Tube DIY Asylum

Do It Yourself (DIY) paradise for tube and SET project builders.

Return to Tube DIY Asylum

Message Sort: Post Order or Asylum Reverse Threaded

Janus Shunt Regulator -- sounds good

75.67.80.214
Posted on July 26, 2007 at 09:13:42
Perseo
Audiophile

Posts: 393
Joined: March 4, 2002
 Herewith a report on the "Janus" shunt regulator, published by John Broskie of Tubecad fame here:

http://www.tubecad.com/2007/06/blog0112.htm

(scroll down to bottom, see 2d and 3d to last schematics).

I don't know where the "Janus" name comes from; it may be JB's appellation. It is a feed-forward shunt regulator akin to the AudioNote idea plus a grounded-grid feedback shunting element of a type JB has enthused about over the years. I think the T-Rex amplifier may have incorporated the grounded-grid aspect, feeding noise to the cathode of a high-mu triode that fed the non-inverted amplified noise signal at its plate to the grid of another triode serving as the shunt element.

Anyhow, my version uses a 12B4A instead of the 12au7, to be able to handle a bigger load, and a 6AY11 instead of the 12ax7, because (i) 6AY11 is a cute little stubby compactron that no one uses and (ii) it actually is a tad stronger in gm (.0019) than 12ax7. My capacitors are 40uf motor runs. It is basically a Cap-Resistor-Shunt Regulator supply, with the Resistor being 750 ohms. Oh, instead of the FW bridge shown in JB's schematic, I have a full wave center-tapped arrangement (2 silicon diodes).

A peek at the output on a scope (a new Stingray - very fun) showed it to be very quiet.

So, I hooked up my headphone amp. This has 2 type 26 tubes per channel, stacked one on top of the other (Aikido-style), feeding a transformer from the plate of the bottom 26 (through a capacitor, parafeed). The novel thing about it is that the 26s' filaments are powered by AC, with 60Hz hum cancelled by hum pots and 120Hz hum cancelled through the stacked arrangement.

The old power supply was a brute force C-R-C-L-C with lots of electrolytic capacitance.

The Janus supply sounds great in comparison. Fill in the audiophile adjectives. Very nice DIY upgrade for short money.

Yes, I know that with solid state components many orders of magnitude better performance could be achieved, etc., but this is easy and tubish.

Steve Root


 

Hide full thread outline!
    ...
RE: Janus Shunt Regulator -- sounds good, posted on July 26, 2007 at 10:40:39
SteveBrown
Audiophile

Posts: 2455
Location: Portland, OR
Joined: November 14, 2002
 Thanks for sharing, great info! Now if only they'd offer a PC board for the final version... that would really be nice!

 

RE: Janus Shunt Regulator -- sounds good, posted on July 26, 2007 at 11:02:40
nullspace
Audiophile

Posts: 237
Location: Philly, PA
Joined: June 18, 2004

Good to hear that it's working well for you. Could you tell me how you picked values for the two capacitors?

Just as an exercise, I drew up a version of the regulator for the 56->45 amp I'm working on currently. Since you have some experience with the design, would you mind letting me know if I botched something up? I picked the 6CM7 because I just happen to have a couple laying around. Thanks.

Regards,
John

 

RE: Janus Shunt Regulator -- sounds good, posted on July 26, 2007 at 12:49:20
Perseo
Audiophile

Posts: 393
Joined: March 4, 2002
John:

I will give you an initial reaction, but want to study your schematic values a bit more so will follow up.

6CM7 has one thing to commend it: 550 volts of maximum capability. The little triode of the two contained in a 6CM7 does not have all that much gain, but it still will provide some benefit. One tube with dissimilar triodes certainly seems a good way to go. Whether the big triode has enough moxey for this task I need to ponder a bit.

