Stanley F. White's Powrtron Amplifier
Robert McLean
amplifier, and inspired by that I went browsing the web for the article
mentioned and so on. During my search I came across this site
http://www.cosmos2000.org/audio/index.htm
which discusses what its author feels are the Williamson's shortcomings.
But it also talks about an amp called the Powrtron (thats how it is
spelled), which apparently came out in 1953. I am wondering if any one has
heard of this amp, or has heard it play, and if so what they thought of it.
Also if they know who Stanley F. White is. Is he the White as in White
Cathode Follower ?
Steve O'Neill
Robert McLean wrote in message ...
White "wrote up" his POWRTRON design in AUDIO , Nov 53. It was a variation
on the Williamson but with three major differences: 1. He used the
cross-coupled (VanScoyoc??) input/phase inverter for both a + and - input
(sort of like an OP amp), 2. he used a combination of voltage feedback off
the 4 ohm tap of the OPT and current feedback tapped off the high side of a
0R47 ohm resistor betw the opt sec common tap and signal ground, 3. He used
an ultralinear OPT. Seems that the goal was to present a constant power to
the widely varying speaker load as opposed to a conventional constant
voltage. Good write-up in Audio Anthology Vol3. Never heard one but the
phase inverter looks very interesting. As an aside, I read somewhere that
D.T.N. Williamson considered the application of the ultralinear output stage
to be an abomination before god...a true Williamson is triode. Based on
limited experience I might tend to agree.
--
Steve
LGeoCole
>White "wrote up" his POWRTRON design in AUDIO , Nov 53. It was a variation
>on the Williamson but with three major differences: 1. He used the
>cross-coupled (VanScoyoc??) input/phase inverter for both a + and - input
>(sort of like an OP amp),
Later work on this inverter showed that the small cathode load
resistors (470ohm) caused distortion with input voltages over
about a volt. Larger (5.6K) proved to be better. See the August
and September 1956 issues of Radio-Electronics magazine.
George C.
Russ Sadd
> There was a post today about building/improving on a Williamson type
> amplifier, and inspired by that I went browsing the web for the article
> mentioned and so on. During my search I came across this site
>
> http://www.cosmos2000.org/audio/index.htm
>
> which discusses what its author feels are the Williamson's shortcomings.
Or to put it somewhat less tactfully, one on which the author indulges
himself on a long rant against the original Williamson design. Let's
just say that his perceptions of the state of domestic and amateur
audio, circa 1945-1955, is rather different from my own!
Best regards,
Russ Sadd
www.griffon.dircon.co.uk
Philip Lawrence
Is the amp as good as he says.? White rambles on about the powtron.
As well as ranting against the Williamson
Some real info and less rant would have been more interesting.
Phil
Patrick Turner
Russ Sadd wrote:
You know Russ, about every second paragraph about the WILLIAMSON
at the above web pages
raise the possibility of being able to steadfastly refute
another human being's mutterings about Olde Audio.
Some rants are so ill founded to be absurd.
For example, it isn't evil to have seperate chassis for
power supplies, and well done umbilical cords are quite OK.
Quad used such cords between power and pre-amps,
and the cords contained signal, 300 vdc, mains, and heater power.
I doubt anyone has been killed by this arrangement over the last 55 years.
I have no problems with power cables between power supplies and
other pieces of equipment.
Dividing up chassis make the amps lighter to move around,
and if the power tranny is off the amp chassis,
then the vibration and noise and radiated fields
are less of a problem.
I like seperate chassis!!!!
The rave about the Williamson's bandwidth and transient ability
of the drive amp is sheer poppycock, and bandwidth
is 300 Khz, not 20 Khz as stated.
Zobel networks at the output of V1 will indeed reduce the bandwidth,
but with the benefit of far better HF stability due to reduced phase shift at
around 100 to 200 Khz.
All good amps should have this feature.
It does not spoil transients.
It simply allows an amp to operate more closely to open loop
conditions at 100 Khz, where low distortion and Rout are simply not
required.
One needs to measure this bandwidth carefully into
a resistive load to find out the true unmolested
bandwidth.
But even when the output tubes cause a roll off at 100 Khz,
due to the miller effect, the phase shift is such that it is not 180 degrees
by the time the open loop gain has fallen to unity.
