I need a bunch of toggle switches of various types for a project. These will have to handle the full DC voltage of the power supply, ie 400-450VDC but I can't find anything at that rating.
Suppose I found a switch rated for 10A and 240VAC (= 2.4kW) and a dielectric withstanding voltage of 1.2kV. Would that also be safe to use with 480V DC up to 5A (=2.4kW)?
Note I don't need to switch anything hot. The switches will only ever be toggled when the power is off.
Suppose I found a switch rated for 10A and 240VAC (= 2.4kW) and a dielectric withstanding voltage of 1.2kV. Would that also be safe to use with 480V DC up to 5A (=2.4kW)?
Note I don't need to switch anything hot. The switches will only ever be toggled when the power is off.
Switching dc is tough because you break the circuit and an arc forms and has to be extinguished. AC makes it easier as the voltage crosses zero volts where as the dc supply does not. When the switch is closed and you have a high voltage on it the contacts hardly notice it. As you seem to realize the metalic bits need to be adequately isolated. If the switches are thrown cold the contacts are no more stressed than a wire under a screw terminal as far as I can see.
PLEASE DON'T MESS WITH 480V!
Most AC switches are rated for only 32VDC, DO NOT use such a switch outside its ratings. If it can be thrown when energized, it will be. The switch can control the coil of a properly DC rated relay. These are not commonly available, or cheap, however.
Thanks for the warning but they absolutely won't be thrown when the circuit is hot, if that makes a difference.
This is an experimental shared power amp build which will only be used by me. I need to choose tube/ss rectifier, choke/resistor, filter cap values - that sort of thing - but the switches won't be touched when the circuit is in use.
This is an experimental shared power amp build which will only be used by me. I need to choose tube/ss rectifier, choke/resistor, filter cap values - that sort of thing - but the switches won't be touched when the circuit is in use.
Hi,
The switches are no big deal if only used when the power is off.
The rating is what they can interrupt, not what they can flow,*
so a 10A mains switch will be more than fine passing 5A DC.
rgds, sreten.
* obviously much higher than the interrupt rating, and is
simply a current limit (loads) as no voltage is involved.
AC switches need derating in voltage to interrupt a DC
current the same as the AC current rating, but it is
absolute nonsense AC switches are mostly 32V DC.
DC voltage depends in the current to be interrupted.
The switches are no big deal if only used when the power is off.
The rating is what they can interrupt, not what they can flow,*
so a 10A mains switch will be more than fine passing 5A DC.
rgds, sreten.
* obviously much higher than the interrupt rating, and is
simply a current limit (loads) as no voltage is involved.
AC switches need derating in voltage to interrupt a DC
current the same as the AC current rating, but it is
absolute nonsense AC switches are mostly 32V DC.
DC voltage depends in the current to be interrupted.
Last edited:
I would use a soldering iron to do the switching. That way it is much less likely that switching will take place while the device is on. Scattering switches around a power circuit more or less guarantees that you can't build it right (e.g. twisted wires, good spacing and grounding) so it won't work right - at the best you could make bad deductions from your experimental results.
Why not use the traditional method? First design it. Then build it. Then debug it. If necessary, build it again using what you have learnt.
Why not use the traditional method? First design it. Then build it. Then debug it. If necessary, build it again using what you have learnt.
The Matchless Spitfire, Matchless Lightning, and Goldtone GA15 have almost identical EL84 power amps. Same with the 5E3 Tweed Deluxe and Divided by 13 CJ11 with their cathodyne PI 6V6 power amp. These are all 15W-ish amps which can use the same set of transformers... see where I'm going? It's possible to refactor all the shared bits into a "hub" box containing trannies, power amps, and rectifiers which would connect to separate preamps. It's a 5-amps-for-the-price-of-one kind of deal.
However, although parts of the circuits are nearly identical, there are still a bunch of things you have to put on a switch such as ss/tube rectifier, choke/resistor, filter cap options, and others. In use, the hub box would be configured to create a specific amp circuit and only then switched on.
So, unfortunately I can't do it the traditional way with some experiments and a final, fixed circuit. The switched options are an integral part of the design.
However, although parts of the circuits are nearly identical, there are still a bunch of things you have to put on a switch such as ss/tube rectifier, choke/resistor, filter cap options, and others. In use, the hub box would be configured to create a specific amp circuit and only then switched on.
So, unfortunately I can't do it the traditional way with some experiments and a final, fixed circuit. The switched options are an integral part of the design.
You used to be able in the 1960's to buy a knife switch in any hardware store. They had a break length of 2.5", which would quench about any arc. My father used one to break the twinlead wire from the TV antenna, which of course was subject to lightning.
