Hacker RP38A Hunter with poor FM sensitivity

[canonman75]

Hi folks

I mentioned this radio over on the Hacker Hunter Owners thread, and Mark H has quite rightly said I should start a new thread on it...

My latest RP38A Hunter is pretty close to being finished, so I thought I'd run it side-by-side with my other Hunter and see how they compare, bearing in mind the latest one has an anonymous but chunky 8 ohm speaker fitted instead of an Elac.

Anyway, to cut a long story short I've noticed that this Hunter isn't anything like as sensitive on FM as the other one. AM does seem as good so I'm not worried about that.

Having at a look at the service manual, I remembered the famous words of that chap from EEVBlog, "Thou must check voltages!" so that's what I did!

I ran the radio off a bench supply, and the voltage measured at the battery snaps was 18.03v. The readings I obtained are as follows, with the expected voltages (according to the manual) in brackets:

FM tuner HT supply:
8.43v on FM (should be 12.5v)
3.53v on AM (should be 4.4v)

IF supply, junction of R7 and R9
13.05v on FM (should be 10.6v)
6.90v on AM (should be 5.7v)

AM osc. supply, juntion of R5 and R1
0v on FM (should be 0v)
6.76v on AM (should be 5.8v)

As you can see, almost all of these readings are out, some quite drastically so! The manual also mentions the VHF RF stage supply on pin 1 of the VHF tuner, but I couldn't even find that to measure it

I've done all the obvious stuff already, including cleaning the switch bank and making sure there's no dry joints on the main PCB. I can't see anything amiss, nothing black and crispy or anything like that. I didn't have time to take the chassis out and check the soldering again but I will do tomorrow night, if I get the chance. I'll also check the individual switch gangs with a continuity tester as well, just in case any of them are faulty.

I've added a few photos of the switch bank and FM tuner connections, just in case anyone spots something I haven't spotted. To start with I thought the FM tuner was wired up wrongly, until I realised that tag D is just being used as an anchor point for the wire from the FM aerial and one lead of C41

That's all for now. I'll report back when (if?) I make any more discoveries!

Thanks in advance for any help and suggestions that may be forthcoming

Cheers,

Dave

[canonman75]

I've just been studying the circuit, and I'm going to stick my neck out and say there's an issue with any of R7, R8 or R9 being way out-of-spec or possibly C43 being leaky.

I think the section of SW3 that switches the 18v supply to either the FM tuner (HT1) or AM osc. (HT2) is working okay.

Will investigate further tomorrow

Dave

[paulsherwin]

Change the BF194/5s. Any HF transistor such as an MPSH10 can be used, or even a jellybean BC548/2N3904. Watch out for the unusual leadout.

Obviously there are other potential causes, but it's now looking as if all Lockfits are going to fail eventually, so you might as well change them now even if it doesn't clear the fault - good preventative maintenance, a bit like changing waxies and Mouldseals in valve radios.

[canonman75]

Originally Posted by paulsherwin

Change the BF194/5s. Any HF transistor such as an MPSH10 can be used, or even a jellybean BC548/2N3904. Watch out for the unusual leadout.

Thanks Paul.

The Lockfit problem had occured to me, but I thought I'd investigate why the voltages are all over the place first. I suppose failing Lockfits could be the cause of that!

[paulsherwin]

Absolutely - if the transistors are internally leaky then the voltages will be way out.

[canonman75]

If I use MPSH10 transistors as replacements, there's a possibility the radio will need re-aligning, isn't there?

I can get N.O.S. BF194's and BF195's but I'm not sure that isn't asking for trouble

[paulsherwin]

No, you don't need to realign unless the transistors are dramatically different. I've replaced the BF194/5s in a Roberts R606MB with BC239s and it's fine, and that isn't an ideal substitution.

Never use NOS Lockfits - it is indeed asking for trouble.

[canonman75]

Ta

MPSH10's seem to be as cheap as chips, so I'll give them a try!

[paulsherwin]

You can use anything you have to hand in the short term at least. These circuits seem to work perfectly well with generic general purpose Si transistors.

