Pros and Cons of Mercury Rectifiers
wi...@my-deja.com
of mercury rectifiers. It seems that they have a more stable and
relatively low voltage drop, but I would appreciate any additional
insight.
Thanks
JD
Sent via Deja.com http://www.deja.com/
Before you buy.
Grover Gardner
> I was hoping to get some insight into the advantages and disadvantages
> of mercury rectifiers. It seems that they have a more stable and
> relatively low voltage drop, but I would appreciate any additional
> insight.
>
> Thanks
> JD
Mercury-vapor rectifiers require one of two things:
1) a choke-input filter, thus increasing the size of your HV
transformer and power supply significantly.
2) with a cap-input filter, the filaments must be pre-heated to
vaporize the mercury before applying HV, and the input cap must be very
small, this reducing the efficiency of your filter, which can be an
issue in SE amps requiring a lot of B+ filteration.
Even if you execute the second option carefully, the turn-on stress is
high and smaller rectifiers, like 83s, are prone to failure. This is
my experience, anyway. Perhaps my 83s were all well-used, but I found
it enough of an inconvenience to avoid them. Others, however, swear by
them and I'd be interested to hear other experiences. They certainly
offer a beautiful sound and significantly better efficiency than other
types.
L.J. Timpert
Mercury Vapour rectifiers have a low voltage drop for tubes, around 15
volts. This drop is fairly onstant over the current range so regulation of
your supply is usually good compared to normal vacuum diodes. On the
contrary, they produce more RF noise because of their quick turn-on and
low impedance. So small RF chokes should be inserted between the plates
and transformer winding. Also, they don't like cap loading, so it is best
to use choke input filter or a very small capacitor at the input of your
filter. They definitely don't like high voltage when they are not heated
to working temperature, so you need a separate HT switch. IMHO they have
little advantage over semiconductor diodes. One big advantage is the
beautiful blue glow they produce.
Jurgen Timpert
wi...@my-deja.com wrote:
> I was hoping to get some insight into the advantages and disadvantages
> of mercury rectifiers. It seems that they have a more stable and
> relatively low voltage drop, but I would appreciate any additional
> insight.
>
> Thanks
> JD
>
Russ Sadd
> I was hoping to get some insight into the advantages and disadvantages
> of mercury rectifiers. It seems that they have a more stable and
> relatively low voltage drop, but I would appreciate any additional
> insight.
Advantages: high power output, good regulation.
Disadvantages: warm up time, availability, RF hash.
Thing is, these aren't being made in huge quantities like popular audio
valves are, and so we'd be drawing from limited supplies. Mercury rectifiers
are also fiddly to implement, as they frequently need minutes to warm up
before you can apply HT (i.e. separate HT and LT supplies and patience), and
inject a good bit of RF noise into the circuit that needs to be filtered out
with RF chokes.
For domestic use, mercury rectifiers aren't ideal. Yes, they're high power,
but generally in excess of what domestic audio equipment requires. It's
cheaper, easier, and safer to use conventional high vacuum rectifiers. On
the other hand, you'll certainly see them in ham radio gear, as you need
some kind of mercury rectifier to deliver sufficient kilovolts to run
transmitting valves off! Indeed, you occasionally see them in audio
amplifiers using things like 845s, but it's rather easier to put in a couple
of 5R4s or 6D22S's instead.
Best regards,
Russ Sadd
Birmingham, UK
E-mail: gri...@dircon.co.uk
WWW: http://www.griffon.dircon.co.uk
Jwarwick
You wrote:
> 1) a choke-input filter, thus increasing the size of your HV
> transformer and power supply significantly.
>
> 2) with a cap-input filter, the filaments must be pre-heated to
> vaporize the mercury before applying HV, and the input cap must be very
> small, this reducing the efficiency of your filter, which can be an
> issue in SE amps requiring a lot of B+ filteration.
.
Is filament pre-heating not necessary with choke input because of a smaller
current spike at turn-on? Also, what's the maximum practical current/voltage
these things can deliver, choke input? Anybody...?
Thanks,
Jim
Jerry Young
rectifiers I tried. However they didn't hold up well. After blowing a
couple I decided it was time to move on. Perhaps a standby switch would
have been the answer.
