"Suflex" type polystyrene capacitors - failure modes

[Phil G4SPZ]

I have just spent three long evenings tracking down an intermittent fault in a communications receiver BFO. I finally traced the cause to a 1,000pF polystyrene tuning capacitor across the BFO transformer primary. The capacitor was hidden inside the can...

The symptoms were peculiar, but uncannily reminiscent of a fault I encountered in the decoder of a Sobell colour TV back in the 1980s. In both instances, the signal gradually faded away, but would sometimes suddenly return. No amount of prodding or poking would induce the fault to occur.

This old thread from 2009 discusses Suflex capacitors and concludes that this type of component should probably be suspected on sight. However I'd be interested to know if anyone is aware of what causes them to behave like this. The faulty component checks out OK on my capacitance meter, but the way the onset of the fault sounded when it did occur - slow drift in frequency, followed by complete fade-out - suggests to me that intermittent loss of capacitance may be one of the causes.

Any suggestions welcome!

[GrimJosef]

I confess that I was surprised to find how easily the wires can be pulled out of this type of capacitor. Perhaps if they're mounted under any kind of stress, particularly cyclic stress, the connection can fail internally, so the capacitor looks OK but the internal contact resistance will vary intermittently and/or with (say) thermal changes ?

Cheers,

GJ

[Nuvistor]

Suflex capacitors were always suspect for intermittent faults in tuned circuits. I rarely had them go short but O/C or low capacity was common.
If the circuit relied on a property of the cap, it was replaced with the same type but usually you could use a sliver mica for replacement.
Has GJ says the wires tended to come out of the cap quite easliy.
Frank

[Phil G4SPZ]

Originally Posted by GrimJosef

I was surprised to find how easily the wires can be pulled out of this type of capacitor

Yes, I inadvertently pulled one wire off the replacement component! Fortunately I had several spares. You're probably right. I have also noticed that, since replacing this capacitor, the BFO is a lot more stable and the frequency is much less susceptible to flexing and mechanical movement.

['LIVEWIRE?']

4700pf + 1000pf Suflex caps. were used in Radiomobile & Motorola car radios of the 1070/1080 (737/838) series in the Colpitts Oscillator circuit. IIRC there was a fairly high failure rate in these due to intermittent connections, going o/c, etc., though in the Dealership where I worked back in the 70s, the policy was to buy spares from the manufacturer concerned, so, until Radiomobile came up with better alternatives, the failed Suflex cap. was replaced with another of the same type. I don't recall S/Mica caps ever being used in these circuits, though.

[Heatercathodeshort]

You need to be a little cautious here. The polystyrene capacitors that were a regular cause of trouble in line oscillator and tuning stages were not actually 'Suflex' brand components.
Somewhere I have a letter from the makers of Suflex capacitors stating that it was not the type manufactured by themselves that went intermittent O/C!
Polystyrene capacitors certainly gave us all a lot of work in the 70's and 80's. John.

[Phil G4SPZ]

Thanks for that John, good point. Perhaps the Mods could consider removing the word 'Suflex' from the thread title? I'd hate to get sued!

[Nuvistor]

That's interesting about the letter from Suflex, I wonder who did make them. Whoever did they did give us work.
Frank

[MotorBikeLes]

I remember talking to an engineer at Tektronix (a Telequipment man) about 30 years ago I think regarding a D67. He said they had a few faults with these transparent/silver caps (notice I avoided legal redress risks there), and the way to check was to gently squeeze with long nose pliers. If the fault appeared or disappeared, you had located the culprit.
Les.

[Mike Phelan]

I think one of the problems with these was that the leads were just wrapped around the last outer turn of the foil, so the oxide corrosion on the latter, being a good insulator, OCd the component.

Shame they came into production at the same time as colour TVs!

[Bazz4CQJ]

Am I right in saying that one "distinctive" property of polystyrene caps is that they have a fairly large negative temperature coefficient and so offer a means of incorporating some degree of temperature stabilisation in a tuned circuit? Replacing a polystyrene cap with another type in some circuits might be problematic.

Worth noting, however, that they are not the most rugged component ever made.

B

[Radio Wrangler]

The tempco of polystyrene capacitors almost exactly cancelled that of Siemens N28 material pot cores. Handy for filter design.

Polystyrene capacitors undergo a value change at about 75 centigrade, and it's irreversible.

David

[Phil G4SPZ]

Thanks to everyone for the informative replies. It does appear that the weak point in these capacitors is the way the lead-out wires are attached to the foils, coupled with the potential for thermal damage due to heat from soldering being conducted up the wires. The wires are very thin, presumably to limit thermal conduction, but it would appear wise to use forceps or tweezers as a heatsink when soldering.

It's probably safe to conclude that the predominant failure mode is 'intermittent loss of capacitance or open-circuit'.

Thanks again all,

[Mike. Watterson]

They also melt easily if the leads are too short or you heat too long!

[John_BS]

As David says, these capacitors have an N150 (ppm) temp coefficient, which is handy for tuned circuits where the L is usually positive. They are also very high Q from audio through to about 20MHz. LCR is another company who supplied these.

John

[Ambientnoise]

I used a lot of genuine Suflex polystyrenes in filters (they temperature compensated ferrite cores as stated above ) and found them very reliable unless overheated on installation (always use a heat shunt) or inadvertently touched by a soldering iron.

Ken

[turretslug]

Originally Posted by Phil G4SPZ

Thanks to everyone for the informative replies. It does appear that the weak point in these capacitors is the way the lead-out wires are attached to the foils, coupled with the potential for thermal damage due to heat from soldering being conducted up the wires. The wires are very thin, presumably to limit thermal conduction, but it would appear wise to use forceps or tweezers as a heatsink when soldering.

Indeed, so thin that they could suffer from "green spot" type fracture if poorly stored or used in a humid environment. Philips made a range of precision polystyrene capacitors with very neat and competent packaging and thicker, less brittle leads, though I was still very circumspect about soldering. Professional audio kit from some very respected names fairly bristled with them, which struck me as a vote of confidence. I once saw several cubic feet of massed NOS Suflex tangle discarded- I was tempted to rescue them for the likes of the hobbyist, but no-one thought much of them. Had I been aware of the likes of this forum, I would have done so. Ho hum.