Discharging Capacitors... (very simple)

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
G

Guest

Guest
I have spent the past few months trying to understand simple Power supplys abd building a few of them
I have used quite a few different designs..
- Gyraf 18-0-18 calrec psu
- group diy psu number 2
- bo hansens 48v phantom psu

I always put a blue LED with a series resistor just before my outlet terminals
One thing I have noticed is that when I power off the blue led takes some seconds to dim (sometimes 5-7 seconds)

Simple newbie question...
Is this the capacitors discharging?
And a stupid question...
If it is a viable way of discharging capacitors - is it sensible to place and LED with a big big resistor it into a High Voltage Valave Power Supply..
The idea being when the LED doesn't shine it's safe to play as all the capacitors have been discharged...
 
[quote author="uk03878"]Is this the capacitors discharging?[/quote]

Yes.

[quote author="uk03878"]is it sensible to place and LED with a big big resistor it into a High Voltage Valave Power Supply.[/quote]

Is it possible? yes. Is it sensible? Hmmm....

Assuming a 300V PSU, the resistor needed would have to dissipate something like 300Vx15mA=~4.5W of heat!!! All that just for an LED? That's anything but sensible in my book...

Put a high value resistor across the output of your HT supply and (like amorris says) break out the meter.

Peace,
Al.
 
> The idea being when the LED doesn't shine it's safe to play as all the capacitors have been discharged...

Or the LED has burned-out. Unlikely, but never trust a single indicator.

At least be sure the LED is "on", and then fades to "off".

> the resistor needed would have to dissipate something like 300Vx15mA=~4.5W of heat!!! All that just for an LED? That's anything but sensible in my book... Put a high value resistor across the output of your HT supply and (like amorris says) break out the meter.

With high-value caps and a DIY need for speedy discharge, 15mA is not an excessive bleeder current. You may as well get some light with your heat.

One other thing to check: with voltages over 300V, in a bright room, or with a leaky LED, the LED may appear "off" while there is still a stinging 50V-100V on the cap. Compare the LED to a meter the first few times and be sure "off" is really a non-dangerous voltage.
 
[quote author="PRR"]With high-value caps and a DIY need for speedy discharge, 15mA is not an excessive bleeder current. You may as well get some light with your heat.[/quote]

Oh, I was assuming the LED was supposed to be there permanently as a "POWER ON" indicator when the unit is working. If you're just momentarily hooking it up to discharge caps when servicing your gear, then by all means.

Peace,
Al.
 
what about a zener/resistor Vreg circuit for this application? I know you are still pissing away the same amount of heat, but using this setup will allow the led to stay on down until it reaches it's conduction threshold (what's a blue led, around 3V or so??) unlike using a resistor alone which as PRR mentioned would turn off around 50-100v..

just a thought..
 
> I was assuming the LED was supposed to be there permanently as a "POWER ON" indicator when the unit is working. If you're just momentarily hooking it up to discharge caps when servicing

In a tube power amp, 15mA bleeder isn't a big deal. My last tube amp had 50mA of bleeder. The original reason was to power the speaker field-coil, but also because if you use a voltage-drop resistor to feed the screens of a 6L6 you need at least 25mA of bleed to handle the fact that a 6L6's screen current can vary wildly. The original 8uFd caps bled-down quick, but the bleeder was handy when I super-sized to dual 470uFd caps. Without the bleeder, the amp would play for 40 seconds after I pulled the plug, and still be holding 200V on the caps when the cathodes were too cool to emit.

So if you want/need a bleeder anyway, and it happens to be about the right current for an LED, why not? You've already planned to waste some heat, why not get some light?

Of course if you build 300V 3mA 1-tube mike-amps, 15mA is an obscene waste.

Let's see: figuring on thumbs, a 300V dozen-mA supply with a dozen-uV of ripple might be 50uFd, 1K, 50uFd. Let's arbitrarily try 100K bleeder. First check power: 300*300/100K= 0.9 Watts, use a 2 Watt part or even two 220K 2W parts: you never want your bleeder to silently fail. Big resistors are cheaper than small funerals. Figure time-constant: 100uFd total capacitance and 100K gives TC=R*C=100K*100uFd= 10 seconds. Still counting on thumbs: in 10 seconds the voltage will drop to one-third of starting voltage. It takes 10 seconds to decay to 100V, 10 more seconds to reach 30V, etc. Even with this 3mA bleeder, you want to wait more than 10 seconds to stick your hand inside.
 
Back
Top