Hacker RP72MB Sovereign III - Low FM Sensitivity

[Bristol603]

My radio is actually a RP722MB (Golden Sovereign), but I think the circuits are the same as the more usual model.

The problem is low FM sensitivity. I have run through the FM circuit adjustments as set out in the service manual, i.e. Optimising the "S" curve at 90MHz and 100MHz and this hasn't helped (at least it didn't make it worse!). The service manual states that a 5 microV aerial input (90MHz carrier, 30kHz deviation) should give 50milliV at the tape output. I measure that I need 500 microV to 1 milliV at the aerial to get this 50 milliV output, i.e. the sensitivity is down by a factor of 100-200. My signal generator hasn't been recently calibrated and so the absolute number may be out, but I think it shows there is a problem.

This is also confirmed by normal use in that the receiver only picks up the strongest local FM BBC stations. It also does something I don't understand and that seems to contradict the above in that it receives better in the centre of the house rather than at the windows (where I would expect the signal to be stronger).

The sensitivity on LW and MW is excellent and the audio circuits are all working nicely. Does anyone have any suggestions as to what might be causing this low FM sensitivity?

Many thanks,

Nigel.

[mhennessy]

Have a search for "Lockfit problems".

The MPSH10 appears to be a good substitute.

Not saying that's 100% the problem, but it's a good place to start.

[Argus25]

Originally Posted by Bristol603

My radio is actually a RP722MB (Golden Sovereign), but I think the circuits are the same as the more usual model.

It also does something I don't understand and that seems to contradict the above in that it receives better in the centre of the house rather than at the windows (where I would expect the signal to be stronger).

Hi Nigel,

That is a very interesting and astute observation and it may well be a clue.

I have the Sovereign II , which is probably similar (I'm not sure). However I have the schematic for that in front of me.

At least you know the FM local oscillator in the radio is fine. With the gain low, most likely the fault is in the IF stages (most of a radio's gain is in these stages) or perhaps the RF stage, which on my schematic is TR1 a BF195 in the tuner unit, like I say I don't know if that matches your radio, but the principles are the same.

The fact that the FM signal is not apparently entering the set well through the telescopic antenna, and it works better in the middle of the house, makes me think the RF stage is possibly not working well. This stage oddly can be the most vulnerable to failure, not so much with the Hackers, though it could happen, but in other radios too. If a person walks on a surface and picks up a large static charge, then touches the telescopic antenna, it can fry the input RF transistor. (This is a big problem in some sets like the Barlow Wadley XCR-30. But it could happen to a Hacker) but it might just be the propensity of the lockfit BF195 transistors to fail and nothing to do with electrostatic damage.

If you can get inside the tuner unit and attach a 1 meter long wire to the emitter of TR2 (the local osc transistor) or the collector of TR1 (the RF amplifier output) or the test point in that area, and if this resulted in much higher signal output it would probably confirm that the RF amplifier transistor is faulty.

If it is not that, then head to the IF amplifier and a good place to start is to check the IF transistor's emitter voltages, as that is an indication of their bias conditions and also if they are ok.

The fault could be in other locations but you need a place to start.

[paulsherwin]

More and more Lockfits seem to be failing now, so it's the first place to look if they are present in a misbehaving radio. I wouldn't advise buying NOS Lockfits as they may also be bad. A surprising number of general purpose types will work perfectly well as replacements, even jellybean 2N3904/BC548 types in the BF194/5 positions, though it's good practice to use proper VHF transistors where such types were originally fitted.

Watch out for the slightly odd BF194/5 leadout if making substitutions (though the MPSH10 is the same I think).

[Bristol603]

Thank you. I will try swapping out the lockfits. Nigel.

[paulsherwin]

Do them one at a time, checking nothing has got worse after each sub.

