ATTEN ATF20B teardown

[hacklordsniper]

So while I'm building second bench to be able working on two projects at the same i was waiting on eBay to literally steal a programable quad power supply which in the end gone for 140 $ and i over slept it while going to rest few minutes. ARRRGH.

SO while i checked other items from the seller he had an ATF20B for sale. Well i wanted to teardown one of these some time ago. So here it comes today. IT came in inappropriate box simply flying from here to there and smashing itself (China packaging)!. With it you get a BNC cable, BNC to croco cable and user manual in Chinese 

I noticed the unit is very heavy, much more heavy than Rigol DS1052E. It is also little bit biger than Rigol, in my opinion it is too long.

There is a USB port on the front and RS232 on the back, but more about that to follow...

After opening it i notice the circuit board fell of from its holder and happily swinged here and there whole way. It moved so much it even damaged the metal case... Luckily it did not seem damaged. But i must admit inside is a most sexy mains transformer i have seen in very long time!

Fellows with sharper eye will notice a very long push rod for a real mains switch,  Agilent did you see that ?!? The unit works fine, the display is little bit confusing and controls are too. The USB port on the front is fake, you cant store any waveforms on external stick, its not connected anywhere and there is no output for it on the board 

Whatever you wish, please ask because i don't know what to rant more about it 

You can see its brains clearly on the picture so im not going to write them up. You can also notice nicely marked JTAG

[hacklordsniper]

.

[saturation]

Nice, more later as I can study the photos.

I note you have both the Atten and the Rigol 1022, could you compare the two particularly how well the waveforms are generated?  Most importantly, if you can discern any harmonic distortion in the sine wave outputs, the rise time and squareness of the square waves up to the maximum supported frequency.

Thanks a bunch!

[hacklordsniper]

Nice, more later as I can study the photos.

I note you have both the Atten and the Rigol 1022, could you compare the two particularly how well the waveforms are generated?  Most importantly, if you can discern any harmonic distortion in the sine wave outputs, the rise time and squareness of the square waves up to the maximum supported frequency.

Thanks a bunch!

Hello,

i quickly checked both on my Agilent DSO-X when i managed to get it boot

The crap part is that the ATTEN can generate only the sine up to rated 20 MHz and others to 1 MHz. . Rigol DS1052 rise time is at square 1 MHz 19.8 ns, while ATTEN at 1 MHZ 22.8 ns.

The waveforms squareness is almost exactly the same, very nice

[Richard W.]

Pictures of the transformer and the amplifier output stage would be interesting.

What about frequency accuracy?

You can do some lissajous-patterns with both of the signalgenerators. Make some screenshots 

[hacklordsniper]

Pictures of the transformer and the amplifier output stage would be interesting.

What about frequency accuracy?

You can do some lissajous-patterns with both of the signalgenerators. Make some screenshots 

There is no separate output amplifier than what you see on the board. There is only 4 relays on a small PCB soldered on the front BNC.

The Atten outputs 19.98 MHz and 20 MHz and Rigol does not have this error. Rigol is far more accurate than Atten, and Rigol menu is so inituive like your pocket, while Atten is a little bit retarded. I will try to generate some and see what i get, now i need to go to bed because i just came from it to photo this transformer which initiated various complaints from my wife 

However here are the pics of the transformer, pure and i mean pure nerd porn. I never seen anything alike and i really opened alot of equipment in my life.

[saturation]

Thanks a bunch hacklordsniper.

So to sum, sine waves to 20 MHz, but square, triangle etc., waves only to 1 MHz?

Hello,

i quickly checked both on my Agilent DSO-X when i managed to get it boot

The crap part is that the ATTEN can generate only the sine up to rated 20 MHz and others to 1 MHz. . Rigol DS1052 rise time is at square 1 MHz 19.8 ns, while ATTEN at 1 MHZ 22.8 ns.

The waveforms squareness is almost exactly the same, very nice
[/quote]

[hacklordsniper]

Thanks a bunch hacklordsniper.

So to sum, sine waves to 20 MHz, but square, triangle etc., waves only to 1 MHz?

No problems, i need to push all the teardowns quickly since i will be rebuilding my lab so they sometimes lack the information.

Yes, and how retarded is that. CHANNEL A - sine 20 MHz, all other 1 MHz. CHANNEL B - sine 1 MHZ, all other 100 KHz 

[hacklordsniper]

Some fun DSO-X3000, Atten ATF20B, Rigol DG1022

Im sorry i noticed only now my scope date is set wrong 

[saturation]

Nice, your scope really is fast update rate to show Lissajous clearly, reminds me of an analog screen!

[ciccio]

However here are the pics of the transformer, pure and i mean pure nerd porn. I never seen anything alike and i really opened alot of equipment in my life.
[/quote]

This is a rather unusual type of transformer, called R-type, that was invented in Japan by KITAMURA KIDEN Co. ltd. http://www.kitamura-kiden.co.jp/english/products_e.html.
It has a semi toroidal core, made by first winding a variable width steel tape (the width is variable to get a circular cross-section core), then mounting on the core  two split bobbins that are clamped around the core and have small indents on their border (like a gear-weel) that allow them to by rotated by the winding machine.
This construction combine the low flux dispersion and high efficiency of toroids with the automatic fabrication possible with standard EI core.
I've seen making them in a exhibition: the machinery needed for making cores (cutting and winding the steel tape) and windings are impressive... and truly expensive, so only a limited number of manufacturer is equipped for them.

[hacklordsniper]

Nice, your scope really is fast update rate to show Lissajous clearly, reminds me of an analog screen!

n

Im amazed by my dso-x and realy think Agilent did a wonderful job (except some minor flaws). For fun i tried it on DS1052E and as expected got only garbage on the screen

ciccio thank you for claryfing the transformer. i will study the link

[alm]

Yes, and how retarded is that. CHANNEL A - sine 20 MHz, all other 1 MHz. CHANNEL B - sine 1 MHZ, all other 100 KHz 

Think about how the frequency spectrum looks for the various types of signals, and it will start to make more sense. The bandwidth needs to be much larger to generate a 20MHz squarewave with a reasonable fidelity than for a 20MHz sine. Most function generators either are the same or everything starts looking like a sine at the high end of their frequency scale.

[hacklordsniper]

I agree but the Rigol can do all other to 5 MHz and i just wanted to note Rigol is much better unit for the same price.

[Bored@Work]

The Rigol DG1022 does a 20 MHz sine, but is spec'ed to internally use 100 MS/s in ARB mode. And it can do a square wave up to 5 MHz. Now, why should they use a different sampling frequency for the build-in waveforms than for the ARB mode? If they don't, you have a 20:1 frequency ratio for the square wave.

The Atten ATF20B does a 20 MHz sine, and a 1 MHz square wave. That is 20:1, too. Coincidence? Could it be that the Atten internally just uses 20 MS/s, and that's it? Or Atten really messed up the analog part, designing it for 20 MHz, and then someone figured out, uhm, that's not good for square waves?

But if we ignore the Atten rubbish for now, what about the Rigol rubbish? If it does 100 MS/s internally, what prevents one from getting more than a 20 MHz sine out of it? Maybe only the firmware.

[Bored@Work]

I agree but the Rigol can do all other to 5 MHz

No, pulse is limited to 3 MHz, and ramp and triangle are limited to just 150 kHz.

[hacklordsniper]

I agree but the Rigol can do all other to 5 MHz

No, pulse is limited to 3 MHz, and ramp and triangle are limited to just 150 kHz.

Did not test those still, i was coninced that pulse is also up to 5 MHz and ramp and triangle to 3 MHz

[Bored@Work]

150 kHz was from the specs, and what I remembered. I can get my hands on a dreaded DG1022 earliest on Monday and have a look if the specs lie or not. But I rather prefer not having to touch it or any other Rigol gear.

[alm]

But if we ignore the Atten rubbish for now, what about the Rigol rubbish? If it does 100 MS/s internally, what prevents one from getting more than a 20 MHz sine out of it? Maybe only the firmware.

Maybe they screwed up the output filter, so they had to limit the bandwidth to suppress sampling artifacts from the DAC? Haven't felt the need to acquire one either, so unable to test.

[bruce]

I have been following this thread with much interest. This seems to be the only ATF20b in captivity in the U.S., The Optional accessories seem to be a mystery, They are,  RS232, a 7 watt amplifier and a frequency counter.  Pages 26-29 talk about PC control, How does that work if the USB is fake and RS232 is optional? and, according to Goodluckbuy, you can get the ATF20b for $319.00 with free shipping, or the ATF20d for $450.00 U.S. with RS232 and the frequency counter pre-installed??? Does the add on Frequency counter only add the ability to check the external frequency? I couldn't find this info on Atten's web site... I think I want to purchase one of these, AFT20b if I can control it externally. Thanks for the help, Bruce

[hacklordsniper]

Lose intrest please, no need to discuss this trash. I use it sometimes but its the worste piece of gear landed in my lab

[bruce]

Thanks,
 do you like the RIGOL DG1022 any better? I can get the Atten for $319.00 delivered, and the Rigol for $440.00. If you prefer thr Rigol DG1022 is it the specs, the operating procedures, or the reliability?Is Attn in your opinion, a bad brand, (with great transformers)? If none of the above, what function generator in the $300-$450 range that does similar functions, do you like?
Again thanks for your time. It has helped me.

Bruce

[hacklordsniper]

I have the Rigol too and i like it verry much. DG1022 is great and i dont regret even little for buying it. Software is little bit anoyying sometims but you can use Agilent software to create waveforms and store them to usb stick. Atten does not have software and usb is fake, Atten instrument software is full of bugs, idiotic and hard to use and whole instrument is piece of trash. I would reccomend Atten only if it would cost 50-100 $

[bruce]

Thanks,Hacklordsniper,
  I will be ordering the RIGOL DG1022  next week! in my case a lot to learn!

Bruce Meisinger

[hacklordsniper]

Thanks,Hacklordsniper,
  I will be ordering the RIGOL DG1022  next week! in my case a lot to learn!

Bruce Meisinger

No problem.

If you will need any help with the Rigol feel free to ask. Im glad i saved you from Atten!