A 100 MHz, 3600 V/us CFB _QUAD_ for two and a half bucks? Where has this chip been all my life?
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com
Four times as much heat to get rid of.
-- John Larkin Highland Technology, Inc trk jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
Yeah, but the rather nice 200-MHz THS3091 that I've been using costs five bucks in reels and ten bucks in onesies.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com
Yes, that or just a DSL driver.
That area isn't on my turf yet but it is a nice amp. Distortion isn't stated past 40MHz though and up there it doesn't look pretty.
-- Regards, Joerg http://www.analogconsultants.com/
5V at 100 ohms is 0.25W. Have you checked some RF amplifiers. Like those in some antenna systems. They use 75R.
That one is old. Sometimes it is good to look at specialty markets since those chips are often not advertized in the regular analog marketplace. In the 80's we had similar problems where the only really fast opamps were from Harris but only came in the super-pricey mil-spec ceramic package, you had to call the utility before turning it one and they almost got hot enough to result in a blister after touching. Then I discovered the unsung heroes which were only marketed to engineers designing video recorders and hard drives. uA733 and such, those were a real bargain in all domains. Then some Japanese ICs that even our distributors didn't know about.
Same with discretes where PIN diodes were almost mil-pricing. Once I had to bypass our purchasing department because they wouldn't bypass "their" distributors which couldn't get that stuff. I ordered the diodes myself, in Malaysia, and that resulted in a VP office discussion where it got loud.
And sometimes you just need this much muscle. Just like a 2-liter engine is more efficient but can't tow the 5-ton trailer.
-- Regards, Joerg http://www.analogconsultants.com/
Unfortunately they have lousy to no DC performance. They also have poor linearity when the output swings +/-2.5V and most can't even do negative without some tricks.
-- Regards, Joerg http://www.analogconsultants.com/
?
le!
Do you know of a wideband analog amp that is stable with a cable load? They all require back termination. It seems paradoxical that the typical amp wi ll break into an oscillation frequency at which the cable impedance is resi stive, so where did the phase shift come from?
Add a transformer to get low output impedance and wrap the feedback loop around that?
Cheers
Klaus
He already stated that he needs to drive several volts, for some strange reason. A transformer just means he has to drive even higher levels out of the amp. Seems counterproductive. I'd look at the assumptions. High line levels are the problem. See if something can be done about that. If you don't like the answer, change the question. ;-)
We need this to work down to DC.
High line levels are customary in industrial environments. You also have those with DSL and such and there are nice chips for it. They are single-supply though (can be hacked if needed) and have huge power dissipation which would be a serious problem in this case. Those are meant to drive at >10Vpp.
Anyhow, it's the customer's wish and the customer is always king.
-- Regards, Joerg http://www.analogconsultants.com/
If it's so easy, what's your problem?
The customer is always the customer but you're the one he's hired to be the expert.
As I have described, heat. Plus space. This thing also has to be very tiny.
If the customer wants something they usually have good reason why and I am the guy that needs to make it happen. If it was easy they wouldn't need me. Consultants tend to only get the difficult projects which is something I thoroughly enjoy.
In essence I have a solution which is brute force. If needed two regular diff-amps get paralleled. The main reason for this thread is my curiosity why most diff drivers are spec'd at unrealistically high Z.
-- Regards, Joerg http://www.analogconsultants.com/
So you want power but no heat. That is an interesting problem. Ya think it's time to change the question, yet?
Usually always. Without understanding the complete problem, you (and we) can't know.
The answer to that is pretty obvious.
Run class D at maybe 500 MHz.
-- John Larkin Highland Technology, Inc trk jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
No, I want linearity with a modest quiescent current. Power during signal peaks is not a problem because they will be short burst transmissions.
Well, I know :-)
Looks like a marketing decision on the part of the mfgs to me. IMO there would be nothing wrong with stating "Here, this would be the max swing and distortion specs if you run it at 100ohms like almost anyone who wants to drive a line does".
-- Regards, Joerg http://www.analogconsultants.com/
Are there any solutions using class G amplifiers? If you have dual-rail plus ground, max heat is at 0.5 V+ and 0.5 V-; if you have six-rails plus ground, max heat is at (V+-) * (2N+1)/6 , and the maxima are smaller.
But, the output circuit gets rather more complicated; does anyone make an IC version?
I doubt it and a complicating issue is that this amp needs a GBW far above 1GHz.
-- Regards, Joerg http://www.analogconsultants.com/
TI has a number of dual rail class-G xDSL drivers, but I think they are all too slow to do 100MHz
So what are people using for the drivers on sig. gens.? I've got a cheap rigol 25MHz. Surely that make a 200 MHz for more money. (DG4202, ~$1k on amazon, maybe there's a teardown vid.)
George H.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.