I chose capacitor values as follows: I wanted to make sure that the AC ripple on the power supply side of the big resistor on top (your 910 ohm) was not too much for my shunting triode's grid to handle, so I modelled a simple one capicitor filter in PSU2. Of the big voltage caps I have on hand for the service in the first position after the rectifier (not shown on your schematic), my choices were 12uF and 40uF, and I felt like at 40uF the remaining ripple was less likely to overwhelm the grid. I found that ripple is sensitive to the load -- higher current throughput led to higher ripple -- but not too sensitive to the resistance of the power transformer. My transformer was a salvage and of unknown capabilities.

For the cap running from the B+ rail to the cathode of my 6AY11, JB's Janus showed 47uF, and I know the low resistance from cathode to ground means that that cap has to be relatively large like that (can't get away with a .1uf). I never calculated it, but just used another 40uF (it should be simple enough to calculate).

The cathode bypass cap on the shunting triode sent me to the books, because I hardly ever use bypass caps. (JB's schematics did not show this value.) I reread the Aiken Amplification article that explains how to calculate a cathode bypass cap value, which is here:

http://www.aikenamps.com/TechInfo.htm

(Good website, by the way).

40uF got me down to 12 or 14Hz Hz, as I recall, which seemed OK if not stellar.

I am supposed to be working now so will add some comments later.

Steve


 

RE: Janus Shunt Regulator -- sounds good, posted on July 26, 2007 at 13:08:56
nullspace
Audiophile

Posts: 237
Location: Philly, PA
Joined: June 18, 2004
 Thanks for the comments Steve; I'm looking forward to hearing what else you might offer up.

As far as a bypass cap value, I've been going by a RofT Paul Joppa suggested: Ck (in microfarads) = gm (in micromhos) / 30, so maybe 150uf for the 6CM7 V2. I have a scope but am still working on how to use it, so I couldn't tell you whether that's a good rule or not (though I think the source is a good one).

Regards,
John

 

RE: Janus Shunt Regulator -- sounds good, posted on July 26, 2007 at 13:42:57
Michael Samra
Dealer

Posts: 36118
Location: saginaw michigan
Joined: January 30, 2005
 Maybe its joint army and navy US. lol

 

Source of Janus name for this Shunt Regulator, posted on July 26, 2007 at 14:26:00
Patrick Kopson
Audiophile

Posts: 140
Location: Boston
Joined: May 4, 2002
 Janus was the dual-faced Roman god of doorways, beginnigs & endings.
His dual faces allowed him to see both forward and backward.
As the shunt regulator uses both feed-forward and feed-back, the name seems fully organic.

 

RE: Janus Shunt Regulator -- sounds good, posted on July 26, 2007 at 20:45:54
Perseo
Audiophile

Posts: 393
Joined: March 4, 2002
 John:

Paul Joppa is one of the EF Hutton guys. When he talks, we should listen. I have been on this and related forums for a long time, and have my own short list of guys who know their onions, stay out of flame wars, and are a huge plus for the hobby. Joppa is in the elite group.

Paul admits that his thumb rule re cathode bypass caps is pretty conservative:

http://db.audioasylum.com/cgi/m.mpl?forum=bottlehead&n=111470&highlight=Paul+Joppa+cathode+bypass&r=&session=

I would think that 60u would be fine for cap bypass service here (within the more liberal gm/100 rule), but I already admitted to not being much of an authority on cathode bypass caps. I have not done enough fiddling with my new O-scope to have any real insight on this metric.

I calculate 854 ohms for the big capacitor in the power rail, and would definitely endorse your 910 ohms. I get 854 ohms as follows:

For 6CM7 triode 2, gm is .0044, so 1 divided by gm is 227.3 ohms (this would be the optimum resistor size if 100% of the rail noise were transferred to the grid of triode 2). The dynamic resistance (impedance) of the 6CM7 triode 1 is rp (11,000 ohms) plus mu times rk, or 20 times 1800, giving a total of 47,000 ohms. The load resistor is 130k, so the grid is getting a fraction of the full power supply noise (on the supply side of the 910 ohm resistor) that is 47k/177k, or 26.6%. Invert that 26.6% and multiply by 227.3, and (allowing for rounding error) voila 854 ohms. You must have done the same ciphering.

