Seeking Valves

[Nowhere-Man]

I've been reading through the thread and an a little worried. You say about safety but I feel that some of the precautions you seem to be taking could actually be a hazard. In one instance you say about using heatshrink over a resistor. This is really silly, the resistor is rated for natural cooling in air, wrapping in heatshrink will cause the resistor to overheat and could result in a fire. Like I've said before try to get the right valve or an equivalent. You'll pull more current off silicon diodes too, a fault downstream could up the current to blow your mains transformer well before you blow a fuse.

It's not totally wrapped. The heat shrink I used is the industrial type and also has a large diameter. It's from the welding shop where I live. Air can still get in but the soldered joints and HT wire is not able to short out to chassis. I also tested the resistor for a while (as well as everything else) and haven't found any sign of appreciable heat. Bear in mind this is a pretty large radio and there's plenty of air and space, plus the resistor is a modern 7 watt. I explained to Maplins what it's use would be.

On the other hand, my small Ultra does get very hot. It's a much smaller Bakelite radio and the tubes are cramped together. I've seen nothing like that kind of heat in my GEC. I don't run the Ultra for too long as it's 1947 and quite small.

However, having said all that, I agree after all this hard work I don't want to take any risks so a proper tube rectifier is probably a good idea. Probably I can get one for a tenner which isn't a King's ransom.

[Murphyv310]

Hi.
Never ask maplins staff about anything related to valve technology, they haven't a Scooby. These forums are here to help, dispel inaccuracy, put you on the right road etc. One thing though is to heed the advice given, years of doing this for a hobby and a living and giving advice onto deaf ears is irritating to say the least.
I'm not though singling out anyone in particular.

[Nowhere-Man]

Like Tevor, I too am worried. Heatshrink over a high power reisstor is a very bad idea. If the resistor gets really hot as they are designed to do the heatshrink won't survive. At best there will just be a nasty smell, at worst it may not be as flameproof as you hoped.

Fuses look simple but are a complicated subject. A 1 amp fuse doesn't blow at 1.1 amp. The more current you take, the quicker it blows. This page will tell you a lot more than you need to know: http://wiki.diyfaq.org.uk/index.php?title=Fuse The important point is that fuses will not protect against small, long term, overcurrent.

In many cases there's no problem if you replace a valve rectifier with silicon diode(s) and a resistor. Yes it can stress the capacitors in the set more than a valve because full HT is available immediately power is switched on and before the valves start to conduct and drag the HT down a bit. Yes you can get higher curernts through silicon under fault conditions. Thes are often acceptable risks. We're here to help you make sensible judgements.

You've clearly been struggling a bit with soldering. We all had to start somewhere and I'm sure most of us made a hash of it at first. Presented with a crowded chassis, wires wound round tagstrips and old solder it can be tricky to do a good job. It can be hard to melt the solder at all at first. One trick is to run a small amount of new solder between the iron and the old joint. Once the old joint is fully molten you can use a solder sucker to get rid of a lot of the old solder. To save disturbing a possibly delicate tag such as on a valveholder you can remove a component by cutting its leads near the body of the component. Since modern repalcements are usually smaller it's then quite easy to join thenew component to the old leads. The purists can make some very neat mechanical joints between new and old leads. I'm not a purist

As I said earlier, we all had to start somewhere and we've all made mistakes. Those of us who are still here obviously didn't make any fatal ones We like to help those who are less experienced but it can be hard to do it from verbal descriptions and even from photos. There's nothing like learning directly from somebody with experience sitting next to you but this isn't always possible.

O.K. a few words about heat-shrink and resistors as today I tried out a simple experiment. Before I continue, I'd also like to say I agree with your comments on fuses. I did fit 2 amp fuses to all my radios and I agree it does provide some safety yet no guarantee. While I was at it, I wired in an RCD closer to my supply and that should trip out at 0.03 amp.

Now, here's what I did with the heat shrink. The results are open to debate: The stuff I used is quality industrial shrink. Basically I loosely wrapped the upper bar of my soldering iron with this shrink and switched on. I allowed the iron to get extremely hot. And I mean, hot. I then unplugged the iron and allowed to cool a bit. I carefully scraped and removed the heat shrink from the iron. It was way too hot to touch. There was no sign of smoulder, blackening or burning.