And at 20 Khz, phase shift is negligible, and all tendency to oscillate with
any sized purely and solely capacitave load can be easily
dealt with by a couple of Zobel networks,
even when 20 dB of NFB was applied around the loop.
The idea about having to use a 2Hz to 200 Khz
OPT is a lofty well intentioned specification that few makers ever adhered to,
but that didn't stop folks from enjoying a good tune.
The best amps do have unconditional stability compare to those that don't.
It is not the only thing to consider though.
I happen to believe that open loop bandwidth is very important,
as it makes it far easier to stabilise a feedback amp.
This article complains about the high Ra of the 6SN7
resulting in the poor driver amp result.
Baloney.
6SN7 is a low Ra tube, well suited to the application.
The author complains about the crummy split load
concertina phase inverter, and its unequal anode and cathode
output impedance at HF, thus causing all sorts of
audible distortions,
Baloney again.
The concertina is a good buffer between V1 and the following
ballanced amp, and reduces the miller effect, and out of balance
signals are negligible at 20 Khz.
The first THD product of 20 kHz is 40 Khz, so where is the
shockingly audible distortion?
If you want to balance the output of the concertina
at 100 Khz, add a small trimmer cap to the cathode resistor,
maybe 15 pF will do, and then the roll offs of each
half of the balanced amp will be equal.
It is not a big deal though eh.
The concertina extends the bandwidth, and reduces phase shift.
There are some fine ideas at the site, and concerns about LF stability
are valid. But Electrolytics in 1945 were NOT all that wonderful,
and the amateur could find oilers easily.
The use and benefits of chokes in the power supply
as well as caps is very valid, It all helped, and
common mode rejection of PP class A
amps meant they were inherently quiet.
Quad used no choke in the plate supply, and had
16 uF only off the rectifier and although the
ripple voltage was 17 volts at the CT of the OPT,
the Quad amps were acceptably silent.
Williamson did have chokes, and it represented better
practice than the Quad II.
I could say a lot more, but why waste my typing????
In the Mullard 520,
The use of the EF86 followed by a 12AX7
in an LTP causes a hugely greater phase shift than in
the Williamson.
Many other amps along the way, Leaks, Dynacos, and many
others failed dismally to reach the spec set out by Williamson.
They sold in their thousands.
I have stabilised a shirtload of old amps, and a few modern ones,
such as Jolida, with OPTs
very much poorer than any Partridge work of art.
Still they are stable, and give pretty decent performance,
once the tricks of reducing the open loop bandwidth
at extremes of frequency has been attended to.
I still think trying to get 200 Hz of open loop bandwidth
is a good idea, as it makes for a better sounding tube amp,
but in practice, it is very difficult to achieve.
One would have to start with a better OPT than the original Williamson.
Its nice to do this, but for most DIY folks, not absolutely necessary.
But then 80 Khz is easily possible, without any Zobel networks.
and from this, we can work towards 65 kHz of stable operation
which will be unaffected by any capacitave loading.
The author above goes on to talk about other aspects of distortion in
speaker cones, records, and other transducers.
Just because we end up with low distortion amps does not mean
the reproduced sound will be wonderful,
as the amp is only one part of the story.
The issue of the Powertron amp of 1951 is mentioned.
I experimented with an amp with such an arrangement of current and
voltage loop NFB.
I thought such an amp made the music sound lifeless.
Speakers are always designed for a low impedance feed,
and a constant power amplifier is one which has an Rout
of an average of about 4 ohms.
Therefore when there was a need for a heavy current
for a transient, it never gets delivered.
And when bass resonance and a high Z load is presented to the load,
say by a reflexed speaker, the high Rout allows the peaks of resonance to go
unchecked compared to a low Rout amp.
It makes Carlifonian boom box type designs boom better than ever,
instaed of sounding tight, detailed, and edible, if the bass was food.
Good bass is the foundation of music, and it gives the authority
and structure for emotion,
so why woffle this up with high Rout?
Having said that, some speakers have passively well damped
action, even at low bass, and well damped mainly
resistive load character all the way to a Mhz, and these
are a delight to send watts to.
You will find that if Rout was kept at a constant 4 ohms,
then when RL was between 3 and 10 ohms, the power
would be nearly constant.
In the real world, most dynamic speakers are not as "reactive"
as some folks suggest,
and are in fact mostly resistive, despite all the effects of crossover filters.
