Author Topic: Sperry Ratio Potentiometer Teardown  (Read 14329 times)

0 Members and 1 Guest are viewing this topic.

Offline amspireTopic starter

  • Super Contributor
  • ***
  • Posts: 3802
  • Country: au
Sperry Ratio Potentiometer Teardown
« on: November 28, 2011, 07:50:27 am »
While thinking about building a Kelvin Varley Divider, I came across this on ebay.

No-one had bid at the time, and it had a magnificent row of decade switches across the bottom. How can you go wrong with that! I bid and got it for $16.50.  I wouldn't bother to post a teardown here unless it had something really special inside, and luckily for me, it did.



So first, some background. It was made by Sperry Piedmont for Lockheed Martin around 1960 so it is over 50 years old. The costs was $2,600 which was half the family average income in the US at the time. It has a serial number "72" - not many were made. This one was used by the Australian Air Force. Last calibrated in 2000 and so it all ties in with the notion that this is used for calibrating avionics for the Lockheed Martin Orion P-3C. Around 2000, the Australian P-3C's were upgraded to AP-3C's with all new avionics, so this instrument probably became obsolete. The Orions are one of those planes that just will not die. They are 50 years old, and right now, the US Coast Guard is upgrading theirs to last for a minimum of another 20 years.

I will not try and use this - I will be pulling it apart. It is very heavy, it needs 120V/400Hz supply and the electronic meter for measuring the potentiometer null has a maximum sensitivity of 3mV full scale. So there is nothing magical about the valve based electronics. It that part of the device, I can salvage some good switches, some nice transformers, a few wirewound resistors, a nice meter movement, some real museum-piece germanium diodes and some General Radio capacitors.

The case is brilliant, and I will be keeping that. Completely weather proof and the rubber seal in the groove on the lid is still soft after 50 years. The clips are really tight. Fantastic!

https://www.eevblog.com/forum/index.php?action=dlattach;topic=5726.0;attach=17488

The inside of the lid had the circuit printed on an anodized aluminium sheet, and I have attached it below. But when I first looked at it, I was puzzled - the schematic had no decade switches on it at all!

Then I saw this down the very bottom right corner:



That must be the switches. What is a Gertsch Standard Ratio Transformer?

I went on to the internet, and discovered that they are basically  exactly the same as the Tegam Ratio Transformers that IET sells today. In fact this Tegam divider probably has the same transformers that my device has, except that I have the better heavy duty switches.

http://www.tegam.com/product.asp?modelNumber=RT-60B

I will go into the details of this transformer in a second post, but for now, lets just say they are amazing - unbelievable really. Gertsch made standalone dividers, but their main sales was to makers of avionics test gear for the military. So the transformers, the switches and the wiring is all 100% Gertsch and it is just put straight into my device.  Being a transformer, we are looking at an AC bridge here and it just uses the 400Hz mains to power the transformer.

The reason I am saying this is there are obviously other old pieces of gear floating around with the exactly same dividers. Keep an eye out for that row of decade switches on old military test gear.

Here is the guts:



You can see the row of Gertsch switches along the top of the photo. 

One of the things that really got me was look at the wiring. The wiring that Sperry has done is all with white wire! The 400V DC supply wire  is the same colour as the 120V mains input wires, 6.3V filament wires and the millivolt signal wiring from the bridge. What could possibly go wrong with that?

The guy who built these obviously had one gigantic reel of white wire, and he couldn't be happier! I am glad I am not trying to fix it.

The Gertsch divider is 5 decades with 3 decades handled by one transformer, and the other two decades by a second transformer. A potentiometer provides the last decade of the output. Here is the primary transformer box:



So I have the Gertsch RT-7.

That Variac under it is one of the nicest I have ever seen or used. A fabulous light wirewound feel to it. 120V in, 0 to 134V out and even though they are using it at 400Hz, it is rated for 60Hz. 165VA. A pity it isn't 240V like our mains. I may have to sell it.



The Gertsch switches are custom built for the job. It is a 10 position switch, but they double up on every contact, so that means 20 main contacts. This is the heavy duty switch that everyone who knows these wants to see.  2 milliohm contact resistance after 40 years of use and the last 10 years doing nothing. The switches are really heavy to use - you really grab these knobs. None of this wimpy fingertip stuff.

