900Mhz 50W amp RF Deck and 2W 1296 PA
Mike Lewis
FYI. I found a 50W 28VDC miniature 900Mhz RF amplifier in what looked like a nickel plated machined brass enclosure. It is 47dBm out (50W ) with 40dB gain therefor about 7dBm input drive. It is very petite and needs to be mounted on a heat sink with 4 bolts. No bias control, just RF in, RF out and 28V at around 2.7A by my measurement.
900MHz 50W 47dBm PA WIFI Signal WLAN High Power Amplifier for 840-940MHz，28V | eBay for $122 shipped.
They have several similar amps for a range of bands and power levels. They even have a 50W 2 to 18Ghz PA for a cool $25,110 😊.
Their 50W 5.7Ghz PA caught my eye, maybe later. 100% New 5700-5900MHz 50W 5.8GHz PA Wireless Signal High Power Ampliifier 47dBm | eBay
$220.
I am still waiting for delivery of my 120W 12V to 28V converter to power the PA which will mount on the outside back panel. The converter and PA are switched on only when 900 band is active. I can also switch the PA ON during TX only if desired. I added in sequencing today using the solid state IF switch as a RF block during RX to TX transition, letting the T/R relay settle before I let the RF flow.
My 1296 2W amp is supposed to arrive tonight. I found a writeup on this module today with details power/gain/current plots. Even a hand drawn schematic. It has 2 stages. The first stage takes 0dBm input and boosts it to near 20dBm. 20dBm is what my transverter board outputs so I plan to bypass the 1st stage and skip attenuators keeping the cabling cleaner. This will bolt on the opposite heat sink end panel next to my 222 transverter board. Chinese 2W Amplifier 40-1200MHz 1v0.pdf
Microwave RF Power Amplifier Board SBB5089+SHF0589 40MHz-1.2GHz 2W Gain 25DB #TZ | eBay $11.62 shipped. Quite the value. These use discontinued 2007 era parts putting them to good use.
I bolted the PA unit on the heat sink end panel of my 3-band transverter box with some grease. The UR3LM 903 Xvtr boards puts out up to 20dBm. I used a 12dB attenuator and 8dBm got me to a solid 47dBm out. Holding it key down with FM signal for a couple minutes the end panel was warm, the amp enclosure was hot to the touch neat the RF output SMA connector but low enough I could keep my finger on it. I used copper jacket UT-085 on the output running to the SMA T/R relay. The input is 12” of RG-316. I have a 10dB and 2dB attenuators visible on the input side. Seems to work well on my test bench. Power held steady over a 2 minute transmission. Saw about 2.7A draw at 28V at 50W.
Mike Lewis, K7MDL
CN87xs
Rick Beatty
--
~The Voice of the Pacific NorthWest VHF Society~
You are subscribed to the Google Groups "PNWVHFS" group.
To post to this group, send email to PNW...@googlegroups.com
To read message history, visit http://groups.google.com/group/PNWVHFS
---
You received this message because you are subscribed to the Google Groups "PNWVHFS" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pnwvhfs+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/pnwvhfs/SN6PR12MB2640BDE1CC32612075209EFEF60F2%40SN6PR12MB2640.namprd12.prod.outlook.com.
Mike Lewis
Follow up on the 1296 2W module. The SHF-0589 is rated for 2W up to 3Ghz. Pretty simple circuit. I cannot get more than 30dBm out across the range of frequencies. Should be 33dMb at saturation. I bypassed the 1st stage to drive the 2W part directly, the most I could generate (conveniently) is 18dBm at 1296 and I got 28dBm out. At 1100Mhz with my signal generator I can do over 20dBm and still got < 31dBm out. Used to drive the SG-Labs 25W PA, I only get 12-15W out. The 1296 Xvtr board is putting out 14dBm, not 20dBm so this 2W amp module part is not going to work.
Decided to move to another solution. I could not decide whether to do a 5W boost amp to pair with the 25W SG Labs PA/LNA which can be mounted near the antenna (requires big attenuator), or do an internally mounted 50W amp and maybe add an internal LNA. And move things around to fit it in. I will likely install the 1296 50W amp in place of the 903 PA and move the 903 PA to the middle or opposite end panel and use some sort of standoff blocks or finned heat sink.
Most of the amps I could find required very low input drive, did not cover 1300Mhz, and few were 12V.
So I ordered both 1296 PAs and will decide when I see them in my hands and how they will fit the space I have.
This one is 12-18V for about 7W at 14V.
1.2-1.6GHz RF Power Amplifier 5W Output 12-18V 40dB with SMA Female Connector | eBay
This one is 28V but 50W.
1.2GHz 50W Power Amplifier Module High Quality RF Amp with SMA Female Connector | eBay
I like the idea of having 50W on 1296, goes well with the 50W on 903, all internal to the box, with 80W 432 and 180W on 2M via external amps. 222 is still 8W for now. I will have to add one more power relay to switch 28V between the 903 and 1296 PAs.
