What would be the point? As a smoothing caps, there is nary a difference between a foil cap and a 'lytic. The difference in ESR and DA are negligible in their effects.Tubetec said:
Rectifiers
- Thread starter 80hinhiding
- Start date
Help Support GroupDIY Audio Forum:
Sure does. Can you elaborate? Decreasing Rs reduces sag, which for me means better load regulation. Indeed, line regulation won't change.CJ said:
Hi Abbey,
The reason for using a small first filter cap is that I end up with a lower ht voltage without resorting to resistor voltage dropping .By selecting different octal 5v rectifier tubes in combination with different first filter caps I can get a range of ht voltages from the same transformer .
Maybe less stress on the rectifier tube also , I can put in big 220uf and 330uf lytics after the two chokes that follow to get very low ripple and good regulation .Its was the lower ESR directly after the rectifier tube with foil caps that I was concerned about .4-8uf would have been common values way back in the paper and oil days but I wasnt sure what effect foil vs paper might have.
The reason for using a small first filter cap is that I end up with a lower ht voltage without resorting to resistor voltage dropping .By selecting different octal 5v rectifier tubes in combination with different first filter caps I can get a range of ht voltages from the same transformer .
Maybe less stress on the rectifier tube also , I can put in big 220uf and 330uf lytics after the two chokes that follow to get very low ripple and good regulation .Its was the lower ESR directly after the rectifier tube with foil caps that I was concerned about .4-8uf would have been common values way back in the paper and oil days but I wasnt sure what effect foil vs paper might have.
No; the choke assumes the same role as a resistor, in teh sense that it reduces ripple, but it has much less voltage loss.80hinhiding said:
The bigger the cap, the higher the current peaks the rectifier has to withstand.
A choke offers good ripple rejection with minimum voltage loss. But a similar level of performance can be achieved with resistors, at the cost of the need to increase the rectified voltage in order to compensate the resistive loss, in the context of a piece of gear that has almost constant current requirement, such as a preamp, or compressor, or EQ..
Different situation with power amps.
A choke has a resistance at Dc ,lets say for the sake of argument 500 ohms ,this 500 ohm appears in series with your big smoothing cap limiting the current flowing through the rectifier . At Ac frequencies the choke appears like a very much larger resistance maybe several kilo ohms , your big capacitor has a low resistance to Ac ,so the benefit of a choke is you loose much less DC volts across it than a resistor of for the same reduction in ripple . A choke also stores energy magnetically in the core material so at a time where there's a peak demand for current some of this stored energy helps stop your HT voltage from sagging.
The bigger the value of the choke in henries and the larger the capacitor you use forms a resonant filter capable of supplying larger amounts of current at low frequencies, so you would expect better bass performance in whatever amplifier your supplying .
The above explanation is a very much simplified version of whats really going on , the maths are quite complicated as shown by the other contributors to this post .
The modern version of the Vox Ac30 (cc2)has a switch on the back so you can change the first filter cap after the rectifier tube,22uf which is the traditional value or 47uf ,theres a small difference in hum levels , 47uf being slightly quieter and also giving cleaner bass response than the 22uf.
Hopefully the above explanation gives you some insight into whats going on
The bigger the value of the choke in henries and the larger the capacitor you use forms a resonant filter capable of supplying larger amounts of current at low frequencies, so you would expect better bass performance in whatever amplifier your supplying .
The above explanation is a very much simplified version of whats really going on , the maths are quite complicated as shown by the other contributors to this post .
The modern version of the Vox Ac30 (cc2)has a switch on the back so you can change the first filter cap after the rectifier tube,22uf which is the traditional value or 47uf ,theres a small difference in hum levels , 47uf being slightly quieter and also giving cleaner bass response than the 22uf.
Hopefully the above explanation gives you some insight into whats going on
Magnetics Voodoo ,nah ,its a science alright ,fairly complicated subject in its own right ,theres many different coil configurations ,then you have various different core materials ,shapes and sizes with different properties for different jobs . Transformer designer is more or less its own speciality . CJ who posts here is very knowledgeable on the subject ,he's made great posts of transformer teardowns ,where he takes a damaged transformer ,carefully unwinds it ,noting the wire gauge used and the no. of turns ,then rebuilds it to factory spec.
People never did use costly tube regulators unless they HAD to.
Which mostly means DC-precision work (voltmeters, pyrometers, analog computers).
Audio works fine with very smooth (no hum/buzz) power. A little drift usually makes no difference.
