HMV radiogram no AM reception

[lesparapluies]

Breakthrough! replaced VT5 and MW roared into life, with all voltages pretty much as they should be. A good learning curve, and very grateful to all those who gave advice, particularly frsimen / Paula - many thanks. A couple of new issues have now raised their heads - LW doesn't work at all, and on MW the last section of the tuning capacitor - from fully closed to about 25% sweep, results in a sharp drop in signal with no stations received, almost as if a switch is being suddenly turned off when the cap reaches this point. The LW thing is a bit of a puzzle, and I'm going to check all the coils on the front-end this weekend, but I suppose the answer could be that 198, which is virtually the only thing on LW now, could be in the last section of the tuning cap. FM is fine over the entire sweep of the cap. Forgive the relative ignorance of radio front-end circuitry but I've noticed that on the schematic there is a MW oscillator coil but no LW one - I presume the MW one acts for both?

[frsimen]

Well done!

Check for the vanes on the AM sections of the variable capacitor shorting together or some metal whiskers between the vanes where the signals disappear. You might need to bend the offending vane slightly for a cure.

There will be some extra capacitance switched across the oscillator coil for long wave, it's a very common method.

Paula

[lesparapluies]

Hello, this is a continuation of an older post (I've been inactive for a while due to family issues). Okay, I still have a dismantled HMV radiogram chassis on the bench; most of the issues are now sorted but I still have the annoying issue of the AM oscillator which ceases to function about half way down the MW band (reception stops at around 800 and nothing on LW at all) I'm virtually certain that the tuning cap is OK, have done exhaustive checks on it and it doesn't display the crackles you'd expect from shorted vanes etc. I have replaced the transistor, to no effect. I'm now going to try replacing every resistor and capacitor around the oscillator circuit. Can anyone suggest any further tests, or anything I'm missing? it seems such a simple circuit yet there must be something going seriously wrong in there. Thanks for any thoughts.

[ortek_service]

If you've got a 'scope, you could maybe try looking at the output of the oscillator (transistor) to see what happens to level / frequency, as you tune it.

I notice this uses an AF116/126, which IIRC were well-known for (Tin-whiskers?) breakdown to the (shield)case.
So maybe that has started to happen on both original (and replacement?) and disconnecting the shield from ground may prove this?

[Radio Wrangler]

You've made the assumption that the oscillator was properly designed to start with.

Many are not.

Lots has been written on oscillator frequency control and stability with many scholarly analyses as a foundation. Almost nothing has been written about oscillator start-up and amplitude stability, yet we need our oscillators to start and to run close to an expected level.

Sometimes you get a bit of a hand-wave explanation... 'When the power rail comes up, the oscillator gets a bit of a kick and it starts from this' Aye, right, as the Scottish phrase of disbelief goes. Usually someone builds one and winds up the DC bias current until it seems to start reliably, then they stick to those values. Not exactly scientific. Get some moisture in cardboard coil formers and the Q falls.

It might just need the bias increasing a bit.

Using a scope on an oscillator with a start or freq coverage problem can easily affect it so you may not see anything at all. Heisenberg.

I had the job of writing a new oscillators chapter for the ARRL handbook in about 1992 and I spent some time looking for references relating to oscillator running amplitude. Apart from Motchenbacher and Fitchen's book, there really is very little.

David

I agree, whiskers could affect both original and stored (NOS) parts, so that's worth investigating.

[Station X]

Originally Posted by lesparapluies

Hello, this is a continuation of an older post

Threads merged.

[ortek_service]

Yes, attaching a 'scope could affect its operation. But hopefully, if using a 10:1 probe then around 10pF capacitance shouldn't affect things too-much at these relatively-low RF frequencies.
And could maybe put in parallel with another existing circuitry's capacitor, that is >>10pF.

It may also be possible to create a small pick-up coil loop on the end of the probe, and use that to prevent virtually any loading of the oscillator. But then coupling efficiency may well be frequency dependent so may not get as constant a level.

Of course at RF (although relatively rather-low frequency in this case), anything near an RF circuit (Like an Engineer's finger or putting a lid over it) can the operation. But hopefully this should be evident when working on it, if it starts acting a bit like a Theremin musical instrument when approaching it to make measurements as it is that-much on -edge of working
(Where cooling or heating could also affect operation).

