Author Topic: The Rigol DS1052E  (Read 622658 times)

CharlieD and 4 Guests are viewing this topic.

Offline Mastro Gippo

  • Regular Contributor
  • *
  • Posts: 55
Re: The Rigol DS1052E
« Reply #325 on: March 19, 2010, 08:06:48 am »

I want ask: What are numbers in Y scale? (My quess (becouse these shown numbers) is that they are scope measured RMS and if they are, please connect this 50ohm terminator and test agen ;) ). Sig gen output is for 50ohm load.)

What are oscilloscope settings in this measure. Same settings in whole frequency scale? (scope mV/div, equtime or realtime? x ns/div? normal memory? Sinx/x on/off?)
What is signal  exactly connection to oscilloscope inputs.  (cable type/termination? calibration or some opinion how accurate leveling over used freq range = also if it is R&S it may be fail.)

If I or somebody tell data/numbers but not tell how I get data/numbers they are nearly or just as garbage... ;)

I ask becouse there is something wrong!
there is minimum 3 posibilities:

1. Test configuration is bad? (bad cable/missing or fail termination.)
2. Your oscilloscope is fail. (not my first opinion becouse both channels equal)
3. Signal generator is fail. (leveling defect)

If all is ok with test: 1052Emod or 1102E frequency response (flatness) is never as bad as in picture. If picture freq response is true for these scopes they are just as toys or garbace stuffs. (but they are not if they work normally)

But in real: Simple, I think test configuration fails ;)

whoah, sorry!
Numbers on Y scale are mVpp, for the setup I don't know exactly but I think he just wired it up with a bnc-bnc cable, w/o terminator. I'll ask him later, but it's all my fault for not telling him.
About oscilloscope settings, I guess he zoomed up as needed to keep the signal well visible. All the other settings should be the default ones, as I factory reseted it after changing SN.
By the way, I remember doing a similar test when I got the scope, and with the 50ohm terminator I got a similar "wavy" curve...
Can you post the correct setup so I can repeat the test? I feel bad for my father, as he spent 3 hours taking that data for me..  :-[
 

Offline rf-loop

  • Super Contributor
  • ***
  • Posts: 4064
  • Country: fi
  • Born in Finland with DLL21 in hand
Re: The Rigol DS1052E
« Reply #326 on: March 19, 2010, 08:53:19 am »
Keep macines on minimum 30 minutes before test and before test run Rigol selfcal and maybe also R&S selfcal.


Sign gen (R&S) out (N or BNC)-------50ohm cable -------- (50ohm GOOD terminator!)-BNC Rigol CHx

R&S out ~106mVrms example 10, 50 or 100kHz and this R&S Vrms level mean that it is TO 50 ohm load IF this instrument is with 50 ohm output (there are also other impedances dependent model options. Temination and cable need match for this)
If you forget termination you see wrong level and may see also bad flatness (cable lenght, input of scope etc play together and SWR make all measurements fail (some points may be just ok but you can not know what) and you see near like double values in voltage (in low frequencies).  (as can see your data)  

(note: I do not know this individual sig gen details so I do not know how is this low freq end and is it trusted level in low freq. So if you want 10MHz as reference point for 0dB vertical... use it and do this initial procedure for leveling with example 10MHz (this case use 2ns/diw also for this)
--------
EDIT: (SMB100A: Level uncertainty f = 200 kHz to 3 GHz <0.5 dB) so, I recommend to take 200kHz for vertical 0dB ref point and for this initial settings for this test and in this case 1us/div is better for horizontal)
--------

Rigol
CH1 (or) CH2 only in use.
DC
50mV/div
5us/div  (use 2 or 5 ns/div if 10MHz as 0dB vertical reference. This case not directly comparable to normal scope specs)
equtime
Average 8
center trig, center vertical

Check rigol with average in use that it show exactly 6div vertical. (exaxtly 300mVp-p +-2mVp-p)
If not, adjust R&S level littlebit so that display in Rigol is 300p-p (106mVrms) It is good practice not use maximum vertical area in scope. This scope have 8div vertical so it is reasonamble to use around 6 div. (of course can use what ever levels but this is example and my practice (maybe coming fron Agilent... oh no...I mean Hewlett-Packard of course and Tektronix... ;) )

After this, do not touch R&S output level.

