Results 1 to 4 of 4

Thread: New User, New TRX project.

  1. #1

    Default New User, New TRX project.

    Been working on this one a while-haven't yet picked up the soldering iron because I don't really know what I'm doing with these multi-legged fuses. I'm used to FETs that survive nuclear EMP blasts just fine, even if they are a little fragile in the "rapid vertical deceleration test".

    Microcontroller board is designed and layed out, 1st mixer board is designed and layed out, and the audio stuff is drawn up on paper but I haven't really done the layout for it yet.

    I got bored with the other sections, so now I'm working on the RF amplifier section.

    Design-Specifications:

    30W output, 40-20-10-6 Meters

    To use parts on hand.

    Parts bin contains:

    9 Motorola M9587 RF power transistors (equal to 2N6081)
    10 MRF630 RF driver
    NE592 Differential Video Amplifier
    Fair-Rite material 61 mini binocular cores and some bigger (little less than an inch OD) toroids
    All kinds of through-hole polystyrene caps, full complement of SMD 0805 resistors/capacitors, 1/4 watt metal film, 1/2 watt metal film resistors, Not much in the carbon-comp resistor bin though.
    Plenty of magnet wire in many sizes (I wind my own power and modulation transformers for tube gear)

    Also, I have copies of Motorola AN593, as well as Solid State Design for the Radio Amateur, and the 1978 ARRL Handbook (most of my literature predates these texts...)

    What is catching me up is the biasing requirements. As it is now, I understand that I need to bias the output transistors to roughly 10-15mA to achieve "an operating point" however I am used to tube data where the operating points are easily picked from characteristic curves and suggested operating data. With the transistor data I have...those pages aren't there, just "do not exceed" and sometimes data at specific points.

    For example, the M9587 transistor data states design Vcc is 12.5V, maximum do-not-exceed Vc-e is 36 volts, power dissipation is 31W, and output power (I assume class C) is approximately 15W. Suggested operating data is for 175Mhz, 15W output, class C-operation at 12.5V supply. (and thats all the operating data has to say about it!) I have read that it is very similar to 2N6081, so I looked up the data on that one and it's very, very close-but there is a little extra, like I would need 2.5W drive to obtain that 15W output, and that it has a power gain of 6.3dB at 175Mhz.

    The bias circuit must be low impedance, I understand-that's fairly easy and by using a two-transistor circuit I can temperature compensate the output device bias decently. However, what base voltage is appropriate? Is it just always 0.7V or so?

    The current plan is to place two of the M9587's in push-pull, and drive those with a push-pull pair of the MRF630s, which are in turn driven by an NE592 differential amplifier. The NE592 is being used as it's already frequency compensated for gain, and I can use an output from the microcontroller board to adjust gain on the fly. The idea being to set stage gain on the MRF630's to be correct for driving the M9587's to full 30W output when the NE592 is operating at "microcontroller max" setting, and if the micro detects an SWR mismatch or if I select a lower power output, then the micro will reduce the gain of the NE592, in turn, underdriving the amplifier and reducing power output.

    Currently, the power supply section supplies +12V and +/-6V, +5V, and the MCU board generates 3.3V to drive the PLL local oscillators.

    If there are a few here who are willing to work with me on the design, I'll post up more details and my preliminary drawings. I am currently away from home for work so will be a few days until I can get preliminary drawings posted.

    I'm good on the mixers, filters, and control board-it's just understanding (instead of blindly following) the application of bipolar transistor amplifiers that I need assistance with.

  2. #2

    Join Date
    Dec 2008
    Location
    Elgin, Illinois
    Posts
    1,619

    Default

    Hello Jacob,

    The M9587 RF power transistors (equal to 2N6081) were designed for 132 through 175 MHz so they will never operate on the HF Bands of 7 through 54 MHz, I don't care how you try to make them work. BTW according to Motorola Specifications that device is rated at 11 Watts Output. The MRF 630 is designed for the UHF Band 400 through 470 MHz so this one wouldn't be good for the HF Bands either.

    Here's a great website on homebrew equipment, I believe Iulian Rosu probably will clear up your questions about solid state RF:

    http://www.qsl.net/va3iul/

    http://www.qsl.net/va3iul/Homebrew_R...sign_Ideas.htm

    This is an excellent website for the homebrewer as far as RF Amplifiers http://www.communication-concepts.com/ Keep in mind that you're looking for RF devices that operate between 1.6 through 54 MHz and has the PEP & CW Rating. Watch the Voltage ratings as they can change to 28 VDC or 50 VDC if you're not careful.

    Motorola Semiconductors was spun off in 2004 which became Freescale and now that product line is known as NXP Semiconductors. So back you go to the drawing board to re-evaluate your RF design

    73,

    Dan
    WA9WVX

  3. #3

    Default

    Just because they're characterized at that range doesn't mean they don't work at lower frequencies...they aren't MMIC's. But, I'll admit it's possible they have internal LC matching to optimize for that 132-180Mhz range, however the 2N6081 does not. I have seen several published designs using the 2N608X at HF, but always single-ended. The 2N6084 will do 40W out, so it gets used single-ended often. (or used to, anyway)

    The big bugaboo I will have to worry about is parasitic oscillations, but if I wrap enough local feedback around the finals and drivers, I should be able to reduce gain enough to solve that.

    Anyway, here's my first go at a drawing. Cribbed from Solid State Design for the Radio Amateur and ARRL Handbook 1978. Resistors R5 and R6 aren't dimensioned properly yet, Nor is VR2, and the input of the opamp needs to be biased to +6V yet. There is no frequency compensation in the driver stage, nor local feedback in the driver stage. All that will come as I develop this. Also, 2N930 was just handy in the drawing software. The drivers are MRF630's and Q1 and Q2 are TIP3055. Q2 is bolted to the heatsink between Q5/Q6 for thermal compensation.

    T2 is a binocular core, just enough to drive the differential inputs on the opamp. TX1 is a pair of 61 material toroids, bonded together as a binocular core. TX2 is a quad of 61 material T82-sized toroids bonded up as a binocular core for the output, or maybe just a pair stacked. I need to look up the power handling for that size core. (fair-rite 61 material, just under 1" OD and 1/2" thick)

    I am thinking there should be some chokes in the bias supply line and maybe the +12v lines feeding TX1 and TX2, but not sure yet. More research due.
    Last edited by KI4YAN; Mon 29th May 2017 at 19:19.

  4. #4

    Default

    Interesting project! I'm sure you'll be able to find the perfect soldering iron to use. Looking forward to your final result!

Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •