New capacitors for old condensers

 The most frequent problem met when restoring old radio equipment is old condensers. Usually the small-value varieties such as moulded or silver mica fail much less often than the tubular wax-covered type used for decoupling or coupling in valve circuits which I'll deal with here. To diagnose a faulty condenser, often one end is clipped and its DC resistance is measured. A really poor condenser will have a low resistance, where low means anything under one megohm. Unfortunately a simple measurement like this may not reveal many problem condensers which will only fail when a high voltage is connected across them. Connect a couple of hundred volts across an old 350v condenser and after a few moments the thing will often exhibit leakage which can then increase dramatically with time. To confuse matters, once the thing has been disconnected from a high voltage its resistance may rise again however if left connected it can eventually become too hot to touch. See the R206 restoration.

A restorer can easily replace an old condenser with a new capacitor, however the view under an old chassis is then marred by lots of shiny new parts. One solution is to detach an old condenser, remove its innards and fill with a modern capacitor, a process known as "stuffing". Usually a smaller tubular capacitor will be fitted inside an old case, but I've adopted a different approach now that sub-miniature high voltage capacitors are available at much lower prices than standard types. Below I've demonstrated how to restore a tubular aluminium condenser used in the R1155 receiver. I've also used a different method here...

Click this picture to see this page which has more detail.

 First the raw materials. A roll of 0.1uF x 500v capacitors with one shown on a 25mm diameter 2p piece. Click the 2p to see the capacitor spec. (C1210V104KCRAC)




 In this exercise I'm going to restore a 3 x 0.1uF condenser which is about 85mm long and 18mm diameter, and rated at something like 350V and made in 1943, using three chip capacitors each with a "1210" case whose maximum size is 3.3mm x 2.6mm x 2.1mm.

You can see that the rubber insulation has perished so that there may even be leakage due to this quite apart from internal problems.

 The first step is to drill out the old leads and around 50mm of the internal condensers using a 5mm drill. The new insulation material is a length of drinking straw about 5mm in diameter into which will be inserted the new capacitor assembly.



 Because the old condenser has a wire common to the three 0.1uF sections I've used a new black wire connected as shown. To solder the wire some sort of clamp will make things easier otherwise the chips will stick to your soldering iron.

Below, I've connected a red wire for each 0.1uF capacitor and then checked the assembly slides into a length of drinking straw with its end sealed with wax to prevent ingress of old condenser particles left from drilling.




 The tube is inserted into the 5mm hole drilled into the condenser.

At this stage the new capacitor assembly can be pulled out of the tube so to secure this I've melted some wax from another old condenser into the gap to seal it in place.

Below is shown the finished triple 0.1uF capacitor refurbished 73 years after manufacture and ready for fitting back to the R1155 chassis.

I measured the capacitance of each section and found all were identical at 0.122uF. The meter shows the reading when the three red wires were connected together.

When you select the red and black wires ensure that they're tolerant to heating when soldering and their insulation and voltage rating is OK for use in the equipment in which the refurbished capacitor is to be fitted.

Parts cost was under 16p which is pretty good for three 0.1uF x 500V capacitors....

I've used fairly thin wires so suitable sleeving can be pushed over them to make the set look original.




 While on the topic of measuring old condensers I'll mention my new Mega328 tester. These things are incredibly cheap compared with say the Peak series of testers which I've relied on for many years. You can get this device from lots of suppliers, predominantly in China, but mine came from the UK (at twice the price = under £9) as I wanted to try it out sooner rather than a lot later. I'd already bought a cheap plastic box for another project so I used that to mount the tester to make it more robust.

 Just out of interest I checked a set of old condensers removed from various radios over the years with the following results (0.1uF = 100nF)

Note that a zero/unmeasurable Vloss figure represents excellent with anything over 1.5% considered poor.


 Marked value


 Measured value

 ESR ohms

 Vloss %


 Dubilier 4423, 1936





 Dubilier 4423, 1936





 Dubilier 4423, 1936





 Dubilier 4423, 1936





 ZA24256, R1155





  ZA24256, R1155





 345, WW2





 Dubilier, 1936





 Modern yellow plastic




 During testing I noticed the battery voltage steadily dropped from 9.5v to 8.9v (probably a well used PP3) so I measured the current taken by the tester in both OFF and ON.

OFF = zero measurable current, ON =24mA. This being so, I'll fit a battery holder and use a set of six AAA cells.

Return to Reception