Jerry Young
Bob Weiss
of the filter topology. Preheating the tube evaporates any liquid mercury that
may have condensed on the plate or filament structure, which could lead to
localized arcing and overheating. When new tubes are placed into service, the
filaments should be run at rated voltage (no plate voltage applied) for at least
30 mins to condense the mercury into the bottom of the bulb. Subsequent
preheating cycles are anywhere from 30 seconds to 5 or more minutes, depending
on tube type. In most transmitters and industrial equipment using these tubes,
the preheating cycle is automatic, controlled by an Amperite or similar time
delay relay in the B+ circuit.
Bob Weiss N2IXK
k6hm
As long as the tubes are warm, meaning the mercury was changed to vapors after
the filament was turned on for more than couple of minutes, it is O.K. to
operate on any position. The famous TV-7 series tube tester use it at
horizontal position.
Best luck.
Young
Gareth Davies wrote:
> In relation to Mercury Rectifiers, I'm planning on using a type 83 on an
> amp I'm building at the moment. The specs state that the tube should be
> operated in the vertical position with the base down... fine. But what
> happens when I want to check out the circuit voltages and have to turn the
> amp on its side or upside down?
>
> Do I have to test the circuit from underneath while leaving the rectifiers
> with bases pointing down? Doesn't seem the safest of options to me.
>
> Any ideas or recommendations appreciated.
>
> Many thanks,
>
> Gareth Davies.
Ned Carlson
I think so. 83's ought last for eons if they're used properly.
Ned Carlson Triode Electronics "where da tubes are!"
2225 W Roscoe Chicago, IL, 60618 USA
ph 773-871-7459 fax 773-871-7938
12:30 to 8 PM CT, (1830-0200 UTC) 12:30-5 Sat, Closed Wed & Sun
<A HREF="http://www.triodeel.com">http://www.triodeel.com</A>
Tube and Tube Amp info on the net...<A HREF="http://www.triodeel.com/tlinks.htm"> The Big Tube Links Page!</A>
Ned Carlson
<g...@ink-blot.freeserve.co.uk> wrote:
>In relation to Mercury Rectifiers, I'm planning on using a type 83 on an
>amp I'm building at the moment. The specs state that the tube should be
>operated in the vertical position with the base down... fine. But what
>happens when I want to check out the circuit voltages and have to turn the
>amp on its side or upside down?
Don't worry about this.....about 10 million tube testers operate 83's
sitting *sideways*, Won't hurt them to do this for a few
minutes while you check the circuit.
Ned Carlson
wrote:
>All MV rectifiers should be preheated before applying plate voltage, regardless
>of the filter topology. Preheating the tube evaporates any liquid mercury that
>may have condensed on the plate or filament structure, which could lead to
>localized arcing and overheating. When new tubes are placed into service, the
>filaments should be run at rated voltage (no plate voltage applied) for at least
>30 mins to condense the mercury into the bottom of the bulb. Subsequent
>preheating cycles are anywhere from 30 seconds to 5 or more minutes, depending
>on tube type. In most transmitters and industrial equipment using these tubes,
>the preheating cycle is automatic, controlled by an Amperite or similar time
>delay relay in the B+ circuit.
The RCA book doesn't note a warmup time for '83 (it just says
voltage should not be applied until the filament is warmed up),
but it definitely does for 866A (15 sec) and 872A (30 sec).
IMA, a lot of tubes being passed off as 83's, aren't, I've got some
JAN-83's here that are definitely *vacuum* tubes.
If an 83 isn't giving off a nice blue glow, it's NOT an '83, it's
a 5Z3.
Some latter day 866A's aren't mercury, they're xenon.
83's don't have time delays in tube testers, because
no plate current is applied until the "TEST" button is
pushed.
Gareth Davies
amp I'm building at the moment. The specs state that the tube should be
operated in the vertical position with the base down... fine. But what
happens when I want to check out the circuit voltages and have to turn the
amp on its side or upside down?
Do I have to test the circuit from underneath while leaving the rectifiers
Grover Gardner
wrote:
> Hi Grover,
>
> You wrote:
>
> > 1) a choke-input filter, thus increasing the size of your HV
> > transformer and power supply significantly.
