1296 Amplifier

1296 MHz AMPLIFIER KIT

These kits are available from Chris Shoaff at the address below. Mine works great!

Chris Shoaff

N9RIN cshoaff@yahoo.com

30w

1296 MHz PA kit $50 + $5 for US shipping

I'm building the amp kit shown below to replace my old Mitsubishi module that let out the smoke. My old amp for 1296 was based around a Mitsubishi RA18H1213G module. I apparently needed a bigger heat sink for it.

The kit shown below is based around the Infineon PTF 10021 LDMOS FET. Gain is rated at 13 db.

http://kahuna.sdsu.edu/~mechtron/1296Amp/1296Photo1a.jpg

The pic below is the box that rides on top of my van when I rove during VHF amateur radio contests. It contains the preamps, amps, and relays for the 903 MHz and 1296 MHz bands.

Mount completed, ready to build up PC board and test:

Here is the amp built up and in the middle of low level testing.

Here is a close up of the amp:

What you can't see is the large Aluminum heat sink under the 1/4 inch copper plate. The roof box also contains a muffin fan driving over the heat sink. I don't anticipate running it at full power for hours on end, so the cooling arrangement should be sufficient.

Here's the amp in the box, all wired up and ready for a full system test.

SWEET! 22 watts out at 1296 MHz! It can do more, but my driver amp is already near 1 db of compression and I'd rather keep everything running at a slightly lower power than to press my luck at full output power. (28 watts) The ICOM IC-706 radio is shown in transmit on 144 MHz. That's the IF frequency.

DUT: Device Under Test, powered up and cabled up

I posted the question of why my amplifier seemed to be low on output power. The low output power was found during the testing in the second picture from the top. Long story short, my calculated 52 db worth of attenuator chain and cheap ebay connectors were more like 56 db of real world loss. The amp was actually running just fine.

Here are the responses I received.

1) Either the device is bad, the RF coupling from the device output connector is faulty, or the attenuator to the power meter is higher than anticipated. - W3UHF

KM3G response: The DC parameters were in spec. (Idq and gate bias) That would seem to indicate that the device is working. The output connectors were checked visually. The attenuator stack and crude power meter I was using were replaced with a Bird 43 meter and the appropriate slug. The bad attenuator stack ultimately was the cause for the apparent low power output.

2) Sounds like you have something wrong in the output. I would double check C3 and C4 to make sure the value is right and the placement of C3 is correct. It is also easy to damage these caps as the plated ends will easily de-laminate making it look like they are soldered in properly when in fact it's only connected on one end. I suspect the FET is OK as you can drive it and get the proper drain current for full power. - N1DPM

KM3G response: C3 and C4 were checked for proper value. Their joints were reflowed to check to see if the plated ends were still intact.

3) The input output connectors connect to the board on the top ( long legs spread) but are they solidly connected at the bottom?

KM3G response: Yes, the SMA connectors are well connected on the bottom of the board.

4) The hole cut for the device is too large..there is a significant gap between the device and the edge of the board. I don't know if it came cutout or not...but this affects the matching considerably with the relatively low impedances involved.

KM3G response: I got a little carried away with the file while cutting out the hole in the board for the device. I wasn't aware the spacing around the device was that critical. (it's always fun to learn something when building kits)

5) My gut feel is that the output match of the device is not optimized. This is when a 3-stub tuner is handy to pull the load line and see what the device can do, then go back and adjust the match on the PCB. - WA1ZMS

KM3G response: I never got this far in troubleshooting before I found my attenuator stack was about 4 db more than calculated.

If you don't have a tuner, then you can try driving the device with about 500mW and try snow-flaking or ATC tuning stick caps to peak the output power. - WA1ZMS

KM3G response: I never got this far in troubleshooting before I found my attenuator stack was about 4 db more than calculated.

If you drive the part VERY lightly you can play around with the output tuning with less of a chance of device damage. That said, I would also use a current limited drain supply so you can help save the device if your hand slips while tuning the circuit. This device is an older LDMOS device designed by Ericsson before that division was bought by Infineon. I'm not sure the device is made any more. But I could be wrong. Even at 26v the old Ericsson datasheet shows you should get 31w out. - WA1ZMS

KM3G response: I never got this far in troubleshooting before I found my attenuator stack was about 4 db more than calculated.

The amplifier/preamplifier box is shown below in it's natural habitat. (small gray box on the top of the van) According to the weather reporter, this is a "light dusting" of snow in the Scranton, PA area. I had to cable everything up in the snow before the January VHF contest.