One of the most reactive Olde Time speakers was the Quad ESL 57.
It was light years ahead of many cone and cone speakers of the day.
DOME tweeters were not used until later, mostly.
So what is so darn bad about a bit of reactive L or C in a speaker?
Probably not that much.
Those 2.83 volts worth will do for now.
Patrick Turner.
Patrick Turner
Philip Lawrence wrote:
> Hi Russ
> Is the amp as good as he says.? White rambles on about the powtron.
> As well as ranting against the Williamson
> Some real info and less rant would have been more interesting.
> Phil
It would seem he thinks an ouput stage with an open loop of about
4 ohms is better than triode, or pentode, so the UL
is OK.
As if most of us didn't already know.
The powertron gives equal power into any RL.
20 watts into 3 ohms would mean 7.7 volts,
and 20 watts into 20 Ohms would mean
20 volts.
The trouble is that speakers have very different acoustic
outputs for the same power input, something designers
take into account when dreaming up speakers.
It would be evr so much more difficult to
design speakers to give a flat response from an equal input
of power, regardless of frequency, or impedance.
So, we are stuck with nearly all speakers requiring a low Z output,
cabable of little voltage variation between 3 and 20 ohms.
Williamson amps have 0.1 % THD, and about 0.2 ohms Ro,
all of which contribute to good sound, IMHO.
Nothing wrong with C-L-C type power supplies,
and best when values of C and L are both nice and large.
So no need for oil caps.
Flame suit being dusted off......
With a few simple mods, and a quad of output tubes to deal with
low sensitivity modern speakers, a well done Williamson is pretty blameless.
I have built a few.
I went on to build very slightly simpler circuits,
( less stages ) seen at my web pages.
But there is very little sonic difference.
All my designs incorporate some respect for the Williamson approach,
which was one heading in the right direction,
and one on which we can improove.
Patrick Turner.
John Byrns
<in...@turneraudio.com.au> wrote:
> Some rants are so ill founded to be absurd.
> For example, it isn't evil to have seperate chassis for
> power supplies, and well done umbilical cords are quite OK.
> Quad used such cords between power and pre-amps,
> and the cords contained signal, 300 vdc, mains, and heater power.
> I doubt anyone has been killed by this arrangement over the last 55 years.
I wouldn't bet on that, the Quad umbilical cord arrangement is actually
quite dangerous, and could easily kill. The problem is that if the system
is turned on, with the umbilical cord to the tuner(s) connected, but with
the audio cable(s) disconnected, as a result of either forgetfulness or
accident, the chassis of the tuner(s) will have in excess of 300 Volts of
B+ on them. This results from the fact that the ground return is carried
only on the shield of the audio cable, and not in the umbilical cord, I
doubt this sort of connection would be allowed today.
[Snip]
> The concertina is a good buffer between V1 and the following
> ballanced amp, and reduces the miller effect, and out of balance
> signals are negligible at 20 Khz.
> The first THD product of 20 kHz is 40 Khz, so where is the
> shockingly audible distortion?
> If you want to balance the output of the concertina
> at 100 Khz, add a small trimmer cap to the cathode resistor,
> maybe 15 pF will do, and then the roll offs of each
> half of the balanced amp will be equal.
Does the 15 pF trimmer cap always go on the cathode of the concertina? I
would think it might sometimes need to go on the plate, depending on the
relative values of wiring capacitance and other stray capacitance in the
plate and cathode circuits. I would think it better to just try and keep
the wiring as symmetrical as possible, and forget the 15 pF trimmer cap.
Regards,
John Byrns
Surf my web pages at, http://www.enteract.com/~jbyrns/index.html
Patrick Turner
John Byrns wrote:
> In article <3CD67774...@turneraudio.com.au>, Patrick Turner
> <in...@turneraudio.com.au> wrote:
>
> > Some rants are so ill founded to be absurd.
> > For example, it isn't evil to have seperate chassis for
> > power supplies, and well done umbilical cords are quite OK.
> > Quad used such cords between power and pre-amps,
> > and the cords contained signal, 300 vdc, mains, and heater power.
> > I doubt anyone has been killed by this arrangement over the last 55 years.
