Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

donate to eham
   Home   Help Search  
Pages: Prev 1 [2] 3 4 5 6 7 ... 9 Next   Go Down
  Print  
Author Topic: Command Technologies 2500 amp aka Dedicated RF aka AN Wireless  (Read 95091 times)
N4RSS
Member

Posts: 346




Ignore
« Reply #15 on: January 12, 2012, 07:01:18 AM »

Tom, I think you and I discussed this before, but for the benefit of the others who may have not seen ANY of the older Command Tech amps.  The HF and 6 meter Command Tech amps take a seemingly unique approach to the tank circuit.  In an effort to eliminate padding capacitors for load and tune on the lower bands the output uses a 4 to 1 (matching transformer) so that the tank circuit only has to match down to 200 ohms impedance, the 4 to 1 transformer on the output brings it down to 50 ohms to the antenna.  This design eliminates the padding caps and the switching needed to handle that chore.  A single plate and load capacitor covers all bands.  It sure makes things simple in the tank circuit.  I don't recall any of the other amp mfgers using this method.  Of course when I first noticed this I asked Tom about it and he had considerable input about such a system.  As I recall I think the new Commander amps still carry that design.

QRO does the same thing
Logged
W1QJ
Member

Posts: 2966




Ignore
« Reply #16 on: January 12, 2012, 08:21:29 AM »

That makes sense since they were once partners in the amp business.  They both kept that design I guess but Pat liked triodes instead of tetrodes.
Logged
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #17 on: January 12, 2012, 09:01:50 AM »

Tom, I think you and I discussed this before, but for the benefit of the others who may have not seen ANY of the older Command Tech amps.  The HF and 6 meter Command Tech amps take a seemingly unique approach to the tank circuit.  In an effort to eliminate padding capacitors for load and tune on the lower bands the output uses a 4 to 1 (matching transformer) so that the tank circuit only has to match down to 200 ohms impedance, the 4 to 1 transformer on the output brings it down to 50 ohms to the antenna.  This design eliminates the padding caps and the switching needed to handle that chore.  A single plate and load capacitor covers all bands.  It sure makes things simple in the tank circuit.  I don't recall any of the other amp mfgers using this method.  Of course when I first noticed this I asked Tom about it and he had considerable input about such a system.  As I recall I think the new Commander amps still carry that design.

Pat lived near me, and called me about that idea. We had a long talk about it.

Pat measured amplifier IMD incorrectly by running a two-tone generator into a regular radio. He looked at amplifier output IMD, and found that adding the toroid reduced the measured IMD. He even wanted to patent the idea. I don't know if he tried or gave up.

What was really happening was the toroid added distortion that corrected distortion earlier in the system. The problem with this, is it is only applicable to ONE test setup and one power level.  We can't add additional distortion Willy-nilly and expect the additional distortion to correct linearity in other systems. All it really means is we added distortion.

This is why two-tone tests should never be done with dirty sources. Extra distortion through dumb luck, might, at times, correct other distortion. Then we change one thing, and it suddenly adds distortion. Pat was convinced he had an invention. A way of removing IMD by adding a simple iron core transformer!

As for the network advantage by running into 200 ohms, there isn't a clear advantage.

For a given harmonic suppression, you have to use about the same value capacitors if the network dumps into 200 ohms or 50 ohms, or even larger. The second harmonic suppression is more a function of the capacitance value than the load impedance. It's a far more complex interaction than just changing the impedance to 200 ohms so we can throw in a smaller loading capacitor.

For example, if we design the network for the same Q (let's say 12) and change load impedance between 50 and 200, on 160 meters we have:
200 ohms
Cp = 508
Cload = 1550
L = 20uH
2nd harmonic -23 dB


50 ohms
Cp=525
Cload =2830
L= 16.75 uH
 2nd harmonic -27dB

So we save one single switch contact and 1300 pF padding capacitance, but we lose 4 dB of 2nd harmonic suppression and add significant additional loss, and require a much larger tank coil, for the same network Q.

This is probably why using the 4:1 on the output is a local trend, rather than something that caught on. When the numbers are run, there is really no advantage at all for the same harmonic suppression, it has a little less efficiency, and it costs more.     :-)

Why would we want to do that, when we could add a few contacts, use the same net hardware, and have a pi-L that had better performance? Sometimes new ideas are not better.

