Nightmare Roberts R707 blowing raspberries

[jonnybear]

This Radio was one of my collection, last year I replaced the tin whisker AF*** in the Mullard module and had not been used for a few months, when it was switched on the sound was awful distorted and blowing raspberry sounds. Checking the AC188 and AC187 I found these also had tin whiskers to the casing, so decided to replace, this only made things worse voltages were all over the place. The culprit in the end was a Philips 470uF capacitor C39 which had decided to become a 0.8 ohm resistor. never had a Philips electrolytic fail this way before.

John

[poppydog]

I have found that the low value Philips ones are never very good.
Regards Poppydog

[crackle]

I find they are normally reliable.
Mike

[paulsherwin]

Audio stability problems are normally down to bad electrolytics.

I agree those Philips parts are usually pretty good, but they are getting very old now.

[MrBungle]

The old Philips axials are horrible regardless of value. I've had a couple of them go totally short. The first one burned out the power transformer in my PM3217. The second one killed another Philips PM3315, fortunately however that had a somewhat better power supply arrangement. Took me a while to find it.

[mhennessy]

This is only the second one of these that I've heard of that has failed short.

It's impossible to make generalisations like "horrible regardless of value" about them based on 2 samples - this is a civilised forum, not social media! I don't believe that their lifespan is any worse than any other decent-brand capacitor of equivalent temperature/endurance rating. I've seen literally thousands of these over the years, and the ones that have failed prematurely did so because the designer didn't choose a better performing part. The 2 that have failed short are most definitely outliers.

On the other hand, Daly and TCC Elkomold types are truly deserving of such labels. Every single one that I have tested has been faulty (usually leaking heavily) - I'd estimate that would be several hundred, so a reasonable evidence base for a valid generalisation.

[MrBungle]

Not so sure. I decided to dig through my repair notes, which I've been keeping OneNote and found some more. Total list:

1. Primary filter caps on a PM2524. One short.
2. Filter caps on Fluke 8600 causing transformer heating.
3. Tektronix 2225 power supply failure to start (2 different scopes)
4. Blown up PM3217 power supply due to short.
5. PM3315 5V rail short causing power supply failure.
6. NOS one ebay 1984 date code open circuit.
7. Telequipment D83 V2 (there's literally one in that and it was duff!)

That's more than a little bit of coincidence. That's not to say there aren't worse but they are as suspicious as a RIFA.

TCC / Plessey ones are the worst - I'll give you that. I don't think I've seen one that works or isn't spewing something for a while. When I rebuilt my D83 recently the Plessey ones had destroyed the power supply board entirely. I had to buy another unit to get a working power supply. Oh and Tektronix tantalum caps.

[vinrads]

Well done John , this is a lesson suspect the unexpected. Mick.

[mhennessy]

Originally Posted by MrBungle

Not so sure. I decided to dig through my repair notes, which I've been keeping OneNote and found some more. Total list:

1. Primary filter caps on a PM2524. One short.
2. Filter caps on Fluke 8600 causing transformer heating.
3. Tektronix 2225 power supply failure to start (2 different scopes)
4. Blown up PM3217 power supply due to short.
5. PM3315 5V rail short causing power supply failure.
6. NOS one ebay 1984 date code open circuit.
7. Telequipment D83 V2 (there's literally one in that and it was duff!)

That's more than a little bit of coincidence. That's not to say there aren't worse but they are as suspicious as a RIFA.

TCC / Plessey ones are the worst - I'll give you that. I don't think I've seen one that works or isn't spewing something for a while. When I rebuilt my D83 recently the Plessey ones had destroyed the power supply board entirely. I had to buy another unit to get a working power supply. Oh and Tektronix tantalum caps.

OK, so assuming each of those failed short, that's 7 to add to the 2 I know about (this thread, and David's R505). Didn't know Tek were using them though... Nor Fluke, prior to the merger with Philips T&M. That means there's more out there than I thought.

I've got 3 Telequipment D83s and 3 Telequipment DM63 (used to have 4!). No faulty ones in those so far. The large ITT smoothers are usually tired and visibly leaking by now.

Still, 9 does not make a pattern. I've had no shortage of Japanese and American caps fail short too - it can happen to any electrolytic, but not in numbers that is indicative of a pattern.

They certainly have a long way to go before they reach RIFA's standards.

Tants have earned themselves a reputation, but at least some of that is down to the designers sailing close to the wind in voltage rating terms. Personally, I'd say I've been quite lucky with them - will probably regret that soon!

[David G4EBT]

Originally Posted by mhennessy

OK, so assuming each of those failed short, that's 7 to add to the 2 I know about (this thread, and David's R505).

There are several forum threads on R505s where R40 - 680uF, has gone short circuit, causing the output pair to get very hot, and the fault sometimes being misdiagnosed as failing/failed transistors. Several threads mention symptoms such as "the set comes on for a couple of seconds then fizzles out". Here are a couple of relevant threads:

Post #8
https://www.vintage-radio.net/forum/...t=Roberts+R505

Post #3:
https://www.vintage-radio.net/forum/...t=Roberts+R505

I don't see it as a failure of Philips or of Roberts that these faults arise - indeed, some might argue that after more than fifty years it's a success story. Neither would I advocate wholesale replacement on sight (a la waxy caps in DAC90As). On Paxolin circuit boards with many components mounted upright it would cause more harm than good. I'd simply say to view them with suspicion and study the circuit to see what role they play, maybe lift one end of the cap if need be to check if it's morphed into a low value resistor.

Most of us know by now that there are other issues than tin whiskers and duff electrolytics that can cause no sound from a transistor radio, not least the earphone socket being open circuit. ('nothing comes out of the speaker but it works on headphones').

As to R505s, If C41 & C42 go short circuit or leaky, though they won't risk killing the output pair, they'll quickly drain the battery, so are worth checking. I replaced mine as a preventative measure.

For anyone who has a scope and a 1 kHz sine wave source, it's also well worth checking and if need be adjusting R36 per the instructions in the Robert's R505 Service data. Pic 2 below shows how my set was before checking and adjusting R36 and pic 3, after adjustment.

Let's not forget of course, that the reason old radios often come into our hands is that they've become faulty. The 505 that I have was from a neighbour who said "before this goes in the bin, I wondered if it's any use to you for bits?"

[tri-comp]

Originally Posted by David G4EBT

indeed, some might argue that after more than fifty years it's a success story.

Absolutely.
But they are starting to pop-up now and again.
I've seen them go duff in some of the Philips Laser Mecha's that Bang & Olufsen used in their players.
Here's a picture of one found in a stereo-decoder belonging in a late 60'ies Philips 22RH786 receiver.