>
> I wouldn't bet on that, the Quad umbilical cord arrangement is actually
> quite dangerous, and could easily kill. The problem is that if the system
> is turned on, with the umbilical cord to the tuner(s) connected, but with
> the audio cable(s) disconnected, as a result of either forgetfulness or
> accident, the chassis of the tuner(s) will have in excess of 300 Volts of
> B+ on them. This results from the fact that the ground return is carried
> only on the shield of the audio cable, and not in the umbilical cord, I
> doubt this sort of connection would be allowed today.
But when the Quad II is used with the preamp 22 unit only,
what is the problem?
How many dead bodies?
I would have thought ANY cable to a device with mains voltage is dangerous.
In Oz we have 240 volts mains, and many mainly
imported appliances have fragile mains leads, yet very few dead bodies.
I have built several amps with remote PSUs, and all have
well grounded metal chassis, and the zero volt line is
connected to the chassis via a low value resistance,
and the cable is very rugged, normally used for
415 v 3 phase application in mobile cranes, etc.
These cablings are no more dangerous than any other.
And to single out the Williamson way of having seperate chassis on the grounds
it
may kill someone, is stupid.
>
>
> [Snip]
>
> > The concertina is a good buffer between V1 and the following
> > ballanced amp, and reduces the miller effect, and out of balance
> > signals are negligible at 20 Khz.
> > The first THD product of 20 kHz is 40 Khz, so where is the
> > shockingly audible distortion?
> > If you want to balance the output of the concertina
> > at 100 Khz, add a small trimmer cap to the cathode resistor,
> > maybe 15 pF will do, and then the roll offs of each
> > half of the balanced amp will be equal.
>
> Does the 15 pF trimmer cap always go on the cathode of the concertina? I
> would think it might sometimes need to go on the plate, depending on the
> relative values of wiring capacitance and other stray capacitance in the
> plate and cathode circuits. I would think it better to just try and keep
> the wiring as symmetrical as possible, and forget the 15 pF trimmer cap.
I have always found the 15 or so pF goes on the cathode,
as the following stage has about uniform capacitave loadings, but
the imbalance at HF always seems to accompany a roll off in the plate
circuit before the cathode circuit..
The 15 pF has a reducing impedance as F rises, thus
boosting HF plate gain a tiny bit.
The net effect is that the concertina is balanced in output up
to a higer frequency than otherwise.
You can ignore this aspect completely, and still get
wonderful music from a Williamson design when well implemented.
Patrick Turner.
Russ Sadd
> Hi Russ
> Is the amp as good as he says.? White rambles on about the powtron.
> As well as ranting against the Williamson
> Some real info and less rant would have been more interesting.
> Phil
Hi Phil,
I'll leave the circuit analysis to Patrick, but would note that the
author doesn't make any real note of the component quality. For
example, the piece de resistance of Williamson's design had been to
specify an output transformer of superlative quality for the time
(1947), and by the end of that decade, the many manufacturers on the
bandwagon, had moved the goalposts even further. So, just to rebuff
our original author a little more, I'd like to add a few corrections
here:
- The Williamson-style amplifier was not a radical departure, but
rather the latest "quality amplifier" in a long series. 'Wireless
World' had a decades-long tradition of publishing superlative designs.
Williamson's design merely pushed the envelope further with its call
for far superior output transformers, triode output stages, and plenty
of tweakability to balance the push-pull design.
- The use of feedback loops in amplifiers was a relatively new
concept, circa 1947, in as much as its use - and theory - was a
regular topic of debate in the said journal.
- A remarkably similar design had already been published in a
Marconi-Osram pamphlet as a suggested design for the KT66 valve - the
same pamphlet giving similar designs for the PX4 and PX25 triodes.
- As was made obvious in the advertisements and write-ups of the time,
the KT66 was intended as a replacement to the PX4 and PX25, hence it
being one of the rare beam tetrodes *intended* to be triode-strapped
(as were indeed the Mullard EL34 and EL37). It is therefore not
surprising that the KT66 isn't too far off from the PX4 or PX25 in
terms of its behaviour as a triode.
- H&K did *not* invent ultra-linear. Top marks for publicity, however.
Ultra-linear operation had been dismissed by Williamson and Walker (as
per their 1952 article, "Amplifiers and Superlatives") by the
beginning of the 1940s. Williamson's stance that using ultra-linear
merely to emulate triode-mode for the sake of saving a few extra volts
of HT now appears to be the accepted wisdom!