Now tst conf is ok and reference is ok for measure BW and also we have 0dB reference point (10, 50 or 100kHz)
Change R&S to 10MHz (and later step up forward)

Change Rigol to

CH1 (or) CH2 only in use.
50mV/div
2ns/div (of course 5 if 1052 original in test)
egualtime
Average 4 - 8

make tests with all desired freq points between 10 to >200MHz (no need touch vertical speed). rememember wait enough averaging if average is use! (for randomnoise I recommend averaging becouse this is not analog oscilloscope)
10MHz level is typically now around 1- 3% low compare to refpoint.

Of course you can also use different test signal levels. (Vp-p = Vrms * 1,4142 * 2)
This give good picture about real BW.
(BW maybe littlebit different in 2 main bands in V/div settings. listen relay click...you know these bands. 2-200mV and 500mV - 10V also maybe some small differents in other settings)

Also if like testing, same with realtime (but sinx OFF becouse sinx in this rigol IS bad)
« Last Edit: March 19, 2010, 09:53:17 am by rf-loop »
I drive a LEC (low el. consumption) BEV car. Smoke exhaust pipes - go to museum. In Finland quite all electric power is made using nuclear, wind, solar and water.

Wises must compel the mad barbarians to stop their crimes against humanity. Where have the wises gone?
 

Offline rf-loop

  • Super Contributor
  • ***
  • Posts: 4064
  • Country: fi
  • Born in Finland with DLL21 in hand
Re: The Rigol DS1052E
« Reply #327 on: March 19, 2010, 10:48:10 am »
I think this question is ok specially in this topic now after all this knowledge.

I know this is littlebit paranoid question. But I also think that this IS relevant question and maybe it is also hard question. So hard that maybe we do not want know all (possible) dirty truth. Maybe too hard...

Peoples buy Rigol DS1102E from china or other country from supplier who do not buy directly from Rigol factory these equipments.

How they can really know that they have genuine Rigol factory made original DS1102E or counterfeit?

How many pictures we can find what are take inside PCB's from both models. If original factory made 1102 or 1052, are there differencies or are these seen differencies only revision differences. (specially in PCB just after BNC connector.)

I can not believe that this is first time when somebody know how to do this modification?

Also inside Rigol there are many peoples..... maybe somebody find good "night job".





« Last Edit: March 19, 2010, 10:53:28 am by rf-loop »
I drive a LEC (low el. consumption) BEV car. Smoke exhaust pipes - go to museum. In Finland quite all electric power is made using nuclear, wind, solar and water.

Wises must compel the mad barbarians to stop their crimes against humanity. Where have the wises gone?
 

Offline flolic

  • Frequent Contributor
  • **
  • Posts: 386
  • Country: hr
    • http://filiplolic.com/
Re: The Rigol DS1052E
« Reply #328 on: March 19, 2010, 11:42:49 am »
I suppose that only way for check if is it genuine instrument is by serial number?
 

Offline rf-loop

  • Super Contributor
  • ***
  • Posts: 4064
  • Country: fi
  • Born in Finland with DLL21 in hand
Re: The Rigol DS1052E
« Reply #329 on: March 19, 2010, 11:48:38 am »
I suppose that only way for check if is it genuine instrument is by serial number?

How?
I drive a LEC (low el. consumption) BEV car. Smoke exhaust pipes - go to museum. In Finland quite all electric power is made using nuclear, wind, solar and water.

Wises must compel the mad barbarians to stop their crimes against humanity. Where have the wises gone?
 

Offline KTP

  • Frequent Contributor
  • **
  • Posts: 512
Re: The Rigol DS1052E
« Reply #330 on: March 19, 2010, 11:49:05 am »
How could it be remotely economical to counterfit such a vertical market product as an oscilloscope and sell it for $400???

My DS1052E seems to have nice knobs and buttons and a good finish, at least on the outside.

If it is a counterfit then I would like this company to make me a $1000 Lexus or BMW please.
 

Offline flolic

  • Frequent Contributor
  • **
  • Posts: 386
  • Country: hr
    • http://filiplolic.com/
Re: The Rigol DS1052E
« Reply #331 on: March 19, 2010, 12:14:23 pm »

How?

I am stupid!  ;D
Hard or impossible. You can check with Rigol if that serial number belongs to 100MHZ or 50MHz instrument, but if
counterfeit it can be multiple instruments with the same serial...
 

