|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
207.7.189.118
I've wondered this since my first encounters with B+ supplies that use voltage doublers. They do not appear to be any more efficient cost-wise, but I presume that there must be technical advantages to using them - many of the great amp manufactures have used them over the years.
Thanks for your thoughts.
Adam
If you find yourself frequently repeating the phrase, "garbage in, garbage out," you may need to rethink your system.
Follow Ups:
...
If you find yourself frequently repeating the phrase, "garbage in, garbage out," you may need to rethink your system.
Adam,
The rectifier winding should be capable of an AC RMS current not less than 4X the DC draw.
Eli D.
...
If you find yourself frequently repeating the phrase, "garbage in, garbage out," you may need to rethink your system.
Gentlemen,
Think about parts availability in the early 1960s, when Hegeman, Marantz, et al developed their "classic" designs. The available SS diodes were nowhere near as tough as the stuff currently available. Also, the best available diodes were costly, back then. 4X diodes, in series connected pairs, wired as a Greinacher doubler got by the "bean counters". 8X diodes, in series connected pairs, wired as a full wave bridge would have caused accountant indigestion. FWCT topology was not in the running, given the PIV limits of the day.
"Follow the money" is sound advice in MANY situations.
Eli D.
Voltage doubler circuits were around in the early days when high capacitance electrolytic voltage ratings did not reach 500VDC. But there were 250VDC ones.
-Kurt
hey-Hey!!!,
That only applies to the doubler input caps; the final B+ still needs the higher voltage, and to get it one needs the HV caps or series-connected pairs. Heathkit did that on the DH-rectified W5m for example, as well as Sheldon Stokes on his cap boards for the Dynaco amps.
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
But when solid state rectification was the new in thing, many high power tube amps just used two large electrolytics from the doubler with no more following filter caps.
Of course there are two kinds of doublers as I think you know: full wave and half wave. Full wave doublers keeps the DC rectified voltage fed with each phase. This is not too far off from full wave center-tapped rectification.
The example I recall was the Music Reference RM-9 that uses a full wave doubler and no more succeeding filter components for the final output tubes. I believe they were 1000 uF 300V caps or thereabouts stacked on top of each other.
I have a versatile "1/3 wave" tripler in my DIY amp with the last stack of caps being 3 1000 uF 250V. The half wave doubler will need the last cap capable of the total end voltage unlike the full wave doubler needing no extra voltage. The tripler will need 200V, 350V, and finally 500V caps to reach a 500V final output voltage. The N-tupler voltage supply as used in electrostatic speakers and headphones need many stages where each succeeding stage of multiplication has to increase its needed voltage rating. The end capacitor may only be 1nF at 3000V as power is not taken from this unloaded supply.
-Kurt
Why use doublers? so you can use old scope transformers to power tube amps of course!"Of course there are two kinds of doublers as I think you know: full wave and half wave. Full wave doublers keeps the DC rectified voltage fed with each phase. This is not too far off from full wave center-tapped rectification."
Not too far off indeed, assuming cap input (choke input is different). By tuning the capacitor values, you can get the same ripple, regulation, and response time out of each (given the same transformer and inductor) as shown above. Series windings with FWB and parallel windings with the doubler are nearly identical. Nearly identical peak charging current per winding of about 750mA also.
Edits: 09/18/09
hey-Hey!!!,
I have a bunch of 'scope TX's, and I've yet to see one with parallelable windings( save for the big 6.3V heater coils ). Besides that, with a series-ed set of coils, one can potentially use a choke-input filter from a bridge( a +critical current one), that draws less peak current. Not judging the doubler, just pointing some features out...:)
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
The comment about scope transformers is they seem to be low voltage high current and suitable for doubler service.
Parallel windings was for the sake of comparing the doubler to cap input, apples to apples (analogy only, I'm not talking about fruit here ;-)
Agreed that choke input has lower peak charging current by a huge factor.
