I'm looking for a low distortion RF amplifier. So far I found the ADL5602 from Analog Devices. It has a 3rd order distortion of 42dB and a gain of 20dB. The gain is OK but I would like an amplifier with less distortion. So far I looked at devices from Avago, Hittite and Mini circuits but so far the ADL5602 seems the best there is. Any other suggestions where to look? BTW I could live with cascading 2 devices.
I have looked at that device from Mini Circuits but the gain rolls off quite sharp with increasing frequency. That seems to be a general 'problem' with the gain blocks from Mini Circuits. Also the output level used to measure the OIP3 numbers isn't specified. I would like to get at least 0dBm from the amplifier.
Meanwhile I found another chip which looks very promising: the TQP3M9009 from Triquint Semiconductor (never heard of them before but the part is available in single quantities from Mouser). The OIP3 is much better than the ADL5602 especially at the lower frequencies which is where I want the least harmonic distortions. The noise figure for the TQP3M9009 is slightly worse than the PSA4-5043 though.
that's gonna be tough, there is quite some research going on in cascode feedback amplifier design. Some research papers mention 51 dBm OIP3 with 250-3000 MHz BW. I don't recall having seen any commercially available but I'm not an active follower. IEEE has quite some papers.
best I could come up with next to the ones you mentioned:
5-2500 MHz, 43 dBm OIP3
SXB2089ZDS (Sirenza)50-4000 MHz, 40.5 dBm OIP3
TQL9048 (TriQuint)DC(?)-6000 MHz, pretty good OIP3 below 500 MHz
ADL5565 (TI)good above 500 MHz-4000 MHz
MMA-054025 (MWT)Maybe it would be better chop your BW up in 2 or 3 frequency regions with some kind of band switch. There are really nice components for DC-1GHz
Hi ntcnico,
It is difficult to find good wideband amplifiers that is working from DC to Ghz.
The high output version (low distortion MMIC) almost al start from 20 to 50Mhz.
Internalli this chips have coupling/decoupling and you can't use them from low frequenties.
A nother problem with MMIC is the sensitivity, there easely blown up!!!
Do not use a coupling capacitor on the input of more than say 10uF, or Pufff gone is your MMIC :-(
I think these IC's cant handel the charging currents through the input connection.
Another problem is the pull up from the collector, you cant use a inductor or inductor+resistor.
Your frequency respons wil not be flat arond a few Mhz.
Only with a resistor i had a flat respons from 10Khz to about 500Mhz +- 0.1dB
But you wil lose a lot of power in this resistor, no problemo for me, its a LAB wideband amp and used somtimes :-)
Kind regarts,
Blackdog
@Neganur: I tried the ADL5562 (a similar chip to the ADL5565) but I could not get that to work properly. Using baluns on a very small PCB proved to be hard. One of the problems is that I only have 13x13mm of PCB area to put the circuit on so bandswitching is also out of the question.
@Blackdog: Ideally I'd like to be able to go from 100kHz but I'm happy if I can push it from a few MHZ. Flatness is not really a concern because there is automatic output leveling. As long as I have about 20dB of gain or more the pre-amplifier stages don't distort too much.
The project is some hands-on tinkering with RF: I'm trying to devise a replacement for the output amplifier of an E4420 series RF generator from HP. The original amplifier is very hard to get but I start to get more and more respect for what the original amplifier is capable of. OTOH I lost my respect for HP because the reason the amplifiers fail often is the absence of a smooth surface on the (cast aluminium) heatsink which is supposed to keep the amplifier cool.