But there is a big problem if you want to switch transformer windings.  You never want to let a winding go open circuit while it is carrying current, or you will get a transient.  But you also never want to short two windings together. So you cannot use a make-before-brake switch, and you cannot use a break-before-make switch.  What is the solution?



The solution is to add another 18 intermediate contacts connected to the windings and to have make-before-break contacts that go from one winding to the resistor to the second winding.  On the first decade the resistors are 2k2, and the other decades have 47 ohm resistors. The resistors are normally open circuit, and they switch in across the windings as you turn the switch.

I am going to strip the guts out, cut the top off the front panel so I just have the Gertsch divider switches, add some 4mm terminal sockets, and built a box  to mount the panel and transformers in. I have some Bakelite sheets, so I might make the box from that for a retro look. When I build me Kelvin-Varley resistive divider, I want it to look the same so I need to work out a common style.

Back to what is special about this transformer divider. Why would you want one?  I will cover that in the second post.

If you want to look at the schematic, it is below. It is basically an AC amplifier that can go from 3mv full scale to 300volts full scale. It can chop the signal to be able to do phase sensitive nulling. You can adjust the phase from -135 degrees to 135 degrees. It only works at 400Hz and has a built in 400Hz bandpass filter to reject noise and harmonics.

Richard
« Last Edit: November 28, 2011, 08:41:05 am by amspire »
 

Offline amspireTopic starter

  • Super Contributor
  • ***
  • Posts: 3802
  • Country: au
Re: Sperry Ratio Potentiometer Teardown
« Reply #1 on: November 28, 2011, 02:09:25 pm »
I will post some details on the transformer, but in the meantime, the Tegram RT-60 manual is almost identical to my RT-7 - it is all based on the same 1956 design that hasn't changed since. This document also contains a lot of the original Gertsch documentation.

http://www.tegam.com/PDF_Files/Resources/RT-60B-901-01%20RevB.pdf

Also if you are interested in a ratio transformer, check out ebay item 170677230036 (end in a day or so). This is for a ESI Dekatran DT45 coaxial transformer. ESI was Gertsch's main competitor and they made superb quality transformers.

 

HLA-27b

  • Guest
Re: Sperry Ratio Potentiometer Teardown
« Reply #2 on: November 28, 2011, 06:01:32 pm »
Looks very nice Richard. It can house your KVD a null detector and voltage reference all in one. You can even put a battery to completely isolate it from outside. 
 

Offline amspireTopic starter

  • Super Contributor
  • ***
  • Posts: 3802
  • Country: au
Re: Sperry Ratio Potentiometer Teardown
« Reply #3 on: November 28, 2011, 10:09:25 pm »
Looks very nice Richard. It can house your KVD a null detector and voltage reference all in one. You can even put a battery to completely isolate it from outside.

It is just too big and heavy for that. The case may be great one day if ever I need to set up some kind of field electronics setup in a weatherproof box. It has room for batteries, and inverter, meters, a digital oscilloscopes, etc.

But I do need to make a null detector, and I will probably use the meter movement.

Richard
 

alm

  • Guest
Re: Sperry Ratio Potentiometer Teardown
« Reply #4 on: November 28, 2011, 10:16:19 pm »
Impressive piece of kit. Looks like it would be fun to take this through the US TSA ;).

How much did you pay in shipping? It looks quite heavy. Or was it local pick-up?
 

Offline amspireTopic starter

  • Super Contributor
  • ***
  • Posts: 3802
  • Country: au
Re: Sperry Ratio Potentiometer Teardown
« Reply #5 on: November 28, 2011, 10:51:27 pm »
What Is So Good about a Ratio Divider Transformer?

It is a simple device. Put 1V 1KHz across the input windings and you can set the switches for any output volts from 0 to 1V with better then 1uV resolution.

But lets first look at the specs of a good modern multimeter. For measuring relative values like the gain of an amplifier, accuracy is not important, but reading uncertainty figures that include non-linearity as a component are signifigant.

If you take the Agilent 34411a, at 1 KHz, reading uncertainty is 0.03% of full scale (1 year calibration), and if you use it in cold or hot temperatures, this can double to 0.06%.

They are pretty good numbers. There are a lot of expensive meters that are an order of magnitude worse.