Here's another thing to think about. The DCR of the power transformer is multiplied by the reciprocal of the conduction angle (roughly) for it's contribution in the total Zout of the power supply. So choke input supplies have a small multiplier (like 1.4?) but the above example C-input presents ~7.5X the power tranny DCR at it's output terminals and the doubler presents ~15X it's power tranny DCR at it's output. So even though the doubler starts with a low DCR coil, the net Zo contribution is the same as an equivalent amount of iron and wire in a bridge configuration delivering the same output.
With a critical inductance choke input one should be able to get away with a higher DCR power transformer, and get equivalent regulation at the output.
The problem I have with critical-L supplies is they are difficult to tune for nice well damped but quick response as in the above examples. Not impossible but difficult. Also without a swinging choke it's difficult to accomodate class AB.
Cheers,
Michael
"The problem I have with critical-L supplies is they are difficult to tune for nice well damped but quick response..."
The more fundamental problem with choke supplies for audio amplifiers is that chokes represent a constant-current source. As a result, high-current transients cause choke output voltage to decrease. Maintaining non-sinusoidal waveform fidelity requires the constant-voltage energy storage of capacitors.
-Kurt
Hi,
The big big advantage by far in my opinion, and perhaps most fully employed by Harmon Kardon, is to have ultra low DCR high voltage windings, typically, under three Ohms DCR ( HK Duece, HK 250, many HK tube receivers of the early era.)
I find a "stiff" power transformer is perhaps the most overlooked (and often lacking) performance gain area in tube audio, for many years now.
Your power trannie IS the heart of your amplifier.
The sub three Ohm DCR power trannie, in practice, more than makes up for the inferior rectification function one obtains with a doubler, (as compared to having a full wave bridge rectifier, which is inherently superior way to rectify.)
Now you know.
Jeff Medwin
Are any of the existing custom winders willing to appease such a request? Who are they? Most of them emphatically state.. " NO DCR REQUESTS"....
Edits: 09/18/09
Ha, I can see why transformer winders might get annoyed :-) Fortunately you don't need to pick DCR... If you request a transformer for voltage doubler purposes at proper current and voltage, it's basically guaranteed to be in the single-digits by necessity of design. I have a pair of Edcor XPWR070 power transformers that we spec'd some time ago, which are rated at 165V 1A. They have 3.4R secondaries, which is good enough for monoblock Citation IIs if you don't mind the slightly lower voltage.
Edits: 09/18/09
Hi Jon,
165 volts times 1 A. is 165 VA.
My favorites in 2009 are in the 800+ VA range !!
Jeff
If you want a 35 lbs transformer, 800VA is always possible, although a gratuitous waste of natural resources ;-)
Hi Jon,
It is actually about 18 pounds, with multiple TCSS lead outs in key places, and Wonder Solder to attach them. I find this to be the best compromise solution known to me, after over three decades of working with low DCR power trannies.
The P.T's multiple TCSS lead outs are TOTALLY audible to me, in SE builds. I presently consider this necessary, in a good SE amp design.
Jeff Medwin
Half the number of turns on the secondary, but you do need to allow for four times the current (because of the low efficiency of the doubler. You end up with very low secondary resistance, as Jeff says.
Does a low step-up ratio typically correlate with low PT DCR?
Please explain the word "stiff" as you used it.
Thanks,
Adam
If you find yourself frequently repeating the phrase, "garbage in, garbage out," you may need to rethink your system.
Hi,
Q: "Does a low step-up ratio typically correlate with low PT DCR?"
A: Not always. Few manufacturers and designers got it right, but Harmon Kardon did - in its day. My friends HK 250, a nice amp, I measured at 2.63 Ohms DCR in the HV secondary !!
It can correlate, because it doesn't need a long length of wire, (with inherent resistive losses) a lower number if turns is needed around the core, to get the (lower, secondary side) target voltage. Besides the wire being shorter, there may be room in the transformer's open window, to employ heavier gauge wire, (with less resistive losses).
Q: "Please explain the word "stiff" as you used it."
A1: This happens when you drink too much Jack Daniels, in too short a time. But that is not the real answer!!
A2: The transformer is SO powerful, in voltage and current capability, that (the continuous transient nature of music) will not disturb it.