> >
> > 2) with a cap-input filter, the filaments must be pre-heated to
> > vaporize the mercury before applying HV, and the input cap must be very
> > small, this reducing the efficiency of your filter, which can be an
> > issue in SE amps requiring a lot of B+ filteration.
> .
> Is filament pre-heating not necessary with choke input because of a smaller
> current spike at turn-on? Also, what's the maximum practical current/voltage
> these things can deliver, choke input? Anybody...?
>
> Thanks,
Apparently with choke-input there is just enough time for the
rectifiers to heat up. As for current/voltage, depends on the
rectifier. There are 83s which are similar to a 5U4 IIRC, and there
are 866s, which handle much higher voltages and currents. Interesting
to me is the 866 Jr., which is a baby version of the 866. Never run
across any, though.
Dave Slagle
lets dispense with a few myths....
all MV tubes, must be prewarmed before applying the HV... and if the tube is
jarred inverted etc... a good 30 minute prewarm should be granted... sure
once the mercury is liquified, it don't matter what position it is in, but
the TV-7's do warm it up, and don't hit it with 500V @ 200ma.
they operate equally well into choke input and cap input circuits, but only
if you supply a gentle ramp up... the choke in filter naturally does this...
the DCR limits the current, a cap in is a short and will cook a mv tube
quickly, a cap in... yes even 100 mics will not kill a mv tube, IF you ramp
up the B+ gradually... but a dozen turnons without rampup to the same filter
will send it to that great hazardous waste site in the sky...
RF hash.... this one is my favorite... some idiot reads an old ham book and
sees an 872 at 1KV and an amp will generate hash... sure it will for AM
broadcasts... but last time i checked, thats a bit higher than we deal with,
and any measures i have taken to combat this alleged hash have sucked the
life out of the sound...
as a parting note... on two seperate occasions, friends have taken mv tubes
into labs, and were unable to measure any noise because the ambient RF was
too high... all other rectifiers showed noise, but somehow the MV tube in
question became 'the cleanest thing since a piece of wire'
oh yeah... they sound good too... well worth the pain in the ass they are to
turn on....
dave
Jerry Young
made a loud noise when flipping the switch. The only place I found that
had silent switching was between the rectifier & first cap which was
still applying voltage to the rectifier. I finally decided to move on to
FREDs as they have a similar voltage drop. I miss the sound of the 83.
Any suggestions?
Jerry Young
Gary Wheeler
is that the tubes either need or at least will last much longer with delayed
application of plate voltage. What can be used to switch the A.C. at the
voltages and currents involved? I'm working on a design that uses a B+ of
400V @ 300ma. At that current I need to use to 83s in parallel (wish I'd
heard about this 866 before now!). I would love to find relay which can be
switched after a 30s delay and the presence of the power tubes' bias
voltage, but I haven't seen any rated for 400VDC. The easiest solution may
be to get a seperate filament supply for the 83's and switch the B+ tranny
at the primary - not particulary elegant, as I have a nice 5V 6A winding on
my B+ tranny. Any suggestions would be very appreciated.
Gary
Richard Matthews
1) Mercury is liquid even at room temparature
2) No "ramp up " is need to operate MV rect. inot cap input filters, what is
need is for the filament voltage to be applied before high voltage. See the
RCA transmitting tube manual TT-5. The 872A is capable of 5 amperes peak
current. In must cases the transformer primary resistance in combination
with other series resistance can keep the peak current below this. It can
even do a fault current of 50) amperes for 0.2 seconds. (Again see TT-5)
3) RF hash.. It is not the frequency that hash is generated at that is the
problem. There are many places in adio circuits that act non-lineraly at RF
frequencies, this means that the "hash" can be "dtected" (ie rectified) and
that you may be able to hear it in the audiono band. Taake for example
interferance caused by CB radio. The emmision is it 27 MHz, but if it is in
close to your amp/pre/etc you may hear the information anyway. Everyone has
heard intereferance from radios from car services, pages, etc.
Furthermore, I have never seen a case when the baseline noise has been
improved in an audio system when the mid-range and top end quality did not
also improve.
4) All rectifiers make RF noise. A perfectly rectified sinusoid by perfect
rectifiers has INFINITE bandwidth. Study your fourier transforms.The
difference is that MV rectifier have an additiona source of noise.