Offline tristan

  • Contributor
  • Posts: 12
Re: The Rigol DS1052E
« Reply #332 on: March 19, 2010, 12:22:59 pm »
I think the problems I described a few posts back (scope locking up, corrupted system info display) might have been due to me hitting return before alt-010. It didn't cause a permanent problem but it's worth being careful. The display corruption looked like random data got into the video buffer RAM - maybe the incorrect S/N or version didn't have proper null termination resulting in a buffer overflow?

Tristan
 

Offline flolic

  • Frequent Contributor
  • **
  • Posts: 386
  • Country: hr
    • http://filiplolic.com/
Re: The Rigol DS1052E
« Reply #333 on: March 19, 2010, 12:23:26 pm »
How could it be remotely economical to counterfit such a vertical market product as an oscilloscope and sell it for $400???

My DS1052E seems to have nice knobs and buttons and a good finish, at least on the outside.

If it is a counterfit then I would like this company to make me a $1000 Lexus or BMW please.

Think about it, you can buy ship cargo container of 50MHz units for 150$ a piece (and probably even less), firmware modified it, change front panel label to 100MHz and sell it for 700$+  8)
« Last Edit: March 19, 2010, 12:25:46 pm by flolic »
 

Offline KTP

  • Frequent Contributor
  • **
  • Posts: 512
Re: The Rigol DS1052E
« Reply #334 on: March 19, 2010, 12:24:13 pm »
Do you guys really mean stolen when you say counterfit?  I don't know if this word translates to something else in non-english language.

I can't imagine the R&D that would be involved in accurately counterfitting the Rigol DS1052E and getting it to perform similar to the true model, not to mention creating all of the injection molds for the various plastic components and buttons.  Then they would have to source all of the analog parts....all of this to sell a few thousand units on ebay????

I could see counterfitting a mop or perhaps a cheap cellphone that would sell millions, but this?

Or do you guys mean that some employee stole a few hundred Rigol scopes from the factory and sold them on ebay?
 

Offline flolic

  • Frequent Contributor
  • **
  • Posts: 386
  • Country: hr
    • http://filiplolic.com/
Re: The Rigol DS1052E
« Reply #335 on: March 19, 2010, 12:33:02 pm »
Counterfeit is probably not the best word, but in my language (Croatian) it translates right...
By counterfeit I mean buy 1052E, modify it, change labels and sell it as genuine 1102E for extra profit.
 

Offline flolic

  • Frequent Contributor
  • **
  • Posts: 386
  • Country: hr
    • http://filiplolic.com/
Re: The Rigol DS1052E
« Reply #336 on: March 19, 2010, 12:36:16 pm »
I think the problems I described a few posts back (scope locking up, corrupted system info display) might have been due to me hitting return before alt-010.

Yes, you are right on that ;)
 

Offline rf-loop

  • Super Contributor
  • ***
  • Posts: 4064
  • Country: fi
  • Born in Finland with DLL21 in hand
Re: The Rigol DS1052E
« Reply #337 on: March 19, 2010, 12:39:02 pm »
How could it be remotely economical to counterfit such a vertical market product as an oscilloscope and sell it for $400???

My DS1052E seems to have nice knobs and buttons and a good finish, at least on the outside.

If it is a counterfit then I would like this company to make me a $1000 Lexus or BMW please.

First, there is no any kind of mind to make counterfeit RigolD1052E. Not of course. I think you miss now some few points. Also maybe not know what are Chinese inside country markets and peoples living there. I know something, I have be there.

If there is any counterfeit business it ALL need happen there.

Do you know how much money is USD400 example inside China.
Do you know how much these kind of labels cost in China.
Do you know that 1052 is same as 1102
Do you know that scope itself can change model with 15 seconds.

Now if you know this all how to do and also you know how to do all perfect counterfeit things for labelling this scope to different model. Oh... maybe 10RMB...

Next you organize some hundreds or thousends of cheap 1052E go to to change after some buy 1102 example from some asian webstore.

If you get example 1000RMB per unit...

And then you are source to some webstore...

This is not much here in western country... with this money one people can easy live one month in China and I can tell that for this kind of business there you can find very easy million people who want competite who can do this...