[lesparapluies]

Thanks for all the recent replies, a lot to take in. I knew about the tin whiskers problem on the old AF116 large bodied transistors but I hadn't appreciated that the AF126 series smaller ones were also affected, I assumed these were a good replacement. When I received the set the radio was completely dead, and replacing all the AF115/6's with AF 125/6's got it going. I may try chopping off the 4th lead and see what happens. I have to admit my knowledge of the RF part of an AM radio is a bit sketchy and I'd appreciate any links to something which explains how the oscillator/ mixer stages work. Unfortunately I don't have a scope but may be able to borrow one. I'm going to do some voltage checks on the oscillator transistor in the working and not working modes. Re adjusting the bias, I did see a YouTube video where someone with a similar problem experiments with bridging the bias resistor with another resistor of high value, I can see how that would slightly increase the current going through the transistor; is that something which is recommended, or avoided? Also, in my naive way, I assumed that if the oscillator failed, it would fail as the frequency increased, not decreased, as the transistor would be working harder, or have I got that completely wrong? Incidentally, this is very much a matter of pride, I fully realise that very few people listen to MW and there's only one station left on LW, and the VHF section works perfectly, but I'd still like to get it working properly!

[music-centre]

Originally Posted by lesparapluies

Thanks for all the recent replies, a lot to take in. I knew about the tin whiskers problem on the old AF116 large bodied transistors but I hadn't appreciated that the AF126 series smaller ones were also affected, I assumed these were a good replacement. When I received the set the radio was completely dead, and replacing all the AF115/6's with AF 125/6's got it going. I may try chopping off the 4th lead and see what happens.

I cant remember reading that there was the same tin whisker problem with AF126 but of course one could be faulty in another way.

[ortek_service]

When I saw on the circuit that both AF116 & AF126 could be fitted for the oscillator transistor, I assumed they were basically the same (package).

But I've not really come across the AF12x series before (only AC12x etc), so not sure if these are also affected by the the Tin-Whisker issue - that I thought had applied to (most?) 4pin metal-cased Mullard Germanium transistors.

It may be possible to measure any leakage to the case with a multimeter, instead of cutting the lead on (now) quite-expensive obsolete transistors (Maybe just desolder the shield lead, to test, if any any doubt).

As these frequencies are relatively-low, you could probably actually buy a new (USB PC-based?) 'scope for around £10 to £20? that some low-bandwidth can be obtained for.
Although, rather than a Digital Storage one, a more-vintage true-analogue CRT 'scope with >= a few MHz Bandwidth may be preferable for looking at low-signal levels / oscillators as DSO's can tend to give a rather-noisy display (especially cheaper only 8bit ones / even 10 or 12bit etc. ones if sample rate is set too-high)
And these 'obsolete' analogue now tend to go for only a few £'s at rallies etc.

Yes, increasing the bias may help, if oscillator circuit is a bit marginal (Radio Wrangler suggested maybe some weren't originally designed too-well)
Although Germanium transistors didn't have much bandwidth, I doubt these ones would be that bad to run out of steam at only 1MHz, so should normally be fine in this oscillator.

The oscillator transistor is essentially an amplifier (With around 180degrees phase-shift, if inverting configuration)
- So in theory just needs to provide an overall loop-gain (including losses in feedback / resonant tank circuitry) of > 1, for oscillator to work.
Although would normally have quite a bit more gain, to ensure sufficient margin.
And this - along with overall phase-shift of 360degrees from positive (rather than negative) feedback, is what makes it oscillate.
So the phase-shift criteria, as well as loop-gain, criteria needs to be met across the whole tuning range (quite a lot for MW that spans 3octaves) for oscillation to always be maintained.
If some of the original capacitors / inductors etc. materials have started to degrade, then their RF-losses may have increased (Q decreased), which could be enough to affect reliable oscillation.

I'm sure others like Radio Wrangler will have some good links to articles on oscillators (Probably a few books on these, at the archive websites)

[lesparapluies]

Again all valuable info, and I'll do a bit of research as suggested. Regarding the oscilloscope, if I decided to take the plunge and buy one, would one of the entry-level digital ones be sufficient for the kind of thing I'd be using it for, i.e. mainly domestic audio and radio "of a certain age". Probably an unfair thing to ask, as everyone will have their differing opinions! I've been looking at the FNIRSI 1014D which seems to have quite good reviews for the price. My reservations about seeking a vintage CRT one is what is likely to go wrong with it, as I certainly don't have the skills to repair one!

[ortek_service]

I'd not heard of that brand before, so did a quick Google.
Where at first glance it seems Aliexpress are selling it for only £40, but you have to be careful, as that listing default to a different product and it is actually around £150 from that seller - So much-cheaper elsewhere.