73 Tom
 

Logged
KK3AN
Member

Posts: 42




Ignore
« Reply #18 on: January 12, 2012, 01:22:15 PM »

For what it's worth, right now Don and I are looking at possibly replacing the current 'classic' style tank coil completely - along with a number of toroids - with a roller inductor in the new HF models.

Some of the current tank tap points are soldered where they are now on the coil purely out of convenience, and not exactly in the sweet spot for amateur bands.

We like the idea a lot that the roller inductor would allow resonance anywhere in the HF spectrum, based on firmware settings permitting this.

Dan  KK3AN
Logged
K6AER
Member

Posts: 5726




Ignore
« Reply #19 on: January 12, 2012, 02:53:43 PM »

Dan,

That is a nice idea. Wold you use a separate coil for 10 and 12 meters to reduce teh stray inductance?

Mike
Logged
KK3AN
Member

Posts: 42




Ignore
« Reply #20 on: January 12, 2012, 03:05:32 PM »

Mike, honestly decisions like that I am leaving up to Don, as he has more RF experience than I do.

The only worry I had with doing this is the gradual accumulation of dust / dirt on the coil surface, but evidently this is a moot point because Henry used the roller inductor also.

Dan
Logged
W1QJ
Member

Posts: 2966




Ignore
« Reply #21 on: January 12, 2012, 05:12:28 PM »

I suppose a good quality roller insductor would work well.  A cheap one may fail.  Sounds like a wonderful idea if everything can come together around the roller inductor.  Full coverage with exact tank settings would produce stunning results on all bands!!
Logged
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #22 on: January 13, 2012, 09:34:26 AM »

I suppose a good quality roller insductor would work well.  A cheap one may fail.  Sounds like a wonderful idea if everything can come together around the roller inductor.  Full coverage with exact tank settings would produce stunning results on all bands!!

I don't know about stunning.  :-)

For some reason, there is a very popular myth or idea that tank circuits and performance hinges on a certain optimum loaded Q.

There is no real "sweet spot", and loaded Q can vary up to 200% or more with very minimal effect on efficiency. Unless tank components are just horrible, and tank power losses dominate the efficiency, tank loaded Q makes very little difference except for lower order harmonic suppression.

Let's look at an example.. if we lost 1% of power at a Q of 15, and moved Q to 30 with no other changes in unloaded Q, we would lose 2% in the tank. If the tube had 60% efficiency and the tank 99%, and power input was 3000 watts, tank heat would go from 15 watts to 30 watts and output from 1782 watts to 1764 watts if loaded Q was doubled. I don't know that I would call a change from 1764 watts to 1782 watts "stunning". :-)

This is why tube type VHF or UHF amplifiers often run with loaded Q's of 50 or more, sometimes into the 100's, and have good efficiency.

I think the myth that Q greatly influences efficiency comes from the fact we read over and over again about a mystical Q of 12 as a design target. I don't even know where the 12 came from, other than it worked OK in most systems. It certainly is not optimal as an arbitrary number, yet I see people all the time struggle and plan for a Q of 12...or some other pet number.

Rollers have been around since the 1900's. There is a good reason they only appear in limited frequency range systems.

Roller inductors are great in limited frequency range applications, but they are a PITA in wide frequency range applications.
Henry somewhat successfully used rollers in 80-10 meter applications, but as general rule they are a disaster waiting to happen with much more frequency range than that in high impedance systems.

The problem centers around all that extra inductance hanging there unshorted as the roller is cranked up to the upper frequency limit.  We cannot easily implement a progressively shorting connection in a normal roller. When the inductor is sized right for 160 meters, and the roller is cranked out to ten meters, virtually all of the turns are hanging open with the only short near each end of the roller. This pushes the SRF of the roller down into upper HF, which not only kills Q, it also makes a little Tesla coil.

This is why so many antenna tuners with rollers just barely work on 160 meters. If they had the correct size inductor for 160, they would have a dead zone up near 15 or 10 meters. The roller size is compromised to a bare minimum acceptable inductance on 160, just so they don't melt down or go dead on upper HF. 

