Racal RA17L, missing tuning hiss on MHz dial

[billy.bates]

Hi, I am new to this forum & have recently bought the above set. Tuning through the MHz scale only produces a noise peak between 0 &1 MHz, I tried feeding the aerial input with a noise bridge & observing the 100kHz o/p while tuning the MHz dial, I could clearly see noise peaks across the whole range. I checked the 1MHz Crystal o/p with an accurate frequency counter & it was really spot on. My personal feeling is that there is most likely a problem in the 2nd mixer cct. I can see from searching other posts that resistor failure is a common problem & worth pursuing, I have checked valve voltages in the IF strip & they appear to be good. Any pointers suggestions would be most welcome.
God bless, Bill.

[G8BBZ]

Hi Bill,
As well as the info available on this forum, you might want to consider joining the "Racal RA17 Forum" on Yahoo Groups wherre the subject of noise peaks at the 1MHz srttings has been dealt with extensively. There are a number of possible causes. Try searching on Mc/s Tuning.
good luck
Peter G8BBZ

[dtvmcdonald]

Originally Posted by G8BBZ

As well as the info available on this forum, you might want to consider joining the "Racal RA17 Forum" on Yahoo Groups ...

I find no such group. Is it perhaps at a UK site not yahoo.com?

[ronbryan]

https://uk.groups.yahoo.com/neo/grou...RA17forum/info

Ron

[billy.bates]

Thanks for the information & pointers.

God bless, Bill G6ATO

[Skywave]

Do you have - or know of - the maintenance history of this set? RA-17s (and their variants) have been around a long time now and if you fix that fault, you'll probably find that it'll unearth another. In my experience, if a complete overhaul of a RA-17 hasn't been done within the last 20 years (and usually much more than that), a wholesale replacement of Rs and Cs, perhaps a valve or two - 6BE6s spring to mind - plus attention to the usual 'mechanical things' and a re-alignment is always worthwhile.

Al.

[billy.bates]

Thanks, I can see this isn't going to be a five minute fix ☺

[Radio Wrangler]

No, not 5 minutes.

It could be any number of things, but steady, progressive working through the receiver structure will get you there.

There's a lot of radio in an RA17, but it's just a large number of fairly basic circuits. Step one is to get it down to a block, then fix that block.

Once you've got it running again, then it's time to consider a preemptive strike against capacitors and resistors which are on the way out. This can turn a receiver that's rather disappointing into a joy to use.

You'll very likely need a reasonable signal generator. If you don't already have one, it's time to start looking.

David

[billy.bates]

Thank you for all your responses / pointers, I am blessed in working in test equipment & repair that I do have access to some top notch Sig gens & the likes ☺, knowing how to use it on valve equipment.... well that's going to be a whole new game ☺

[Skywave]

Signal generators and their connections to valve radios, especially the RA-17.
There are many points in this radio (and some are none too obvious) where you will need to inject a signal, but those points may have HT+ on them. So measure first and if necessary use a suitable blocking capacitor. The last thing you need is to damage your sig. gen.!
(And yes, I do know that some generators have RPP fitted - but why risk it?)

Al.

[billy.bates]

A very important point and noted, I have also come across blocking capacitors that have failed, letting HT voltages through! At this stage I think it's a case of measure twice & cut once.

[Skywave]

"Blocking capacitors that have failed": yes, I've had that. When doing a job - such as discussed here - and you suddenly discover that such a capacitor is needed, the temptation to grab one - say 0.1 µF, 1kv. d.c - from the bits box is quite strong. But that must be resisted! Get a brand new, unused cap. - e.g. as above and typically 630 v.d.c. - and use that. There is a case that can be made for having a couple of dedicated test leads with such a capacitor permanently connected to one end of that lead, or, failing that, such a capacitor kept to one side and reserved for that purpose.

Al.