The Gertsch RatioTran Transformer Ratio Divider (like my RT-7)  is 0.001% accurate - 30 to 60 times better then the 34411A.  But that is only part of the story.

The divider is accurate from -15 degC to 85 degC.

Most importantly of all, as long as the switches are in good condition and no-one has damaged the windings with an overload, the accuracy is always the same as the day it was built.  No 24 hour - 30 day - 1 year accuracy stuff. 

I have checked out my divider and everything is in good condition, so its accuracy will be the same as it was 50 years ago. Here is a quote from the Gertsch documentation:

Quote
A careful study of the distribution of resistance and reactance in RatioTran leads us to believe that it is extremely improbable that the specified values of error will ever be exceeded. The accuracy of RatioTran does not change appreciably with age and periodic calibration checks are unnecessary. Any periodic checks should be directed chiefly at detecting malfunctioning switches or potentiometers.

I cannot think of any other precision electronic device that is  10 ppm accurate, and you don't need to bother about calibration - ever.

It is in effect the AC equivalent to a Kelvin-Varley Divider for DC but with some big advantages.  The KVD has a very high output impedance so that the load of a 10 Mohm meter is enough to cause errors that can be up to 1%.  A 10 Mohm meter load on a Transformer Ratio Divider will cause an error less then 1 part per million.

If you start looking into metrology, all the most accurate readings  for resistance, capacitance and inductance usually end up involving ratio divider transformers.

When you look at the equivalent model for the transformer, the specs are unbelievable.

  • Inductance: 500H.
  • Output leakage inductance: < 75 uH
  • Output resistance: <2.5 ohms.
  • Phase error of output at 1Khz:  < 0.003 degrees

These are really amazing numbers.

So if you want a KVD for DC, you probably want to get a transformer divider for AC, and they are getting harder to get. Very few are being made any more, and if you have one, it can easily last another 50 years.
« Last Edit: November 28, 2011, 11:24:38 pm by amspire »
 

Offline amspireTopic starter

  • Super Contributor
  • ***
  • Posts: 3802
  • Country: au
Re: Sperry Ratio Potentiometer Teardown
« Reply #6 on: November 28, 2011, 11:05:21 pm »
Impressive piece of kit. Looks like it would be fun to take this through the US TSA ;).

How much did you pay in shipping? It looks quite heavy. Or was it local pick-up?

It was sent by post and I think they charged $34. So it ended up costing me a total of just over $50.  If I start selling off the parts I do not need like the variac and some of the very nice transformers, I could probably make that back.
 

HLA-27b

  • Guest
Re: Sperry Ratio Potentiometer Teardown
« Reply #7 on: November 28, 2011, 11:39:23 pm »
What Is So Good about a Ratio Divider Transformer?
...
...
So if you want a KVD for DC, you probably want to get a transformer divider for AC, and they are getting harder to get. Very few are being made any more, and if you have one, it can easily last another 50 years.


As you describe it, it is perfectly serviceable and very accurate. In that case why would you want to dismantle it? If it can measure reactance impedance etc. to those accuracies wouldn't you like to keep it as it is?
 

alm

  • Guest
Re: Sperry Ratio Potentiometer Teardown
« Reply #8 on: November 28, 2011, 11:55:20 pm »
It's quite plausible that these transformers are very accurate, not a lot of variation in the turns ratio after all, but where are you going to find an AC source with 10 ppm accuracy, or even 10 ppm stability? A Fluke 5522A calibrator designed for calibrating DMMs with up to 6.5 digits does 120 ppm on its best range within 90 days after cal, no 24h or transfer accuracy is specified. And you'd have to improve at least ten times on this very expensive piece of equipment (granted, you're also paying for all the other functions). With DC it's feasible to get a reference stable within 10 ppm short term for a modest cost (eg. LM399), but I don't see a cheap way to do the same for AC. What's the point of a divider if you don't have an accurate reference to divide?
 

Offline amspireTopic starter

  • Super Contributor
  • ***
  • Posts: 3802
  • Country: au
Re: Sperry Ratio Potentiometer Teardown
« Reply #9 on: November 29, 2011, 12:25:53 am »
Ratio dividers are for accurately generating and measuring ratios. In my mention of DMM specs, I was careful to only talk about reading uncertainties.

So they are used for comparing parts - say you have a 0.01% 1nF standard capacitor and you need to make a very accurate 5.21nF capacitor.