Mike Lewis
I received 2 1296 amps (50W and 8-12W) I ordered off eBay today. Still waiting for the 200-300Mhz 50W amp. I opened them up, inspected, then put them on the test bench. I have now received 4 of 5 PAs that I ordered in late November and December. 1 is a jammer-converted-to-PA unit.
- 1280MHz @ 50W at 28V – Drive 0-5dBm. Jammer-converted-to -PA. Looks like a 3-stage design with about 42dB gain total. There is a metal wall isolating the final stage. No VCO installed. They claim it does 47dBm output with 0-5dBm drive. This is true at 960MHz, not 1296.
The final is a NXP BLF6G10S rated for 700-1000Mhz service at 45W and 23dB gain at 960MHz. Curves show 35W CW and 45W PEP on 2-tone is the sweet spot before gain falls off. It was introduced in 2007 and last updated in 2015 targeting the CDMA sector at 960Mhz. It has internally matched 50ohm inputs and outputs. Spec sheet says good to 12A drain current. Lots of tuning capacitor pads around the input and output stripline copper. It has a +5VDC bias enable wire and a red TX ON LED visible through a hole in the lid. A CPU can control it easily with TTL levels.
The 2nd stage driver is a NXP MW6S004N 28V 4W PEP, 18dB gain. 1-2000Mhz. Gate threshold about 2V. The 1st stage looks like a YG602020 which I believe is a Chinese foundry, looks like a MMIC class device.
Before tuning, best RF power I could get out at 1296 was 21W . I added 2x 1pF MLCC caps on the input and gained a couple watts, adding more made no change. The output wanted 1.2 to 1.5pF more. I installed a 1-6pF ceramic trimmer on the output in an empty pad. I could peak it up to 29W at 2A draw. At 30V I get 30-31W.
The output at 1296 is capped at 30W likely due to the internal matching. If I reduce the output to 20W I can put a cap on the input of the 2nd stage device input and raise the RF output by several Watts. If I then increase drive I end up at 30W and the cap makes no difference. The final is just not going to put out more.
If I use my HP 8640B as the source, it is 1100Mhz max. As I tune down to 960MHz I can easily hit 50W. That is what this chip was designed for. It is an internally matched device so I doubt I can get much more out of this at 1296. Adding caps around the output stripline edges where there are empty cap landing pads, makes no change. Later when I went back to test 960 and 903, I could only get 30W. Strange.
There is a bias adjust pot VR3, It ranged from 0 to 1.9V at the pot wiper. It passes through a 1K resistor to the Gate stripline. The spec sheet calls for typical gate threshold at 2.15V. Adjusting the pot made no difference. At idle I see no current on my bench supply 3A scale meter. I only see 0 to 0.16V at the gate. Shorting the resistor did not change anything. The LDMOS must be drawing enough current to drop the bias voltage to a few tenths. Does not seem right. The resistor was good.
The Bias enable (Ven) line appears to run into a DC-DC converter. It will switch on at around 3V input and put out 5.2V to the bias pot and does not vary with input up to 5.5V that I tried. VR1 and VR 2 adjust the non-existent VCO and a NE555 that likely sweeps the VCO for jamming. This is the same as observed in a similar family 5.8GHz PA.
This will work for 903 or as a 1296 beacon amp at 25W. For now, it is going onto a shelf. My SG-Labs 25W is about the same power but runs on 12V and has a LNA and I can put it out at the antenna so I am not going to use this 1280/50W amp in my 705 Xvtr box as I hoped. I do need a 50mW to 1.6W boost amp to drive the SG-Labs amp. That is next.
- 1200-1600Mhz @ up to 8W at 18V with 5dBm drive – main amplifier part is 1500-2200Mhz MHVIC2215N 2-stage LDMOS rated for 28V and 15W CW, not sure about the driver stage part voltage yet. At 15V and around 0dBm drive I get 3W. I only need 1.6W. Very small in milled box. Looks good on the spectrum analyzer. No VCO or traces for one.
- 900MHz @ 50W at 28V - Got 50W easily at < 3A . Very compact size. No VCO or traced for one. Also very compact. Mentioned in previous email.
- 200-300 @ 50W. 28V and features an Internal isolator on the output. – Awaiting delivery. Equally compact.
- 2W wideband 100-1200Mhz claimed. 2-stage, the final is rated for 2W up to 3GHz. 10-30VDC power. I never got more than 1.1W. Could try raising the 7V supply to the final to near the rated max 9V. I tried bypassing the 1st stage with ~20dBm drive and that was the best at 31dBm. I need 32-33dBm. A write up in Germany got 1.4W out at 1296. I needed 1.6W to drive my 25W SG-Labs PA/LNA. I then discovered the 50W 1280M unit and got the 12-18V 5-8W unit as a backup plan. $11. Now in the spare parts bin.
I will install the 5W PA into my 705 Xvtr box and drive the external SG-Labs 25W. 4W less but has the bonus of a built in LNA and I can put it nearer the antenna getting rid of cable losses making up much of the measly 4W deficit. I am happy with the 50W 900 amp so far. Will see what the 200Mhz amp looks like.
- Mike K7MDL