I did discover this review-article on it: https://hackaday.com/2022/06/07/cheap-oscilloscope-is-well-cheap/
- Where it doesn't seem to come out too well. So whilst this 3-in-1 unit has additional features like a DDS generator, it seems some compromises (Like Low sensitivity, so not as ideal for low-signals audio work) have been made (Mainly as a result of its budget price?)
And there are probably other budget ones that are rather better - In particular, Rigol & Siglent (possibly also Owon) ones seem to be fairly well regarded despite being fairly-cheap.
So it's worth checking some of the reviews of other models, linked from that article etc.

I wouldn't discount just getting a fairly-vintage 5MHz / 10 MHz / 20MHz Analogue CRT 'scope, as they are generally quite reliable - especially the most basic Analogue-only ones, as not too much to go wrong.
And a true-analogue 'scope seems most suitable for what you would like one for.
If you can get these for only around £10-£20 (sometimes even less / free!), then it's no great loss if it does fail as can just get another (even though it is always nice to be able to repair things). Maybe a fairly-late 80's or 90's model, before they really stopped making these, would be best as not too old.
It is probably worth getting a dual-channel one, to be able to make comparisons of signals through audio chains.
There are some analogue ones with more channels, but rarer and generally not so necessary for audio etc. work.

Really-old all-valves models are likely to be large, heavy, and also generally low-bandwidth, so probably best to avoid any pre 1970's models that are now really only for collectors of these.

[lesparapluies]

I've been looking at either a Tektronix 2205 or a Gould OS250... around £75 2nd hand

[Leon Crampin]

I would not favour a Gould 'scope, from personal experience. In my last job, we had a mix of Gould digital 'scopes and Teks which were vastly superior in every respect.

The Gould 'scopes would exhibit random noise on the display - with a grounded input - and had lamentable reliability and poor service turnaround from the maker. The analogue Teks just functioned to spec. and worked.

From earlier discussions, the AF116 family is electrically identical to the AF126 series, but the 12x devices differ in using a nickel plated 4 lead TO-18 sized can, which completely avoids the whiskering problems of the earlier devices which are in larger tin plated cans. It is the tin which forms dendrites (whiskers) on the inside, leading to short circuit failures.

Sadly, the AF126 family of devices have become fairly expensive these days, but odd devices are generally available. All tin plated can devices are prone to this problem - I had a failure of a Grundig xx boy radio the other day, caused by a whiskered output transistor - AC187 I think.

Leon.

[ortek_service]

£75 does seem rather a lot for a 2nd hand basic 'scope these days.
Whilst it was common for used <=20MHz 'scopes to fetch around £100 back in the 80's (when those models were > £300 new, and struggled to get even a single-channel 10MHz 'scope for < £200 new), more-recently I've seen them at rallies for £10-£20
- Especially considering what you can get new for a little more from many Chinese brands like Rigol, Siglent, Owon etc. (Most with UK Reps / distributors like Telonic etc)

(Advance)-Gould are a rather old 'scope Manufacturer, and the OS250 is only 10MHz (OS250B=15MHz). Whereas Tektronix are still around, and were often regarded as one of the leading 'scope manufacturers producing a wide range inc. the most-sophisticated models.
The Tek 2205 is 20MHz (with a 40MHz version available) so is also a better 'scope than the Gould, if similar price.

Modern 'scopes will be somewhat smaller and lighter (Plus some maybe DC powered). And you can also get 'scopemeters' that are a DMM integrated with a basic 'scope, which are generally even-smaller (but maybe too small in some cases?) - I did see some quite nice looking ones at the recent National Hamfest in Newark, that hadn't seen before, and seemed to have quite high bandwidths.

Here's an example of a 10MHz (50MHz Sample rate) dual-channel handheld 'scope for just over £30 with deliver:
https://www.aliexpress.com/item/1005006426182452.html
- Plus a few other similar ones, on the page below it.

Virtually all of the main Test Equipment manufacturers produce a basic 2-ch Colour LCD 'scope at a fairly-competitive price, with the likes of Rigol etc. And it seems many of this are really a much higher bandwidth model but with that deliberately reduced in software! (And so some of these can be upgraded wi5th an unlock code - Sometimes at zero-cost, if this 'hack' gets publicised)

[lesparapluies]

Again, all advice appreciated and taken on board. I think I've sort of made my mind up to go for either the Tektronix 2205 at about £70, or just to dive in with a cheap Chinese one.. I've never used a 'scope before so it'll be learning from scratch either way. Just one more question - previous replies have suggested that a CRT (but solid state) one would be more suitable for the kind of thing I'm doing, than a modern LCD display - any particular reason for this?. Thanks again