[julie_m]

Later Mullard semiconductors (after the first OA diodes and OC transistors) use an extension to the Mullard valve code. The first letter is A for germanium or B for silicon; then C for small signal AF transistor, D for power AF transistors, F for small-signal RF transistors, L for power RF transistors, A for signal diodes, Y for rectifier diodes and Z for special (usually Zener) diodes. Unfortunately, the polarity of transistors is not coded into the number. The same piece of silicon was usually offered in several different encapsulations including TO18, Lockfit and TO92 packages (sometimes more than one TO92, with the leadouts ordered differently).

The important parameters, when choosing a replacement transistor, are gain (hFe), collector current (IC), collector-base voltage (VCB) and turnover frequency (fT). The battery voltage and the collector load resistance (then, by making use of Ohm's Law, the maximum current) can be obtained from the wiring diagram for the set. Too high a gain or fT could result in instability; too low could result in loss of gain.

[Argus25]

Originally Posted by paulsherwin

More and more Lockfits seem to be failing now

That is very interesting, it would be good to determine the mechanism of that. I have so few items with the lockfits in them, I think my Hacker Sovereign is the only one in my collection with them.

Interestingly lockfits were also used by Lucas in some of their early alternator regulators, so presumably these would be prone to failure too. I have also seen them in some Philips test gear. Now I will look upon them with suspicion.

[paulsherwin]

Most British manufacturers used them extensively after switching to silicon at the end of the 60s. You find them in all sorts of 70s stuff including pro gear like mixing desks. There is no definitive explanation as to why they are failing, but the best guess seems to be that the hermetic seal breaks around the legs allowing moisture ingress over many years.

Like Rimlock valves before them, I don't think they were used outside Europe.

[mhennessy]

Originally Posted by Argus25

That is very interesting, it would be good to determine the mechanism of that. I have so few items with the lockfits in them, I think my Hacker Sovereign is the only one in my collection with them.

Interestingly lockfits were also used by Lucas in some of their early alternator regulators, so presumably these would be prone to failure too. I have also seen them in some Philips test gear. Now I will look upon them with suspicion.

Much has been written about them on this forum. If you haven't already seen it, I've attempted to bring it all together here: http://www.markhennessy.co.uk/articl...rs.htm#lockfit

Earlier I mentioned the MPSH10. This is a suggestion from Paul - I've ordered a batch, but haven't had a chance to try them myself (that's the only reason I've yet to add them to that article). However, they appeared to be OK (as in, didn't make it worse) in the Hunter discussed in this thread: http://www.vintage-radio.net/forum/s...d.php?t=138623

One key point to take away from my article is that the actual failure rate is likely to be much higher than we realise, simply because the partial failure mode doesn't always give problems in the particular circuit in question. In something like an alternator regulator, a bit of extra noise from the transistor is likely to be academic. In the RF or IF or early AF stages of a radio, that's different. And it's worth saying that they were even less reliable in TV sets, possibly because of higher temperatures, or possibly because of the more demanding job they did?

[paulsherwin]

There seem to be fewer cheap MPSH10s around than there were a year ago - maybe forum members have been stocking up They're still available at reasonable prices though.

[Argus25]

Originally Posted by paulsherwin

the best guess seems to be that the hermetic seal breaks around the legs allowing moisture ingress over many years.

I have never been enamored with epoxy/plastic cased transistors.

Looking at the design of vintage metal cased devices such as the 2N3055, or TO-5 such as the 2N3053, it seems so superior to have a metal case with glass like lead in seals and no plastic. Though with small resin filled metal cases like the typical BC107, they seldom give any trouble. But epoxy plastic case devices can end up with some quirks. In power devices they can de-laminate from the metal tabs.

Of course plastic cased devices are the new standard and have been for decades with IC's, though it is interesting that most mil spec IC's and wider temp range IC's are ceramic or resin filled metal cased. Metal & ceramic cased semis can survive higher temperatures than epoxy cased ones and are clearly more stable.

I was actually somewhat devastated to learn that the TO-3 case is obsolete, as it doesn't lend itself well to modern surface mount techniques, but I still believe it is a superior package to an epoxy case device.