Do you know how big business it is example make "genuine" labels for counterfeit batteries... and money in these are just only coins...  there are so many peoples who are ready to make lot of work for get 1RMB. With 2RMB I find tens of peoles who are ready to pickup me with bicycle...  

If this small business for change cheap 1052 to 1102 is possible to somebody without (or just cheap) money and can get part of this money what is price difference between these models... you can be sure somebody have done it or do it all this time.

Who really think make counterfeit oscilloscope... no one.
But buy ready made cheap model and change it to expensive model...

Some people in China only open his window and there rains money in... they are not stupid.
But we... we think that we can check if qenuine by looking serial number... just we talk that mod need sernum change and it takes just second with whatever RS232 terminal (no need PC). Whole type change take not more than 15minutes per unit and this time include also eating, tea and toilet time. After this time normal customer can not know that it was 1052E originally. These cheap labels and papers with scope... how many these you want? Walk an hour around "small" 10million peoples town in right place in China and you have (but be sure that do not try if somebody can quess you are not Chinese)
« Last Edit: March 19, 2010, 12:54:08 pm by rf-loop »
I drive a LEC (low el. consumption) BEV car. Smoke exhaust pipes - go to museum. In Finland quite all electric power is made using nuclear, wind, solar and water.

Wises must compel the mad barbarians to stop their crimes against humanity. Where have the wises gone?
 

Offline jone

  • Contributor
  • Posts: 15
Re: The Rigol DS1052E
« Reply #338 on: March 19, 2010, 12:39:28 pm »
I think the problems I described a few posts back (scope locking up, corrupted system info display) might have been due to me hitting return before alt-010.

Easier than alt-010, configure your terminal software to use only LF for newlines (if using the serial port)... I guess some windows software is configured for CR/LF by default?
 

Offline rf-loop

  • Super Contributor
  • ***
  • Posts: 4064
  • Country: fi
  • Born in Finland with DLL21 in hand
Re: The Rigol DS1052E
« Reply #339 on: March 19, 2010, 12:47:24 pm »
Also when you mod... insert RS232 terminal cable before you powON. So you can be sure that buffer have nothing "scratch". And before all, do these right settings for comm parameters. (avoid machines what have not RS232 standard signal levels!)

Remember that in Sysdata displ (R/S red) after *RST it preboot with new "data"... (R/S go green but all is freeze -- do not touch) and....... then powoff/on. also do NOT this mod if Rmt on dspl.

--------


Do NOT try try this modification with RS232 communication if you do not really know how to do this kind of RS232 communication what works perfect. It need work perfect in HW level and it must do without any mistake in data. If you are not really sure how to do reliable trustworth and robust serial connection do not even try. You may seriously damage your oscilloscope. Remember: this use undocumented commands! fail in data may damage your machine nearly as permanently.

Before you do anything with undocumented commands you need (minimum) be sure that communication have no any kind error what may product unwanted transmitted data inside oscilloscope.

Also do not trust Windows/PC RS232 port. It may be out of specifications by many meanings even with voltage levels.
Any missing, extra or wrong data may damage your scope with these undocumented commands.


This is only one small example about RS232 problems with PC and Microsoft:
http://www.home.agilent.com/agilent/editorial.jspx?cc=US&lc=eng&ckey=131609&nid=-35204.0.00&id=131609

"Problems Using RS-232 on Agilent Instruments with Laptop PCs running any Microsoft OS
Symptoms: Timeouts, corrupt data, missing data, error messages while uploading data
"
« Last Edit: April 09, 2010, 01:54:56 pm by rf-loop »
I drive a LEC (low el. consumption) BEV car. Smoke exhaust pipes - go to museum. In Finland quite all electric power is made using nuclear, wind, solar and water.

Wises must compel the mad barbarians to stop their crimes against humanity. Where have the wises gone?
 

Offline KTP

  • Frequent Contributor
  • **
  • Posts: 512
Re: The Rigol DS1052E
« Reply #340 on: March 19, 2010, 12:52:29 pm »
Ok, I see what  you meant by counterfeit Rigol scopes.  You guys were talking about DS1102E scopes that were former DS1052E models before a quick flash of the model and serial number.

Since there is not a model below a DS1052E, we must assume *those* scopes are not counterfeit.  Then the ebay email that person received that said the DS1052E was being removed because of being counterfeit is probably untrue.
 

Offline Mastro Gippo

  • Regular Contributor
  • *
  • Posts: 55
Re: The Rigol DS1052E
« Reply #341 on: March 19, 2010, 01:06:12 pm »
Sign gen (R&S) out (N or BNC)-------50ohm cable -------- (50ohm GOOD terminator!)-BNC Rigol CHx

So... you're not talking about a T splitter with the signal from one end and 2 100ohm resistor in parallel wired to the other, right?  :-\

I think the problems I described a few posts back (scope locking up, corrupted system info display) might have been due to me hitting return before alt-010.
Yes and no (and maybe  :D); I happened to do the same mistake, but the scope didn't lock or corrupt; only difference was, when sending the first *idn? command, it replied
Code: [Select]
Rigol Technologies, DS1052E, DS1ED112404432
, 00.02.02.02.00
instead of
Code: [Select]
Rigol Technologies, DS1052E, DS1ED112404432, 00.02.02.02.00
 

Offline flolic

  • Frequent Contributor
  • **
  • Posts: 386
  • Country: hr
    • http://filiplolic.com/
Re: The Rigol DS1052E
« Reply #342 on: March 19, 2010, 01:43:38 pm »

So... you're not talking about a T splitter with the signal from one end and 2 100ohm resistor in parallel wired to the other, right?  :-\

I don't have good experience with T splitter, but maybe there is a problem with it, it is cheap LAN splitter. And you definitely need good BNC 50 ohm coaxial terminator. Two paralleled 100 ohm resistors can do it, but are far from ideal...

IMHO, best way is to use feed through 50 ohm BNC terminator, like that on bottom of this page: http://accessories.picotech.com/attenuators.html
 

Offline rf-loop

  • Super Contributor
  • ***
  • Posts: 4064
  • Country: fi
  • Born in Finland with DLL21 in hand
Re: The Rigol DS1052E
« Reply #343 on: March 19, 2010, 02:33:18 pm »
Also 2 channels can not test parallel without very special knowledge how to do and without also special things what this need.

Test only  single channel at once!
(first I look this picture and numbers I think that oh... there is two cable and without terminator... ;)  )

In emergency situation (with these quite low frequencies) some kind of termination can do so that there is T in the scope end.
50 ohm cable to T and free end of T can put small LOW inductance resistor with as short wires as possible from center to ground. But also can use ethernet terminator if nothing else is in hand. (but both of these need remember that in the high end of frequency there may be some errors. But still these are better than nothing. Most bad of course is try channels parallel.)
I drive a LEC (low el. consumption) BEV car. Smoke exhaust pipes - go to museum. In Finland quite all electric power is made using nuclear, wind, solar and water.

Wises must compel the mad barbarians to stop their crimes against humanity. Where have the wises gone?
 

Offline KTP

  • Frequent Contributor
  • **
  • Posts: 512
Re: The Rigol DS1052E
« Reply #344 on: March 19, 2010, 05:29:57 pm »
Hey, so how are you guys getting this fast rise time pulses into the DS1052E?

I just tried to build a fast rise time circuit to test my newly modified DS1052E and it was fail.

I had a fairchild tinylogic NC7NZ14 inverting schmitt trigger that I built deadbug style on a piece of copper circuit board (which was an amazing feat if you consider how small this smt part is.  It works, but the scope displays a horrible ringing and the risetime is more like 20ns than the 4ns I was expecting.  I have the little copper board connected to the bnc input of the scope through about 1/2 inch of cable and I am feeding the schmitt trigger with a 1khz square wave from a wavetek model 21 FG.  Of course the signal looks the same on my TDS 210, so I know the fault is with my setup and not the DS1052E...
 

Offline flolic

  • Frequent Contributor
  • **
  • Posts: 386
  • Country: hr
    • http://filiplolic.com/
Re: The Rigol DS1052E
« Reply #345 on: March 19, 2010, 06:26:18 pm »
Did you see this link?  http://www.i9t.net/fast-pulse/fast-pulse.html
This is fastest pulse generator you can easily built, imho.

And this is how I built it, supply circuit is on the other PCB:
 

Offline djsb

  • Frequent Contributor
  • **
  • Posts: 886
  • Country: gb
Re: The Rigol DS1052E
« Reply #346 on: March 19, 2010, 06:26:42 pm »
Hi,
I've been following this thread with interest. Can someone please summarise the steps needed to convert the DS1052E to 100Mhz?
Thanks.

