Faults on Racal RA117A receiver.

[Racalrob]

All.

I thought I would share my latest experience with you all on repairing a faulty RA117A receiver.

Recently I was asked to repair a partially re-furbished receiver from a local enthusiast, but he did warn me that it crackled a lot
and was a tad deaf.
The IF strip had been re-built with new components and the 37.5 & 40Mc/s BPF had been re-built with new silver mica caps.
On powering it up, it crackled loudly and only strong local MW stations were received.

I identified 3 faulty valves and replaced them with working ones. Fault number 1.
Still no signals.

Signal injection and tracing identified that the fault lay in the 37.5mc/s chain.
Injected signal into TP1 only resulted in a couple of mv at TP3.
The same result with an injected signal at V8 (g).
However, on injecting a signal at V10(g), got nearly 2v at TP3.
Obiviously the fault lay in the BPF boards.
Removed the C91-C64 board and replaced it with a known working board.
Injected a signal at TP1 and got nearly 2V at TP3.
The fault lay in the faulty board.
The faulty board workmanship was okay and no dry joints.
Removed the plastic covers on the trimmer caps, un-meshed the vanes and sprayed IPA on them.
Moved the trimmer caps several times to work in the IPA.
Let them dry off and re-installed the plastic covers.
Upon re-installing the board back into, injected a signal at TP1 and after re-trimming the caps, I got nearly 2V at TP3.
This time signals were being received over all bands, however there was now a fault with the AGC in that it was inoperative. Fault number 2.

After a few minutes I identifed that the AGC fault lay around V18 circuitry.
Instead of being 27V on Pin 1 of the valve there was 127V.
R120 and R121 were fine, however R122 was measured as being 68K instead of 6K8.
The resistor was brand new and had the correct colour banding for a 6K8.
Replaced the offending resistor with a correct 6K8 and this cured the AGC fault. Fault number 3

However, there was still a lack of sensitivity. this was traced to the LC filter and after examination a dry joint was discovered
on one of the wafer switches. I dismantled the entire LC filter and re-soldered all the connections and this cured fault number number 4.

The constant crackling still remained. I went away for 10 minutes and upon returning was greeted by silence.
The set was not working at all.
Checked all the fuses, the continuity between the mains plug and the set. That was okay.
There was no short between the primary and secondary of T1.
Upon proving there was continuity between the mains input tags and the mains switch, i identified that the live side was okay.
However, on checking the neutral side I discovered an open circuit.
It appeared that the black wire had NOT been soldered at the factory. It was only the black insulator which was melted onto the centre
tag of the mains switch. Looking at the scorching on the tag it appeared that only the end of the conductor was touching the centre tag
and vibration on the set caused the wire to move with the resultant of constant crackling. Resoldered the black wire onto the swithch and the
set jumped into life.

To summarise there was the following 5 faults:

1. Faulty valves.
2. Faulty 36.5Mc/s board.
3. Faulty AGC system.
4. LC filter fault.
5. Non soldered neutral wire on mains switch.

All the components which were replaced were brand new and NOT New Old Stock (NOS).

The set is now fully operational on all bands and It was returned to a happy customer.

Regards

Rob.

[G8BBZ]

Hi Rob,
Well done with that one.
Your fault #5 the not soldered wire on the mains switch is an echo of a fault I had also on an RA117A where the HT output wire from the rectifier tagboard was not soldered to the terminal but just touching end on. The wire was hidden under other wiring so not immediately obvious. The set worked correctly most of the time but on occasion would be dead at switch on. As soon as I moved the set to investigate the problem it would burst into life. Took me quite a while to find the problem.
Seems the production line inspection on the RA117A was not as good as it needed to be!

[G6Tanuki]

Nicely traced! Multiple simultaneous faults are always frustrating to deal with.

That non-soldered mains connection parallels my experience with an Eddystone 840A: sometimes it seemed to take longer to wake-up after switch-on than other times, and there were occasional crackles. At first I suspected the Brimistor in the heater-chain but it always checked-out just fine so I put this down to the radio 'just being quirky' - until one day there was a distinct smell of burning followed by silence.

One of the two fuseholders in the mains-line [remember this is an AC/DC radio so both sides are fused] had suffered a meltdown: autopsy suggested that the soldering of the wire to the end-contact had been - honestly - nonexistent - is a joint with *no* detectable solder the perfect example of a 'dry joint' ?) and continuity had only periodically been maintained because there was a black rubber sleeve fitted over it which concealed the original assembler's shoddy workmanship.

A new fuseholder - properly soldered/sleeved - and the problem has never recurred.

[trh01uk]

I think the pros would have labelled that one BER (beyond economic repair)!

Richard

[Colinaps]

I've seen a few non-soldered decoupling caps in 117s and 17s. On the plus side, they're easier to replace!

Colin.

[turretslug]

If only serial numbers indicated whether something was made on a Friday afternoon....