- As regards component quality, using ex-military paper-in-oil
capacitors is hardly the cheaper option! These had the advantage of
high voltage tolerance and tremendous robustness (unlike
electrolytics). Again, this is hardly going against current wisdom in
design. A substantial pi filter before a balanced push-pull *triode*
output stage is adequate, with a second pi filter for the critical
driver stages. Besides, Williamson himself revised the power supply
about two years later, to feature indirectly heated rectifiers (thus
making the author's complaint spurious at best), Langford-Smith
advocated choke-loading this double-pi filter for even better
regulation, and domestic Williamsons, by the mid-1950s, opted for a
mixture of paper-in-oils and electrolytic filter capacitors.
- Two-chassis designs do actually help to minimise power supply
induced hum. Distance, as an engineer friend of mine always insisted,
is the best insulator. Besides, it makes the earthing arrangements a
hell of a lot simpler! But this is still a typical design choice, and
Williamson *at the time* suggested layouts for both single and dual
chassis options!
- Finally, Partridge was merely one of a dozen or more companies which
later jumped onto the popular success of the Williamson design, in
that it produced and marketed a set of dedicated transformers. As did
Savage, Woden, Gardners, Parmeko, etc. As time went by, these
furthermore surpassed the original Williamson criteria (which, as I
keep insisting, came first!). As a further rebuff to our author,
Williamson kits were extremely popular, and you could buy chassis from
manufacturers as well as the transformers, capacitors, and sundry
resistors.
End of rant.
Patrick Turner
Russ Sadd wrote:
Well, I reckon you are about right Russ.
Electros were pretty fragile if over currented and this happened when a tube
began to fail, and draw a lot of current, hence ripple went beyond the low rating.
The oil cap was not affected.
But not much capacitance to anchor down the CT.
But then not much class B action, so no need for a big C at the CT.
In class A, you would not need a cap at all at the CT.
It could even be fed by a CCS.
However, LF stability was an issue at the time, and adding a preamp
with bass boost in a tone control stage was deadly if the preamp
had the B+ power drawn from the power amp's psu.
I wonder how much skill and art and technique has been lost
over time with winding OPTs.
I would say a heck of a lot.
Many of those old makers really knew a lot about OPTs.
They also had auto traverse winding machines, and the lads
and the ladies using them were well skilled, having had a lot of practice.
Now, I don't get enough practice, and nobody much around me
wants to wind anything except basic power trannies.
Patrick Turner.
John Byrns
<in...@turneraudio.com.au> wrote:
> But when the Quad II is used with the preamp 22 unit only,
> what is the problem?
I didn't say there was a problem with using the Quad II with the preamp
22, the problem is with the tuner(s) and the preamp 22, try it and see.
> How many dead bodies?
That's a good question.
> I would have thought ANY cable to a device with mains voltage is dangerous.
> In Oz we have 240 volts mains, and many mainly
> imported appliances have fragile mains leads, yet very few dead bodies.
I'm not sure what mains Voltage has to do with the issue, the B+ is
independent of the mains Voltage.
> I have built several amps with remote PSUs, and all have
> well grounded metal chassis, and the zero volt line is
> connected to the chassis via a low value resistance,
> and the cable is very rugged, normally used for
> 415 v 3 phase application in mobile cranes, etc.
The problem with the tuner(s) is that the "zero volt line" is not carried
through the unbilical cord.
> These cablings are no more dangerous than any other.
> And to single out the Williamson way of having seperate chassis on the grounds
> it
> may kill someone, is stupid.
No, Quad's omission of the "zero volt line" from the tuner(s) umbilical
cord is what is stupid, and dangerous.
junk room
news:jbyrns-0705...@216-80-74-166.d.enteract.com:
>> I would have thought ANY cable to a device with mains voltage is
>> dangerous. In Oz we have 240 volts mains, and many mainly
>> imported appliances have fragile mains leads, yet very few dead
>> bodies.
>
> I'm not sure what mains Voltage has to do with the issue, the B+ is
> independent of the mains Voltage.
He merely meant to demonstrate that voltages nearly as high as some amps'
B+ are commonly run all over the place in extension cords etc in Aussie
homes. We are somewhat blase about the high voltage we use.
Patrick Turner
John Byrns wrote:
> In article <3CD79B14...@turneraudio.com.au>, Patrick Turner
> <in...@turneraudio.com.au> wrote:
>
> > But when the Quad II is used with the preamp 22 unit only,
> > what is the problem?