Measuring the additionalnoise spectrum frum something like an 872A i a piece
of cake. I suggest a Stoddart current probe, and even an ancient spectrum
analyser like a HP 141T. I have done it. It is simple, but you need a
hygenic set up. See "Electromagnetic compatibity Handbook" by Violette,
White, Violette
Before I owned Leeds, I worked as an electrical engineer, in EMI/RFI
compliance and design for an aerospace contractor.
Please excuse all typos, and syntactic errors, I am late for work!
Richard
Leeds Electronics
http://www.leedselect.com
----------
In article <INIv4.357$M83....@newsread2.prod.itd.earthlink.net>, "Dave
Alan Douglas
Richard Matthews wrote (in part):
>4) All rectifiers make RF noise. A perfectly rectified sinusoid by perfect
>rectifiers has INFINITE bandwidth. Study your fourier transforms.The
>difference is that MV rectifier have an additiona source of noise.
>Measuring the additionalnoise spectrum frum something like an 872A i a piece
>of cake. I suggest a Stoddart current probe, and even an ancient spectrum
>analyser like a HP 141T
Well said. I've looked at a few 83s on a curve tracer, and saw a
lot of instability and hash during warmup, for some samples.
No one has mentioned yet that the 83 is prone to arcing if the
plate-to-plate voltage rating is exceeded, and if it does arc from a
transient, it is essentially a short across the transformer and will
take out the line fuse.
The 83 was developed for class-B audio in the 1932 radio model
year. The fluctuating current of a class-B output stage demanded a
low-impedance supply to avoid feedback into the RF stages. When
class B died a quick and deserved death the following year, 83s went
out of fashion. It's only because they were ideally suited to the
Hickok tube-tester circuit that they exist at all.
73, Alan
Alan Douglas
>O.K., what are the brass tacks of implementing these?
> I'm working on a design that uses a B+ of 400V @ 300ma.
Forget 83s at that voltage, unless you like replacing fuses.
Forget 866s too: even hams don't use these any more. They should be
enclosed in metal cages to prevent contact with the exposed top caps.
And don't rely on the ceramic caps to be insulators when they get hot.
If ya gotta be retro, xenon rectifiers like the 3B28 are safer for
the home environment. You really want a room full of hot mercury vapor
if a tube breaks? It's just not worth the risk.
73, Alan
wi...@my-deja.com
mercury would solidify and come back to earth almost instantly if a tube
broke and the temperature dropped to 70ish.
> You really want a room full of hot mercury vapor
> if a tube breaks? It's just not worth the risk.
>
> 73, Alan
>
Alan Douglas
>I don't want to inhale vaporized mercury, but I would think that the
>mercury would solidify and come back to earth almost instantly if a tube
>broke and the temperature dropped to 70ish.
Not at all. Mercury is liquid at room temperatures, and in fact is
fairly volatile.
Alan
Ned Carlson
Young) wrote:
>I used a standby switch for awhile in the CT of the power xfmr but it
>made a loud noise when flipping the switch. The only place I found that
>had silent switching was between the rectifier & first cap which was
>still applying voltage to the rectifier.
Try a hefty DPDT switch between the transformer secondary and
the plates of the 83. This is basically what's done in Marshall
guitar amplifiers, tho Marshalls use a bridge rectifier.
Da Man
at room temperatures.
Jeff
"Alan Douglas" <adou...@gis.net> wrote in message
news:38c084b8...@news.newsguy.com...
David Ginsberg
gonna modify my Marantz 8B's to run with 83 rectifiers. Any
comments, warnings? Note, these beauties are only run between 1
and 8 kHz (6db/oc. slopes) and are already
mylared/strapped/trioded.
BTW, what is an 83V? Got some of them too.....
* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!
Jerry Young
Jerry Young
Douglas Diermann
F. Gagliano, A. Walker
AW
Jerry Young
Jerry Young
JW
Mallory (the cap company, not the author) has a good apnote on the use of their Quencharc snubbers and the why's
and wherefore's.
David Ginsberg
Steven Kirkpatrick
Just thought that I'd throw my two cents in regarding mercury rectifiers.....
