Looks like I may be replacing some valve sockets, so I wondering about best kind of wire for different uses.

Over the years my radio has had many repairs and replacements, and acquired almost every kind of wire, from bare plated copper, to multistrand plastic insulated copper, rubber-insulated multistrand steel, shrink-insulated single-strand copper, and a few surviving cloth-insulated connections.

Should multi-strand be avoided for fixed internal connections?

What's the right guage and type of wire for valve connections?

Are bare "air-insulated" wires really advisable?

When undertaking a substantial re-wiring, how do you route your wires, and what order do you approach things, so as to minimise tangles and visual confusion?

When I look at a typical radio chassis from the 1940's there seems to be a bit of everything. VIR-insulated stranded wire seems to have been used a lot for valve heater wiring and hook-up wire for various top-mounted components like transformers and coils. In later years the insulation was more often PVC or fibreglass though back in the 1930s a product called busbar wire was used. This was a thin-gauge square-section wire either soldered at various points to the chassis or mounted on insulators, as needed. Busbar wire was also used on fibre breadboards in the 1920s.

Most of the high tension circuit generally comprised of components being soldered to each other and to tag strips and valve sockets to provide the necessary support. For this reason, the components of yesteryear had longer leads than you'll find on their modern bretheren.

There is nothing wrong with continuing with what manufacturers came up with for their radios. Some, like the 'Empire State' Radiolettes were birds nests and a little more care needs to be taken, particularly with the cluster of condensers at one end of the chassis.

As a rule, a thicker gauge wire needs to be used for the heaters - they draw more current than the tuned circuit. I wouldn't think that anything greater than 0.75mm² would be required for most jobs though.

Coming back to this thread because I think I made the wrong choice when replacing some perished original wiring a couple of years back. I used mains-rated multi-strand plastic-insulated copper for all the HT DC connections. Because it conducts heat so well, I had to pour an awful lot of heat onto the valve socket and IF transformer tags to solder the connections. I'm thinking this might have contributed to weakening some connections inside the IF transformers.

I'm wondering if I could go the opposite direction and use some single-strand copper hook-up wire of the type used in telephone wiring systems. I know it carries 80v in the phone system, and hope it might be robust enough for the small currents in the B+ and audio connections.

Opinions?

Maven

I'm wondering if I could go the opposite direction and use some single-strand copper hook-up wire of the type used in telephone wiring systems. I know it carries 80v in the phone system, and hope it might be robust enough for the small currents in the B+ and audio connections.

Telephone wire has thin low voltage insulation -- not what you want for B+.

Thanks for the warning - I'll look for a lightish gauge single-strand with solid insulation. Or maybe settle for bare copper with spaghetti tubing.

Any known supplier for this type of wire?

Maven

I generally use the plastic insulated 300 volt multi-stranded stuff from Altronics and I have some modern cloth covered stuff that I have used on occasion for authenticity when plastic would look out of place.

As has been discussed in another thread on wiring, you can sheath 300 volt stuff in spaghetti to raise its voltage rate a bit more when and if needed.

As for bare wire, it's more usual to use tinned copper wire -- available from Jaycar/Altronics.

Single core insulated is more often found in filament circuits for high current rectifiers.

However, I'm not sure that the wire choice is your real issue here. If I'm dealing with a situation that provides natural heat sinking (such as a lug connected to the chassis), I use the hottest iron that I can but only briefly enough to get enough solder on the thing that will allow a well-tinned multi-strand wire to be joined to it quickly thereafter.

Otherwise, apart from other problems, too much heat for too long will cause the insulation to melt.

If you only need a few scraps of wire you could grab some from old fluorescent lamp fittings. It is something like 0.5mm² and has V105 insulation so if there is a meltdown inside the radio at a later date the insulation will hang on for longer.

I've noticed a big difference in the time it takes to bring solder to melting point on the IFT lugs if I am using the heavy multi-strand copper. Like about six seconds, compared to 1-2 seconds if I'm attaching a finer wire. I reckon that 6 seconds gives too much time for the other end of that lug INSIDE THE CAN to get to, or close to, fuse temperature.

I think next time I'll try using that solid-copper phone wire, and sheath it with spaghetti tubing for extra insulation. The potential is high but the current must be pretty trivial unless there is a catastrophic short, when all bets are off. The thin copper has way more capacity than the coil windings in the IFT. It's just the connections to the IFTs that bother me at the moment, I'll leave the rest with heavy multistrand in place.

Maven

I've noticed a big difference in the time it takes to bring solder to melting point on the IFT lugs if I am using the heavy multi-strand copper

What type of solder are you using? Is it the old tin/lead or the newer lead-free stuff?

I have found that an iron tinned with the lead-free stuff doesn't melt the old lead stuff very well. I still use tin/lead myself -- at least while my stock of it lasts.

I'm using tin/lead resin-core solder. A dab from a flux pen seems to help the spread of heat on old solder, I guess by getting the tip to start sinking in and increasing the contact area.

Anyway, the amount of heat sinking into the copper wire is proving to add significant time to the contact between iron and lug.

Maven

Anyway, the amount of heat sinking into the copper wire is proving to add significant time to the contact between iron and lug.

Are you separately tinning the copper wire first -- that is, before using a dob of solder to join the tinned wire to the tinned lug?

Yes, but with thick multistrand, "tinning" soaks up quite a weight of solder and you can end up with stiff, thick end that won't go through a lug hole. Several of those lugs have more than one wire or lead connected, so re-flowing the joint for another connection (say replacing a component) takes another heavy dose of heat.

Maven

If there's no room in the lug hole due to other wire or pig tails in it, I simply solder across the lug making sure it's a good clean joint.

That is unless I have to replace the part entirely which provides the opportunity to make a clean fresh joint.

I think I've fallen in love with 1mm copper wire and spaghetti tubing.

The 1mm copper gets to solder fuse temperature much more quickly than the multi-strand flex sections I had been using before. It also can be bent and will hold a shape to route tidily around other components. And there are no loose strands hiding under joints or getting mixed up with melted insulation.

For all tasks recently I've been soldering one end of the copper, setting up route to the next point by bending then trimming to length, then sliding on the spaghetti tubing, before soldering the second end.

Spaghetti tubing is also handy to colour-code different circuits. I've taken to adding it to component tails as well as point-to-point connections.

Best of all, I have a lifetime supply of 1mm copper thanks to scavenging a few metres of offcut HV mains earth wire as used in building installations. This consists of six strands of 1mm copper in a sheath.