Hacker RP25A - Missing MW/LW reception

[Bristol603]

I have a new acquired RP25A and I have a MW/LW problem with which I could use some help.

I have done the normal switch cleaning and audio amp adjustments, which has got the set singing sweetly on FM - it was very scratching and the amp noisy to start with. However, the AM side is giving me troubles.

The is no audio signal being output on either AM band. I have tapped the signal at the volume pot and I can see the audio signal there when playing FM, but not when AM is selected. I have checked the voltage across the big electrolytic C21 next to the volume pot and it is showing 8.6V when MW/LW is selected and zero when FM is selected. I have also probed that board with my simple oscilloscope looking for what I believe should be a 470kHz oscillator signal. I can find an oscillation at what I measure as 37kHz which is present when AM is selects, but not when FM is selected.

Are there any common problems that I should check for that effect AM on an RP25A.

Thanks you.

[AC/HL]

Hello and welcome.
The Manufacturer's manual is here: http://www.service-data.com/section....overeign-mk-ii. The radio has completely separate RF/IF stages for AM and FM, and the AM one uses AF117 transistors. These are likely to be the source of the problem, search for "tin whisker" in the box at the top of the page.

[paulsherwin]

Yes, it sounds like a classic AF117 fault.

[Paul_RK]

It does, except there aren't any AF117s - Hacker switched to silicon in the course of RP25 production, and this set's an RP25A.

Paul

[AC/HL]

Ah, same manual but the addendum (serial numbers 13,000 onwards)
In which case it may be lockfit syndrome, although both amplifiers do use them.

[paulsherwin]

Agreed, it could be bad Lockfits, though the RF ones are less likely to fail in my experience. You can replace them with 2N3904s or other jellybean transistors, but beware of the leadout, which from memory is C-E-B for the BF194/5. The MPSH10 has the same leadout if you have some to hand.

[GeoffK]

Originally Posted by AC/HL

Hello and welcome.
The Manufacturer's manual is here: http://www.service-data.com/section....overeign-mk-ii. The radio has completely separate RF/IF stages for AM and FM, and the AM one uses AF117 transistors. These are likely to be the source of the problem, search for "tin whisker" in the box at the top of the page.

Does this set have totally separate IF stages for FM and AM, the usual arrangement is the FM and AM IF transformers are in series with the collector of the relevant IF transistor, as FM is working it would tend to rule out the IF transistors and IF transformer coil connections, unless they are totally separate circuits. It could be that the AM oscillator is not running though this transistor is usually the first FM IF transistor as well. Worth testing the AM detector diode.

[Paul_RK]

Hi Geoff,

Completely separate circuits up to the audio amplifier board, like other Sovereigns except the last (Sovereign IV). All else aside, the arrangement simplified Hacker's offering a full range of its larger portables: the Heralds left out the FM tuner and IF board, the FM-only VHF Herald omitted the AM sections.

Paul

[Bristol603]

Thank you Gentlemen - problem solved.

I replaced T2 (BF194) with a 2N3904 and the AM section came to life. I carried on and replaced T3 and T4 with 2N3904 as a preventative measure.

Considering it is the middle of the day, the AM reception is above my expectations.

[Station X]

Well done!

Fixed in only nine posts!

[paulsherwin]

Lockfit failures are becoming as common as AF11x tin whiskers nowadays. They have really become change on sight items. Fortunately substitutes are plentiful and cheap.

[Bristol603]

Gentlemen,

With your help this RP25A is now working on all bands and isn't showing any obvious faults. My daughter is away at university and a vintage style fan. This radio is for her to use at university.

With this use in mind, what are your views of the preventative swapping out of potentially unreliable transistors or othe components? Once the RP25A is away with her at university, it will not be possible to readily fix a future transistor (or other) fault.

On the one hand I think that if the original transistors, and other components, have worked happily for nearly 50 years, then I should leave well alone. This is also good not for "messing" with the radio any more than absolutely necessary. On the other hand, as an "ambassador" for the worth of well engineered old kit with the younger generation, I would be sad if it was less than reliable.

What are your views?

Nigel.

[julie_m]

Lockfit transistors die of old age. The seal breaks around the leads, moisture gets inside the package and damage results. The other transistors in your set are all the same age as the dead one, and have been exposed to the same environment .....

If I was going to do a transistor swap for the sake of reliability, I would take advantage while each transistor was unsoldered to measure its neighbouring resistors to be sure they were still within tolerance, and be merciless with any suspect capacitors. Of course, I would power up the radio after replacing each component, so as to be sure I had not introduced any new faults by miswiring anything. (Hands up who hasn't been there .....)

BC547 / 557 ought to be a straight substitute for BC14x / 15x, and likewise BF5xx for BF1xx ("pentode"-looking numbers, with an F or L, indicate RF transistors). If you have some 2N3904 / 6, ZTX300 / 500 or similar in your parts drawer, they probably also can be used. Having the base in the centre will make for a neater job; if you do need to cross over the leads, slide a length of sleeving over one of them. The worst that can happen is that the replacement transistor won't have enough gain at the frequency it operates at (up to about 2 MHz. for the oscillator / mixer in a MW set, and a quarter of that downstream of there), in which case the radio will just not work. Even if the transistor overheats suicidally, it needed replacing (with a higher fT) anyway .....