[dtvmcdonald]

I would not even think about trying to get an RA-17 working without a very complete recap, resistor check, and visual inspection. I restored
one a couple of years ago and while I could not test all the caps
(because all the visible paper and electrolytic ones were simply not there),
ALL the hidden ones were in fact bad. Every last one! And lots of resistors were bad. And worse ... numerous mica caps were bad. Its very unusual in USA made gear to see a bad mica cap except ones that are inside IF transformers and have exposed silver. A large number of bright red ones in my Racal had
red flaking off, and many (but not all) of those were bad. Especially the ones inside the two big 37.5 and 40Mhz filters. Some of those had been replaced
earlier ... and the replacements were bad! Some that were not bright red were bad.

[vosperd]

It's worth making a thorough check with a dvm to see if the HT is getting everywhere it should and all the valves have the correct voltage before ripping the set apart and replacing components.
I have had problems with V3 in the first vfo. You cannot get to the pins without making up some sort of valve base extender. In the end I had to remove the first vfo as some components had either failed or gone well out of tolerance.
Possibly there is someone in your area with a valve tester if you don't have access to one yourself.
Don m5aky

[G8BBZ]

A couple of things to try before you startripping the set to bits.
The level of the 1MHz signal into the harmonic generator is quite critical. Look at the signal on the 1MHz output on the rear panel (scope or voltmeter) and adjust L2 (the coil in the anode of the 1MHz oscillator) for maximum output - should be 2v rms minimum.
You can connect an antenna to the 2nd IF input on the rear panel and tune stations between 2 - 3MHz. If this is OK then that proves all of the set after the 2nd Mixer.

[dtvmcdonald]

This thread reminds me of an oddity in my restored RA17C-12.

The tuning of the megahertz dial is strange. At normal line voltage
and most frequencies there are three peaks close together, not one.
At some frequencies there are only two. Tuning to the center one, or either if there are just two, and the radio works just fine. Its perfectly stable.

However, if I raise the line voltage so that the filament voltage rises above
6.55 volts the multiple peaks go away and there is just one broader one,
roughly the width its supposed to be. And if I turn down the line voltage, it remains this way, at least for 30 minutes or more. The voltages in the
37.5 MHz amplifier chain are correct. The big 37.5 MHz filter had to be realigned due to mica cap replacements. Its pretty good when sweeping by hand with the radio off.

So I simulated it in the 5Spice program. The behavior I see simulates well
as I vary the input and output loading in the simulator and the coil resistances and couplings. Note that I'm simulating by using transformers with extremely small secondaries as coupling elements.

Thus, I expected the effect to be caused by changing tube (valve) (they are of course "tubes" in a C12) transconductances. This does not simulate well.
Perhaps the filter is getting hot and changing characteristics? Note that there are no explicit load resistors!

Has anyone else's set behaved like this?

[billy.bates]

I got around to having a look at the RA17L this afternoon, most of the 100k resistors seem a bit high (114k), I took the resistor off V2 and replaced it but I think there must have been a dry joint because when I plugged it in it appears to work! I ran through the MHz dial against a Sig gen & it looks pretty reasonable, I couldn't pick up any broadcast stations with no suitable ariel & a substantially metal building but we seem to be making some positive progress, I'll need to check out the alignment but I am obviously really pleased & very grateful for everyone's help, I also realize I'm not done yet ☺

[Skywave]

It's well worthwhile removing the 2nd. VFO from the chassis, removing that VFO's screening and examining the components inside the box above the VFO's sub-chassis. You'll probably find many components are defective and possibly a badly burnt tag-board too. Usually a total replacement of all the Cs and Rs will be worthwhile and a replacement (ideally new) 6BE6 is often required (they go noisy with age). A subsequent adjustment of the variable Cs and Ls for the band-pass coupling is rarely necessary. kHz calibration: don't try to adjust the tuning core of the osc. coil (located in the base) unless absolutely necessary: it's usually locked anyway. (Adjusting the trimmer is O.K. - if required). However, if the calibration is so far out that the osc. coil might need adjustment, (which is very rare) contact me: it's a tricky procedure. That's why Racal say "Leave it alone!"

Al.

[billy.bates]

Once again, thanks for the wise advice ☺ I'll keep an eye out for the 6BE6 or suitable alternative (if there is one).