I've been working on a SE audio amp for a couple of months now which uses two of
the 866A's per channel in the power supply. Contrary to what Alan's experience
has been, I find the 866A's to work fine for my application. My SE amps are
based upon a 304TL transmitting triode which is very power hungry. Currently,
two 866A's configured as full wave rectifiers are delivering 1530 VDC at 240 mA
from a center tapped HV transformer rated at 2520 VAC @ 280 mA. The two 866A's
are feeding a capacitor input "Pi" type filter with the following configuration:
the cap closest to the 866A's is 4 microFarads rated at 3500 VDC followed by an
8 Henry choke rated at 400 mA DC which is in turn followed by 44 microFarads
rated at 2400 VDC. I initially tried to use more than 4 microFarads for the
initial capacitor but found that higher values caused the peak anode current
within the 866A's to become too large resulting in some internal arcing.
Reducing the capacitance to its current 4 microFarad value eliminated the arcing
problem (this is the effect referred to by Grover in his reply to the
newsgroup).
Based upon my experience, obtaining 400VDC @ 300 mA from these stout rectifiers
would be no problem at all.
One of the nicest characteristics that I have found by using the 866A's (besides
the
"retro look" mentioned by Alan when they glow a pretty shade of blue) is that
your output tube can swing alot of current without drastically changing the B+
voltage due to the constant 15 V drop across mercury vapor rectifiers regardless
of anode current (try that with a 5R4GT.......).
Next, Gary mentioned the problem of needing to preheat the filaments of the 83
or 866A prior to applying the B+ voltage. This is because the mercury in the
866A's need to vaporize before they will properly operate. I'm afraid that I
haven't come up with an elegant solution for this situation other than by
providing seperate filament transformers for the 866A's. The filament
requirements for two 866A's is 2.5V @ 10Amps so you need good quality filament
transformers. I just switch the filament transformers on 10 minutes before I
apply voltage to the B+ transformer and everything works great.
Finally, I must say that I am a little bit more anxious about using mercury
filled rectifier tubes after reading about Alan's concerns over breathing
vaporized mercury when a tube breaks, but I guess that if I am willing to go to
work on a daily basis and measure air pressure with a laboratory grade barometer
which contains about 50 cc's of mercury, I shouldn't be too worried about the
866A's. ( Just call me stupid, I guess............)
Steven Kirkpatrick <lau...@bellsouth.net>
chris...@minesprings.com
22 Ohm (typical? from memory) in series. They are then dipped to
insulate the junction, and give only 2 legs on the thing. They are
available in different values. I've got a noise book that tells how
to calculate the optimum values for a given current and load
inductance. I think it said use 1u for every RMS steady state AC amp,
and calculate R so the circuit is over-damped, and doesn't oscillate
(I can't remember how to do that). You may need a higher voltage cap
if you are putting it where it will see greater than 450V or so.
Also, put the snubber in parallel with the inductive load. (Allen
Bradley PLC instructions for relay coils off AC hard contact outputs.)
Now that I've written all this, it doesn't seem to speak to the origin
of this topic. It seems you said there was a snap on turn-on. All
these snubbers are really just for turn-off surges, I think. So maybe
you need an Inrush current limiter/thermistor. Especially if that
rectifier is hot switching into a significant capacitance that is
uncharged. Or maybe just 5 or 10 ohms to kind of damp the surge a
little bit. Rectifiers love that resistor padding going to their
plates. And you probably won't even notice the 5 or 10 volt drop in
B+. Use big watters, cause they do get hot from capacitor charging
surges on every cycle. Even though you may be pulling 100 ma DC, the
rectifier may be dumping spikes of 1 amp or more twice every cycle,
and 3 or 4 amps on turn on.
Chris
James Cross
The point of Alan's discussion was that hot mercury would certainly go into the air
as vapor from a hot tube. I suspect your barometer is not heated.
--
Jim Cross - Vacuum Tubes, Inc. - 513-738-8823 (738-TUBE)
http://www.vacuumtubesinc.com - email to j...@vacuumtubesinc.com
Check us out for new, used, audio, antique, and collectable types.
Vacuum Tubes, Inc.
10995-2A State Rte 128
Harrison, OH 45030