What TX amplifier do you plan to use?

  • Firstly you really must put it on a heatsink or plate so the grounding is correct, properly correct for microwaves. You can't rely on a jumper, I am not sure I can emphasise this enough but people still just don't believe it. These devices can easily oscillate and self destruct. So many people destroy devices in what they think is a safe test environment and perhaps would be for HF devices. It's not safe to not properly ground the sources and PCB. I'll say that again. It's not safe to not properly ground the sources and PCB. And again It's not safe to not properly ground the sources and PCB. I was really unhappy so see these amplifiers delivered like they are with no source connection, not say at least wrapped in aluminium foil, but that's where we are. I am certainly not going to test mine without first doing the work to make the heat spreader.

    This tendency to oscillate is the normal reason for failure apart from over-drive. I suggest you do the proper grounding and then somehow try to increase bias - the fact it is now 0.5V suggests to me there is a diode to ground but it might just be set there. Perhaps that's the minimum and it's adjustable but if it resolutely stays there until you disconnect the device and acts normally without the device then the device is dead. I don't think it is because it managed to take current at one point.


  • Thanks G0MJW

    I will stop testing this amp until I make proper heatsink.

    Out of curiosity: is it also important to have the proper ground/heatsink for a quick passive power-on test? Will it still do this self-oscillation? My initial assumption was that it's safe just to turn on the amp without driving it.

  • Hello Florian,

    After observing the same error as @ DF2HF (# 279) and @ DK4FT, as a beginner I now dare to make the necessary modifications to get the bisonelectronics PA to work as described by you.

    Your posts # 136, # 137 (doherty mod with picture), # 146, # 161, # 163, # 232 and especially # 286 are extremely helpful.

    Now to my question about the Changed IDq to manual setting on both LDMOS:

    do you have the connection on the BLD6G22L-150BN2, as in # 148 BLD-closeup.jpg, lifted up and connected to the Zener diode etc., or are you using another access for the IDq voltage?

    How are you connecting the IDq for the preamplifier (BLC6G22L)?

    If the questions are too nonsensical, please excuse me.

    73 Walter dk1za :/

  • I received a different Amp with the "same" problem.

    The problem was quick found . It has nothing do to with the doherty or what ever. There are small 10Ohm R between the 28V and the Bias of each Transistor and they were "burned" . They are not complete dead, they are 50kOhm and so the voltage is wide too low.

    I soldered in new 10Ohms (5 on this one here) and the complete amp is working excelent. Nothing on the doherty was changed!

    Measure them on yours and i hope it´s the same on your side. You can measure them inside the AMP, no need to desolder them . 10Ohm or 50kOhm is a wide range and your Multimeter can this :)

    Have a nice weekend

  • Thanks Sandor,

    I will check tomorrow and will report. I have an idea what Rs you mean. To be sure it would be helpful if you could make circles around the Rs in the picture of your PA.

    Have a nice weekend



  • PA3GUO

    Looks very good. :)

    By the way, the directional coupler at the output can be omitted and thus may bring a few tenths of dB more output power to the antenna. The amplifier already has a coupler built in and you can simply connect the power meter to the second MMBX socket. The corresponding calibration value has to be determined of course, but this would be necessary for the external coupler as well.

    But one could also realize a deluxe version and use the external coupler to measure additionally the reflected power.

  • I first have to make a picture from the repaired one.

    The "old" one is no longer in use while the PCB traces burned after more than 35W on DATV....

  • Hallo Mike,
    I think, there is no need for copper. The idle current is around 800 mA. With rf the current reaches approx. 5 A. If one interrupt the supply to the idle current circuit, the standby mode is easy. This can be done disconnecting the copper on the lower left side below the huge electrolyt capacity C136. Then the idle current is off complete. The PTT-switch on can be done with electronic, there are a few Milliamps flowing only. I like "high side switches" for example BTS4140 from Infineon. They have a control input pin. Pulling it to GND, the input supply is given to the output pin.

    With the normal doherty operating no dangerous heating can be seen. The transistors are not at the same mechanical hight as the print. It's recommend to resolder before mounting. With a good thermal connection like "Artic Silver" or similar, save operation with an aluminium cooler is possible. With driving power around one to two watts, the output is more than enough for NB-operation with very good linearity.

  • I also have two of these boards with a combined shielding cover. One amp is behaving normal, the other one is starting to oscillate after applying the small input signal. Without signal it is stable. Amplifiers are on 50 Ohm dummy loads. Allready tried to reduce any feedback. Any hint what I should do to get this amp stable?

  • I am testing this pallet driver amplifer: Andrew with MW7IC2240 chip. I have applied supply voltage 28V and both gate voltages (2,9V nominal), but did not get any quiescent current. Even an increase of the gate voltages to 4,6V did not show ANY drain current. Checked all the voltages on the chip itself. Any recomendations? I think the chip is dead.

  • @UA1CUU

    Hello, I tested a second amplifier and the quiescent current was adjustable nicely. I tuned the amplifier to 2,4 GHz using a Z-shaped brass foil soldered to ground. Signal in DATV was 2,5W out with shoulders -30 dB. So it is a nice driver amplifier for DATV or a quite linear amplifier for NB modes. The efficiency of this amplifier is quite bad, for 2,5W out the input was about 1A at 28V.