Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

donate to eham
   Home   Help Search  
Pages: [1] 2 Next   Go Down
  Print  
Author Topic: Best MMIC amplifier?  (Read 2240 times)
9H1FQ
Member

Posts: 194




Ignore
« on: May 11, 2019, 08:02:58 AM »

So far, some MMIC amplifiers  that I have built , were not very stable. They blew up after a few uses! Cheap Chinese stuff?
Which is the most stable, for general amplification use ?
Logged

An unknown gem of the Mediterranean sea is the Island of Malta. blessed with warm sun , all year round.
In just one day, you can see places which will make you travel in history from the Prehistoric temples throu the middle ages, the Knights of Malta, all the way to world war two. All the major civ
DL8OV
Member

Posts: 1057




Ignore
« Reply #1 on: May 11, 2019, 12:00:45 PM »

Look at the Minicircuits catalog and you will find  dozens of different MMIC devices, some high gain, some high power. some low noise and some low IMD. You pick the right MMIC for the job, for example.a device that is used as the first amplifier in your transmit chain would probably make a very poor receiver preamplifier.

If your MMIC devices keep on dying then either you are overdriving them, your resistor in the supply line is too low or your power supply voltage is too high.

Peter DL8OV
Logged
9H1FQ
Member

Posts: 194




Ignore
« Reply #2 on: May 11, 2019, 10:27:16 PM »

The thing with Mini- Circuits is that they have such a large selection that it’s hard to select the proper one for the job. The resistor and the voltage was always correct and carefully monitored . Overdriving is a possibility!
Logged

An unknown gem of the Mediterranean sea is the Island of Malta. blessed with warm sun , all year round.
In just one day, you can see places which will make you travel in history from the Prehistoric temples throu the middle ages, the Knights of Malta, all the way to world war two. All the major civ
VK6YR
Member

Posts: 15




Ignore
« Reply #3 on: May 12, 2019, 03:09:26 AM »

Thanks re the MMIC questions. I also have an interest and would appreciate recommendations on Minicircuits MMIC's suitable for HF use, specifically Rx pre-amp, transmit chain and MOSFET driver. Thanks in advance.
Geoff VK6YR
Logged
W1VT
Member

Posts: 3394




Ignore
« Reply #4 on: May 12, 2019, 03:32:45 AM »

Sometimes the device has too much performance for the application.  If an amplifier has plenty of gain into say, 6GHz, that means it as to be designed using 6GHz techniques, even if you only want to use it at 14 MHz.  Using HF design  techniques and implementation will almost certainly result in oscillations somewhere in the GHz region. You can have waveguide coupling from output to input when you put it in the box!
« Last Edit: May 12, 2019, 03:35:06 AM by W1VT » Logged
KE0ZU
Member

Posts: 469


WWW

Ignore
« Reply #5 on: May 12, 2019, 05:39:05 AM »

As mentioned Mini-Circuits make a tremendous variety of amplifiers and distribution products.   They can be found on the bay for very reasonable prices.  They are NOT CHI-COM stuff.



Block diagram of my antenna distribution network pictured above.  This provides all 6 receivers on the operating bench connection to the antenna simultaneously. I often run 1500W amplifiers with this configuration and have absolutely no problems with receivers, be they tube or SS designs.


« Last Edit: May 12, 2019, 05:58:08 AM by KE0ZU » Logged

Regards, Mike
https://mikeharrison.smugmug.com/
Pics and bold print are usually links.
WD4HXG
Member

Posts: 379




Ignore
« Reply #6 on: May 12, 2019, 06:32:32 AM »

If you confine your selections to 1 GHz and below the choices
for Mini-Circuit MMIC's reduces to a handful of numbers. Take a
look at the outtake from their website below.



This series operates mostly from DC to 1 GHz so you are not
inviting gain bandwidth problems with devices that can operate
up to 6 - 20 GHz.

Also when using the device as a new user, "DO NOT" use dead
bug construction techniques. The devices are designed to be
driven from a 50 ohm source and drive a 50 ohm load. To that
end use printed circuit board with microstrip widths that provide
50 ohms. Using dead bug techniques can be done, but it is a
really good way to invite Murphy to the party.

The devices provide the performance as advertised. If you drive
it with a source that is not 50 ohms resistive or load with other
than 50 ohms resistive you can potentially create conditions for
it to oscillate. That is the reason for emphasizing using 50 ohm
tracks on your board layout.

If for some reason you cannot drive with a 50 ohm source or load
the device with 50 ohms and you determine that is causing a
problem you can insert a resistive attenuator pad in the signal
path just prior to the amp or just after.

As an example, you lay out the board using a MAR-1+. When you
fire up the circuit it works ok but the input is twitchy, moving your
finger near it, or waving your hand over it causes it to oscillate or
go regenerative. Insert a 3 dB PI attenuator in line with the input.
The 3 dB attenuator will immediately improve the input match by
6 dB. So if the input match is say 5 dB, adding the pad will improve
the return loss to -11 dB. The device is now sourced by a more
resistive signal source and most of the time any instability will
not be an issue. Yeah you sacrifice 3 dB of gain but with these
devices you can cascade stages easily enough to recover it.

Lastly do not use 'cheap' (low quality) caps for coupling or
decoupling. Many of the low cost surface mount caps have a
propensity to experience termination failure when using hand
assembly and soldering. I avoid this using Johanson and ATC
surface mount parts. Any place I can I use NPO caps and only
use X7R cap material or similar where forced into a corner. The
NPO material caps have proven much more rugged in hand
assembled devices for me.

Just my 2 cents and YMMV.

73

Chuck WD4HXG

73

Chuck WD4HXG

Logged
VK6YR
Member

Posts: 15




Ignore
« Reply #7 on: May 12, 2019, 07:58:53 AM »

Thank you for the responses and thank you Chuck WD4HXG for your comprehensive reply. I shall keep your valuable advice in mind when next donning the close-up glasses and wielding the fine tip iron!
Geoff VK6YR.
Logged
N5EG
Member

Posts: 347


WWW

Ignore
« Reply #8 on: May 12, 2019, 08:31:12 AM »

Hi Geoff and ??,

Even though you are using the MMIC device at low frequency, it is
inherently capable of UHF-Microwave performance, and you
still have to treat it like a microwave part or else it can
oscillate.

One of the most important things when using a MMIC amp is
SHORT grounding length of the ground tabs on the package. These
connect to the emitters of the transistors inside. Even
3 or 4 mm of lead length can cause the device to oscillate.
You need to count the length of the through-holes if you solder
the part to the top side of the board but the ground plane
is on the bottom side.  One through-hole per ground pad is
nowhere near good enough in my experience.

On the older pill package style, it was best practice to drill
a hole in the board that the package can just drop into with small
clearance. That way the leads did not need to be bent, and could
connect straight to the board with minimal length and maximal
straightness.

You may want to read this HP app note:
http://www.hp.woodshot.com/hprfhelp/4_downld/lit/iclit/ans001.pdf

-- Tom, N5EG



Logged
G3RZP
Member

Posts: 1313




Ignore
« Reply #9 on: May 12, 2019, 10:03:56 AM »

Although it is some years since I used one, my experience mirrors that of Chuck WD4HXG. I figure he has very good advice there...
Logged
W1VT
Member

Posts: 3394




Ignore
« Reply #10 on: May 12, 2019, 02:25:53 PM »

The old MAR-X series is pretty well behaved except for the MAR-8, which has no internal gain feedback, and thus has way too much gain to be stable unless carefully applied.
Logged
VK2TIL
Member

Posts: 648




Ignore
« Reply #11 on: May 12, 2019, 03:06:59 PM »

As N5EG said, grounding is vital to success; I learned that years ago when I made MMIC PCBs with a simple microstrip down the centre.

I got them to work after a lot of fiddling with Z-wires, copper-foil edge wrap-arounds and even copper rivets but, today, you can buy a PCB peppered with tiny vias or a complete MMIC amp on ebay for a few dollars.

Out of curiosity I bought this recently;

https://www.ebay.com.au/itm/LNA-50-4000-MHz-RF-Low-Noise-Amplifier-Signal-Receiver-SPF5189-NF-0-6dB-EW/263930044259?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2057872.m2749.l2649

I didn't need it, I was just curious.

It is actually very good, plenty of fairly clean gain from near-DC to some GHz; I am in the process of modifying its bias circuit, encasing it and putting it to work on the bench.

It uses an SPF5189Z MMIC; the device is spec'd down to 50MHz but delivers gain at much lower frequencies.

The bias circuit for MMICs is very important.