[Julesomega]

To avoid going off-topic for this radio repair, I started a thread on the modern technology section, which has brought some helpful replies. You may like to look at the discussion Modern LCD 'scopes for the vintage repair bench

[Boulevardier]

Originally Posted by lesparapluies

previous replies have suggested that a CRT (but solid state) one would be more suitable for the kind of thing I'm doing, than a modern LCD display - any particular reason for this?. Thanks again

I was given a Rigol DS1052E, two-channel 50Mc/s scope a few years ago. It had been supplied FOC to someone who was going to review it, and return was not required once the review had been done; he didn’t want it, so gave it to me. So I didn’t have to justify to myself any investment in buying the thing, and I already had a decent Hitachi 20Mc/s two-channel scope, so I had a choice of which one I used.

The practical difference between the two scopes hit me in the face. The analogue scope had all functions visible and selectable on the front panel (lots of knobs!). On the Rigol (very few knobs), and many (most?) of the functions could only be found and switched by burrowing down through layers of (often non-intuitive) menus. I spent more time in consulting the very large user-manual to find the functions I wanted than in actually making any measurements.

If I used the Rigol all day every day (or if my memory was better!), I’m sure I would have quickly mastered the menu locations of the fantastic number of facilities it offers, but for the relatively infrequent use I have for a scope I find myself using the analogue Hitachi every time – since all switching and functions are immediately visible and available on the control panel without relying on memory of where to find them in the nests of screen menus.

If I were you, Lesparapluies, I would try and find someone close to you with a digital scope who would let you have a play with it and gauge whether you’ll get on with one.

Mike

[ortek_service]

Originally Posted by Boulevardier

I was given a Rigol DS1052E, two-channel 50Mc/s scope a few years ago.
>> and I already had a decent Hitachi 20Mc/s two-channel scope,

The practical difference between the two scopes hit me in the face. The analogue scope had all functions visible and selectable on the front panel (lots of knobs!). On the Rigol (very few knobs), and many (most?) of the functions could only be found and switched by burrowing down through layers of (often non-intuitive) menus.
>>
>>
Mike

(Hopefully last comment on 'scopes'! just to clarify)
The real advantage of a True-Analogue scope is that the signal is written directly to the screen.
Note: Early Digital Storage 'scopes (plus there were also some Hitachi etc. Dual Analogue & Digital modes hybrids) also used CRT's
So True-Analogue involves no digitisation, which can produce a variety of artefacts on the screen (like noise, due to limited number of bits etc, response-time lag due to the digital processing and 'aliasing' if sampling rate isn't set high-enough) that are not really on the signal being measured and could mislead you.
And on DSO's you generally have to be more-careful on the settings to get the best waveform, with these effects minimised.


'Scope User-Interfaces (UI) is rather separate from this.
And even some old Analogue models could have many switches and knobs with 'Stability' etc. adjustments that all needed to be set right.
Whereas most DSO's do have an 'Autoset' button, optimising all settings for what part of the waveform it thinks you are trying to look at
- But might sometimes get it a bit wrong (like 'AI'),

DSO UI's can vary quite a lot even between different generations of models of the same make. Early HP CRT and Tektronix LCD Mono-colour DSO's often only had one adjustment knob and everything buried in menus!
But manufacturers have generally learnt from this, and later models are generally much-more user-friendly - And now virtually all have colour displays, with individual traces colour-coded to help distinguish things.

However, cheaper models will probably have less individual adjustment-knobs controls to cut costs / size (Plus also trend for widescreen displays has meant even Tektronix have had to condense controls into a small vertical area at edge of screen, so back to need for more-menus on these!).
So may find that other Rigol models are much more User-friendly, as many on here seem to have these. Plus they do appear to be rather-similar to ones from Tektronix etc., but generally much-cheaper.

[lesparapluies]

Once again a long gap before I could do any more work on the radio. Tonight I made a breakthrough (kind of) in the "silent" part of the MW band. I discovered that if I tune to the low frequency end of the MW band (the part beyond the threshold where I suspect the oscillator fails) - if I then touch the ferrite rod aerial the radio springs to life and I receive the stations in the low frequency end of the band - as I bring my finger away the reception continues for a couple of seconds then abruptly stops. Does this imply that the extra strong signal at the aerial is somehow kicking the oscillator into life, and I should look at increasing the gain of the oscillator transistor to stabilise it? Or is that naïve? As I've said before, my knowledge of RF / IF is limited so I'm kind of stabbing in the dark