Posts by 4Z4LS

    in any case issue with ISC5XX series i was test in 2014. for simple explane final PHASE NOISE value will lower than -40dBc/Hz at 10 kHz offset. from datasheet page of IDT_511 device: One Sigma Clock Period Jitter tjs 25 ps. (This parameter of ICS512 is 70 ps!!!) see datasheet pages of devices calculator link

    LTC6950 is better solution, is 1.4GHz Low Phase Noise,Low Jitter PLL with Clock Distribution device

    Hello Geri! is not exact brave or not, but BU500 is only mixer with separated stages. i'm familiar with this product, and read page 4. example of 50 MHz is given. also, you can trylowest frequency, but conversion rate and gain will drop. 100 MHz is promissed condition. some amateur was try to use LNB in direct conversion, also same result up to dead point of receiving.

    That depends on the amplifier response but I doubt it will be clean. If it is that's clearly good but a mixer that has no image or carrier leakthrough is a rare thing indeed. Also its advisable to check all spurious, not just ones less than 180MHz away. You might get away with a saw filter to clean things upI

    I think is other way to get clean signal w/o any filtering stages. i can use in gate-filter, but is not necessary in my case.

    i use AppCAD from HP for a long period, from 1997, is never makes mistakes.

    Righ now i'm build one like this,but total transverter for other HF range. it's for RX and TX simultaneously. can't send circuits and PCB gerber's now, in the future i will do it.

    please see attachments. BOTH STAGES are UP-CONVERTER. now let's do some calculation: 740 x 3 = 2220; 2400 - 2220 = 180. In any case mixer reproduce products in both directions. Any questions?

    about spurs - no closer one than 180 MHZ . see what's amateurs use of for feed and amplifiers stages. feed is not so simple to adjust. amplification power stages have very narrow band with. efficiency of +/-150 MHz not so high.

    Any ideas for an band pass filter, that is usable, if LO=2372MHz and IF=28MHz? Should be steep.

    Regards, Mike

    why did you need any filter, if frequencies map is ok? there is many other ways to get 2400,xx without 2372+28.

    Hi Artur. All written on the cover not more than COMMERCIAL GIMMICK. In the article you are present is about all and about nothing. In any case, you can use this device, but remove DRO disk and inject external LO frequency 3,25 GHz, where LNB working on 3rd harmonics of external LO. Is not new, first time making in 1992. Todays radio amateur was testing this issue, it's correct.

    Hi Artur!

    Full duplex in sat range is not exact,name of, but it's mean TX and RX modes run via same path - AERIAL of OWNER’s TRX. May be is not exact name , but is it. RX signal from LNB is really strong,so 20dB pad is not a problem. Few radioamateurs add cable to rerduce signal level up to 40 dB! If you compare,many constructions include attenuator on RX path to reduce a lot of power from LNB. In that case sensetivity still at desired level. Supression of conversions or unwanted products. Because in every point or in every mixer is single frequency signals,try to calculate where you can see unwanted products.

    My oldmen teacher (RIP few years ago), was teach me 35 years .ago and i really remember many lessons and themes, but one geniuos phrase i'm remember perfect: "every RF circuit will working ideal on the paper even mistakes is in".

    I will explane you all part and part from this design, and because this project is totaly my from first block diags up to synth calc include phase noise,loop filter, program and other blocks. Also, you can immaginary about fantastic parts, but if you have not compact, low power consuption and RF level VCO, your project will on paper only.

    Off course i can make any manupulations with frequency sources, multiply, divide, add or substruct, but first all i need choosing parts with cost-efficiency line. In thats case must to see what's market have too.

    If i talk about design for HF transverter, in any case it must to give to owner working (by request) in one of HF bands from 3,5 to 28 MHz.

    I will attache screenshots of Appcad 's run spurs of mixer. I open scale of 5-th product of every mixer combinations.

    Now about choosed parts - my opinion is VCO from Maxim series MAX262x & VCO from Alps E-D07A.

    Now show me please VCO for 720 MHz mid range with same parameters. Believe me, I testing market before making first step.

    You are change my circuits, but have not parts. I receive 711,5 MHz by product of 1660,5-949 MHz. In the next mixing stage I’m add 1660,5 to 711,5 MHz, there additional TRX freq of 28 MHz will give 2400 MHz. To change final product to desired HF band, I reprogram & lock PLL 949 MHz to 935 MHz e.t.c. For working on 20 m band, mixing must to give 2386 MHz. Also other ranges of HF.

    It’s good or not I don’t know. Now I wait for mixers, so important parts is tested – PLL lock and software test.

    I hope that you have all answers for your questions.

    Hi Armin! thanks for your responce! This proto really have not filters, for only one reason - power level of VCO very low, so frequencies of products is sufficiently spread out. Also, mixer's upper range close to TX & VCO freq. I need to make deeper testing for units, now only lock PLL's and adapt loop filter values. I know divider's out level is much high than VCO and harmonics. Will be continue.

    Short preface.

    All new is well forgotten old.

    Over 25 years ago I was working on project of synthesizers in local company, where one day boss ask me if is possible to pick up one day project for single frequency synthesizer with use of VCO from mobiles. In those time Nokia was use Murata MQE series VCO models. I try to use of LMX1511 PLL from National Semiconductors, really simple schematics for experiments and not only.

    The memories of this issue return me to development for multipurpose up/down converter. My version is not for one day project, but reduced one is it. PCB routed for all possibilities RX or TX path, where TX path can up-converting from any radio amateur band’s owners transceiver.

    Basic idea is assembly some casing of standard form and replace only few parts of loop filter and choose one of two types of VCO. Board can working in only one mode – RX or TX at same time.

