EMRFD Message Archive 11168

Message Date From Subject
11168 2015-06-12 11:58:18 ashhar_farhan Low IMD driver IC

I just chanced upon this 

http://cds.linear.com/docs/en/datasheet/643115f.pdf


The chip/module draws less than 100mA and provides 45dbm OIP3! Which brings me to the question. How does one design for higher intercept points? Is increasing the device current the only handle we got?


- f

11169 2015-06-12 12:25:00 Lasse Moell Re: Low IMD driver IC
Interestingly the "brother", LTC6431-20 do offer 20dB gain and slightly less NF, and almost 1/3 the cost at Digikey!

BTW it seems they are reluctant to tell us how they design this very low distortion circuit...

/Lasse SM5GLC

12 juni 2015, farhanbox@gmail.com [emrfd] skrev:

 

I just chanced upon this 

http://cds.linear.com/docs/en/datasheet/643115f.pdf


The chip/module draws less than 100mA and provides 45dbm OIP3! Which brings me to the question. How does one design for higher intercept points? Is increasing the device current the only handle we got?


- f

11170 2015-06-12 12:39:12 Dana Myers Re: Low IMD driver IC
11178 2015-06-13 15:13:29 billw77aaz Re: Low IMD driver IC
> I just chanced upon this
> http://cds.linear.com/docs/en/datasheet/643115f.pdf
> http://cds.linear.com/docs/en/datasheet/643115f.pdf
>
>
>
> The chip/module draws less than 100mA and provides 45dbm OIP3! Which
> brings me to the question. How does one design for higher intercept
> points? Is increasing the device current the only handle we got?
>
>
> - f
>

Too bad it doesn't go down to HF.
W7AAZ
11179 2015-06-13 16:04:52 Todd F. Carney / ... Re: Low IMD driver IC
11180 2015-06-13 19:41:48 Dana Myers Re: Low IMD driver IC
11181 2015-06-13 20:29:58 Todd F. Carney / ... Re: Low IMD driver IC
11182 2015-06-13 23:01:12 vasilyivanenko Re: Low IMD driver IC
Hi Todd et al:

I found parts of the datasheet confusing and would contact LT about it.

I wonder if the statement "......... However, if the  LTC6431-15 is preceeded by a low frequency termination,
such as a choke, the input stability network is NOT required ........"

Should go something like ---- However, if the  LTC6431-15 is preceeded by a low-impedance input termination  (like a ~50 ohm resistor), the input network may be omitted.

Shouldn't be difficult to determine on the bench.

Best
Todd(2)
11183 2015-06-13 23:38:24 billw77aaz Re: Low IMD driver IC
>
11184 2015-06-14 05:29:33 Thomas S. Knutsen Re: Low IMD driver IC
Since the design is based on a SiGe process, making capacitors that work at HF would be expensive. That's probably the reason for the 20MHz low frequency cutoff. 
The way around that is either gain reduction like in figure 1 on datasheet page 11, or negative feedback from output to input. 

SiGe BiCMOS, and most all SiGe devices should have the same 1/F noise as Silicon. For those with access to IEEE Xplore, try Lukyanchikova, N. et al, 1/f noise and generation/recombination noise in SiGe HBTs on SOI.,  10.1109/TED.2005.850697
SiGe may have a reduced 2. order intercept, that could lead to some problems, measurements should be preformed before the device is used. This is based on my experience with a couple of amplifier designs on SiGe, but I have not found any article reference that backs up this claim. Some more experiments are in order. 

S-parameters are avaible from Linear, some experiments in a simulator like QUCS or SPICE would do a lot to the understanding of the device, just make sure you don't simulate outside the frequency limits in the s-parameter set. With a decent S-parameter set, and a good simulator, you should be able to plot the Rollet factor and do stability analysis, leading to more insight into this device. 

The pricepoint on more advanced semiconductor compositions have dropped quite a lot over the last couple of years, making a spin of your design is now below the $100k mark, so we should hopefully see more interesting design emerging over the next years. 

73 de Thomas LA3PNA.


2015-06-14 8:38 GMT+02:00 bcarver@safelink.net [emrfd] <emrfd@yahoogroups.com>:
 

>

11190 2015-06-14 08:28:30 Dana Myers Re: Low IMD driver IC
11191 2015-06-15 19:12:51 biastee Re: Low IMD driver IC
An investigation into the use of external feedback and input terminati
11192 2015-06-15 21:56:54 Cecil Bayona Re: Low IMD driver IC
Good find, thanks for the link.

11193 2015-06-15 22:23:24 Todd F. Carney / ... Re: Low IMD driver IC
11194 2015-06-16 05:54:24 timshoppa Re: Low IMD driver IC
$7 for a sample. The 4mm QFN package, I might be able to dead-bug it and see how it serves as a small-stick 160M preamp/impedance normalizer in presence of my local AM broadcaster. If that works out OK, it could serve as a 160M directive RX array, at my home QTH, just hypothetically.

(I have had the chance to use much bigger 160M 8-circles at contest stations. Wow - they are like a 5 element yagi in directivity. Amazing. I do not have the  real estate to do a full-size 160M 8-circle, but I could see a EU-centered array and hope the house being in the middle doesn't mess it up too much.)

The low end frequency will almost certainly go way below the specced 20 MHz if I use coupling capacitors bigger than the 1000pF shown in the test circuit. Will try 10 times bigger coupling capacitor.

MMIC's with several watt dissipation (hundreds of mA) often spec 50dB IMD.

For comparison... some 160 meter pre-amps, DXE RPA-1 has been measured at 43dBm OIP3 and is specced at 200mA (nominal 12V?). Internal construction of that seems to be two Norton-like negative feedback 2N5109's in push pull.

Tim N3QE
11195 2015-06-16 08:04:32 farhanbox@gmail.c... Re: Low IMD driver IC

the imd specs are a moving target.   i discovered this a few days ago that all bets of imdr increasing by 3db for every 1db are off when you allow the device current to increase with drive.

this is how class AB1 amplifier biased with a few milliamps can report 50db current!

11196 2015-06-16 14:33:11 Thomas S. Knutsen Re: Low IMD driver IC
I knocked out a PCB for the amplifier, avaible in quantity of 3 for $10.3 here: https://oshpark.com/shared_projects/5qq6sXUX

Using Coplanar waveguide for the transmission lines, SMA connectors (you can get BNC in the same form) an SMD 7805 voltage regulator and 1206 size resistors and capacitors. For better UHF preformance, 0805 size or even 0603 should be used. By careful soldering, all of the suggested sizes should fit the PCB. The inductor is 1812 size, due to the better selection of high Q inductors in this size. 

I'm hoping to get around to testing the amplifier during the summer, if no-one beats me to it.

73 de Thomas LA3PNA



2015-06-16 14:54 GMT+02:00 timshoppa@yahoo.com [emrfd] <emrfd@yahoogroups.com>:
 

$7 for a sample. The 4mm QFN package, I might be able to dead-bug it and see how it serves as a small-stick 160M preamp/impedance normalizer in presence of my local AM broadcaster. If that works out OK, it could serve as a 160M directive RX array, at my home QTH, just hypothetically.

(I have had the chance to use much bigger 160M 8-circles at contest stations. Wow - they are like a 5 element yagi in directivity. Amazing. I do not have the  real estate to do a full-size 160M 8-circle, but I could see a EU-centered array and hope the house being in the middle doesn't mess it up too much.)

The low end frequency will almost certainly go way below the specced 20 MHz if I use coupling capacitors bigger than the 1000pF shown in the test circuit. Will try 10 times bigger coupling capacitor.

MMIC's with several watt dissipation (hundreds of mA) often spec 50dB IMD.

For comparison... some 160 meter pre-amps, DXE RPA-1 has been measured at 43dBm OIP3 and is specced at 200mA (nominal 12V?). Internal construction of that seems to be two Norton-like negative feedback 2N5109's in push pull.

Tim N3QE

11197 2015-06-16 15:14:56 Dana Myers Re: Low IMD driver IC
11198 2015-06-16 16:07:24 Todd F. Carney / ... Re: Low IMD driver IC