In cars, it is similar to : "there is no substitute for cubic inches".
The power trannie is the HEART of the audio amp, everything comes from it, and it is better when overbuilt and low in DCR, high in current capability, relative to the circuit's needs. Think in dynamic terms, and as in "instant recovery".
Cheers. The power supply IS "the good battle" in audio amps.
Jeff Medwin
Or you could just move to europe and immediately get a improvement in all the things you speak of :-)
Only thing that improves is the turns ratio. You still need the same absolute number of turns per volt (assuming the same core size), so the advantages of a lower secondary voltage still apply.
Hi Lurcher,
Not quite. You didn't think of one thing.
The 230 or 240 VAC primary's voltage requirement, and hence, DCR, is twice as high as a 120 VAC USA one. Ideally, DCRs should be low on both sides of the audio power transformer's primary and the secondary.
No free lunch, "ideal" requires a really good, purposely-thought-out, balanced design.
Jeff Medwin
Don't forget the current requirements are HALF so you can reduce the quantity of multiple paralleled runs of TCSS from the IEC socket to the mains transformer by a factor of two .
Al
Hi Al,
Regarding lead outs on power transformers, I have noy yet A-Bed wire on the primary side, and so, I just make it conventional heavy gauge lead out wire, typically 12 AWG stranded.
It is only on the high voltage secondary, and the tube rectifier filament winding (which is intimate with the B+) where I employ better lead out wire, which now a days, is multiple TCSS runs.
Jeff
Don't you understand multiple paralleled runs of humour ?
It was aimed as a light-hearted reply ;)
Al
Ahhh...critical stuff. Inch of bad wire and all that. good catch.
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
Hi,
Another consideration is if you use a choke with a voltage doubler, All you want to do with the choke is to kill any hash generated by the diodes, the choke needs to be low DCR also which means a low Henry choke, and that will kill the hash. Here again the H/K Citation II and McIntosh's got it right.
If I was to design and build a tube amp, the voltage doubler power supply would be all I would consider. It is more costly but better. Back in the early 60's large electrolytics were realtively more expensive than today, and silicon diodes need for a voltage doubler cost much more than a couple of tube rectifiers, so back then the voltage doubler PS done right cost considerably more than a tube PS. But the results is what lead several mfgs. to employ it.
If all you use is a low Hy choke after a voltage doubler, the ripple waveform will not be a sine wave.That may or may not matter to you.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 09/18/09
and PSRR tends to deteriorate with increasing frequency too...:)
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
.
I've looked for good primers on the subject, but so far - zip.
Adam
If you find yourself frequently repeating the phrase, "garbage in, garbage out," you may need to rethink your system.
Adam, good question. Maybe, it's just a simple case of saving space for application where a lot of current is not needed. Making big transformers for voltage sake only, prolly was $$$ even in their heydays of the 1950-60's.
I've seen doublers in "good" amps. Like Macs and high-end Marshalls. I know that big amps (100 watts+) that have no doublers, like that Hiwatt I'm going to tackle have BIG PT's. Looks like 2/3rd of the cost of the amp is in the PT and OPT.
One great use of a voltage doubler is to get a screen voltage that is half of the plate voltage in pentode/tetrode amps. The power transformer might also be a bit easier to wind as you have a lower ratio.
There is nothing evil about a voltage doubler as long as your power transformer is designed for the task. It is not fashionable because some amps gave it a bad name, maxing out the power transformer in a voltage doubler and casuing the power supply to be really saggy.
___
Long Live Dr.Gizmo
Also, the secondary DCR of transformers for voltage doublers tends to be much lower than conventional winding. If that kind of thing matters to you. ;-)
I know of no real advantage of a full-wave voltage doubler. Each filter cap sees 60Hz.
I would prefer a HV bridge rectifier, but tube type rectification demands full-wave, center tap for reasonable design/cost thus my main use for DIY. The filter cap sees 120Hz with FWCT or bridge rectification & choke input filters are the best design for overall efficiency and regulation.
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: