10-06-2014, 10:01 AM
That's a good question as to why Philco changed from 8 uF to 4 uF filter caps. Perhaps it was because they were having reliability issues with the 8 uF? It is quite unusual to go to a lower value cap as the ripple will increase. If they wanted to reduce the B+ supply voltage, Philco most likely would have changed the power transformer secondary winding rather than reducing the filter capacitance. I do know that they went back to 8 uF in the next years model 38-116.
Another factor is that the actual capacitance of the early electrolytics was usually higher than the marked value, so the later 4 uF might have been closer to 8 than to 4 uF.
I would still recommend the 8 uF polypropylenes. They have a tight tolerance so their value will not exceed the marked value by more than 5%. If you use 4 uF 500V miniature electrolytics, they will not have a ripple current rating high enough to survive for long in this application.
In general, increasing the value of the first filter cap right after the rectifier will increase the DC voltage and decrease the amount of ripple. How much depends on the DC current drawn from the supply, the higher the current, the greater the effect. Looking at the original 37-116 schematic with the 8 uF filters, the DC voltage readings seem to be right in line with what would be expected, so I don't see why they would want to reduce the B+ supply voltage by using lower capacitance input filters.
Another factor is that the actual capacitance of the early electrolytics was usually higher than the marked value, so the later 4 uF might have been closer to 8 than to 4 uF.
I would still recommend the 8 uF polypropylenes. They have a tight tolerance so their value will not exceed the marked value by more than 5%. If you use 4 uF 500V miniature electrolytics, they will not have a ripple current rating high enough to survive for long in this application.
In general, increasing the value of the first filter cap right after the rectifier will increase the DC voltage and decrease the amount of ripple. How much depends on the DC current drawn from the supply, the higher the current, the greater the effect. Looking at the original 37-116 schematic with the 8 uF filters, the DC voltage readings seem to be right in line with what would be expected, so I don't see why they would want to reduce the B+ supply voltage by using lower capacitance input filters.