The modern equivalent is the FET transistor, which comes in ratings up to 1000 volts for about $1. You'll need to provide a power supply limited to the gate maximum voltage, say 12 vdc, to switch the gates. FETs are known to fail shorted through, so protect the gate circuit with zeners or something. I'm installing some FETs now on NEMA CE laminate (drilled) for a Class AB transistor amplifier rail voltage shutoff.
The modern equivalent is the FET transistor, which comes in ratings up to 1000 volts for about $1. You'll need to provide a power supply limited to the gate maximum voltage, say 12 vdc, to switch the gates. FETs are known to fail shorted through, so protect the gate circuit with zeners or something. I'm installing some FETs now on NEMA CE laminate (drilled) for a Class AB transistor amplifier rail voltage shutoff.
Sounds like a very interesting project. The FET idea is a nice one, but for 'change over' switching (such as ss / tube rectifier, etc.), would you need to include some simple digital logic, so that when one FET is conducting the logic guarantees that another is blocking? Sounds very professional, but possibly a little complex.
Do you want the amp to be just a test-bed on the bench, or a real everyday working amp?
As sreten says, the 240V mains toggle switches would be fine, as long as you can be sure they are not switched when the amp is powered up. If it is to be a working everyday amp, could you cover up the switches with some kind of screw-on lid?
It should be very informative to try out all those circuit variations in a single amp - let us know how the different variations sound when you have it working.
Do you want the amp to be just a test-bed on the bench, or a real everyday working amp?
As sreten says, the 240V mains toggle switches would be fine, as long as you can be sure they are not switched when the amp is powered up. If it is to be a working everyday amp, could you cover up the switches with some kind of screw-on lid?
It should be very informative to try out all those circuit variations in a single amp - let us know how the different variations sound when you have it working.
Last edited:
The amp(s) will be used in a home studio so I can get away with some extra complexity which might not work so well in a gigging amp. At the same time I'm trying to avoid a full-blown prototyping rig. The user-interface can't be simple exactly but I'd like to keep it as simple as possible with just the essential options needed to re-create specific circuits - my favourite 15W amps.
No-one will be using it much except me and even then I'll still be around to warn them what not to do.
I need to update some schematics & layouts but once that's done I'll start a new topic with all the details.
No-one will be using it much except me and even then I'll still be around to warn them what not to do.
I need to update some schematics & layouts but once that's done I'll start a new topic with all the details.
a 50v winding would blow a fet gate right through without regulation. Most fets are limited to 30v on the gate (to drain voltage), some fets have an internal gate protection zener diode. A discrete zener and resistor can make a regulated power supply out of a voltage too high, but having had a solder failure on the regulation zener that caused my first fet to blow, I reccomend a wall transformer making 12 VDC or so. Available at Salvation Army/goodwill resale for about $1.
A one out of 5 switch can be a 1. washingmachine or electronic organ multibutton mechanical interlock switch 2. a cmos binary counter IC & decoder IC and an up pushbutton momentary, with a one shot timer IC for switch debounce 3. a rotary centralab switch like for a PAS3 preamp. For the IC idea, debounce is critical, the counter will count every time the switch bounces, which is typically 2 to 10 times.
fets conduct when the gate to drain voltage is > 7v or so (rating depends on fet you buy) and no current so very compatible with cmos. DIP IC's fit best on a project board from mcmelectronis, $2, and TE connectivity phosphor bronze sockets, $1.5 each.
You want Nfets, enhancement mode, the most common ones. To use the fet drain as the B+ to tube plates, you could use optoisolator ICs to float the gate drive up off the top voltage, instead of down at ground the way switcher power supplies work.
A one out of 5 switch can be a 1. washingmachine or electronic organ multibutton mechanical interlock switch 2. a cmos binary counter IC & decoder IC and an up pushbutton momentary, with a one shot timer IC for switch debounce 3. a rotary centralab switch like for a PAS3 preamp. For the IC idea, debounce is critical, the counter will count every time the switch bounces, which is typically 2 to 10 times.
fets conduct when the gate to drain voltage is > 7v or so (rating depends on fet you buy) and no current so very compatible with cmos. DIP IC's fit best on a project board from mcmelectronis, $2, and TE connectivity phosphor bronze sockets, $1.5 each.
You want Nfets, enhancement mode, the most common ones. To use the fet drain as the B+ to tube plates, you could use optoisolator ICs to float the gate drive up off the top voltage, instead of down at ground the way switcher power supplies work.
I ended up choosing this Kycon connector for 6.3V heaters, and a Universal Mate-N-Lok connector to pass B+ around.
Full details of the project in this other thread.
Full details of the project in this other thread.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.