Question on using a MHL-21336 small chip amp.

  • Just checked the datasheet (I use such a module as driver) and indeed .. no reference is given concerning DC coupling, that is . . . sloppy.


    Anyway, in my design (an old UMTS/3G pallet) there are caps (take 56 - 100p or so)

  • G8UGD

    Changed the title of the thread from “Question on using a MH21336 small chip amp.” to “Question on using a MHL-21336 small chip amp.”.
  • G8AGN I saw your response this morning so I set about putting something together as yours was the second one indicated with direct connection. I eventually etched a board, not done that in a long time, I did not follow any rules as I have no software for this just made something up. This is the board mounted on a heat sink.



    I have not put +5 dBm into the unit yet as the Hackrf I have limits at around -2 dBm but it looks like it has 26 or 27 dB gain. So for a first attempt at this in a very long time I am happy.

    Thanks all for your assistance.

  • Your pcb looks fairly similar to mine except that I have a few more grounding screws.

    I included a 6dB attenuator at the input as an insurance policy and am driving the module from a Lime Mini via a SPF5189 board from eBay. The Lime is normally set at 80 or 85% and this seems OK.

    I see you are powering the MHL21336 from a dc dc module; I do likewise and this works well. With about 18v I can get at least 500mw out and 26V will easily give more than 1w. I haven't pushed the MHL as I'm going to use it to drive a SG Labs 20w PA and this only needs about 0.5w for full output.

    Re the discussion about input and output caps. PA3FYM is correct in using caps as the spec sheet doesn't mention whether some are in the MHL module. I haven't used them in my application because the modules on either side of the MHL already contain them.

    The heat sink I'm using at present is just a piece of Ali plate 65 x 100 x 10 mm. This is definitely not enough when the MHL is running at 26V but can cope at 18V.


    If you want more info, pictures, spectrum plots, contact me via pm (see QRZ.com)


    73 Barry

  • Cutting small groves in the track and fitting caps or even a resistor would still be possible. I measured the input current at just over 15 Volts in to the step up unit when transmitting it was 0.96 Amps so the unit is dissipating 15 Watts. I have a small fan I will connect to the fins if I can figure a mounting method.


    I have no larger amps at present, so just hope to get on SSB initially, need to test the POTY next.

  • Hello Matthias, Thanks for the information. I am using my MHL21336 to drive the Bisonelectronics amp. It seems to work OK for now until I feel the need to get something better. Very similar results on output to yours. I do not have a VNA the gear to test mine fully, but it seems to be OK.


    Adrian

  • 292/5000

    Note:
    I ordered modules from three different Ebay sellers and received two modules with the correct housing but the wrong chip. The wrong chip delivers about 500mW at 1.8GHz, but does not work at 2.4GHz. Has 5V bias on the left and UB 28V on the right !

    The right module on the left, fake on the right.


    Last correct modules are from here:

    https://www.ebay.de/itm/1PCS-M…ksid=p2057872.m2749.l2649



  • Thanks DC2TH! Unfortunately i have the fake one from aliexpress......

    I was already looking in my junkbox for an enclosure and some sma connectors but now i have to find another solution.

    I have ordered a lime sdr mini to replace my tx: icom 7300/ ukrain 144MHz transverter/ dx-patrol up-converter.


    My plan is to use the lime-sdr mini inside the shack and with 25 meters RG214 the signal goes to a watertight box near the dish with a chinese driver and homebuilt amplifier kit from DK2DB.


    I will have 12 dB loss in the coax and i need 20 dBm at the input of the watertight box.

    The lime sdr mini will maybe deliver 5 dBm, so i need a total of approx 27 dB gain.

    I was thinking about 2x the Opa design F10PA amplifier. One near the Lime and one inside the watertight box but that is an expensive experiment as one F10PA wil cost 49 euro.


    Any suggestions for a cheaper solution PA without the MHL21336 or 2x the F10PA ?


    73's Gert PE1RTC

  • 292/5000

    Note:
    I ordered modules from three different Ebay sellers and received two modules with the correct housing but the wrong chip. The wrong chip delivers about 500mW at 1.8GHz, but does not work at 2.4GHz. Has 5V bias on the left and UB 28V on the right !

    Can I just ask if you are saying you have 5Volts on the left Vdd (pin 2) and not 28 Volts? Did you get a datasheet provided with it? On my MHL-21336 both Vdd pins are at 28 Volts.


    Adrian

  • Hi Adrian,
    I also got some fake chips. The correct MHL-21336 is powered at both Vdd pins with 28V.
    At the fake chips Vdd1 is the gate bias with lower voltage, Vdd2 is the drain bias with 28V.
    The drain current will rise steeply when exceeding the gate bias voltage of 7V.
    The fake chips most likely contain modules of the type MHPA18010N, MHPA19010N or MHPA21010.

    They are 10W modules for cellular networks in the range 1800, 1900, 2100 MHz.

    I have not tried retuning them as I think it is not worth given the low price of the original MHL-21336.

    I can confirm that the modules from the vendor given by DC2TH are the real ones.
    Other vendors in China are selling the fake units.

    I will update my description shortly.

    DC2TH, is it ok to use your posted pcitures in my description ?

    Kind regards

    Matthias

    http://www.dd1us.de