YO5PBG

For just about 10 USD including postal fees you can have the extraordinary ultra-wide band almighty QRP power amplifier: the only one you need for all the bands, HF to UHF.

Ordered two pieces via xab3.ro. Came in about 2 weeks. Looks very nice, small, but the soldering is kind of bad. First module tested died in the first day. The module was heating very rapidly if applied voltage was higher than about 9V, and did not applied more than 11V even if 15V is recommended for obtaining 3W output power. I inserted less than 10mW at the input- as specified, 5 to 145MHz.

I decided to investigate a bit the problem, and only some days after to test the second module. I first suspected that the first RF IC was dead (at the input). Later I discovered that the LDMOS output transistor (U2) has around 50ohms between Gate and Source when measured with a ohmmeter. So this transistor was lost. I removed it from the PCB, than measured the bias voltage. It was 3.55V. I think that is way too much (see why bellow).

In the following figure you can see the schematic of the module (i hope i did it well). Extremely to no information is present about this module on the net. I found that very probably U1 (marked on the first module as .BA5 and .BA5Z on the second) could be a SBA5089Z RF IC power driver with 13dbm output and 20db gain from DC to GHz, and U2 – marked K2 could be a RD01MUS1 (used also in FT991 PA) – a 1W/520Mhz/7.2V LDMOS. If true, U1 is quite well biased. In the first module, marking .BA5 it had 91mV drop on R2, so around 60mA. Datasheet specifies optimum of 80mA. In the second module we have a drop of 110mV on R2 which is supplying U1 (here is marked .BA5Z), that means ~ 73mA. Now moving to the U2, LDMOS, datasheet, application notes, FT991 PA schematic refer to an Idq of 35 to 100mA, as appropriate. That means a Vgs voltage of around 2.3~3V. The 3.55V is maybe a killer Vgs for this chinese implementation, moving the Idq to around >0.5A. So i’ve changed the bias resistors, as described in the following figure. This way i obtained an Idq of cca. 40mA. Also inserted a -1.3dB ATT pad at the input for better stability?! Now the module heats, but not more than abt. 30C.

At 11V supply, input of max. 1mW i can get on a 50 ohm load about 10V @ 5MHz (1W), 8 to 9V (700-800mW) till 55Mhz, abt. 9V at 145Mhz, and only 4V (200mW)  at 432Mhz.

Fig. 1 – NWDZ RF PA v2.0 schematic and the first module tested

Update1:  Measured C12 – 6pF; C11 – 270nF

Update 2: Just found some links about this or similar amps: http://m0taz.co.uk/2016/02/0dbm-1mw-output-wideband-amplifier/

Update 3: 432Mhz power increase

C11 was replaced by a ~ 12nH coil. Output voltage increased from 4V to 7-7.6V / 50 ohm, so more than 0.5W (@P.sup. = 12V). Input power less than a mW! Pictures below show the modified PA.

Tags: FT991, LDMOS, NWDZ, power amplifier, QRP, RD01MUS1, RF PA 2.0, RFIC, SBA5089Z, wideband