When the 14-pin and especially 3-pin chip regulators came out, and fell in price, suddenly we were using regulators just-because, no-good-reason. I always found that 24VAC CT and C-R-C filter powered audio op-amps real nice, but fashion moved to making way too much DC (bigger PT!) and throwing away the excess to hit some silly number "exactly".
Which mostly means DC-precision work (voltmeters, pyrometers, analog computers).
Audio works fine with very smooth (no hum/buzz) power. A little drift usually makes no difference.
When the 14-pin and especially 3-pin chip regulators came out, and fell in price, suddenly we were using regulators just-because, no-good-reason. I always found that 24VAC CT and C-R-C filter powered audio op-amps real nice, but fashion moved to making way too much DC (bigger PT!) and throwing away the excess to hit some silly number "exactly".
That is true for most discrete designs. But...PRR said:
I would think there is a reason. A majority of opamps are not very overvoltage tolerant, typically limited to +/-18V. Not that they would not work/fry if slightly overpowered, but still, a manufacturer can't take the risk of putting on the market a product that uses components out of their comfort zone * . In order to maintain headroom, designers have to run these opamps as close as possible to the limit, which implies regulation in order to ensure line and load variations are minimized.
* (or can they? cf. Fender Blues Jr. or Phase Linear 700)
Perfectly fine as long as you don't care about published specs, which unfortunately most designers have to, in order to please the mktg dept.
related info, here is a chart showing how much voltage the rectifier tube is going to chew up as your load current goes up,
for solid state devices the curve is gonna be close to a straight line relative to the amount of voltage chomp eaten up by tube rects,
so now we have a way to tune, let's say, a guitar amp. Pedal Steel players want a quick response so maybe a tube with a more horizontal curve would be the ticket, or maybe even a solid state rect if they want ear piercing quickness,
a crunch monkey might want something in the middle of the tube chart, where as the 50's blues junkie is gonna want that 5Y3 with cathode bias for even more mush, no doh on that,
"Upload Folder Full" error, wtf?
for solid state devices the curve is gonna be close to a straight line relative to the amount of voltage chomp eaten up by tube rects,
so now we have a way to tune, let's say, a guitar amp. Pedal Steel players want a quick response so maybe a tube with a more horizontal curve would be the ticket, or maybe even a solid state rect if they want ear piercing quickness,
a crunch monkey might want something in the middle of the tube chart, where as the 50's blues junkie is gonna want that 5Y3 with cathode bias for even more mush, no doh on that,
"Upload Folder Full" error, wtf?
For +/-15V rails at 100mA current and 1000uF, you get 1Vp-p ripple. this without an additional RC. The additional R is a significant source of sag. Every 10 ohms drops the rail voltage by 1V.80hinhiding said:
If you add R=10r C=1000uF, the voltage drops by 1v and ripple drops by about 20dB (/10).
In that case, using a 3-pin regulator is less bulky, equal in terms of cost and offers better line/load regulation, at the cost of having a slightly bigger PT.
Typically, with SS rectifiers, a 24Vac source delivers about 32Vdc, unless the cap is undersized. In fact, the peak voltage is almost independant of the cap value, but ripple increases as the cap value decreases, decreasing the average voltage, but in the absence of additional filtering, the circuit will be submitted to the highest voltage, although the performance will be that of the lowest voltage, i.e. if you have 10V ripple, the performance will be as if the voltage was 22V.80hinhiding said:PRR,
Seems that 24VAC transformer provides too much excess if I'm aiming for 25VDC using an unregulated supply. A 24VAC transformer seems to be a good amount for +-18 or +-15 though. Just wonderin'
I know this is a little off topic from rectifiers, sorry about that.
A
if there is no load, that filter cap will charge to peak voltage, so 26 VAC x 1.414 = 36.764 minus whatever the cap is leaking.
When the only thing that limits peak current is the DCR of the power transformer, the head capacitor value must not exceed a certain value. With 500r (or a choke, or both) before the head cap, this is not an issue.80hinhiding said:
Adam, I suggest you install PSUD2 from duncanamps and read the help info, and set up your power supply circuit using measured resistances from your power transformer.
That app allows a good estimate of the turn-on peak and continuous rms current levels that can then be referenced to a valve diode's datasheet. That app provides some help to identify when datasheet levels are exceeded, and also helps to identify LC induced start-up ringing, and LC resonance frequency and damping when a step change in output load current is applied.
That app allows a good estimate of the turn-on peak and continuous rms current levels that can then be referenced to a valve diode's datasheet. That app provides some help to identify when datasheet levels are exceeded, and also helps to identify LC induced start-up ringing, and LC resonance frequency and damping when a step change in output load current is applied.