>
> I didn't say there was a problem with using the Quad II with the preamp
> 22, the problem is with the tuner(s) and the preamp 22, try it and see.
I cannot try it right now, as I have not the gear with which to try it.
You are probably right on this nit picking issue.
Generally, not too much wrong with power carrying cables, if they are
well concieved.
>
> > How many dead bodies?
>
> That's a good question.
>
> > I would have thought ANY cable to a device with mains voltage is dangerous.
> > In Oz we have 240 volts mains, and many mainly
> > imported appliances have fragile mains leads, yet very few dead bodies.
>
> I'm not sure what mains Voltage has to do with the issue, the B+ is
> independent of the mains Voltage.
I am concerned about any cable with a lot of volts in it with respect to ground.
It could be deadly, and this includes all mains leads in Oz.
Quad II also have the mains wires going to the preamp 22 so
the mains can be switched.
So Quad is the double whammy.
Two lots of cables.
If the dog starts chewing either of those cables,
its a gonna findoutabout voltage and current in one short last lesson.
I doubt that happens often.
> > I have built several amps with remote PSUs, and all have
> > well grounded metal chassis, and the zero volt line is
> > connected to the chassis via a low value resistance,
> > and the cable is very rugged, normally used for
> > 415 v 3 phase application in mobile cranes, etc.
>
> The problem with the tuner(s) is that the "zero volt line" is not carried
> through the unbilical cord.
I have not looked at the schematic, and the socket config
for the 22 for awhile.
Maybe some bean counter at Quad, said,
"Now gentlemen, we must be concerned about costs,
so let us leave out all unecessary wiring."
> > These cablings are no more dangerous than any other.
> > And to single out the Williamson way of having seperate chassis on the grounds
> > it
> > may kill someone, is stupid.
>
> No, Quad's omission of the "zero volt line" from the tuner(s) umbilical
> cord is what is stupid, and dangerous.
Agreed, entirely.
Like a lot of things, there was good and bad aspects of the ways in which
some electronics gingerbeers went about their business
years ago.
Electronics is still killing and maiming people,
mainly in third world cheap labour countries where they have no rules
for work conditions, and solvents getting into the water supply,
and so on.
They went from the perils of thorium poisoning from making cathodes, and
other risky stuff, like pcb oils, back in the bad ol days of tubes, to
present complex processing, all of which companies try
to have done in some place where the workers
cannot sue for deformed babies, etc.
I don't like ungrounded metal chassis.
Quad just float their gear.
If ever there was a short between the mains active
and the chassis, the chassis would be at mains active potential.
But then I have electric drills I use every day with active and neutral,
and a metal part I touch all the time, and no earth wire.
All legal and approved.
But it would be difficult for the active to short to ground,
due to the extensive use of plastic within which the motor in my drill is housed.
I think Metabo got it right.
Patrick Turner.
BEAR
> Philip Lawrence <hitr...@iinet.net.au> wrote in message news:<3CD67753...@iinet.net.au>...
> > Hi Russ
> > Is the amp as good as he says.? White rambles on about the powtron.
> > As well as ranting against the Williamson
> > Some real info and less rant would have been more interesting.
> > Phil
>
> Hi Phil,
>
> I'll leave the circuit analysis to Patrick, but would note that the
> author doesn't make any real note of the component quality. For
> example, the piece de resistance of Williamson's design had been to
> specify an output transformer of superlative quality for the time
> (1947),
<snip>
> - As regards component quality, using ex-military paper-in-oil
> capacitors is hardly the cheaper option!
At the time they were cheap as dirt. After WWII there was literally
tons and tons of cheaper than dirt surplus electronics - the remains of
which can still be found in places, that's how much of it there was.
It was all cheap then if you went surplus military.
At least that's how it played out in the US - what we call "domestic" ; _)
:- )
_-_-bear
> These had the advantage of
> high voltage tolerance and tremendous robustness (unlike
> electrolytics). Again, this is hardly going against current wisdom in
> design.
<snipped>
>
>
> End of rant.
>
> Best regards,
>
> Russ Sadd
> www.griffon.dircon.co.uk
--
_-_-bearlabs
http://www.bearlabsUSA.com
- Silver Lightning Interconnects -