MFJ tried to fix that with a "resonance killer" that tripped in as the roller moved to low values of inductance, but the mechanics of the system are extremely difficult to reliably implement.

My last roller inductor homebrew all-band amp was in 1972.  I converted it to a band switch and never looked back, although I suppose I could have switched in a 160 meter coil and minimized the inductance of the roller.  That probably would have worked.  :-)

You couldn't sell me on an all band HF amplifier with a roller inductor, unless you could convince me of some realistic advantage. Since there is virtually no performance change with a doubling of loaded Q,  it is pretty simple to do multiple bands with a switch. This is probably why in the entire history of HF amplifiers, everyone does things a similar way. Manufacturers gravitate toward what works acceptably for the cost.

There is no sweet spot for loaded Q. There is a wide range of very gradual performance change as loaded Q is changed.

By the way, if you load your amp to 1500 watts and it has a Q of ~12, and then reload it with less peak drive to 500 watts, now the Q is ~40.

Does it suddenly quit working?    :-)

73 Tom
Logged
N4ATS
Member

Posts: 1303




Ignore
« Reply #23 on: January 13, 2012, 11:05:51 AM »

I repair commander amps as well as most other amps , I just finished a Commander-2500 this week.   http://www.eham.net/reviews/detail/5836
Logged
G3RZP
Member

Posts: 1280




Ignore
« Reply #24 on: January 14, 2012, 08:18:08 AM »

Tom,

I have a 1963 conference paper on an amp that used a variable inductor for tuning, as well as vacuum variables. OK, it was continuous 2 - 27 MHz and 80kW. The coil was made of iridium plated copper tube about 3/4 inch diameter with cooling air forced down it. There were two contacting arms arranged to get rid of the Tesla coil effects at one end and the short circuit turns effects at the other, with special silver loaded carbon brushes to make contact, and some complex parasitic suppressors - they had a problem with a parasitic resonance around 70 MHz enhancing harmonics around there.

Marconi used vacuum variables and switched taps on a pi-L: the 'bandswitch' used cams of Teflon and pushed contacts hard against the coil. Rather better, I feel: variable inductors seem to have a long term problem at high power.

The most succesful variable inductor schemes seem to have been the ones which wind a conductor from a metal drum to an insulated drum: they seem less affected by dirt and dust. Collins, Marconi and Plessey South Africa and probably others have used that scheme.
Logged
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #25 on: January 14, 2012, 09:31:22 AM »

Tom,

I have a 1963 conference paper on an amp that used a variable inductor for tuning, as well as vacuum variables. OK, it was continuous 2 - 27 MHz and 80kW. The coil was made of iridium plated copper tube about 3/4 inch diameter with cooling air forced down it. There were two contacting arms arranged to get rid of the Tesla coil effects at one end and the short circuit turns effects at the other, with special silver loaded carbon brushes to make contact,

There you go. The dual contacting arms shorting the unused turns at mid unused area would solve the problem with broad bandwidth systems using roller inductors. A single shorting bar does not make a good broadbandwidth roller.

MFJ tried to handle this with a trip level that pushes a shorting contact into the 1/2 way point when the roller rolls up below a certain minimum inductance, in effect adding the second shorting bar a long distance from the primary bar, but it is a mechanical nightmare.

Quote
Marconi used vacuum variables and switched taps on a pi-L: the 'bandswitch' used cams of Teflon and pushed contacts hard against the coil. Rather better, I feel: variable inductors seem to have a long term problem at high power.


They sure do. Not only do they have problems with Tesla effect and series resonances sucking out signal on some frequencies, any payback is extremely low. Even a 2:1 change in loaded Q is generally unnoticable, so there is no real sweet spot. There is a wide range of Q where things work about the same.
 
Quote
The most succesful variable inductor schemes seem to have been the ones which wind a conductor from a metal drum to an insulated drum: they seem less affected by dirt and dust. Collins, Marconi and Plessey South Africa and probably others have used that scheme.

I've got some rollers like that here. It is a good scheme, but they never looked like any cost or performance advantage over just adding a few taps to a regular system.

The roller inductors main advantage is getting rid of an expensive switch, but the switch allows a progressive pick up and hold contact that prevents unwanted resonances.

73 Tom
Logged
VE7RF
Member

Posts: 212




Ignore
« Reply #26 on: January 15, 2012, 03:17:18 AM »

This is what happens when engineers that don't even realize it is impossible to build a 4:1 current balun on a single core start dealing with more complex systems like amplifiers. A buddy of mine bought a deluxe expensive auto-tuner, and it had a forced 3:1 voltage imbalance on the balanced output terminals. It created terrible RFI.

Who would miss that problem that understood RF?

It all turns into a mess when things are not looked at carefully.

Anyone who has dealt with a certain person who goes in and out of business and through partners like water already understands what probably happened. Even his own relatives have been shafted in business deals.


## QST tested several linear amps  in one issue. This was years ago. One of them , [ I think it was qro], used the 4:1 un-un..and the 200 ohm output at the end of the pi net.   If u read the qst write up.. the 4:1 un-un blew up.  Can you really get -45 db 2nd harmonic suppression using a simple pi and this 4:1 un-un scheme?

##  I  thought PI nets were good for like -35 db max?   Even if u doubled the loaded Q from 10 to 20.. the harmonic suppression only goes up 3db.  Meanwhile, you cook the coils and bandswitch cuz of the higher circulating current.  Ameritron manages to get > -45 db or more using pi nets on 40-10m..which baffles me.

later... Jim  VE7RF
Logged
VE7RF
Member

Posts: 212




Ignore
« Reply #27 on: January 15, 2012, 03:21:47 AM »

For a given harmonic suppression, you have to use about the same value capacitors if the network dumps into 200 ohms or 50 ohms, or even larger. The second harmonic suppression is more a function of the capacitance value than the load impedance. It's a far more complex interaction than just changing the impedance to 200 ohms so we can throw in a smaller loading capacitor.

For example, if we design the network for the same Q (let's say 12) and change load impedance between 50 and 200, on 160 meters we have:
200 ohms
Cp = 508
Cload = 1550
L = 20uH
2nd harmonic -23 dB


50 ohms
Cp=525
Cload =2830
L= 16.75 uH
 2nd harmonic -27dB

So we save one single switch contact and 1300 pF padding capacitance, but we lose 4 dB of 2nd harmonic suppression and add significant additional loss, and require a much larger tank coil, for the same network Q.

This is probably why using the 4:1 on the output is a local trend, rather than something that caught on. When the numbers are run, there is really no advantage at all for the same harmonic suppression, it has a little less efficiency, and it costs more.     :-)

Why would we want to do that, when we could add a few contacts, use the same net hardware, and have a pi-L that had better performance? Sometimes new ideas are not better.

73 Tom
 


[/quote]

##  and how are u supposed to meet the >44 db FCC spec with either the 200 or 50 ohm example you depict Huh

Later... Jim  VE7RF
Logged
VE7RF
Member

Posts: 212




Ignore
« Reply #28 on: January 15, 2012, 04:25:27 AM »

I suppose a good quality roller insductor would work well.  A cheap one may fail.  Sounds like a wonderful idea if everything can come together around the roller inductor.  Full coverage with exact tank settings would produce stunning results on all bands!!
[/quote]

I don't know about stunning.  :-)

For some reason, there is a very popular myth or idea that tank circuits and performance hinges on a certain optimum loaded Q.

There is no real "sweet spot", and loaded Q can vary up to 200% or more with very minimal effect on efficiency. Unless tank components are just horrible, and tank power losses dominate the efficiency, tank loaded Q makes very little difference except for lower order harmonic suppression.

Let's look at an example.. if we lost 1% of power at a Q of 15, and moved Q to 30 with no other changes in unloaded Q, we would lose 2% in the tank. If the tube had 60% efficiency and the tank 99%, and power input was 3000 watts, tank heat would go from 15 watts to 30 watts and output from 1782 watts to 1764 watts if loaded Q was doubled. I don't know that I would call a change from 1764 watts to 1782 watts "stunning". :-)

This is why tube type VHF or UHF amplifiers often run with loaded Q's of 50 or more, sometimes into the 100's, and have good efficiency.