Checking linearity, so it is good enough to see how good the AC linearity of any DMM on the market actually is.

If you have a Fluke 540a, you can set a generator to output an AC sinewave to a 0.01% accuracy, but the Fluke can only measure down to 0.5V RMS.  With a ratio divider, you can take that 0.5 V and dial up any voltage you like down to below 0.5 uV. It tends to be more accurate then resistive dividers for AC as it does not have the thermal component drift if you are using a low impedance divider, and it is less sensitive to component and stray capacitance and inductances  for a high impedance divider.

If you do make a precise resistive divider for AC, you can use the ratio divider to check the calibration and to check for minute phase angle errors that would indicate stray capacitance or inductance issues.

You might want to make a very precise x10 amplifier.  How are you going to calibrate it for AC gain if you are struggling to get 0.1% relative AC accuracy if you use a DMM for gain measurements? Sure you can set up the DC gain super accurately, but you will not know how accurate the AC gain is.

Of course if a DMM does everything you need, then you do not need a ratio divider which is good as there are not that many to go around. There seems to be less and less affordable dividers available.

Richard
« Last Edit: November 29, 2011, 12:58:06 am by amspire »
 

alm

  • Guest
Re: Sperry Ratio Potentiometer Teardown
« Reply #10 on: November 29, 2011, 12:58:21 am »
You fail to address my point. I'm not arguing that it's not more accurate, but I'm wondering what source you have that is stable enough that the 10ppm accuracy is of any use. To verify linearity, you need a source that's very stable, otherwise the stability of your divider is meaningless. Uncertainty of the signal from the divider is the (square root) of the sum of the (squares of the) uncertainty of the divider and the uncertainty of the voltage source. If the latter is 0.1%, what good is a 10 ppm divider?

What source is even 0.1% stable? Is a good function generator that stable? Maybe for a very short time. I might hook one up to a DMM one day to check. I tend to use horribly unstable sources (eg. LM723-based power supplies) as voltage references for comparing equipment because I don't want to spend the $$$ for a proper single purpose references from 0.1 to 100V or so. Getting anything useful becomes an interesting exercise in statistics :).

The problem of a stable source gets worse as the voltage and frequency goes up. Since you can only divide down, a higher voltage is quite welcome, unless you want to measure a higher voltage with a null detector.

A more common way of AC calibration would probably be to use a thermal AC-DC transfer standard, these also go down to the 10 ppm range. I wouldn't know where to get one for $50, but I don't think I'd be able to get a divider transformer for $50 either. Still doesn't solve the problem of the source.
 

Offline amspireTopic starter

  • Super Contributor
  • ***
  • Posts: 3802
  • Country: au
Re: Sperry Ratio Potentiometer Teardown
« Reply #11 on: November 29, 2011, 01:37:35 am »
You fail to address my point. I'm not arguing that it's not more accurate, but I'm wondering what source you have that is stable enough that the 10ppm accuracy is of any use. To verify linearity, you need a source that's very stable, otherwise the stability of your divider is meaningless. Uncertainty of the signal from the divider is the (square root) of the sum of the (squares of the) uncertainty of the divider and the uncertainty of the voltage source. If the latter is 0.1%, what good is a 10 ppm divider?
For measuring relative values like comparing a device to a reference part, or measuring gains needs no signal stability at all. The Sperry just used the 120V/400Hz mains as its signal source. When you are looking for nulls, all signal level errors have no effect.  In terms of signal level, many signal generators have enough short term stability.  I find the HP 33120A pretty good for about 1V RMS output.   If I crank the output up to maximum, I do see thermal drift.

It is not that hard building pretty stable AC sources for a fixed frequency. A squarewave generated from a DC reference bus going through a passive bandpass filter tends to be very stable. If instead of a squarewave, you have a properly crafted PWM signal, you end up with no harmonics until the 9th or above, and a filter can result in a very low distortion and very stable sinewave.
Quote
What source is even 0.1% stable? Is a good function generator that stable? Maybe for a very short time. I might hook one up to a DMM one day to check. I tend to use horribly unstable sources (eg. LM723-based power supplies) as voltage references for comparing equipment because I don't want to spend the $$$ for a proper single purpose references from 0.1 to 100V or so. Getting anything useful becomes an interesting exercise in statistics :).

