I've ordered a bunch of caps to replace all papers and electrolytics in a Philips 179 radiogram chassis.

Meantime, voltages seem fairly close to the specs and the radio tunes both MW and SW bands. A signal input from pickup terminals is perfectly amplified, so I think the audio and power sections are in good shape. Two issues though -

1. Radio signal is audibly biased to the treble, whereas PU signal is well balanced.

2. The tuning range seems to be compressed compared to the spacing on the dial. I guess this is alignment somewhere.

Can I expect cap replacements to fix these?

Maven

As far as I am concerned, in a case like this, the time spent fault finding in a set with tired electrolytic & wax paper caps, is better spent replacing them. Do not forget to check the resistors.

There are various reasons for the band spread to change. Some Astor sets do that if you sub the 6A8 with a 6J8 (JJ). Coils can change with age.

If I take one end of a Mica out of circuit whilst changing something I leakage test it, the odd one will fail. The same applies to resistors if one end is free, check it. Not all can be accurately tested "in circuit", however, if a resistor is high in circuit it won't improve out of circuit.. replace it.

Once you change all of that you realign it. That is in a nutshell "Mandatory" if RF components are changed.

More often than not, I find that changing out all of the dodgy bits saves time & frustration later, as they are usually the problem.

Marc

I would go along with Marcc's observation, best to rebuild the set stage by stage.

Modern resistors and capacitors are far better quality in any case than the originals.

I take photos of everything with a digital camera before doing any changes so as to minimise the chance of making a mistake.

After a rebuild many of these old sets have an amazing clarity and tonal balance.

Thanks for the advice. It'll be a week or two before the caps arrive, so meanwhile I'll test whether staring at it helps at all.
By staring I can also mentally translate the schematic onto the physical connection points.

This is the first multi-band I've worked on (MW + 2 x SW bands). The band-selector caps are all micas. I won't touch those unless replacing the papers and electros further up the chain fails to fix things.

My anti-error precaution is to replace components not just one at a time, but one end at a time (having checked against schematic to see whether any previous work may have introduced a misconnect).

Maven

Hope it all goes well for you Maven. Have found that sometimes the schematic does not agree with what the radio was built as there may have been production changes as time went by.

Which is why I prefer to take copious digital photos and save them on the computer.

Even print them out in colour which makes it even easier.

As for those old mica capacitors they can cause lurking intermittant faults on occassion.

A while back was working on a Ferris M94 and a mica cap was completely shorted which took a lot of probing about with the multimeter to unearth it.

Evatco in Queensland have a good range of micas and it is my practice now to replace the lot.

Maven,

Good advice from Marc and Simplex .... and I don't know the level of your technical skills .... but if you feel capable ... and inclined ..... and have an oscilloscope, you might investigate a few of the obvious culprits. Yes, Caps are frequency dependent and the first suspect.

Personally, I like to 'problem solve' these old radios ...... then replace the old caps .... . Clearly identify each problem, then use the circuit diagram and an understanding of what a failed cap might do to a circuit to identify culprits. You will need to identify the ac signals. Finding the fault can be extremely rewarding ...

Cheers,
Ian

"Personally, I like to 'problem solve' these old radios ...... then replace the old caps".

And therein lies danger.... If the set is working in some sort of fashion, fine.

But if its an unknown set, you need to know if it is not of dubious integrity. Turning it on without doing some basic checks is looking for trouble, not finding the trouble.

Frequently I will not bother to power up a set; By looking before I leap, you avoid loud bangs, smoke and burning out expensive transformers. These sets were often abandoned due to breakdown.

Our transformers are generally more robust but in the US upping the mains voltage 15V making everything a bit more marginal, turning the set on to see if it goes, has left a litany of posts on the antique radio forum, of burned out transformers. So far here in forty + years I have only had two.

It only takes one dodgy 6X5, a shorted wire (& I have seen plenty) or a filter cap to fail shorted (not uncommon) and you have disaster.

Many of these components are known ferals and should not be powered, many parts can be checked (valves; transformers) roughly without applying mains.

Marc

Personally I would would not bother trying to fault find individual components.

In all of the old sets I have worked on nearly all of the old carbon resistors have drifted high in value, well out of their tolerances.

So even if a faulty cap is found the radio will never work to its optimum.