I think the myth that Q greatly influences efficiency comes from the fact we read over and over again about a mystical Q of 12 as a design target. I don't even know where the 12 came from, other than it worked OK in most systems. It certainly is not optimal as an arbitrary number, yet I see people all the time struggle and plan for a Q of 12...or some other pet number.

Rollers have been around since the 1900's. There is a good reason they only appear in limited frequency range systems.

Roller inductors are great in limited frequency range applications, but they are a PITA in wide frequency range applications.
Henry somewhat successfully used rollers in 80-10 meter applications, but as general rule they are a disaster waiting to happen with much more frequency range than that in high impedance systems.

The problem centers around all that extra inductance hanging there unshorted as the roller is cranked up to the upper frequency limit.  We cannot easily implement a progressively shorting connection in a normal roller. When the inductor is sized right for 160 meters, and the roller is cranked out to ten meters, virtually all of the turns are hanging open with the only short near each end of the roller. This pushes the SRF of the roller down into upper HF, which not only kills Q, it also makes a little Tesla coil.

This is why so many antenna tuners with rollers just barely work on 160 meters. If they had the correct size inductor for 160, they would have a dead zone up near 15 or 10 meters. The roller size is compromised to a bare minimum acceptable inductance on 160, just so they don't melt down or go dead on upper HF. 

MFJ tried to fix that with a "resonance killer" that tripped in as the roller moved to low values of inductance, but the mechanics of the system are extremely difficult to reliably implement.

My last roller inductor homebrew all-band amp was in 1972.  I converted it to a band switch and never looked back, although I suppose I could have switched in a 160 meter coil and minimized the inductance of the roller.  That probably would have worked.  :-)

You couldn't sell me on an all band HF amplifier with a roller inductor, unless you could convince me of some realistic advantage. Since there is virtually no performance change with a doubling of loaded Q,  it is pretty simple to do multiple bands with a switch. This is probably why in the entire history of HF amplifiers, everyone does things a similar way. Manufacturers gravitate toward what works acceptably for the cost.

There is no sweet spot for loaded Q. There is a wide range of very gradual performance change as loaded Q is changed.

By the way, if you load your amp to 1500 watts and it has a Q of ~12, and then reload it with less peak drive to 500 watts, now the Q is ~40.

Does it suddenly quit working?    :-)

73 Tom
[/quote]

## I don't get it. I tried using high loaded Q's in the past, and the result was always sky high circulating tank current. [ + it had to be tweaked more often and was narrow banded].  A Q of 30 ??   IMO..u may as well douse it with lighter fluid and toss in the zippo.   With  1.5A  to 2.0 A of plate current..and a Q of 20-30..implies   30-60A of rf current.  The poor bandswitch would be fried way before then. 

##  Load ur 1.5 kw amp to only 500W..and Q will triple..and eff goes straight to hell every time.  You would have to reduce the plate V + plate current..such that the plate load Z remained the same.  Single digit loaded Q..like 8-9..or 10..is good enough for my hb amps.  Nothing gets cooked, eff is good, and it's more broad banded.  I use monoband ants, so harmonic suppression is a bottom priority on the design..and also why I never use a PI-L.  However I use a L-PI. I install a  typ .66 uh  tubing coil, between plate block caps and main PI net.  The 33 pf of anode to grid C on the 3CX-3000A7 + the .66 uh forms an L network..and transforms the plate load Z way down on 15M.  Then the Q is down to 8 on 15M. [which then rises to Q=12 on 17M, since the 15 M position gets re-used in that set up].

##  henry used a cam activated assy, to add the extra short on the roller it used on some of it's amps.  On the the 8K ultra, they center tap the coil, and short out the low freq end with  a hv solenoid.  Now that only works on rollers  where the coil DOES NOT ROTATE.  That high freq suck out problem started up once the 160m band is added to an amp.

##  On my buddy's  YU-148 amp, we used a model 85 switch..and it's a 12 x position, and NOT a shorting type. [ multi-tech calls a shorting switch a .." multiple pole rotor assy".  It works fine on 160-15m [ no 10+12m].  Instead of adding more load cap padder's, I suggested we add a small coil at the 160m end..and lower the tank Q.  That worked..BUT the entire upper hf end, 20-17-15M got sucked out..when doing the mfj test on the tank. [ resistor from anode to chassis, and mfj on output].