Now, I elected to go high in the value of this resistor, for two reasons. First, it is much easier to add a little parallel resistance to bring it down, once the circuit is up and running. Second, my experience is that gm turns out to fall short of spec (.0044) much more often than not. On my implementation, I guessed high on the resistance, but when I tried to trim it (with parallel resistance) toward the calculated ideal resistance, I could not measure any effect with my scope. Could be I don't know what I am doing in measuring, or the feedback part of the circuit is tamping down any residual ripple below the level I can measure, or something else. I left it with my first guess of 750 ohms for the 12B4A (taking into account the voltage division in my own circuit).

So, I like your choice of 910 ohms.

For cap C2, because your cathode resistor is a highish 1800 ohms (compared to 470 for my 6AY11), in parallel with the dynamic resistance of the tube and plate resistor it is still more than 1400 ohms, which I calculate would require only 20uf for C2 to be good down to 5.4 Hz (the -3db point).

My only concern with the design is that I think you will need a cap of some size between the rectum-frier and the top of triode 1's plate resistor, so that the whittled down power supply ripple signal (26.6% of the gross ripple) that gets to the grid of triode 2 does not exceed the 7 volts of grid bias. My guess is you will need a cap after a full wave center tapped supply or full wave bridge of at least 20uf for this purpose.

With that caveat, based on my experience I would expect good results from your one-tube implementation.

By the way, I use 56 tubes in a little Aikido preamp, and like them. How will you get enough voltage out of them to drive 45s?

Steve

 

That explains it, and . . ., posted on July 27, 2007 at 04:20:52
Perseo
Audiophile

Posts: 393
Joined: March 4, 2002
makes it pretty clear to me that John Broskie dreamed up the name. It just sounds like him.

Oddball names are certainly handy for search engine purposes.

 

RE: Janus Shunt Regulator -- sounds good, posted on July 27, 2007 at 06:33:45
nullspace
Audiophile

Posts: 237
Location: Philly, PA
Joined: June 18, 2004
 Thanks Steve; I certainly appreciate the feedback.

However, I think you're giving me too much credit in terms of how I calculated the 910 ohm resistor. On JB's original schematic, he has a 1520 ohm jobbie, dropping 47v at 31ma. I simply assumed that I needed to drop ~50v across that resistor, plugged in what I expected the current to be (~58ma), and bam! 910 ohms for the resistor. I've used 0D3 and the like before, so I saw that resistor the same way I see the big voltage dropping resistor before a VR tube, and ~1k would as small as I'd want to go in that position. Had the value come out any smaller, I would have bumped up the voltage drop -- probably dumb luck that our calculations line up. It would seem I have some more work to do...

As far as the cap after the rectifier, you're right and I just didn't bother to draw it in there yet.

MQ was having a sale on the EX0-173 at some point in time, and I snatched up a pair. A bit of an overkill for an input 1:2 autoformer, but they weren't doing anything else. I have a 27 preamp which I like quite a bit, so I thought I'd keep the driver tube in the same family.

Regards,
John

 

Type 56, posted on July 27, 2007 at 06:53:06
Perseo
Audiophile

Posts: 393
Joined: March 4, 2002
Ah, the EXO-173s were just sitting there. . .

My 56s give a bright glow from the heater, which I like. I also use 76s in the same preamp, but they don't glow.

It is not as easy to find power for 2.5 volt heaters, but I use a 5 volt winding in series (and for a tube rectifier use 6.3 volt tubes).

Steve

 

RE: Janus Shunt Regulator -- sounds good, posted on August 15, 2007 at 13:09:22
JoshK
Audiophile

Posts: 822
Location: NJ/NYC
Joined: August 3, 2001
 I am working on a 6EM7/6EA7 version of the Janus regulator for my 6sn7 Aikido. JB thought this would work out well. I need to work out the operating points and adjust the values when I get some time and then I'll post a schematic if I can find something to draw one with.