But to be honest, if the set was working, I probably would wait until the next transistor packed up before taking a soldering iron to it again .....

[paulsherwin]

It's a difficult decision given the circumstances you describe, and to some extent it's a matter of personal preference. It's unlikely that another Lockfit will fail within the next 12 months or even 3 years, but certainly not impossible. They are easy to change in these sets - you can replace all the silicon with 2N3904/6s or BC548/558s, though ideally you would match the originals more closely (so MPSH10s for BF194/5s, BC549s for BC149s, etc).

If this was my radio I would probably change all the Lockfits, but others will feel differently.

Remember that you will need to adjust the quiescent current of the output pair if you change the driver transistors.

[Bristol603]

Thanks Julie. On reflection, I have a feeling that I could well cause more trouble and introduce more unreliability by disturbing it. Perhaps I should just give it a good hard soak test. Nigel.

[mhennessy]

Personally, I've never tried substituting BF194/5s in the FM sections of one of these, so I'd be cautious about doing so until I knew what the potential issues were. In the IF amplifier, 10.7MHz is obviously quite a bit higher than 470kHz, but you might get away with using "jellybean" types suggested above (but until I know differently, I'd expect to have to realign the IF stages afterwards). In the FM front end, it's rather more tricky. For that, you'd want to use dedicated RF transistors - not least because access is more tricky, so I'd want to put something in there that I know will work! The equivalents at the time - in the more reliable TO92 package - were BF394/5, but those are only available NOS today. More thoughts about this can be found here.

In my experience - which must be hundreds of these by now - complete failures are much less common. It does happen - as was the case here - and funnily enough, each time it happened in a Sovereign II, it was T2. But perhaps only twice or 3 times at most? If we get many more, I'll give serious thought to investigating why that might be, but that's another day...

Usually, they succumb to a peculiar partial failure mode where they get noisy. In audio amps, the effect is obvious and has been discussed many times before. In RF/IF applications, the symptom is low gain; probably because of increased noise, but I've yet to prove that conclusively (I've got plans to build a simple jig to test them). The only useful indicator I've found is that they are usually temperature sensitive when unhappy, so a can of freezer and a hair-drier is useful.

In short, a set with low FM sensitivity is my worst nightmare! The cause is often the Lockfits, but check DC conditions first - in the Hunter, it's sometimes caused by high carbon-comp resistors, and on several occasions, it's been caused by corrosion inside the AM IF cans (that was a puzzler at first, but was easy to understand). Thinking about it, while low FM sensitivity is common on the Hunter, I'm struggling to remember an occurrence with the RP25/37 that wasn't solved by realignment. So yes, on balance, I come down on the "if it ain't broke" side...

[Bristol603]

Thank you for your thoughts. I found Mark's transistor article very informative. Nigel

[paulsherwin]

Originally Posted by mhennessy

Personally, I've never tried substituting BF194/5s in the FM sections of one of these, so I'd be cautious about doing so until I knew what the potential issues were. In the IF amplifier, 10.7MHz is obviously quite a bit higher than 470kHz, but you might get away with using "jellybean" types suggested above (but until I know differently, I'd expect to have to realign the IF stages afterwards). In the FM front end, it's rather more tricky. For that, you'd want to use dedicated RF transistors - not least because access is more tricky, so I'd want to put something in there that I know will work!

I have replaced the BF194/5s in a Roberts R606 with jellybean types without problems. I didn't realign the IFTs, and performance seemed identical to an unmodified R606. The Ft of these transistors is well over 200MHz.

However, this was as an exercise, and I would normally use VHF transistors simply because it seems good practice to do so. As I mentioned earlier, I've found MPSH10s work well, and they're very cheap from the usual sources (though not as cheap as the jellybeans).

[mhennessy]

I'll be honest: I've not come across the MPSH10 before. It looks promising. But it doesn't seem to be especially available from the big guys - discontinued at RS, not listed at Farnell (though US arm Newark have it). Mouser have stocks. I've not looked at the smaller suppliers and places like eBay yet.

Sadly, as David has pointed out elsewhere with his dual-gate MOS-FETs, it's a bad time for discrete components - especially in non-SMT packages. Jellybean stuff will be around for a while yet, but stock up if you need anything that's a little bit special. High voltage small signal audio transistors - the sort used at the front end of typical power amps - are giving me problems at the moment.

Funnily enough, the humble BC548 has a higher Ft than a BF194. But Ft is about current gain, which isn't the same as voltage gain. In practice, it's the lower Cob of the BF194 that makes it perform better at HF.

[paulsherwin]

The MPSH10 is a legacy transistor, but still churned out in large quantities in China. It's always available from eBay sellers for around 10p per unit.

I got to know it a decade or so ago when the RoHS rules were coming in and lots of people were selling off non compliant stock for peanuts. I bought a bag of 100 MPSH10s from somewhere and am still working my way through them.

I freely admit to having a hobbyist rather than a professional attitude to component sourcing