David.
David
Hertfordshire,UK
University Electronics Technician, London PIC,CCS C,Arduino,Kicad, Altium Designer,LPKF S103,S62 Operator, Electronics instructor. Give a man a fish and you feed him for a day. Teach a man to fish and you feed him for a lifetime. Credited Kicad French to English translator.
 

Offline SimonTopic starter

  • Global Moderator
  • *****
  • Posts: 17728
  • Country: gb
  • Did that just blow up? No? might work after all !!
    • Simon's Electronics
Re: The Rigol DS1052E
« Reply #347 on: March 19, 2010, 06:33:02 pm »
Hi,
I've been following this thread with interest. Can someone please summarise the steps needed to convert the DS1052E to 100Mhz?
Thanks.

David.

here here, I know it's simple but I'm sure many will be in fear of screwing up their scopes if they do something wrong. I'm just off to the shops to get a cable. unfortunately the deadly maplin being the only ones to stock them, maybe it won't fall apart on me
 

Offline jahonen

  • Super Contributor
  • ***
  • Posts: 1052
  • Country: fi
Re: The Rigol DS1052E
« Reply #348 on: March 19, 2010, 07:05:03 pm »
Direct link to that LTC application note for your convenience: Linear Technology Application note #47, page 93.

Cyclone II FPGA LVTTL outputs are still faster, around 160 ps what I have measured from them (without any capacitive loading).

Regards,
Janne
 

Offline KTP

  • Frequent Contributor
  • **
  • Posts: 512
Re: The Rigol DS1052E
« Reply #349 on: March 19, 2010, 07:13:39 pm »
I'll take a stab at what I did, although I don't make any guarantee you will have success and I do not accept any liability if you choose to make these modifications to your scope.  I seem to have converted my DS1052E to a DS1102E, at least it says so under the utility menu and when I send *IDN? to the RS232 port.  Also my minimum time/division is now 2ns.

First I built a straight through 9 pin serial cable although I only used pins 2,3 and 5 (ground).  If you want to test your cable with hyperterminal you can short pins 2 and 3 on the end not plugged into your computer and you should see what you type echoed back to you (or echoed twice if you have the autoecho thing enabled in hyperterminal).

I then read through the how to set up hyperterminal document that Rigol published.  

http://www.rigol.com/upload/accessory/20102/2010221051162027445.pdf

Actually I didn't read this until I had spent 10 minutes trying to figure out why I couldn't send commands to the DS1052E..it was because I was trying to use the enter key and instead should hold down alt and press "0","1","0"  ;)

I then connected the scope to the pc and ran hyperterminal setting it up exactly as in the instructions.  I verified I was talking to the DS1052E by typeing *IDN? (followed by alt "0","1","0") and the scope relayed back to me the model number, serial number and firmware version.  I copied this stuff down in case I wanted to go back to my original setup for some reason.

I then used the not so super secret serial command:

:INFO:MODEL DS1102E (followed by alt "0","1","0")

to write a new model number to the flash on the scope.

I then used the other serial command:

:INFO:SERIAL DS1EB104702974 (followed by alt "0","1","0")

to write a new serial number to the flash on the scope.  I found this serial number on the documentation that came on the CD with the scope under programming examples.  It was an example of what the output might look like when you type *IDN? on a DS1102E so I figured it would be a valid serial number :)

After this I just pressed power off and then powered back on, but other people have done a software reset through the serial port I think.  I am not sure what that command is... *RST or something?  When I powered back on I reconnected via RS232 and issued a *IDN? command (followed by alt "0","1","0") and it showed the Model as being DS1102E and the serial number being the new one.  I checked the time/div and it now went down to 2nS.  I then went to the utility menu and ran the self calibration (takes about 5 minutes and you should have everything but the power cord disconnected from the scope...also make sure the scope has been on for at least 30 minutes).

Well, that is about what I did to the best of my recollection.  I don't recommend trying any of this until someone confirms my steps as I may have done it wrong.  I don't see the reason for people to be smug or try and hide the information (or make vague references to how "easy" it is but not describe anything).  hopefully I have been of some help.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf