Author Topic: Advice please, used 34401a  (Read 2067 times)

0 Members and 1 Guest are viewing this topic.

Offline stdTopic starter

  • Contributor
  • Posts: 25
  • Country: 00
Advice please, used 34401a
« on: November 28, 2024, 06:47:30 am »
Hello guys! 

I just going to buy used 34401a (6.5 digits multimeter), which I want to power on/test (not remote like ebay).
One guy sells few HP 34401a, another sells bunch of Agilent 34401a. I'm absolutely newbie related to buying used equipment or related to 34401a.  How to test and pick best one?  What should I do to avoid potential problems?

Give me advice please.
Happy New Year 2025!
 

Online tatel

  • Frequent Contributor
  • **
  • Posts: 609
  • Country: es
Re: Advice please, used 34401a
« Reply #1 on: November 28, 2024, 11:04:31 am »
Not 34401 owner but I looked for one some time ago. It seems you can boil it all to:

a) You'd probably want to get one with red and black connectors. There are VFD replacements for these, but not for the all-red models, if I got it right.

b) of course, you should check the VFD display is bright, not dim.

c) then, if you turn the device on and you see it "PASS" the initial test all should be good

Also, you could use the search feature and look for 34401-related threads in the forum

Have a nice day
 
The following users thanked this post: std

Online factory

  • Super Contributor
  • ***
  • Posts: 3215
  • Country: gb
Re: Advice please, used 34401a
« Reply #2 on: November 28, 2024, 01:36:55 pm »
And ask the seller to run the full self test & report back with any error codes, if any are shown.

David
 
The following users thanked this post: std

Offline mawyatt

  • Super Contributor
  • ***
  • Posts: 4120
  • Country: us
Re: Advice please, used 34401a
« Reply #3 on: November 28, 2024, 01:42:13 pm »
The AG 34401A are from Agilent and newer than HP 34401A. The all Red terminals 34401As are the oldest and some can have the VFD display replaced (we just did ours).

Also, our older HP can't be "enabled" with the TEMP function, our AG can and we did such. Not sure how you can tell which old HP 34401A display can be replaced and which can't tho, same for added TEMP Function. Maybe best to stay with a newer AG version.

These are amazing DMMs, but can take ~4 hours to completely stabilize, however no annoying fan noise :-+

Best
« Last Edit: November 28, 2024, 01:44:08 pm by mawyatt »
Curiosity killed the cat, also depleted my wallet!
~Wyatt Labs by Mike~
 
The following users thanked this post: std

Online Kean

  • Supporter
  • ****
  • Posts: 2468
  • Country: au
  • Embedded systems & IT consultant
    • Kean Electronics
Re: Advice please, used 34401a
« Reply #4 on: November 28, 2024, 02:37:14 pm »
Check the revision info.  It will be three numbers, like 10-05-02 on my two units, which relate to firmware revisions of the measurement, I/O, and front panel sections.

As mentioned, the early HP models will have all red banana jacks.  I would generally recommend avoiding them unless the display is good and the price is even better.
They will likely have a revision 1 display board which if it fails likely cannot be repaired.  And the main board wont work with a rev 2 display either.
The older models are also possibly not going to have had all the service note fixes applied.

See these threads as well, but it covers much the same points discussed here.
https://www.eevblog.com/forum/testgear/fair-price-for-an-agilent-34401a/
https://www.eevblog.com/forum/testgear/agilent-34401a-revisions/
https://www.eevblog.com/forum/testgear/q-agilenthp-34401a-versions-is-any-particular-one-preferable/
 
The following users thanked this post: std

Offline coromonadalix

  • Super Contributor
  • ***
  • Posts: 7016
  • Country: ca
Re: Advice please, used 34401a
« Reply #5 on: November 28, 2024, 08:59:26 pm »
red black plugs for sure

check for inflated used prices ...

self tests results / errors codes

and see if the vfd is dim or not, if it is not powered on photos and the seller doesn't want to power it up and show you, stay away

i had 3 of theses, one with an brand new cloned vfd, recapped and re-certified after many years, and was rock solid

you have the successors (more connectivity)  with available schematics, the 34410a 34411a  models, and cloned vfd's i would invest in theses instead ... sure the price can be a deal breaker

and for theses too i would say the same previous advice ...
 
The following users thanked this post: std

Offline mawyatt

  • Super Contributor
  • ***
  • Posts: 4120
  • Country: us
Re: Advice please, used 34401a
« Reply #6 on: November 28, 2024, 09:29:28 pm »
For our (older folks) use, there's something nostalgic about these old 34401As. Maybe it's the beautiful Blue/Green VFD Display, or the quick startup, or the acoustic noise free measurements, or just the amazing long-term stability (the LM399s are well aged!!), but whatever it is they bring a grin when we use them :-+

When we just replaced the fading VFD Display in our old HP version, then powered up, this brought a nice smile for this old guy ;)

Best 
Curiosity killed the cat, also depleted my wallet!
~Wyatt Labs by Mike~
 
The following users thanked this post: JohanH, Kean, std

Offline stdTopic starter

  • Contributor
  • Posts: 25
  • Country: 00
Re: Advice please, used 34401a
« Reply #7 on: November 28, 2024, 09:35:43 pm »
I'm a big slacker of course, downloaded 34401a manual to my phone 10 minutes before leaving home. No LTZ1000 board in pocket, no precise resistors :)  But it looks like all these guys who bring used 34401a from other countries are even bigger slackers. The guy who promised to sell got lost. Instead there was some kind of pretzel who doesn't know anything about HP 34401a. He put all of their remaining HP 34401a's on top and turned them all on. Then I immediately rejected two because almost dim display. Guy thought that turn-on test was enough, so they not checked it at all. I quickly searched manual for the phrase "self-test". (Hold shift, power on). Then one consistently failed self-test. The other has the highest readings like 000.256 mVDC on open terminals and high readings on closed too. Its alarmed me (maybe I'm talking nonsense). However, all other 34401a's readings was like ~000.009 mVDC ...  ~000.004 mVDC. 
So all that's left for me is the worst looking HP 34401a, however its electrical condition after closer examination at home seems +/- satisfy me (still not checked I, AC, nor checked all/high bands).
Shame for me I didn't think to ask forum earlier. It's all red banana jacks.
8: REVISION menu: REV 03-01-01   
Now I'm start worrying about repairability.
© "There was no sadness, a woman bought a pig"..  :-DD
Happy New Year 2025!
 

Offline coromonadalix

  • Super Contributor
  • ***
  • Posts: 7016
  • Country: ca
Re: Advice please, used 34401a
« Reply #8 on: November 28, 2024, 09:39:43 pm »
for sure  the 34401a fanless models are good once repaired, new VFD, or re-capped, and or calibrated

but as technology evolves, it depends of the OP  needs,   gpib and rs232   can be a downside ...


good luck

03-01-01  is very old, maybe  the unobtanium  vfd drivers who can segment leak  ....   i would avoid unless the price is very low

« Last Edit: November 28, 2024, 09:41:27 pm by coromonadalix »
 

Offline mawyatt

  • Super Contributor
  • ***
  • Posts: 4120
  • Country: us
Re: Advice please, used 34401a
« Reply #9 on: November 28, 2024, 09:56:19 pm »
Our old HP is Rev 05-01-01 (the AG is 11-05-02) and we able to replace the VFD display, but it wouldn't accept the TEMP enhancement.

We have 3 KS34465As, a DMM6500 and a SDM3065, which are all technically superior DMMs with many more useful features, but still like to use the 34401As once in awhile for some reason ;)

@std,

Where do you live where you can go to a place and evaluate all these DMMs? That's certainly a big benefit!! We were lucky awhile back and a local surplus house (Clearwater, Florida) that doesn't generally carry TE had an old Tek 577 which wasn't working. These command a high price even when not functioned (we paid $500), but were able to restore it and use it :)

Best
Curiosity killed the cat, also depleted my wallet!
~Wyatt Labs by Mike~
 

Offline Hydron

  • Super Contributor
  • ***
  • Posts: 1085
  • Country: gb
Re: Advice please, used 34401a
« Reply #10 on: November 28, 2024, 11:24:47 pm »
The older ones at least match the service manual schematics, newer ones replace the input buffer dual JFET + OP27 with a bootstrapped OPA130. Not sure what the performance implications are, the older version might be better (there are tempting replacements for the OPA130 with better specs, especially lower noise, though others have mentioned that the ADC is the limiting factor in a 34401A, so might be a futile change).

In a perfect world I'd pick a FW 07-05-02 or newer, as it has the new display board and hidden options to enable in the v7+ FW. Also would do a self test and enable 6 digit mode to check that the last digit isn't way brighter than the rest (can be the case if it's been run in the default 5 digit mode most of it's life - I'm using the store/recall hidden FW function to make mine boot into 6 digit volts mode by default).

Definitely happy with my 2007 vintage Agilent meter, especially after changing the VFD for a bright new clone. ~3 seconds to boot (settings recall adds about a second), almost totally silent (just a little transformer hum) and with a lovely VFD display, it's a great complement to my DMM6500.
 

Offline indeterminate

  • Regular Contributor
  • *
  • Posts: 97
  • Country: au
Re: Advice please, used 34401a
« Reply #11 on: November 29, 2024, 10:48:32 am »
Mine has the  all red banana jacks. but is a rev 2 display running FW 10-05-02  with Temp function etc
so defenetly look at the firmware version
 

Offline coromonadalix

  • Super Contributor
  • ***
  • Posts: 7016
  • Country: ca
Re: Advice please, used 34401a
« Reply #12 on: November 29, 2024, 11:13:32 am »
you need to search at the serial number level, there's HP Agilent  papers on witch versions and fw you'll know that it is the newest vfd drivers and associated FW version(s)



thread about this
https://www.eevblog.com/forum/testgear/q-agilenthp-34401a-versions-is-any-particular-one-preferable/
https://www.eevblog.com/forum/testgear/agilent-34401a-revisions/

i did wrote in theses

Brief History  on the   34401A-06     service note  attached here

Up to the fw 06-04-01  you will have the old unobtanium display chip,       pcb: 34401-66502

Past the  fw 06-04-02  you'll get the newest more repairable display pcb   pcb: 34401-66512

You can still order the newest pcb version 34401-66512, but the quantity is low ...  and the price is high,  around 200$ in my country, dont know how much you've paid for it ?

There is a version 10 leaked in the attachement, never tried it, it is supposed to work if you reprogram your (U502) fw chip and put the newest display pcb ... never heard of anyone doing it ???


For the VFD (vacuuum display) pn: 2090-0296 it is not orderable from Keysight anymore, you have to rely on chinesium parts around the web.
« Last Edit: November 29, 2024, 11:58:59 am by coromonadalix »
 
The following users thanked this post: std

Offline stdTopic starter

  • Contributor
  • Posts: 25
  • Country: 00
Re: Advice please, used 34401a
« Reply #13 on: December 09, 2024, 10:52:11 pm »
It appeared resistance measurement in my HP 34401a is incorrect. I discovered this a few days ago but only now had time to write.
When I tested it in seller place, I shorted probes. Readings were low, so I didn't pay attention, I thought these were crappy probes, dust, old, no-contact, etc... It is impossible to figure out what is going on while standing in the passage between the door and the counter and when a guy is waiting for you. To do this, you need a normal place, sit down and at least 30 minutes to figure it out. Maybe in fact everything was fine, the defect was floating. I called them, they offered to come and look at another HP34401a, but I know that there are devices with even greater defects there. So I decided to forget about their two-week warranty and repair the HP34401a myself.

To test I made two banana jacks shorted with 1.62mm copper wire one inch long and began to compare with (oher) Agilent 34401a taken as a sample.

Reference Agilent 34401a ohm readings when input shorted: 
~   000.000,3  OHM ....  000.000,1  OHM

My HP 34401a ohm readings when input shorted:     
000.564,3  OHM    000.701,1 OHM    000.912,8 OHM    even   1.7 ohm,  1.3 ohm,  2.5 ohm, 4ohm, 6 ohm  ! 
The readings change every time I switching the [DC V] and [Ω 2W] modes, when relays click.

K101 relay problem?
Happy New Year 2025!
 

Online Kean

  • Supporter
  • ****
  • Posts: 2468
  • Country: au
  • Embedded systems & IT consultant
    • Kean Electronics
Re: Advice please, used 34401a
« Reply #14 on: December 09, 2024, 11:30:33 pm »
It appeared resistance measurement in my HP 34401a is incorrect. I discovered this a few days ago but only now had time to write.

What do you get if you use the rear terminals?

A common problem is dirty switch contacts for selecting between front/rear terminals, and operating the switch a few times can help.
 

Offline stdTopic starter

  • Contributor
  • Posts: 25
  • Country: 00
Re: Advice please, used 34401a
« Reply #15 on: December 10, 2024, 02:33:22 am »
What do you get if you use the rear terminals?
A common problem is dirty switch contacts for selecting between front/rear terminals, and operating the switch a few times can help.

When I short K101 with jumper wire, the readings change by the amount of some ohms.
Now multiple clicks on [DC V]-[Ω 2W] buttons, or multiple pressing [Front/Rear terminals switch] the readings are the same down to 1-2 milliohm.

For rear terminals resistance readings is correct. 
For front terminals the readings are biased by 0001.417 OHM. Increasing input resistance on front terminals first reduces readings. After this bias point it starts to increase.
Again, after multiple clicks on [DC V]-[Ω 2W] buttons, or [Front/Rear terminals switch] the readings is stable. Just biased.
(It appears when the K101 relay was broken, someone tried to set the zero offset for front terminals which is now incorrect).

« Last Edit: December 10, 2024, 02:48:32 am by std »
Happy New Year 2025!
 

Online Kean

  • Supporter
  • ****
  • Posts: 2468
  • Country: au
  • Embedded systems & IT consultant
    • Kean Electronics
Re: Advice please, used 34401a
« Reply #16 on: December 10, 2024, 03:24:08 am »
For rear terminals resistance readings is correct. 

OK, well K101 comes after S1 (the front/rear switch) so I would have expected a problem with K101 to show up in both positions.
But as you suggest, someone may have done a zero offset calibration adjustment for the front input while a K101 fault existed.

Out of interest, do you have (or can you make) a 4W short and test zero in 4W mode.

K101 is rather tricky to source.  :--

You might want to try this procedure:
https://www.eevblog.com/forum/repair/recovery-of-the-relay-k101(0490-1914)-in-hpagilent-34401a/

Thankfully both of my 34401As have not shown this issue, and 2W short reads about 4.5mohm.
AFIAK the 34970A DMM doesn't have K101 (I have 3 x 34970A and 1 x 34972A).
« Last Edit: December 10, 2024, 03:28:05 am by Kean »
 
The following users thanked this post: std

Offline KungFuJosh

  • Super Contributor
  • ***
  • Posts: 3166
  • Country: us
  • TEAS is real.
Re: Advice please, used 34401a
« Reply #17 on: December 10, 2024, 03:40:41 am »
Radwell sells them for $20 with a 2-year warranty. I've never dealt with them before, so I dunno how they are.

https://www.radwell.com/Buy/COTO/COTO/0490-1914
"Right now I’m having amnesia and déjà vu at the same time. I think I’ve forgotten this before." - Steven Wright
 
The following users thanked this post: std

Online Kean

  • Supporter
  • ****
  • Posts: 2468
  • Country: au
  • Embedded systems & IT consultant
    • Kean Electronics
Re: Advice please, used 34401a
« Reply #18 on: December 10, 2024, 04:01:53 am »
Radwell sells them for $20 with a 2-year warranty. I've never dealt with them before, so I dunno how they are.

https://www.radwell.com/Buy/COTO/COTO/0490-1914

The page doesn't mention them being in stock, just an option to request a quote.
ISTR someone asking them and getting a reply that they didn't have any - although that may have been a different COTO model...

Aha, found the post.  It was a COTO relay for a K196 - and probably a very similar part.
https://www.eevblog.com/forum/repair/some-help-troubleshooting-a-keithley-196-with-bad-ohms-reading(s)/msg3525264/#msg3525264

The suggested COTO 3501-05-511 is the same footprint as the HP 0490-1914 with low thermal EMF, but has a lower voltage rating.
It is available at DK & Mouser (~30 USD) and should work as a replacement if you avoid measuring over 200V.  Usual caveats apply.
 
The following users thanked this post: std

Offline stdTopic starter

  • Contributor
  • Posts: 25
  • Country: 00
Re: Advice please, used 34401a
« Reply #19 on: December 17, 2024, 07:54:35 pm »
Unsoldered K101 (T17231 COTO-9231). 

K101 coil DC resistance = 500 Ω. 
Inductance = 348mH, open reed switch contact capacitance  = 1.2pF.

Next I measured reed switch resistance applying different coil voltages.
Operate voltage 3.33V, release voltage 2.3V.
As in the DMM, reed switch resistance floats between 1.5-15 Ω. With small dispersion in row.

I also found strange thing. When the coil voltage (current and magnetic flux) decreases, the contact resistance does not remain constant. As the voltage decreases, contact resistance increases to hundreds ohms or 3K-7K until it release.  Also when coil operated at higher voltage it makes resistance much lower. Of course, relay is clearly faulty. There should be almost no resistance at all. I thought the problem was in bad solder joint, not reed switch contacts (same as here). Now I'm not sure. How bad solder joints would depend on magnetic flux? And if the reed switch itself is faulty, then it's a fiasco. Even if I disassemble the relay, it won't help.


« Last Edit: December 17, 2024, 08:02:32 pm by std »
Happy New Year 2025!
 

Offline AVGresponding

  • Super Contributor
  • ***
  • Posts: 4931
  • Country: england
  • Exploring Rabbit Holes Since The 1970s
Re: Advice please, used 34401a
« Reply #20 on: December 18, 2024, 06:29:09 am »
In the newer, TrueVolt models, the relay isn't necessarily a COTO anyway, they've used Meder Electronic for example

nuqDaq yuch Dapol?
Addiction count: Agilent-AVO-BlackStar-Brymen-Chauvin Arnoux-Fluke-GenRad-Hameg-HP-Keithley-IsoTech-Mastech-Megger-Metrix-Micronta-Racal-RFL-Siglent-Solartron-Tektronix-Thurlby-Time Electronics-TTi-UniT
 
The following users thanked this post: std

Offline stdTopic starter

  • Contributor
  • Posts: 25
  • Country: 00
Re: Advice please, used 34401a
« Reply #21 on: December 19, 2024, 11:59:53 pm »
I'll continue here my investigations about K101 relay.

Thought it all over:
I also found strange thing. When the coil voltage (current and magnetic flux) decreases, the contact resistance does not remain constant. As the voltage decreases, contact resistance increases to hundreds ohms or 3K-7K until it release.  Also when coil operated at higher voltage it makes resistance much lower. Of course, relay is clearly faulty. There should be almost no resistance at all. I thought the problem was in bad solder joint, not reed switch contacts (same as here). Now I'm not sure. How bad solder joints would depend on magnetic flux? And if the reed switch itself is faulty, then it's a fiasco. Even if I disassemble the relay, it won't help.

It turns out, reed switches contacts just oxidized. Reed switches sealed hermetically. It would seem that they should not oxidize. But practice shows that this is exactly what it is.  Yesterday, I tried to control relay coil from 30 volts DC.  This shouldn't scare anyone because the coil has a resistance of 500 ohms, so the current is 60 milliamps for 1 sec.  I heard before that for microswitches and buttons there is such a thing as minimal current. In correct circuit design (as for microcontroller circuits) you must maintain a minimum current so that it breaks through the oxide layer of the contact. Otherwise, the contact stops working over time. I think this is exactly what happened here inside the HP 34401a.

So yesterday I read the datasheet for the relay, installed a light bulb, looked at the current on the power supply, and started passing current through the reed switch contacts, controlling the coil from 30 volts. Then I disconnected the wires and checked the resistance of the reed switch contacts.
I thought that it would be correct to control the coil with one or the opposite polarity. For this, a controlled transistors bridge needed (or a computer-controlled power supply).  In theory, just need Increasing the current through contacts and to measure voltage drop across reed switch plotting graph. All this must be done by sending pulses through coil so it clicks contacts hard enough. I think even +/-50v is ok (ten times more, 100milliamps but this just short impulses, not holding current).  The main task is to clean the contacts, to remove a very thin layer of oxide. And when the graph becomes +/- flat, it means that the relay is cured.

The 34401a can be used without K101. It does not pass the test with a shorted relay, but other is ok.  It seems K101 only function is to completely disconnect  de-energized 34401a's input. Not sure if I'm right.