Therefore, it would see any danger would have to relate to the actual resistor body and not the shrink. On that score, I really have no idea. I can only assume if the resistor heated up internally due to lack of air around the body, there would be an issue. If anyone knows specifically any heat stats for these ceramic resistors I.e. internal heat tolerances, I'd be interested to know. How do they compare with vintage resistors? I've seen damaged resistors due to heat and have replaced them at times.

And to put minds at ease, I did decide it would be a good idea to allow full air to get to the resistor body, just to be safe and cautious. I'll do that tomorrow. I did today run the radio for a while and I took the temps of the resistor. It was hot but nothing comparable by any standards to the soldering iron test temps. As I say, I simply don't know whether there are risks of internal heat within the body of the component itself. All I can say is if the body of the resistor got way too hot, the heat-shrink wouldn't burn but maybe the resistor could itself be damaged.

I don't ever run my valve radios unsupervised. I enjoy using a 1947 Ultra at nights on my boat but I'd never go off the boat and leave it on (or even plugged in). Even despite my best efforts, it's still a Forties set and very old.

[ppppenguin]

While I was at it, I wired in an RCD closer to my supply and that should trip out at 0.03 amp.

While an RCD is a very good thing to have in the mains supply of an old radio I'm sorry to say that represents a misunderstanding of what an RCD can do. An RCD does NOT trip on overcurrent. What it does is detect whether the current in the live and neutral are the same. If they differ by more than the (usually) 30mA rating then it will trip. If the current that's going astray is passing through you, then it can save your life. If you have a faulty mains transformer (or rectifier or smoothing cap for that matter) and things start to get hot in the set then an RCD will not trip. Except if things get bad enough for the insulation between the primary and frame of the mains transformer fails, and only then if the frame of the transformer is earthed.

I'm sure there's a tutorial thread on RCDs somewhere on GVR.

The whole subject of RCDs, fuses and isolating transformers is quite complex and prone to a lot of misunderstanding.

Additional special note about electricty afloat. If you're connected to a shore based supply then that should be protected by a 30mA RCD as well as overcurrent protection by fuse or MCB. If you're using an onboard supply (generator or inverter) the generator neutral should be connected to the hull (assuming it's metal) and that will give an excellent earth. Then protection can be as per usual for a domestic supply. If the neutral is NOT connected to the hull then this is a specialised arrangement (known as a IT supply, nothing to do with information technology) which could be very hazardous.

[Nowhere-Man]

While I was at it, I wired in an RCD closer to my supply and that should trip out at 0.03 amp.

While an RCD is a very good thing to have in the mains supply of an old radio I'm sorry to say that represents a misunderstanding of what an RCD can do. An RCD does NOT trip on overcurrent. What it does is detect whether the current in the live and neutral are the same. If they differ by more than the (usually) 30mA rating then it will trip. If the current that's going astray is passing through you, then it can save your life. If you have a faulty mains transformer (or rectifier or smoothing cap for that matter) and things start to get hot in the set then an RCD will not trip. Except if things get bad enough for the insulation between the primary and frame of the mains transformer fails, and only then if the frame of the transformer is earthed.

I'm sure there's a tutorial thread on RCDs somewhere on GVR.

The whole subject of RCDs, fuses and isolating transformers is quite complex and prone to a lot of misunderstanding.

Additional special note about electricty afloat. If you're connected to a shore based supply then that should be protected by a 30mA RCD as well as overcurrent protection by fuse or MCB. If you're using an onboard supply (generator or inverter) the generator neutral should be connected to the hull (assuming it's metal) and that will give an excellent earth. Then protection can be as per usual for a domestic supply. If the neutral is NOT connected to the hull then this is a specialised arrangement (known as a IT supply, nothing to do with information technology) which could be very hazardous.

My understanding of RCD's is that under normal conditions the voltages between live and neutral cancel and the vector sum is zero. If there is some leak between live and neutral, the vector sum isn't zero. In the device is a toroidal transformer that will detect and trip the circuit.

On boats, there are significant misunderstandings of electrical systems. I am currently rewiring a boat the electrical system of which has had me baffled. The shore line supply was actually taken to an inverter/charger and then the A.C. from the inverter was taken to the Consumer Unit. What happened was the inverter blew (I guess the internal fuse) and that disconnected the entire supply. I couldn't get the inverter out to inspect the fuse as it has been cabled into a small cupboard and couldn't be freed up. So, I took it out of circuit and connected shoreline straight to the Consumer Unit.