As I mentioned, the Fluke 540A can be used to set up an AC output to better then 0.01% and there are a lot of stable sources. A function generator using a a digital to analog converter rather then an amplitude stabilized oscillator should be stable as a source.

A number of people in the forum, including myself, have a 540a. When you use it, you see straight away what sources are stable. If the source is not stable, you can never get a proper reading. The needle just keeps drifting.

There are newer technology thermal transfer devices from Fluke and others that can easily do the 10ppm, but the 540a is the one that can be picked up cheaply, if you are lucky. I think mine was about $100 and it is accurate without calibration.
Quote

The problem of a stable source gets worse as the voltage and frequency goes up. Since you can only divide down, a higher voltage is quite welcome, unless you want to measure a higher voltage with a null detector.

A more common way of AC calibration would probably be to use a thermal AC-DC transfer standard, these also go down to the 10 ppm range. I wouldn't know where to get one for $50, but I don't think I'd be able to get a divider transformer for $50 either. Still doesn't solve the problem of the source.
The recent prices of the Fluke 540a seem to have skyrocketed recently. I think 5 years ago, people thought they were an old bit of useless old junk and I saw some crazy low prices in the US. Now they seem to be $400 or more. Yuck!

The ratio transformers are good for low frequencies only, so they have full accuracy at 1KHz and by 10KHz, they are down to 0.1% accuracy. But having said that, when they have errors, they always have the same error as the error is the result of the physical location of wires and that never changes. So if I used the 540A or something better to calibrate a particular divider setting accurately at 20KHz, then the calibration factor would always be accurate.

The ratio divider does not do everything, but what it is good at, it is exceptionally good at. A DMM needs thousands of things to work right to give a good reading, and unless you calibrate regularly, you can never be sure if some small part of the meter is not working right. The divider only needs the switches to work and a simple resistance test with a DMM can find any switch problem.

Richard
« Last Edit: November 29, 2011, 02:00:17 am by amspire »
 

Online lowimpedance

  • Super Contributor
  • ***
  • Posts: 1281
  • Country: au
  • Watts in an ohm?
Re: Sperry Ratio Potentiometer Teardown
« Reply #12 on: November 29, 2011, 02:51:59 am »
Gotta love this old gear, built to last or become obsolete before wearing out.
I'll bet the wire was bought from the Ford motor company cheap as it was the wrong color :D.

Those are serious switches, and what a score with the Gertsch transformer. well done.

Again here at work there are a number of this manufacturers transformers, ie
RT-60 5 decade and the RT-7A 6 switches (built into a traditional looking decade resistor box!)
John
The odd multimeter or 2 or 3 or 4...or........can't remember !.
 

Offline amspireTopic starter

  • Super Contributor
  • ***
  • Posts: 3802
  • Country: au
Re: Sperry Ratio Potentiometer Teardown
« Reply #13 on: November 29, 2011, 03:36:50 am »
Gotta love this old gear, built to last or become obsolete before wearing out.

That is what is impressive. Nothing cracked or broken. A near airtight case so there is no corroded metal. The front panel and brackets looks like it is 4mm steel. I am used to seeing valves and HT connections covered in dust in old gear, and there is not a spec of dust in this one.

From the condition, it could have been made a year ago. This one supposedly wore out due to an "undiagnosed fault" diagnosed early this year. Since it has probably been obsolete for the last 10 years, I think the fault was that no-one knew what it did any more, and so why bother plugging it in?

Richard
 

Online BravoV

  • Super Contributor
  • ***
  • Posts: 7550
  • Country: 00
  • +++ ATH1
Re: Sperry Ratio Potentiometer Teardown
« Reply #14 on: November 29, 2011, 03:53:16 am »
Thanks for sharing, great photoshoots & explanations, really love it.

Just curious when seeing the Gertsch switches, are those golden contact pins inserted/secured at a wood plank ? That thick brown thingy looks like wood to me.




Offline amspireTopic starter

  • Super Contributor
  • ***
  • Posts: 3802
  • Country: au
Re: Sperry Ratio Potentiometer Teardown
« Reply #15 on: November 29, 2011, 04:06:11 am »
Thanks for sharing, great photoshoots & explanations, really love it.

Just curious when seeing the Gertsch switches, are those golden contact pins inserted/secured at a wood plank ? That thick brown thingy looks like wood to me.



