|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
101.99.29.215
In Reply to: RE: Tube voltage doubler posted by Eli Duttman on May 02, 2016 at 08:37:35
Ok thanks.
So the doubler stack capacitors must be of sufficient value and you think 100uf is a reasonable value that often works for people.
But given B+ is 1150v-1200v roughly, which 100uf cap that has 700V plus rating (main trains secondary is 500VAC) and still fit into the chasis?
I guess 2 pair of 100uf 450V electrolytic caps in series should do ok, right?
When dealing with this type of supply, have you had issue relating to charging pulses of the bigger caps causing noises?
Anyway, is regulation important in a class A amp which draws about only 90ma-100ma from the power supply? I have always had a preference for very small first cap in my other amps with cLC filter.
Thanks.
Follow Ups:
I came up with that 100 muF. value, after perusing a 5U4G data sheet. A 40 muF. part is given as "typical" in the commonplace FWCT setup. As the inrush current limiter provides a margin of safety, I opted for 50 muF./5U4 plate.Using 100 muF. caps. in the stack of a "full wave" doubler is equivalent to 50 muF. in a FWCT configuration. Cap. generated "hash" should not be beyond your choke's suppression capability.
When really big caps., like the 820 muF. parts found in a "McShaned" H/K Cit. 2, are in a doubler stack, I favor inserting a "hash" filter LC section made from a high current RF choke and a 1000 pF. mica or NPO ceramic cap. between the doubler stack and the main filter choke. Here, RF energy that could pass through the main inductor's winding capacitance is a matter of concern.
BTW, a "hash" filter is never a bad idea. Even if cap. associated noise is not an issue, suppression of garbage riding on the AC mains can only help the cause.
8X Mouser stock # 647-LGN2H101MELZ40, plus voltage equalizing resistors, would make a suitable doubler stack. That part is less than 1 inch in dia. and less than 2 inches tall. Perhaps mimicking Sheldon Stokes, by etching a PCB for those caps., would be a good idea. S/S frequently engineered his cap. PCBs to mount on stand offs attached to power trafo hold down machine screws.
Eli D.
Edits: 05/02/16
The data sheet specs 85 ohms per plate "effective plate supply impedance" for 40uf.
It's not the inrush current but the repeated peak current each half cycle as the cap charges that is the limit here.
The max allowable uf value for the input cap is dependent on the impedance (DCR) of the power transformer.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
I would say that 5U4's are cheap...but that was once said by folks who abused 6DJ8's in pre-amp designs. It just isn't a good idea to beat on them too hard.
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
I agree with the sentiment of not beating the guano out of 5U4s or any other type. However, we have to remember that rectifier data sheet limits were "settled", when PSU 'lytics tolerance was -20/+100 %.
I'm not a simulation guy. I believe that the 1st thing to do is taking all steps needed to prevent arc over, at power turn on time. Of course, arcing is very bad, at any time. JMO, taming turn surge "demons" goes a long way towards obtaining overall good behavior. NTC inrush current limiters can contribute towards satisfactory outcomes.
Eli D.
"...prevent arc over..."
By all means.
"...beating the guano out of..."
Any cap input filter will cause the diodes and the first cap to be beaten.
Maybe not beyond their capability but they are being beaten on, repeatedly.
A critical inductance input choke puts a stop to all that.
The whole supply calms down.
I like calm.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Quite so...but the doubler performance is largely a function of peak repetitive current capacity, and given that tube diodes are at a severe disadvantage to the modern SiC Schottky types.
take the sim data and use it with a few tsp of salt, and be prepared to revise the circuit...:)
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
Here is a sim of the Typical Operation from the datasheet, 450vrms per plate and 85 ohms effective plate supply impedance per plate with the input cap at 40uf and a load current of 225ma.
The peak repetitive current maximum of 800ma., called "current forward repetitive maximum" by the simulator program and "Steady State Peak Plate Current Maximum" on the datasheet, is being exceeded by 11.51ma. repeatedly.
On the other hand if I decrease the DCR of the secondary winding to 50 ohms (25 ohms per half winding...per plate) the peak current goes to 916.89ma.
If the load current is decreased from 225ma. to 100ma. then the peak repetitive current drops to 507ma. with the DCR of the winding at 25 ohms per plate. and 456ma. when the DCR is 85 ohms per plate.
The point is, the DCR of the power transformer winding, the value of the input cap and the load current are all part of the equation when trying to discern if the peak repetitive current rating of the rectifier tube is going to be exceeded.
The capacitor value given in the datasheet is only the max value for the operating condition given. It's not a one size fits all value.
It's not the capacitor value given in the datasheet that shouldn't be exceeded, it's the peak repetitive current maximum rating of the tube that can not be exceeded without causing harm to the tube.
The maximum cap value will be different for each different set of operating conditions.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
" 85 ohms effective plate supply impedance per plate with the input cap at 40uf and a load current of 225ma". 800ma peak repetitive, ouch! Ive juggled input cap values and their ESR in psud and noted the peak repetitive power transformer current. What really hit it home is 1sr hand experience feeling the transformer operating temp. Of a cap input vs choke input supply. It is not subtle. Needless to say I've been sticking to tiny c input supplies, just enough capacitance they snub and save sand rectifiers. JH
It's not front page NY Times news that the full VA capability of power transformers is unavailable, when cap. I/P filtration is employed.
I 2 R heating associated with cap. charging pulses must accounted for. I frequently call for LARGE caps. in the stacks of SS diode doubler B+ PSUs. If the winding being rectified is rated for n mA. RMS, only n/4 mA. is available as DC. TANSTAAFL!
Choke I/P filters are current rich and voltage poor. Cap. I/P filters are voltage rich and current poor. Did you ever get the impression that someone "has it in for us"? ;> )
BTW, if power magnetics are warm to the touch, but are not too hot to remain in contact with, things are OK. If the hand has to be withdrawn, correction(s) are required.
Eli D.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: