This might be a simple question. But I am new so be patient with me. If your "rail voltage" is lets say 85vdc, does that mean any capacitors must 85vdc or higher? Very simply "rail voltage" determines your capacitor voltage? And is that voltage determined by the secondary windings on the power supply?
First, it depends on where the capacitors are used. If they are power supply reservoir/filter caps the general rule is to choose a rating about 25% greater than the rails. This is because at power-on there is usually a short surge that may exceed the rating. Also if discharged quickly for any reason the voltage may exceed the rating. To complicate maters, it is not precisely the rail voltage one should look at but voltage that can be present across a cap. This is equivalent to the rail voltage where one terminal is connect to the rail and the other to Ground (0V). If the cap were connected to +85V on one terminal and -85 on the other, you would wnt a rating of 170V + whatever additional margin is appropriate.
In the other direction, a rating much lower than the rails may be acceptable if the max voltage to be seen across the the cap is lower. For example, caps are ofeten used to bypass bias setting diodes. Suppose a pair of 1N4148s are being used for this purpose. the voltage drop is about 1.4V. The noise spikes that are the reason for bypassing are somewhat higher but rarely as great as the rails, so you will often see a cap with a 16-35V rating used for this even if the rails are much higher. However, using a cap with an even higher rating does no harm except to the space required on the PCB and to cost.
In the other direction, a rating much lower than the rails may be acceptable if the max voltage to be seen across the the cap is lower. For example, caps are ofeten used to bypass bias setting diodes. Suppose a pair of 1N4148s are being used for this purpose. the voltage drop is about 1.4V. The noise spikes that are the reason for bypassing are somewhat higher but rarely as great as the rails, so you will often see a cap with a 16-35V rating used for this even if the rails are much higher. However, using a cap with an even higher rating does no harm except to the space required on the PCB and to cost.
Mermprin said:
Yes
The transformer secondary voltage will be measured RMS, so after rectification, the rail capacitor will charge up to 1.414 X RMS voltage, minus the voltage drop across the rectifiers.
A friend had a Musical Fidelity MOSFET power amp, not sure of the model but "Dr Thomas" rings a bell... It had 63v rated caps and I measured the rail volts as +/- 64v.
I had to replace the caps, but they had lasted about 10 years, and I suspect that the failure was mainly due to the heat dissipated by the power amp. They had leaked electrolyte.If the applied voltage is too high, the dielectric breaks down and they go short circuit, usually with explosive results
I would use a good safety margin though, as Sam9 suggests.
It had 63v rated caps and I measured the rail volts as +/- 64v.
A couple of possabilities:
The amp may have had some kind of in-rush limitation such that the caps were never subjected to a turn-on spike.
and/or
You may have measure the rails while quiescent, with a signal present and connected to a load, the rails may have been quite a bit lower (I don't think this applies to Class-A?).
and/or
The datasheets of some caps will show a higher short-term voltage rating to allow for in-rush, etc.
and/or
The maker of the amp may have determined somehow that these particular caps consistently exceeded ratings; if purchased in enough numbers (and at the right price) deals to screen for above spec units can be made with the supplier.
If you are just buying ordinary caps for DIY use, it's best to stick to a conservative selection rule.
A couple of possabilities:
The amp may have had some kind of in-rush limitation such that the caps were never subjected to a turn-on spike.
and/or
You may have measure the rails while quiescent, with a signal present and connected to a load, the rails may have been quite a bit lower (I don't think this applies to Class-A?).
and/or
The datasheets of some caps will show a higher short-term voltage rating to allow for in-rush, etc.
and/or
The maker of the amp may have determined somehow that these particular caps consistently exceeded ratings; if purchased in enough numbers (and at the right price) deals to screen for above spec units can be made with the supplier.
If you are just buying ordinary caps for DIY use, it's best to stick to a conservative selection rule.
This is a good question..
I had this happen to me when I bought my first quantity
of my Apex "Senior" Sub-Amps. They had 80V 6800uf
caps. 4 of them. They were coming back and I wasn't
happy. A customer told me about the rail voltage surge
and the problems they can cause and we replaced all
the rest in stock and now have them ordered with the
new 100 V caps already installed and offer a retrofit
kit to those who have bought the 1st batch...
Steve @ Apex Jr
I had this happen to me when I bought my first quantity
of my Apex "Senior" Sub-Amps. They had 80V 6800uf
caps. 4 of them. They were coming back and I wasn't
happy. A customer told me about the rail voltage surge
and the problems they can cause and we replaced all
the rest in stock and now have them ordered with the
new 100 V caps already installed and offer a retrofit
kit to those who have bought the 1st batch...
Steve @ Apex Jr
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.