##  I found that using the mfj + resistor test, that on a normal amp, once the tune/load caps tweaked for min swr, that you can then move the freq of the mfj 259 up/down quite a bit..and also note the 2:1 swr points.  These 2:1 points would double every time u doubled the freq.  Like 67 khz on 160m, 140 khz on 80m, 300 khz on 40m, 600 khz on 20m etc, etc.  When the top end gets sucked out, these 2:1 points become extremely narrow, like 50 khz on 20m..instead of the usual 600 khz.   We found that by just shorting out 1/3 of the new coil, the problem vanished.  We ended up tossing the new coil..and just used more load cap padders on 160m.

##  On my latest project, I use 3 wafers of a model 88 switch in parallel.  tank taps done on wafer #3..and the common of wafer #1 goes to the vac load cap..[+lf end of coil].  That way, the current divides up equally between the 3 x wafers.  I modified it into a shorting switch. I had multi-tech build me.. " 5 x pole rotors".  Installed one on each of the 3 x wafers.  Now this is only a 10 x pos switch, [36 deg index] so I end up with only 6 x usable positions.   This is for a 160-80-40-20-17-15m amp.  When switched to 160m..the 5th rotor  on the 5 pole rotor now sits adjacent to the 15 m contact.  Nothing I can do abt that.   Their model 85 has 12 x positions [30 deg index] .

##  I use a roller coil on the 3CX-3000A7 amp. +  fixed tubing for 20-17/15m.  I used a vac relay to provide the extra short on the roller coil...when on the higher bands.  A modified hv solenoid does the combo 17/15m tap.  It all works, but I would not use a roller coil again in a HF amp, pita.

##  Alpha plans to use their own in house edge wound roller + 2 x vac caps + balun on the  input of the their new T type ant tuner. Their own version roller coil is similar to  the sson RI-40, in that it uses a pair of contacts on either side of the ribbon, to pinch it firmly.  The new alpha roller coil is built like a tank, best roller I have ever seen.  It's also the most expensive automatic T type tuner ever made, bar none. However, they must have got a bargain on the pair of 7500 V vac caps.  It's on their site.

Later..Jim  VE7RF
Logged
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #29 on: January 15, 2012, 06:07:54 AM »

## I don't get it. I tried using high loaded Q's in the past, and the result was always sky high circulating tank current. [ + it had to be tweaked more often and was narrow banded].  A Q of 30 ??   IMO..u may as well douse it with lighter fluid and toss in the zippo.   With  1.5A  to 2.0 A of plate current..and a Q of 20-30..implies   30-60A of rf current.  The poor bandswitch would be fried way before then.
 

Certainly higher Q reduces bandwidth, and it also increases circulating currents, but there isn't the narrow peak of efficiency people sometimes think.

With a good tank system quite large changes in Q are tolerable.

Why do you think plate current times Q equals tank current? Where does that come from?

Tank current is related to loaded Q, anode load impedance, and RF power. With 1500 watts and 2000 ohms, you'd have sqrt of P/R as current into the tank, times Q, for currents.

True enough extra Q increases circulating currents and decreases bandwidth,  but there will be very little change in efficiency unless the tank is lousy. The only way to find optimum Q is to measure harmonic suppression, and set the Q as low as possible for the target suppression that still allows matching the lowest possible load impedance.

Quote
##  Load ur 1.5 kw amp to only 500W..and Q will triple..and eff goes straight to hell every time.  You would have to reduce the plate V + plate current..such that the plate load Z remained the same.


That's not because the Q tripled, Jim. That is primarily because the anode load impedance does not match the optimum anode voltage and currents. The anode load impedance is too low, and so the tube has higher current in proportion to voltage, resulting in more heat in the tube. Secondarily, it can be because conduction angle increases.

It has very little to do with Q.

« Last Edit: January 15, 2012, 06:10:21 AM by W8JI » Logged
Pages: Prev 1 [2] 3 4 5 6 7 ... 9 Next   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!