Recently acquired at a hamfest. I was attracted by the nice milled aluminium box with goldplated SMA connectors, terminated with 50R SMA terminators.
The single SMA is labeled RX_IN, the 4 SMA on the right are labeled LNA. The 2 grey devices are Anaren microwave directional couplers.
I can identify some power regulator chips. Plus (?) MMIC or GaAs devices(?).
Anyone recognise this one?
Wilko
no mmic's, all gaas fets, the big ones are some kind of mgf1801 types.
i have several boards that look like this but are made for the 900mhz band and the big fets do 300+mW on 10ghz if properly tuned and on a custom pcb.
Working from left to right the signal comes in and is filtered, it then passes through two MMIC amplifier stages before hitting the main power amplifier assembly. The main power amplifier uses 2 x 90 degree hybrids and two more higher power amplifier chips (the yellow devices), it's done like this to improve transmit IMD and ensure a 50 ohm impedance on both the amplifier input and output. Finally the amplified signal is split into four using three 180 degree hybrids.
The circuits at the bottom are either for bias or transmit/receive switching.
That's a cellphone basestation receive amplifier with four outputs. Noise figure should be sub 1dB, and possibly better. The input stage will need to be pHEMT to meet a basestation spec.
A couple of detail pictures. Maybe that helps. The big FETs(?) look like ceramic material, maybe BeO(?), topped with a gold plated cap.
That's a cellphone basestation receive amplifier with four outputs. Noise figure should be sub 1dB, and possibly better. The input stage will need to be pHEMT to meet a basestation spec.
The label on the aluminium box says 1710-1785MHz.. hm, could be cellphone although I think in .nl it is 1800-something. But then again,
it does not need to originate from .nl
Why 4 outputs? That puzzles me a bit I must admit.
no mmic's, all gaas fets, the big ones are some kind of mgf1801 types.
i have several boards that look like this but are made for the 900mhz band and the big fets do 300+mW on 10ghz if properly tuned and on a custom pcb.
FETs look quite similar to what is shown in the datasheet:
http://f5ad.free.fr/Docs_ATV/MGF1801.pdf
That's a cellphone basestation receive amplifier with four outputs. Noise figure should be sub 1dB, and possibly better. The input stage will need to be pHEMT to meet a basestation spec.
The label on the aluminium box says 1710-1785MHz.. hm, could be cellphone although I think in .nl it is 1800-something. But then again,
it does not need to originate from .nl
Why 4 outputs? That puzzles me a bit I must admit.
Four outputs because it's a basestation, which wants to deal with more than just one carrier at a time, hence why as a receive amplifier it has a decent IP3.
In Eu, the 1800MHz band is 1710-1785 for a mobile to transmit and a basestation to receive. The basestation transmits between 1805 and 1880.
That's a cellphone basestation receive amplifier with four outputs. Noise figure should be sub 1dB, and possibly better. The input stage will need to be pHEMT to meet a basestation spec.
The label on the aluminium box says 1710-1785MHz.. hm, could be cellphone although I think in .nl it is 1800-something. But then again,
it does not need to originate from .nl
Why 4 outputs? That puzzles me a bit I must admit.
Four outputs because it's a basestation, which wants to deal with more than just one carrier at a time, hence why as a receive amplifier it has a decent IP3.
In Eu, the 1800MHz band is 1710-1785 for a mobile to transmit and a basestation to receive. The basestation transmits between 1805 and 1880.
Ah I see. Thank you! I yield to superior knowledge
it was a hard sell to get my cat interested in that.
I've just had one of these Ah! moments, my original post was wrong.
The four connections on the right are INPUTS, one from each receive antenna on the mast which (presumably) has a low noise amplifier built in. These are then combined and fed through a high IP3 amplifier which is constructed using a pair of 90 degree hybrids and two FETS. There is then two more amplifier stages before the signal exits on the left side.
What fooled me is that the unit has been built backwards with the inputs on the right and the output on the left.
I've just had one of these Ah! moments, my original post was wrong.
The four connections on the right are INPUTS, one from each receive antenna on the mast which (presumably) has a low noise amplifier built in. These are then combined and fed through a high IP3 amplifier which is constructed using a pair of 90 degree hybrids and two MMIC amps (or maybe FETS). There is then two more amplifier stages before the signal exits on the left side.
What fooled me is that the unit has been built backwards with the inputs on the right and the output on the left.
You were right first time, the input is on the left, with four outputs on the right.
Sofar I have not found any docs on this device, specifically the subD9 connector. I would love to power it up to see if it still works.
Any suggestions?
I've just had one of these Ah! moments, my original post was wrong.
The four connections on the right are INPUTS, one from each receive antenna on the mast which (presumably) has a low noise amplifier built in. These are then combined and fed through a high IP3 amplifier which is constructed using a pair of 90 degree hybrids and two MMIC amps (or maybe FETS). There is then two more amplifier stages before the signal exits on the left side.
What fooled me is that the unit has been built backwards with the inputs on the right and the output on the left.
You were right first time, the input is on the left, with four outputs on the right.
+1 The DC feed is on the right of each amplifier, so unless they have some backwards MMICs that take DC on the input, outputs are on the right.
all amplifiers are gaas fets.
i have put them into my own preamps and transmitters and put them on the correct bias via negative gate voltage.
the fets work up to 12 or 18ghz, maybe even higher. i have build 24ghz amps around simular looking gaas fets.
I hacked it up a bit to remove the 2 splitter circuits from the output to obtain a single input/single output amplifier. A sweep with the SA shows the attached image.
With -6dBm input power.
Goes without saying that my SA cannot get to the frequency this amplifier is designed for (i.e. 1.7-1.8 GHz). The directional couplers Anaren 1A1305-3
it uses are designed for 1.5-2.2GHz.
Wilko
Just think, now you have an excuse to buy a new spectrum analyzer that goes to 3+ GHz.
Just think, now you have an excuse to buy a new spectrum analyzer that goes to 3+ GHz.
Don't rub it in. Couple of weeks I bumped into a nice portable HP 22GHz incl TG. €900
😂😅
But you can spend a € only once
I would say Wilkinson dividers and not 180degree hybrids.
Can't really get a lot of power from those transistors though so I'm wondering if this was some kind of measurement amp?
Normally you have two inputs on the BTS antenna, +/-45 deg polarization. Here, you're splitting into 4 equal outputs. Why?
Good find!
Working from left to right the signal comes in and is filtered, it then passes through two MMIC amplifier stages before hitting the main power amplifier assembly. The main power amplifier uses 2 x 90 degree hybrids and two more higher power amplifier chips (the yellow devices), it's done like this to improve transmit IMD and ensure a 50 ohm impedance on both the amplifier input and output. Finally the amplified signal is split into four using three 180 degree hybrids.
The circuits at the bottom are either for bias or transmit/receive switching.
Normally you have two inputs on the BTS antenna, +/-45 deg polarization. Here, you're splitting into 4 equal outputs. Why?
A big GSM basestation will be running four carriers on both transmit and receive on each sector, hence wanting four outputs from the pre-amp to go to four receivers, one for each carrier.
I've occasionally seen eight carrier set-ups, but very very rare.