Hiend Audio, DIY, Hi-Fi, Stereo, Electronics site, for lovers of high-fidelity music reproduction. High-End Vacuum Tubes. Silicon. Do-It-Yourself Audio Systems. Schematics: amplifiers, speakers, horns, CD, DAC. Circuits, Topologies, Acoustics, Cables, Speakers, Voltage Regulators, Equipment Upgrades, Modifications, Problems, Solutions, Tips, Tricks. ------ Dla pasjonatów Hiend Audio, najwyższej wierności odtwarzania muzyki na sprzęcie budowanym, modyfikowanym samemu. Wiedza: elektronika, układy lampowe i krzemowe, wzmacniacze, głośniki, kolumny, tuby, DAC, kable, DIY, porady, sztuczki, problemy, rozwiązania, tor audio.



Sep 30 2014

Print this Post

Hush Hush Bridge Photos

Here are some photos of the “Hush Hush Rectifier Bridge”, that I wrote about some time ago.

P1120035 b

Executed on a piece of pre-perforated prototyping board, it actually performs quite nicely.

The board actually is comprised of four sets of Rectifier half bridges, each destined to work with a separate center-tapped transformer winding.   There are two “+” voltage pushing networks, and two “-” voltage pulling networks.

P1120032 b

Each “functional diode” within each respective bridge has been expanded upon …

P1120033 b

Each functional diode actually comprises of two physical diodes in series, each of the BY 228 type, each capable of withstanding 3A of forward current, 1500V of reverse voltage, and a reverse recovery time of 2 us.  (V_r = 1500V; V-rrm = 1650V; I-f(av) = 3A; I_fsm

P1120034 b

Each diode is “enhanced” or “modified”,  by two of three ferrite sleeves / beads threaded upon the connection legs of either side of the diode.

P1120030 b

Each diode is bypassed by two series connected RC groupings.

Each R = 910k;  Each C = 47nF/630V.   The equal valued R constitute an eqal valued voltage divider.. This is necessary to even out the split of the reverse voltage on the 630V rated capacitors. This shall essentially provide immunity of an aggregate reverse voltage of 1260V DC, but practically a tad more, as the caps used are known to be of a somewhat higher resilience to voltage (“reserve slack”).  The RC network, directly at the legs of the noise source (ie. the Diodes), constitute, together with the ferrite sleeves, a first order LC snubber network, and one very close to the noise source at that.

{ The resistors are not visible, because they are deep down, laying flat on the PCB, at ground level, between the caps and the diodes. As they are in aggregate 0,9M + 0,9 = 1,8 Megs in value, they do not dissipate too much heat.  At peaks, they need to handle only 450V * 450V / 910K = 0,22 Watts peak. But the average amount of power dissipation is actually less than that }.

P1120031 b

Each group of two series connected blocks of  ”(ferrite+diode+ferrite) || (RC+RC)” are additionally terminated from either end with a ferrite choke.  The black little cylinders.

These are 1,5 mH ferrite chokes, with 0,3A current handling capability and 2R7 series resistance.  Such “choke encapsulated” groupings are finally embraced by series connected 15nF/2000V capacitors  (the blue ones).

Therefore, in essence we have a first order LC snubber network, directly at each physical diode, and then we have a second bastion of an LC snubber network, directly visible from the green cable screw-connectors. Each such bastion of second order snubbing is build up by the pair of 1,5mH/2R7 shunted by the 15nF/2000V || 15nF/2000V,  which effectively translates to 7,5nF/4000V.   No resistor divider network here, as we have some lavish voltage slack with these series connected caps.

P1120036 b

This setup should effectively and handle peak-to-peak reverse voltage polarity of up to (safely) 2520V, but most probably up to 2820V (probable)

This, divided by two square roots of two, would imply an RMS AC input voltage handling capability  of  893 V AC (safely), or respectively 1000 V AC (probable).

P1120037 b

After rectification, on a C front loaded filter, which would imply DC operating voltages of: 1260V DC (safely), or 1410 V DC (probable).

As the unit will be pumping a two storey capacitor bank, build up of 450V DC rated electrolytic capacitors (with even valued bleeder / voltage divider resistor network),  I do not anticipate working with voltages higher than 900 V DC on each of the units.

With center-tapped windings, each half winding producing 630V AC, I come out with 1,4142 * 630 = 890 V DC being pushed into the “+” sided bank of the filter capacitors.

But … that is only half of the story. That would be +890V as referenced to ground. The “Level 2″.  But what about the basin/cellar, the “underground” ?

Independently, by using the same center-tapped windings, but using the “oppositely directed” diode modules, I can also “pull” 890V DC from the “-” sided bank of filter capacitors.

Using the same ground as a common reference, I can also “produce” -890V DC.


In such a manner, I get 2 x 890 = 1780 V DC from “end-to-end” of the respective filter banks, with a common “ground” in the middle, at “mid-height” between them.

Essentially, I get a pair of +890V and -890V symmetrical rails, which are referenced to a ground running through the middle.

But Hey !

Ground is only but an exercise in relativity. Ground is where I shall draw it. No more, not less.

From the point of view of three wires, running from an ultra high voltage power supply,  it is basically a choice of where you “envision it”.  It all comes down to “perceptions” and total value of ripple voltage, which, considering the filter bank that I shall be implementing, will be peanuts anyway.

So, I can just as easily draw the ground symbol at the “-890V DC”  end.   This would essentially provide me with a two voltage power supply, providing +890V and 1780 V DC, respectively.

I am not sure which topology I shall use for the GM-70 project, but I am inclined to use the “symmetrical” approach.

It is much more safe to die from electrocution from a fairly low 890V, rather than die from a very high 1780V DC.   But then again, some people say that it is not always true that all good things take time.  So, maybe, instead of suffering prolonged cardiac arrest agony from the 890V electrocution shock, maybe I shall indeed go for the quickie version, with the 1780V.

OK, jokes put aside.   1780V is lethal, when viewed from “ground” perspective. That is why I am seriously considering using the “symmetrical rails” concept, with the ground at mid-level between the extremes.

This will complicate the actual “design” of the GM-70 amplifier, but I think that working at “Sub-1000V” levels is well worth the extra design effort.

At this point – I would like to express my great thanks to a certain John B., who provided a lot of eye opening insights as to the possibilities and the potential topologies that could use such symmetrical rails, for a GM-70 push pull stage, into some ultra-empathic and other non-standard solutions, as well as into what I originally perceived as the “ground-shifting” problem.



Permanent link to this article: http://hiend-audio.com/2014/09/30/hush_hush_bridge_photos/