Consider that the DC supply is bypassed and is therefore at ground with respect to RF; the bias choke & resistor are therefore connected from the output to ground, ie shunting the (usually) 50-ohm load.

If the resistor is low-value and the choke is ineffective, the output is almost shorted and a very poor match to 50-ohms, not to mention reduced output, results.

The SPF5189Z in the above ebay amp is a 5v device and the complete unit, sold as designed for 5v operation, has a bias resistor and a choke in series.  Investigation reveals that the bias resistor is actually a 0-ohms type and the choke alone is therefore across the output.

Chokes are tricky things and I'm not sure that the choke in a six-dollar amp is going to be of the finest broadband quality.

I want 12v operation in any case so I will use an approx. 80-ohm bias resistor and a home-made choke which I have already made and tested.  At about 90mA, power rating of the resistor must be considered.

Coupling capacitors are also important; many, even most, of the SM caps that we see are X7R or similar, with fairly high dielectric loss; in my rebuild of the "six-dollar amp", I have used NPo types which will have less effect on loss and input & output match.

Some years ago I wanted a lot of gain for experiments with noise so I built a two-stage GALI-39 amp from a kit sold by MiniKits in VK5 (the kit was sold as a single-MMIC device but the PCB has provision for two MMICs);




It gave good gain with the lid off but also received all the local FM & TV stations from the tower about 6km away; with the lid on, the gain dropped to almost zero.

Re-group & re-build; the second version;




with a close-fitting shield box inside the main case, works very well;




and I'm in the process of building another with SMA connectors.



.
Logged
VK2TIL
Member

Posts: 648




Ignore
« Reply #12 on: May 17, 2019, 08:57:22 PM »

A duplicate of the above amplifier, but with SMA connectors, has been unfinished for some time; inspired by this topic, I have (almost) finished it;



A couple of points;

The blocking capacitors are paralleled 22nF NP0 types; this was done mainly because I didn't have the desired 47nF types on hand but there may be a minor benefit in reducing total inductance.

Two paralleled 0805-size resistors are used in each DC bias line; the required resistance is 240 ohms and power dissipated is 300mW.  I have a "stash" of 1206-size resistors but not in 240 ohms so two x 470 ohm 0805s in parallel were used at each bias point.  0805 resistors are rated at about 1/8 watt so these may be slightly over-run but I used the same technique in the BNC version which has operated reliably for some years now.

Each bias line has an RF choke; these connect to DC via another choke.

The chokes are just 12T of 0.7mm ecw on an FT37-43; testing with a VNA shows good impedance and no resonances up to 1 GHz which is the range in which this amplifier will be used;




The amplifier draws 69.3mA; MiniCircuits recommend 35mA for the GALI-39 so that is fine. Care should be taken to keep within the MMIC maker's limits for device voltage & current; MMICs are very robust but there are limits.

A quick check shows about 43dB gain up to 500 MHz; I will now seal it in its case before final testing.

Here it is in its inner case; not soldered-in yet, just fitted;




The outer case work, machining and painting, is done; the final product will resemble the BNC-connector model in the previous post.


.


Logged
W1BR
Member

Posts: 4194




Ignore
« Reply #13 on: May 19, 2019, 09:58:49 AM »

Why use a MMIC for HF?  A 2N5109/LT1001A using current feedback will provide decent gain and bandwidth with stable operation.  Why use a GHz device for HF, especially if it needs a 3dB pad to ensure stability?  Losing 3dB of gain is not the issue, but adding 3dB to the NF might be. There are many sources for blank PC boards for 2N5109 general purpose RF amplifiers.

I understand MMIC devices are cheap and tiny, but...

Pete
Logged
VK2TIL
Member

Posts: 648




Ignore
« Reply #14 on: May 19, 2019, 02:53:38 PM »

Agree; it's "horses for courses".  Considerable gain where it's not required should be avoided.

G4AON has done a very good design for the 2N5109 or similar;

https://www.qsl.net/g4aon/pdfs/aonrx_pma.pdf

It could be built in "ugly" style.

SV1AFN sells a nice kit or just the PCB;

https://www.sv1afn.com/2n5109.html

I use this little amplifier for HF work on the bench;





It's a little more elaborate than a single-2N5109 device;



but not greatly so.

Correcting an error in my previous post the chokes are wound on FT23-43 toroids.

.
Logged
Pages: [1] 2 Next   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!