 

Updated Version of Janus Shunt Regulator, posted on September 21, 2007 at 19:55:19
Snoopyma


 
Hi there,

Last night, I have homebrew a more updated version of Janus Shunt Regulator (see the last schematic shown on http://www.tubecad.com/2007/08/blog0117.htm)for my Aikido-based cathode coupled WOT line amplifier.

In this +300V power supply, I have used 12AX7 as the cascode tube and 6EM7 (stronger half) as the pass tube. Load is 23ma, the current through the 6EM7 and 12AX7 is 20ma and 0.6ma respectively.

6BY5G as the rectifier tube and choke input design. The prelimninary result is: it works really great. Very quiet and dynamic.

Some questions:
1. I have temporaily set the resistor Rs as 270 ohm. Can anybody teach me how to measure the noise of power supply by digital volt DC/AD meter (or oscilloscope) that I have so that I can adjust the resistor to suit?

2. There is a capacitor/resistor combo (connected in series) at the output of the power supply. Is it used as a filter or to damp the oscillation? How can I calculate their suitable values?

Any advice or information is appreciated. Thanks.

Regards,
TC MA
HONG KONG

 

RE: Janus Shunt Regulator -- sounds good, posted on May 6, 2009 at 21:06:13
mythuat818
Audiophile

Posts: 2
Location: 84
Joined: December 4, 2008



Hi All
Anyhow, I want uses a 6C19Pi instead of the 12au7, to be able to handle a bigger load

Some questions:
1. I have temporaily set the resistor R5 =??, R6+ ??. Can anybody teach me how to measure the noise of power supply by digital volt DC/AD meter (or oscilloscope) that I have so that I can adjust the resistor to suit?

2. There is a capacitor/resistor combo (connected in series) at the output of the power supply. Is it used as a filter or to damp the oscillation? How can I calculate their suitable values?

Any advice or information is appreciated. Help mee Thanks.
How can I calculate their suitable values?

Regards,
MT

 

RE: Janus Shunt Regulator -- sounds good, posted on October 31, 2010 at 10:52:52
phi70


 
I have 2 of the Janus shunt regulator kits, built according to suggested values, and been using them for a while now. Lately I started verifying its effectiveness using a scope and realized that while it filters out ripples real well, it does not deal with the drops and spikes caused by line noise at all.
Set the sweep of the scope to 100 or 250ms and look at the slow sweep,
a good regulator should show a straight line in the B+, not so with the Janus. It spikes and dips up to 10mV at a time albeit a slow dip and a slow rise every few fractions of a second.
Anyone tried to really scope the output at all?

 

RE: Janus Shunt Regulator -- sounds good, posted on October 31, 2010 at 11:45:39
phi70


 
Per Broskie:

The Janus regulator is not, however, a DC voltage
regulator, as it cannot maintain a fixed B+ voltage; instead, it works to eliminate any
AC perturbations on its output voltage. In this way it is similar to a large inductorfilled
power supply that bucks quick changes in output voltage, but allows slow DC
voltage to vary with the wall voltage variations.

*** HOW SLOW IS SLOW? If the AC fluctuates a few times a second by say 1V,
then will the Janus deal with it? I think not. It sure deals with ripple OK, but the slow spikes and dips I see when I hooked my scope to sweep at 250ms sure are still there. Even a hashy string of zeners seem to work better!

 

RE: Janus Shunt Regulator -- sounds good, posted on November 11, 2010 at 15:21:19
phi70


 
Upon closer inspection its even more disappointing.

1. no line regulation: any fluctuations in AC will cause dips and spikes in the output. Yes it filters out high freq hash but so does a good choke.

2. no load regulation: load variation greatly affects B+
You change the plate and cathode resistors on the Aikido (just by even 10%) to change quiescent current and the B+ changes accordingly! Yikes.

Any other regulator in the market including a crappy zener based emitter follower wont do that!

 

Page processed in 0.038 seconds.

[ Top Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2025, Audio Asylum, all rights reserved ]
[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]