While I was thinking this, an incident happened. My power supply has two channels. There are two double wires coming down from the shelf above. As a result, I applied 30 volts to the relay contacts, but a limited current of 0.5 amps (by power supply current limiting). According to COTO 3501-05-511 datasheet, switching current 0.5 amps, so if HP custom relay like COTO 3501-05-511 it should withstand 0.5amps. But I discovered that the light bulb was now constantly on, contacts were welded to each other. I wasn't too upset, since the relay seemed hopelessly broken to me anyway. However, since there is nothing fragile inside, just reed switches in the filling, I took and lighyly hit relay both sideways and vertically on ceramic tiles floor. It started working again. After this incident when it closed it shows same resistance 213-216 milli-ohms. It doesn't float like it before. However, when I close and open the relay contacts, It can be different each time within a small range of 1-3 milliohms. I don't know if this is normal. Maybe new relay should always have a strict constant resistance? Common sense says that it seems to be. It would be great if someone studied behavior of their new and not new reed relays. We need at least a few of them to accurately determine behavior.  Why do I have doubts? Because internal 34401a ADC input is high-resistance, so change in resistance by ohms will be unnoticeable.  This only becomes noticeable in a 2-wire resistance measurement. This is my assumption. Having also studied these issues, I believe that the aging contacts of reed relays in precise 6.5 - 8.5 digits instruments is the cause of various kinds of troubles including long-time stability.

By the way, I think Panfnutiy here is incorrect:
Recovery of the relay K101(0490-1914) in HP/Agilent 34401A 
He tested the reed switches, but he did it with a powerful neodymium magnet. This is incorrect. In my case, relay contact resistance changed when I applied 5V it became significantly less for 30V, vice-versa. Contacts resistance depends on the magnetic flux created by the relay coil.  With a powerful magnet it will certainly work. So it is unlikely that the Panfnutiy cause was poor soldering contact. The reason was the same: formation of "something" (oxide, ultrafine powder due to wearing, etc..) on the reed switch contacts. One of the reasons why these relays behave poorly is the low voltage of the control coil. More precisely, high coil resistance (500Ω) and very low current <10mA. It is clear that the coil heats up very little and therefore the thermal EMF is small. But magnetic field is very weak and the contacts apparently are not cleaned properly. They just don't press together hard enough.

For now I soldered the relay back, I want to see how it behaves. Of course, I plan to find a replacement for a new relay. The only reason that made me do all this is the price in local stores, which is from a third to half the cost of the multimeter itself.

Happy New Year 2025!
 
The following users thanked this post: Kean

Online robert.rozee

  • Frequent Contributor
  • **
  • Posts: 326
  • Country: nz
Re: Advice please, used 34401a
« Reply #22 on: December 20, 2024, 04:25:27 am »
reed switches are very cheap, has anyone thought of making their own replacement reed relay? and at the same time, making it so that the reed switch(es) are easy to replace if/when they wear out?


cheers,
rob   :-)
 

Offline stdTopic starter

  • Contributor
  • Posts: 25
  • Country: 00
Re: Advice please, used 34401a
« Reply #23 on: December 20, 2024, 06:29:30 pm »
Yes, I have had this thought. There is (almost) nothing inside a reed relay except a coil and two reed switches cancelling out thermal EMF. See U.S. patent #4,084,142 (1976, expired) https://patents.google.com/patent/US4084142A/
The patent shows a nice connection of the ends of the reed switches, but if you look at the relay teardown, these are just soldered (or welded?) ends of the reed switches.
COTO seems produces its own reed switches. I think thermal EMF rating/grade is determined either by sorting the final relays, or by selecting reed switches before installing into relay (or both). Therefore, we need nanovoltmeter to sort reed switches by EMF, see "Understanding Thermal EMFs in Reed Relays | from Pickering Electronics | The Reed Relay Experts"

My search for reed switches with low thermal EMF has not been successful so far. Do they even exist?  Seems low EMF is relay property, not reed switches. In any case, sub-microvolt measurements are needed.
Happy New Year 2025!
 

Offline stdTopic starter

  • Contributor
  • Posts: 25
  • Country: 00
Re: Advice please, used 34401a
« Reply #24 on: December 20, 2024, 06:50:31 pm »
This seller https://www.ebay.com/itm/325250068326
sells a Standex-Meder SHV05-1A85-78L3K with a coil resistor as a replacement for the stock relay. I couldn't find any information about the Standex-Meder SHV05-1A85-78L3K having low thermal EMF. Am I missing something?
Happy New Year 2025!
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf