| Search | Login/Register
   Home » Audio Discussions » Input choke for Low Currents? (6 posts, 1 page)
  Print Thread | 1st Post |  
Page 1 of 1 (6 items) Select Pages: 
03-05-2006 Post does not mapped to Knowledge Tree
Romy the Cat


Boston, MA
Posts 10,049
Joined on 05-28-2004

Post #: 1
Post ID: 2155
Reply to: 2155
Input choke for Low Currents?

Today I was reading Jim Hagerman's interview about his Trumpet phonostage and learn that he uses LCLCRCRC filters for B+ in his design. I certainly do not question Jim’s decision; in fact I never heard this phonostage, but I would like to comment, collaborate and perhaps to learn about the subject.

I do not know a lot about the Trumpet but looks like it has 4 stages of 12AX/12AU7 (~.5mA each), that makes the total plate current extremely low. Approximately 4-5 years ago when I built from scratch my beloved 2x843PT (that essentially has also “virtually nothing current” in B+) I went over the ceremony of discovering how to power the thing. The filaments were unavoidably LCLCRC but when I tried to put input choke on the 843PT’s anode I had quite poor results (sound). Being kind of religiously-input-choke-guy for whatever that has draw more then 10mA-15mA, I found that for ultra low current it did not worked. I mean it worked but it did not sound well. I remember the I got the first choke of stupidly-high inductance as well and I drew 30mA through the filter using bleeding resistor but regardless what I did the phonocorrector did not sound OK. To my surprise, the default 843P’s voltage doubler with CRC works absolutely the best and it was how I finished it up.

Did anyone have success to power ultra low current with LC and get correct sound (bass for instance)? Why my bleeder did not help to Sound if the limitation was presumably only with the critical inductance for the given current? I have my interest to ask because I’m now drastically modifying my analog and if any better scenarios to power the low currents do exist then I would like to learn about them.

Rgs,
Romy the Cat


"I wish I could score everything for horns." - Richard Wagner. "Our writing equipment takes part in the forming of our thoughts." - Friedrich Nietzsche
03-07-2006 Post does not mapped to Knowledge Tree
hagtech


Hawaii
Posts 117
Joined on 02-13-2006

Post #: 2
Post ID: 2169
Reply to: 2155
Re: Input choke for Low Currents?
Hi Cat.  I run the 12AX7s just a hair over 1mA each, and the 12AU7s at 6mA.  With differential pairs and bleeders, it ends up at 39mA on each +/-300V rail.  That works out to 7H minimum inductance at 60Hz, using the standard formula.  You can use my online web calculator as a shortcut (www.hagtech.com/theory.html#choke). 

Without enough inductance (the first L), then current can become discontinuous and output voltage a bit uncontrolled.  As you point out, it takes a solid load to make these things regulate properly.  I do the same thing on the heaters.  In addition, I add series resistance with the inductors to tune them for damped response with the loading capacitors.  This is very important if you don't want an LC tank oscillation - unstable output impedance.  Basically, you have to tune everything simultaneously for current, voltage, and damping. 

I added yet another trick with shortened tails on the differential pairs.  They match the plate voltage such that if there is a change in line voltage (short duration), the dc voltage out the plate stays constant.  I guess you can call this passive regulation.  Don't think anyone else has come up with this yet.  It's amazing to see it in action.  I can put a scope probe on the plate (output) of each 12AX7, vary the ac with a variac and the plate voltage doesn't move.  This will not happen with normal SE or traditional balanced stages (worst with current source).

jh
01-16-2007 Post does not mapped to Knowledge Tree
Romy the Cat


Boston, MA
Posts 10,049
Joined on 05-28-2004

Post #: 3
Post ID: 3503
Reply to: 2155
Input choke for High Currents?

Here is another question that almost from the DIY-Audio universe.
 
Dima is finishing Zaratustra II amplifiers. It will be quite interesting beast and I’m slowing begin to think about the power supplies – it will be my responsibility. The Zaratustra II will be up to 175W of pure class “A”, without an ability to switch into class “B”. I said “up to” because it will have quite a few interring innovations that make taking about the Zaratustra II power…  hmmmm...complicated.

Anyhow, sine the output stage sucks constant high current and since I would like it... to sound good we are planning to go for Input choke, with my usual LCRbigC filter. So, I am hunting for the choke.

The output stage will be operating between 3A and 5A. At 3A the critical inductance should be around 30mH and at 5A it should be around 18mH. I usually use triple of critical inductance. Trying few different chooks during the construction of Melquiades (where I was buying them and listening them) I learned that it is very hard to beat the pane-vanilla Hammond chokes. They are silent in Input choke application and they sound better then any other chokes I tried. For high currents I usually order a custom “Ripley” version where a choke submerges twice into Ripley Epoxy - it make it dead silent.

So, the Hammond has two 100mH chokes:

195T5  -  5A  -  .64R - 14Pouns - $69

and

195T10  - 10A - .42R - 35Pound - $260

Both of them can do the job, but I’m a little concern that the first choke might be too small if I drive though it 5A. Also, those chokes run this max current they… kind of vibrate… The 10A is 7x8 inches… it is big. In addition, I’m conserved the with lover current it should give up lower inductance, would it be?

It will take for Hammond 2 month to” Ripleynize” my choke and I need to make decisions if it will be 195T5 or 195T10.  I personally inclined to go for 10A version as if for whatever reasons the filter will, not work as I expects then I could always drive it in necessary operation by shunting the PS and driving more current via the filter. The alternative version would be use of two 5A chokes, combining them in series or parallel connection, depends of what I need…

Any bright ideas?
The caT


"I wish I could score everything for horns." - Richard Wagner. "Our writing equipment takes part in the forming of our thoughts." - Friedrich Nietzsche
01-16-2007 Post does not mapped to Knowledge Tree
hagtech


Hawaii
Posts 117
Joined on 02-13-2006

Post #: 4
Post ID: 3504
Reply to: 3503
dueling reactors
I wonder what the difference is between class A and B insulation is.  Never noticed that parameter before.  The 10 amp choke (I like their word reactor) must have some very large guage wire, as the resistance is really low.  Clearly, both chokes will do the job.  The question you ask is which will perform better.  Fortunately, it seems the usual issues with cost and size can be ignored for the moment.

Personally, I like to over-rate cores when possible.  As a core reaches magnetic flux saturation, the inductance will drop like a rock.  Everything heats up.  So what you will see is that at low currents, inductance will rise slightly.  They used to sell "swinging" chokes for this application in that the inductance was purposely non-linear with current.  It helped maintain continuous current operation over a wide range of operation.  The larger core also runs cooler.  And it looks menacing (that's good).

Another difference is in the windings.  How much to they stick out from the core?  How many layers of copper?  The more, the worse.  Outer windings do not couple magnetically as well as the inner ones, resulting in external stray fields.  Smaller guage wire actually works better, as the gaps between windings is less.  Square or flat conductors are a great thing to have. 

One advantage of using two smaller chokes is that you can wire them up such that the external fields buck each other. 

My opinion is to buy two of the small ones.  Design the chassis such they fit side by side.  You can then try three possible combinations for sonic and measurable performance parameters.  Single, series, or parallel.  Also bucking or not.  I'd guess parallel is best, but ya never know!  If the singlet works fine, then you have a spare part for later.

jh
01-17-2007 Post does not mapped to Knowledge Tree
Romy the Cat


Boston, MA
Posts 10,049
Joined on 05-28-2004

Post #: 5
Post ID: 3508
Reply to: 3504
What can I do… let it be big.

Well, I also would like to have a spare ability to care current but 10A is 35 pounds monster and… it has just 100mH. I would prefer to have extra inductance then exacta current. The current wise is juts the thickness of the wire as I understand…I prefer the chokes gets hotter then they stop to “regulate”….

Yes, I know about the swinging chokes and I was searching for someone who would do 7A total, 200mA as 1.5A but it was not successful. There were places that wiling to do it but I never dealt with their chokes and I was afraid to order form then. The problem that they clams that no one among their customaries use their chokes right after rectifies and therefore they might have no idea how to build chokes that will be working within high ripples environment and to be silent at the same time. I have an unfortunate collection of the chokes that I made when I built Milq, the chokes that do wonderful job filtering with leading cap but they all too noisy with input choke.

Ironically the Hummonds, being the cheapest turned out to be the most silent, particularly the Ripleyed-epoxyed. What I inclined to do is to put two 5A, 100H chokes (still – 28 pounds of stupid metal ballast!!!), one on play and one on minis and they shunt them with a cap. This way each shoulder will buffered with choke… that “might” even prevent me form making a start delay for charging the 550.000uF caps.

The caT


"I wish I could score everything for horns." - Richard Wagner. "Our writing equipment takes part in the forming of our thoughts." - Friedrich Nietzsche
01-17-2007 Post does not mapped to Knowledge Tree
Bud
upper left crust united snakes
Posts 87
Joined on 07-07-2005

Post #: 6
Post ID: 3511
Reply to: 3504
The difference between class A and B insulation
Jim,

UL has applied a compatability test to insulation materials since 1976. This test places representative amounts of each material, found within a potential transformer coil, in a test tube. This tube is then evacuated and subjected to infrared heating for a specific number of hours. One of the materials is a specific length of wire, made from two tightly twisted pieces, that has a coating of some specific chemical nature.

After this ageing test, and with the wires still hot, a dielectric withstand test is performed. Materials collections that pass this strenuous test are included in various reccomended temperature classification for usage in transformer construction. Class A materials are limited to 105 deg C total temperature and are expected to last at this temperature for about 300,000 hours without material change in properties. class B is 130 deg C, class F is 155 deg C and class C is 220 deg C. All of these classes allow the manufacturers of the various materials found within the coil of a transformer, to have their particular material added to an already known and tested collection for an ageing test. If the newly tested wire sample does not fail then their material is considered compatable with that particular group of materials and can be listed under one of the classes as rated for that level of operational temperature.

Transformer companies can buy into one of these materials collections, usually originally paid for by some major material manufacturer like Dupont, and use the "system" of materials to wind coils that will operate at total temperatures within the limits of the particular class of mateials. Total temperature includes the ambient temperature plus the temperature rise above the ambient testing temperature of the transformer in question. A class B transformer can have a 90 deg C rise in a 40 deg C ambient and be expected to have a 300,000 hour MTBF. This assumes that it is a well constructed device, i.e. it meets the various tests, both mechanical and electrical, from one of UL's overall classifications. These overall classifications cover all of the safety parameters applied to the entire device that the transformer operates within. These requirements reflect upon the transformers physical construction, i.e. creepage and clearance distances for spark suppression to ground or surfaces of opposite voltage polarity and dielectric withstand from the mains supply, provided by the transformer, to  the rest of the device. All of these various tests for usage, in a particular device, occur at the maximum rated temperature of the insulation system and are quite strenuous. These final usage tests confirm that the previous ageing tests have provided a stable insulation system and that the physical construction of the whole transformers is robust.

Bud
Page 1 of 1 (6 items) Select Pages: 
Home Page  |  Last 24Hours  | Search  |  SiteMap  | Questions or Problems | Copyright Note
The content of all messages within the Forums Copyright © by authors of the posts