That is high quality Bakelite. It has to be machined - you cannot mold it and get the same quality and strength as you get in this laminated sheet. Using it was not a cheap option. The light is just picking up the saw cut.
 

Online BravoV

  • Super Contributor
  • ***
  • Posts: 7550
  • Country: 00
  • +++ ATH1
Re: Sperry Ratio Potentiometer Teardown
« Reply #16 on: November 29, 2011, 04:13:24 am »
That is high quality Bakelite. It has to be machined - you cannot mold it and get the same quality and strength as you get in this laminated sheet. Using it was not a cheap option. The light is just picking up the saw cut.

Apparently those cuts make it appeared like a wood's texture. :D

Looks like this switch will last for a very long time, a refresh on the dried lubrication at the shaft or bearing, it will be as good as new again.

Offline amspireTopic starter

  • Super Contributor
  • ***
  • Posts: 3802
  • Country: au
Re: Sperry Ratio Potentiometer Teardown
« Reply #17 on: November 29, 2011, 04:53:13 am »
Looks like this switch will last for a very long time, a refresh on the dried lubrication at the shaft or bearing, it will be as good as new again.

With these heavy duty switches, if the wipers get worn, you can dismantle them, move the wiper disk on the shaft closer to the contacts and reassemble. But it doesn't look like I have a problem there. I probably need to clean and relubricate the contacts. You cannot do that apparently with the smaller switches on the other Gertsch models.
 

Online BravoV

  • Super Contributor
  • ***
  • Posts: 7550
  • Country: 00
  • +++ ATH1
Re: Sperry Ratio Potentiometer Teardown
« Reply #18 on: November 29, 2011, 05:12:29 am »
With these heavy duty switches, if the wipers get worn, you can dismantle them, move the wiper disk on the shaft closer to the contacts and reassemble. But it doesn't look like I have a problem there. I probably need to clean and relubricate the contacts. You cannot do that apparently with the smaller switches on the other Gertsch models.

What kind of lubricant that you're going to use at those gold plated contact points ? They're gold right ?

Really curious and like to know, cause all this time since I don't have any clue and worry doing that without knowing whether the lubricant might affect the contact resistance.
« Last Edit: November 29, 2011, 05:15:20 am by BravoV »
 

Offline amspireTopic starter

  • Super Contributor
  • ***
  • Posts: 3802
  • Country: au
Re: Sperry Ratio Potentiometer Teardown
« Reply #19 on: November 29, 2011, 05:46:02 am »
What kind of lubricant that you're going to use at those gold plated contact points ? They're gold right ?

Really curious and like to know, cause all this time since I don't have any clue and worry doing that without knowing whether the lubricant might affect the contact resistance.

There may be silver plate on the top of the gold contacts - it was a bit hard to see.

I am not sure about the best lubricant. Does anyone know?

Richard
 

Online BravoV

  • Super Contributor
  • ***
  • Posts: 7550
  • Country: 00
  • +++ ATH1
Re: Sperry Ratio Potentiometer Teardown
« Reply #20 on: December 01, 2011, 04:33:14 am »
There may be silver plate on the top of the gold contacts - it was a bit hard to see.

I am not sure about the best lubricant. Does anyone know?

Richard, maybe for these kind of precious metal contact points, they don't need lubricant at all, cause I've been thinking you're going to use them for the kvd and expecting a consistent resistance at all poles, while even a nano meter thin film of oil sandwitched among two contact points could make some deviations among those poles, but again, I'm not really sure.

Anyway, once you've dismantled those switch, please, take few shots of them, really love to have a closer look.

Offline amspireTopic starter

  • Super Contributor
  • ***
  • Posts: 3802
  • Country: au
Re: Sperry Ratio Potentiometer Teardown
« Reply #21 on: December 01, 2011, 05:17:47 am »
The switches do need lubricating, but Gertsch do not recommend a lubricant.

I found the Gertsch manual: http://128.238.9.201/~kurt/manuals/manuals/Other/GERTSCH%20RadioTran%20Instruction.pdf 

Apparently lubricating a contact does not affect the contact resistance negatively.

Before I set to work removing dismantling the Sperry unit, I just have to decide how I will make a box to put the divider in. But I will post the results when I have done it.

Richard
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf