Consider a 600 V to 1000 V rated power transistor. Solves the control voltage problem.

100 uF is not a tremendous amount of energy storage, even at 500 V. If the load is continuously connected, why do you need a bleeder? They are becoming more common in MIL requirements - is this for a similarly specialized market?

ak

 

Some hayseed reasoning.

"Yes after reading other Fujitsu relays I found that relays rated for 440VAC are rated for 300VDC max"

440 is about 615 peak. So manuf would probably go 700-800, maybe even 1000 on their insulation.

You have 400-500 volts. For what you are using it for, and for 3 bucks..........I'd throw one in and see.

 

I like ak's suggestion. It looks like you changed out the tube rectifiers. Just use a solid state switch.

To me, it's amazing the voltages they can do now.

 

In this particular line of resistors (flame proof, metal film) you have to use 5 watt size to arrive at 500 volts arc limit.
I could swear I have seen some 2 watt resistors that are rated for 500 volts.
Read the spec sheet to keep this job safe!
http://www.mouser.com/Passive-Components/Resistors/_/N-5g9n?P=1z0wo8tZ1yzeg3vZ1z0z819

Thank you, I will look for this

questilong,

I have no idea of what you are doing. If you are refurbishing an old rig, and you already have HV at the control panel (S2 for instance)........then by all means, get a HV switch.

This used to be common.

But if you are building something new, it's just good safety to keep HV away from panel.

I'm not trying to rag on you.

It's a good project.

Hi I am rebuilding something new from an old design I want to improve on that, while still keeping the sound unchanged (it is a one off, not a future product, it was a request from a friend of mine, I don't want him to die, so I came to you guys).
It is a tube amplifier used by guitarists

That's a good point. Front panel switches should have no chance of killing the customer when they fail.

Yeah BR-549 is raising an interesting point ! Thanks a lot !
S2 is a standard feature of all guitar amplifier, that is why I said I just copied the design (and used the same switch as every guitar amp manufacturer)
I did not think too much about this switch, do you think I should not use it ? (because it is on the front panel, it's not safe at all)

(Everything I design is safety driven, do I didn't miss this point, but the lack of relays made me look for other options)

My main concern was to improve safety of this old design, that is why I wanted to add a bleeder resistor, because this is a tube amplifier and you have to change tubes, I wanted to avoid "arcing" when the user will have to change the tubes by himself.
So I wanted to improve safety as much as I can, but I tough adding a bleeder will put too much load on the power transformer, because it is already "under-speced" power supply (it's a feature for guitarist... they like the "sag" they feel when the power transformer has to push current to the tubes as they play... it is not safe, the power transformer gets hot and it drop it's output from 470VDC to 370VDC !!!!)
SO I wanted to make the bleeder connected only when the power if off

"for S2 it's a Carling SPDT (rated for 6A 250VAC) this on was not a lot of thinking, because this switch is used by Marshall shince the 1960 (for 650VAC)"

That does not make sense to me. Are you using a switch or a relay at S2? If it's a relay........what are the specs?

Hi, it's a switch, it is rated for 10A 250VAC :
http://www.mouser.fr/ProductDetail/Carling-Technologies/2GK51-73/?qs=sGAEpiMZZMsqIr59i2oRcsA3mViJTB/Qs7CiR0G3FcU=

Consider a 600 V to 1000 V rated power transistor. Solves the control voltage problem.

100 uF is not a tremendous amount of energy storage, even at 500 V. If the load is continuously connected, why do you need a bleeder? They are becoming more common in MIL requirements - is this for a similarly specialized market?

ak

Thank you very much sir, power transistor is the solution !!!!!!!!!!!!!!
It is to avoid arcing when the user has to change the tubes, so the load is not continuously connected, it's to allows the user to discharge caps before changing tubes.

I like ak's suggestion. It looks like you changed out the tube rectifiers. Just use a solid state switch.

To me, it's amazing the voltages they can do now.

Thank you very much for this guys, I think it's the solution !!!!!
I will experiment with a fixed bleeder, then if I want to make it switchable, this will be the best technique !
Thank you guys. It was fun

 

Starting with 470 V and 100 uF,
Total energy is 0.5 x C x V^2 = 11 W. This is not the continuous dissipation of the discharge resistor, so normally I would say a 5 W part would be fine in this application. However, the peak power is so high that a 10 W part would be prudent.

For a 1 second discharge time:

3 time constants = 95%. 3 x R x C = 1 second.
3.3K will discharge this to 23 V in 1 second. Peak current is 143 mA. Peak power is 68 W.

5 time constants = 99%. 5 x R x C = 1 second.
2.0K will discharge this to 4.7 V in 1 second. Peak current is 235 mA. Peak power is 111 W.

ak