I think the MPSH10 was originally a Fairchild transistor, but it seems to have become the de facto standard cheap HF transistor in China and is made in vast quantities there. You can buy Chinese examples for a few pence each, but even genuine Fairchild examples are pretty cheap. It's a useful transistor for anyone who works on old transistor radios and tuners.

[mhennessy]

I can dig out my notes later, but for now, don't worry about the discrepancy between your voltage measurements and the manual - the manual isn't accurate, sadly.

Having said that, this set is sensitive to supply variations, so carbon-comp resistors drifting high can cause a lack of supply voltage which leads to a reduction in sensitivity. It's rarely just that, but it's often a factor.

Lockfits are often a problem, but often not. Later examples seem to be more reliable. I'm waiting for a chance to try the MPSH10 that Paul recommends, but I'm hopeful these will be a good replacement. If a Lockfit is playing up, it'll often be sensitive to temperature, so a can of freezer is useful here.

Just to add to the "fun", when the Lockfits begin to fail, they remain good at DC, but the noise shoots up, which causes the apparent lack of sensitivity. So they don't generally upset the DC conditions in the way normal transistors failing might. Eventually they might reach that point, but in the small-signal stages, the noise will have stopped the set working properly long before that. Later on - perhaps towards the output-end of an audio amp - the noise can go unnoticed until the degradation has affected the DC behaviour. So because the failure might not cause obvious problems, depending on where in the set the transistor is, it's likely the true failure rate is higher than we might think.

With these sets, I usually start with alignment. I know that goes against the usual advice, but normally they can be improved, and often it's enough.

Next, to narrow it down, I inject 10.7MHz to see how sensitive the IF strip is.

On more than one occasion, the problem has been inside an AM IFT. Yes, I know that sounds illogical, especially as the set worked perfectly well on AM, but there are capacitors inside the cans in parallel with the windings, and if one of those fails O/C, they no longer present the required low impedance at 10.7MHz. The cause is the leads corroding, seemingly a reaction with the glue on the tape used to secure the windings.

As I say, I'll look at my notes later. I've had dozens of these through here, and low FM gain is a common and annoying problem with them. Occasionally I have to admit defeat with them, sadly. Funnily enough, the other Hackers are much, much more reliable in this regard.

[paulsherwin]

Mark has far more experience with Hackers than I have, so always take his advice over mine if there's any conflict

[mhennessy]

Just found my notes.

Originally Posted by canonman75

FM tuner HT supply:
8.43v on FM (should be 12.5v)
3.53v on AM (should be 4.4v)

I've got 8.3V as a typical value. The AM value is irrelevant, so I've not recorded it.

Originally Posted by canonman75

IF supply, junction of R7 and R9
13.05v on FM (should be 10.6v)
6.90v on AM (should be 5.7v)

I've got 13V typically (7V on AM)

So, nothing to worry about. Frankly, the higher the better really.

As well as R7 and other rail resistors going high, I've also had R14 go high.

If you're able to check it, the typical IF sensitivity is about 200uV for 50mV audio out (30kHz deviation)

[canonman75]

Thank you both

Nice to know that my voltage readings match yours, Mark, although it does mean that's not the cause! I was going to do the same checks on my minty Hunter but it's probably not worth the effort now!

I had an idea this wasn't going to be easy to sort out! The whole thing's been a bit of a pig so far, what with the battery leak damage, missing hardware and wrong speaker, but at the end of the day it's a Hacker so it's definitely worth saving.

At this rate, I might have to get the Hameg out. It's still in it's box from moving house, so that gives you some idea how often I've needed to use it! Generally speaking, I mess about with microcontrollers and the like, so the 'scope doesn't get much use. Going back to dealing with analogue stuff is a welcome distraction from that. I don't have an RF signal generator here, although I might be able to borrow one from work. Sounds like I might need it!

[paulsherwin]

I really would change those transistors, because it's cheap and easy to do. It's difficult to diagnose Lockfit problems just by circuit measurements. If you change them then there's one less variable to consider when fault finding.

