05-01-2012, 12:13 AM
Let's just run through the basic factors that determine the ohmic value of a bleeder resistor in a valve PSU such as this.
Its primary function is to provide a discharge path for the filter capacitors when the a.c. supply is switched off. It also provides a minimum load on the PSU, so that in the event of the load becoming disconnected (whilst the rectifier is delivering volts to the filter), the output voltage does not soar to a possibly dangerous value. Depending on the L and C values, it also provides a minimum value of load resistance that will optimise the filter's performance. Of these three factors, in most PSUs of the type we have here, it is the capacitor discharge function that is usually the most predominant and important. Ideally, we want a low enough ohmic value for that resistor so that the filter capacitors are discharged as fast as possible. On the other hand, the lower that value, the more current 'goes to waste' from the rectifier - and more heat is dissipated in that resistor. A given rectifier valve has a maximum value of mean current, so that sets one limiting factor. Then there is the consequent increase in the PSU transformer temperature and the rectifier valve to consider. And finally, the greater the bleed current, the greater the ripple on the HT rail and the lower the d.c. voltage of that rail becomes. So a compromise has to be found. An appropriate value can be calculated, but I usually find that trying a few values within a limited range, (that range based on experience), is usually faster and a lot easier. And in fact, I did try 47k-ohms: it produced a current demand on the rectifier that was well within its capabilities - but I considered that the ripple was excessive, the HT drop excessive and the discharge time too short in relation to those factors. So I traded discharge time against HT drop and settled on 100k-ohms.
Since wire-wound resistors are much more reliable in maintaining their nominal ohmic value than carbon composition and other types, I always use a wire-wound R for this task - and make sure that it is mechanically secure where its lead-out wires join the rest of the circuitry. A bleeder resistor is a component that you fit and forget! And I always use a resistor with a very generous wattage rating - purely to maximise reliability. The very last thing you need in a HV PSU is an open-circuit bleeder - a false sense of security - which could be fatal. :omg:
Al.
Its primary function is to provide a discharge path for the filter capacitors when the a.c. supply is switched off. It also provides a minimum load on the PSU, so that in the event of the load becoming disconnected (whilst the rectifier is delivering volts to the filter), the output voltage does not soar to a possibly dangerous value. Depending on the L and C values, it also provides a minimum value of load resistance that will optimise the filter's performance. Of these three factors, in most PSUs of the type we have here, it is the capacitor discharge function that is usually the most predominant and important. Ideally, we want a low enough ohmic value for that resistor so that the filter capacitors are discharged as fast as possible. On the other hand, the lower that value, the more current 'goes to waste' from the rectifier - and more heat is dissipated in that resistor. A given rectifier valve has a maximum value of mean current, so that sets one limiting factor. Then there is the consequent increase in the PSU transformer temperature and the rectifier valve to consider. And finally, the greater the bleed current, the greater the ripple on the HT rail and the lower the d.c. voltage of that rail becomes. So a compromise has to be found. An appropriate value can be calculated, but I usually find that trying a few values within a limited range, (that range based on experience), is usually faster and a lot easier. And in fact, I did try 47k-ohms: it produced a current demand on the rectifier that was well within its capabilities - but I considered that the ripple was excessive, the HT drop excessive and the discharge time too short in relation to those factors. So I traded discharge time against HT drop and settled on 100k-ohms.
Since wire-wound resistors are much more reliable in maintaining their nominal ohmic value than carbon composition and other types, I always use a wire-wound R for this task - and make sure that it is mechanically secure where its lead-out wires join the rest of the circuitry. A bleeder resistor is a component that you fit and forget! And I always use a resistor with a very generous wattage rating - purely to maximise reliability. The very last thing you need in a HV PSU is an open-circuit bleeder - a false sense of security - which could be fatal. :omg:
Al.
