C. They all need to be in the open position for the buttons to work.

If that hasn't helped, you could go looking on the A9 board for any sign of the display strobes (OP32-OP37 on the circuit diagram). Be warned that the strobes will only be a pulse of a few microseconds and probably will only happen if there is a change of state. OP36, which drives the s-meter, is probably the best bet.

The other place to look for activity is the on the keyboard scan. The lines marked Switch Matrix column should show a regular pulse, but it will only be quite short and might be difficult to see on an analogue 'scope. If you have a digital 'scope, use it in single shot mode and that will capture the pulse easily if it is present.

Paula