[canonman75]

I'm tempted!

The two in the FM tuner look like they'd be fiddly to get to, though. I'd have to desolder the PCB from the tuning cap.

I wouldn't normally go for shotgunning transistors like that, but then again it fixed my Autocrat

[mhennessy]

In fairness, while the faults tend to be on the IF panel, I have occasionally found faulty transistors in the FM front end (but no other faults there that I can immediately recall). There's only 2, and it's not that hard to free up the PCB, so I'd be tempted to try it. I'm only sorry that I haven't yet tested the MPSH10 in there yet. I picked up a batch of 50 after Paul first mentioned them to me, and I'm confident they will be fine, but as I say, I haven't had cause to try them just yet. If I get a chance at the weekend, I've got a set that I could try them in.

If you need to realign the FM tuner, it's pretty easy, and can be done without specialist tools. The oscillator adjustments on Hackers are notorious for drifting, so at the very least you'll probably have to attend to that anyway...

[Hybrid tellies]

Withy my Hacker Hunter RP38A somebody had a go at re aligning on both the FM and AM RF stages presumably to extend the upper end of the FM band and made a right mess of it. The result was calibration on all bands was hopelessly out and the sensitivity was also poor. Thankfully the IF's had been left alone.
Just armed with the service manual I had the RF alignment back on all three bands in a few hours and with that the sensitivity was very much improved. On FM I can pick up all the local stations without extending the telescopic rod aerial and it even tries to pick up the much weaker national stations with the rod aerial un extended.
I certainly think it might be worth looking at the FM RF alignment first before changing the transistors out. My RP38A still has the original lockfit transistors fitted . As mhennessy says you don't need any specialised equipment or tools to do it.

[canonman75]

Thanks Simon

I've got a bit of time off from work next week, so I can have a proper look at it then.

All I've done since my last post is blanket resolder the IF board, which had a fair few dry joints, and order some MPSH10's just in case. Unfortunately, the resoldering didn't do the trick

EDIT: I thought an RF signal generator was required for doing alignments? I haven't got one, yet...

[mhennessy]

An RF generator certainly makes it easier, but if you know the frequency of the stations you're tuning to, then that's usually good enough.

Just to expand on this, the aim of RF alignment is twofold:

1. To optimise the accuracy of the pointer relative to the scale
2. To optimise the sensitivity

On pretty much all Hackers that I've seen, point 1 usually needs attending to. Start by getting the "pointer datum" right - the pointer should align with the right-hand end of the window in the scale. Sometimes, that's all you need to do with that, but often you do need to tweak the oscillator adjustments. The core is for the LF end, and the trimmer is for HF. The two interact, so expect to repeat for a good half-hour!

Regarding the second aim, I find this drifts much less, and improvements are normally minimal in practice - though occasionally dramatic (in which case, I assume that someone has "fiddled"). Again, there's cores for LF and trimmers for HF, and again there is lots of interaction so an iterative approach is needed. For the effects of the adjustments to be obvious, you need to tune to a weak station, and/or do your best to attenuate the signal - if not, the limiting action of the IF strip will disguise the action of the adjustments. This is where an RF generator is useful, but again, it is possible with off-air broadcasts.

I hope that's useful - give it a go, as you won't make it worse! Probably

Good luck,

Mark

[canonman75]

Originally Posted by mhennessy

On pretty much all Hackers that I've seen, point 1 usually needs attending to. Start by getting the "pointer datum" right - the pointer should align with the right-hand end of the window in the scale. Sometimes, that's all you need to do with that, but often you do need to tweak the oscillator adjustments. The core is for the LF end, and the trimmer is for HF. The two interact, so expect to repeat for a good half-hour!

So I'd need to wind the turning dial as far as it'll go clockwise and make sure the pointer is just seen on the right? If it's not then I need to adjust it's position on the dial cord slightly. Does that sound right?

Sorry for the daft questions but I've not had to do an alignment before, hence the lack of equipment and skills!