The kicker was the lighting system functioned as follows: DC from battery bank fed to Inverter/Charger and emerges as 260 A.C. 260 AC then fed to Consumer unit and to lighting MCB. Outgoing circuit wire then fed to a drop-down transformer and back to 12 volts again. With 6 amps current draw on the batteries just to have the inverter on, result is a huge loss of battery amps reserve. What could have been simpler than just taking 12 volts from the batteries to the halogen lights?? Household cable was also used but on boats we use arctic cable which is more resistant to vibration (domestic cable isn't allowed on boats).

[ppppenguin]

My understanding of RCD's is that under normal conditions the voltages between live and neutral cancel and the vector sum is zero. If there is some leak between live and neutral, the vector sum isn't zero. In the device is a toroidal transformer that will detect and trip the circuit.

RCDs, as their name suggests, measure currents, not voltages. The toroid is a current transformer that has bothe live and neutral as its primary. If the difference between the L and N currents exceed the rating then the current in the secondary will cause a trip. I'd need to think about whether a vector or scalar difference is involved.

As for "leak between live and neutral" this is normal current consumption! I think you meant between live or neutral and earth.

If you're working with an independent supply such as an inverter and the output is floating (ie not referenced to earth) an RCD is completely useless.

[Nowhere-Man]

My understanding of RCD's is that under normal conditions the voltages between live and neutral cancel and the vector sum is zero. If there is some leak between live and neutral, the vector sum isn't zero. In the device is a toroidal transformer that will detect and trip the circuit.

RCDs, as their name suggests, measure currents, not voltages. The toroid is a current transformer that has bothe live and neutral as its primary. If the difference between the L and N currents exceed the rating then the current in the secondary will cause a trip. I'd need to think about whether a vector or scalar difference is involved.

As for "leak between live and neutral" this is normal current consumption! I think you meant between live or neutral and earth.

If you're working with an independent supply such as an inverter and the output is floating (ie not referenced to earth) an RCD is completely useless.

Some boaters have run earth to the battery negative and then to an earth point on the hull - not  a good idea but still done. I always endorse a floating DC circuit and a hull ground for the mains earth.

[Alan Douglas]

Here they're called Ground Fault Circuit Interrupters (GFCI or just GFI) and I thought they tripped at 6mA though mains voltage here is of course only 120V.

A fair percentage of heat in a resistor exits through the leads.

[ppppenguin]

In the UK they are called RCD, RCCD or sometimes the older term ELCB (Earth leakage circuit breaker). In the US GFI/GFCI is the same thing. There are also voltage operated ELCBs which are very different beasts and totally obsolete for domestic and ordinary commercial use.

The mains voltage has nothing to do with the trip current. 30mA is useful for many purposes. 10mA is the most sensitive usually available in the UK and is good for protecting against shock in high risk areas. They are also a pain in the neck due to spurious trips. I've not heard of a 6mA RCD.

100mA and 300mA RCDs are sometimes used for protecting large installations. These often have a time delay before they trip because there is no inherent discrimination with RCDs. If you have a 30mA RCD and a 100mA RCD cascaded and you get a 100mA fault then either or both could trip. This is unlike fuses and MCBs where the lower rated device will normally trip on a fault if the system is designed correctly.

[Nowhere-Man]

In the UK they are called RCD, RCCD or sometimes the older term ELCB (Earth leakage circuit breaker). In the US GFI/GFCI is the same thing. There are also voltage operated ELCBs which are very different beasts and totally obsolete for domestic and ordinary commercial use.

The mains voltage has nothing to do with the trip current. 30mA is useful for many purposes. 10mA is the most sensitive usually available in the UK and is good for protecting against shock in high risk areas. They are also a pain in the neck due to spurious trips. I've not heard of a 6mA RCD.

100mA and 300mA RCDs are sometimes used for protecting large installations. These often have a time delay before they trip because there is no inherent discrimination with RCDs. If you have a 30mA RCD and a 100mA RCD cascaded and you get a 100mA fault then either or both could trip. This is unlike fuses and MCBs where the lower rated device will normally trip on a fault if the system is designed correctly.

It's not a particularly complicated device. I recently wired from shoreline straight to the RCD box and am still baffled as to why the former electrician took the shoreline to the inverter/charger (and also used domestic cable).

The point where a problem may arise is that tube radios aren't grounded like normal appliances. Of course, we all know this to be the case. I imagine if you did ground the chassis to the RCD box, it would simply keep tripping as voltage tends to find its way onto the chassis, even if small.