Other caps may be leaky somewhat but not shorted so will work to some degree but again not to their optimum.

Other damage may unknowingly occur such as a leaky coupling capacitor cause excessive grid current to flow in a valve causing the valve to be damaged.

Best and safest thing to do in the long term is to replace the lot.

Okay .... I am headbutting an old chestnut here .... ..... a safe but technically limiting approach.

Every time I turn on my old valve radios, I marvel at the genius of the people who developed the technology about 100 years ago ......

Cheers,
Ian

After over forty years of fixing, one often just has to look at the set so see where things have gone awry and have a good idea of the cause when a fault manifests.

So if its a commercial fix full of known ferals which I know will fail, if they have not already, the time is better spent as said testing the parts that may make it a non viable repair, before eliminating the ferals from the equation.

There is a vast difference between refurbishing & being a serviceman.

The STC 59 chassis that a circuit cant be found for, when looked at after removal from cabinet; instantly said only a fool would power it.

Mysterious wires with no owners and a tag trip full of non original resistors, do tend to ring alarm bells.

We all have different approaches. One must be more ruthless with a commercial repair. Failures are not good for business.

Marc

Mica caps that have a fair amount of DC voltage across them tend to become leaky. LIke a mica connected between the plate of the audio driver triode and ground (around 60VDC on average). Found one there that caused noise.

But micas without much voltage across them usually don't get leaky.

Some IF transformers used a single sheet of mica with silver printed on it. To form 2 caps, one for the primary (with B+ on both ends), the other cap near ground. Silver can migrate between these 2 caps, as the voltage potential between these caps attract ions. But if your IF transformers use completely separate caps, this won't happen.

I'm taking a middle path on this, since my level of expertise does not extend to being confident that I could identify issues from, say, an oscilloscope trace on an AC signal.

I'll be replacing paper caps first, one by one, starting from the antenna coil and working towards the source switch (since the power amp circuit seems clean).

As I replace each cap, I'll re-test the related resistor and voltage.

If that doesn't identify the problem, I'll start checking the mica caps, working back towards the antenna coil and focussing on any cap that is always in circuit, because the problem affects all bands and all tone ranges.

The IF transformers are in rectangular cans that include the caps. Is there a way to test those caps without demolishing the cans?

Maven

Generally I start at the power supply replacing the filter electros first.

Then the audio stage onwards to the I.F's and then the front end.

Replacing rotted wiring as necessary.

Refering constantly to the digital photos already taken as mentioned earlier.

As for the mica caps in the I.F cans if there are still lingering problems try replacing them as well.

I don't know how the mica caps could be tested, don't have any measuring instruments for these.

Perhaps a voltage test to see if they are breaking down when H.T is applied ?

Have never done this.

Proper convention is to start at the power supply.

The Mica caps here are not like Micamold and several others in the US. Over there Mica caps tend to suffer from silver migration (Silver Mica disease).

Micamolds are not always Mica and I have had dealings with them in Midwests. These are so reliable that I treat anything with Micamold on it with contempt and replace it on sight.

Mica cap failure does occur here but it is rare, to the point that I only leakage test them if one end is out of circuit.

IF's here rarely fail, it is mainly rewiring them that causes breakages. The only one I have seen in an Australian set with a cap problem, was due to a mechanical issue.

The caps on the primary may have B+ on them, but as they are on a very low ohm coil, the DC volts across them is low. If a superhet is working, albeit badly IF's would have to be somewhere near frequency, or it would not work. The most likely Mica to fail I find is on the det audio valve, perhaps delayed AGC (rare), or a rail that has a lot of DC voltage.

Remember to check resistors as you go: Per last comment on them

Marc

Progress: I've replaced all the papers but not yet the electrolytics (there are only three), because the main power filter caps are both in above-chassis cans and I haven't decided whether to bypass them with new caps in the pan or stuff them. I suppose I could test them in the pan and stuff them if replacement makes a big improvement.

Replacing the papers has fixed tone control problem and improved overall sound quality, but other symptoms remain.

There is a low-level hash audible under the tuned signal. This hash is not audible on other radios used in the same room. And there's no hash when source is switched to pickup.

Also, AGC doesn't seem to function properly and weaker stations are not detected. Next plan is to re-check all voltages around the detector, mixer and IF transformers.

Tuning range alignment also not corrected yet.

Maven