    Synthesizer is based on LMX1511 PLL with 3rd order loop filter. Calculation was made on XL program from G4JNT (many thanks for a wise investment), compare to datasheet of PLL no mistakes found and time saving. Because this project is single frequency lock purpose, not so strong emphasis on loop filter. Usually is most important part of synthesizer and many important parameters like phase noise and settling time is depend on correct calculation.

    VCO – Maxim company hase producing a line of VCO’s; series MAX2750~2752. Low RF output (and consumption) power devices help to reduce some battle with spurs. Loaded RF level almost -3dBm, and for any purpose in my synth is enough. 2 of 3 devices is good for converting purpose. Working level windows of PLL is run from -10 to +6dBm.

    MAX2750 – freq. range 2250~2600MHz.

    MAX2751 – freq. range 1920~2350MHz.

    Experienced user can say is something wrong here – PLL maximum input from VCO is 1,1GHz, and you are set VCO with twice in frequency and little more. I use of MC12095 frequency divider with two switches – div by 2 and div by 4. Is old, but very stable device.

    Now let’s do some calculations for every range.

    UHF range for converter, there I set TRX on 435,05MHz. For conversion to 2400,05MHz I need VCO freq on 1965MHz. VCO is MAX2751. Division by 2 is meets the requirements of the PLL’s work window – 982,5MHz. This what PLL will see. Loop filter BW=25kHz and phase detector freq is 250kHz. This is big step, but for this case is good because the synth not in jog mode.

    982,5MHz ; 250kHz = 3930. This is N divider’s value. Internal prescaller chosen for 64/65.

    Now I set value of B & A counters. B= 3930 : 64 = 61; Here must be an integer values, so 61 * 64= 3904. A= 3930 – 3904 = 26. A-counter value must be equal or less than B-counter value.

    N=3930; P=64; Fpd=250kHz; B=61; A=26; R=40.

    Reference counter value is depend on frequency REF oscillator. In any case is 2 modes – oscillator or XTAL. If I choose XTAL=10MHz , R-counter = 10MHz : 250kHz= 40. Also, must get integer value of division.

    VHF range for converter, there I set TRX on 144,05MHz. For conversion to 2400,05MHz I need VCO freq on 2256MHz. VCO like in UHF is MAX2751. Division by 2 not suitable here, where I can use div by 4 and final value will 564MHz for PLL’s input on VHF range.

    N=2256; P=64; Fpd=250kHz; B=35; A=16; R=40.

    HF range for converter. Here is must use other VCO - MAX2750, because coverage of 3,5 to 28MHz is close to 2400MHz conversion product. For conversion to 2400,05MHz I need VCO freq on 28,05MHz. Can be chosen other HF range too. Here I’m use in div by 4 also. fVCO = 593MHz.

    N=2372; P=64; Fpd=250kHz; B=37; A=4; R=40.

    If TRX is tuned on freq 14,05MHz, Fvco after div by 4=596,5MHz; N=2386; P=64; Fpd=250kHz; B=37; A=18; R=40.

    All parameters is saved into PIC12F509 compiled and burned HEX file. Why PIC ? – Because to make damage with wrong code or diff voltages very hard, not like Atmel group of micro’s, where wrong code or fuse drop micro and must be restored to factory level of use.

    Frequency converter. Based on MMIC ERA-3 or same amplifier and RF mixer ADE-25MH.

    Required adapt RF level on LO port of mixer, where on PCB is ready to use place for Pi attenuator’s resistors.

    Supply voltages. On board is 3 voltages converters for 5V, 6V (MMIC supply) and one more for use of Si-590 additional pre-programmed source for HF range in RX path.

    Primary was tested one more VCO from Maxim MAX2609, but I have not positive feedback from this device. Better device for RX purpoces is MAX2624, but PCB not routed for. Is possible try to set MAX2624 on place of MAX275X on board. V-tune voltage must up over datasheet values (~4,5V).

    Loop filter calculated with Kv parameter after division. For example, if VCO parameter Kv = 120MHz/V, and VCO’s frequency at working point is divided by 4, Kv also divided by 4 and equal 30.

    PCB actual size is 56X42mm.

    Attached schematics diagrams can find more qualified in PDF format.

    Hi, 4Z4LS

    Also I wanter how you can archive suppression of Lo and image. A less complicated Lo could be a ADF4350. Do you have capability of external frequency reference from GPSDO? You have an amplifier before the LPF at 2.4 and you amplify also the 739 signal.

    George, please see head of theme after additional explane for all questions. Later i will set schematics diagramm, now is in "crude" form, even though the PCB is ready.

    Actualy i'm not amplify 739 MHz signal from LNB, also i have not switch over 20 dB attenuator to the TRX, because RX signal so strong and must be attenuated. TX out is amplified to get some level after mixing and filtering stage.

    Actual schematics have possibility for external OCXO source (it written before) 8~20 MHz range. You can see it clear in the picture of outside PCB "BOARD7". Now in use separated 10 MHz XTAL's for each REF of source - only for my experiments and testing.

    ADF435x - not shure is less complicated, that my point with PLL's. First all you have not program in simple form to make few steps : calculate new value of counters in PLL, change parameters of source, compile & burn micro with HEX file. I spend a lot of time for program form to set 4 parameters and compile it. this was take a time. Now progam is tested and actualy i can can reproduce any single carrier from each PLL.

    Hi people.

    My new construction and design "full duplex" transverter to operate via HF TRX.

    Based on frequencies mixing. 2 separated VCO units, each one controlled by single PLL & reproduce single frequency. Basic control program for PLL2 may changed by re-calculation of counters value, there may working on any amateur HF-band from 3,5 to 28 MHZ.

    The design allows the use of SDR dongle and an external reference OSC in range of 8 ~ 20 MHz. design does not include driver stage for transmitter path. Every chnge required re-programming PLL controllers.

    In attachments seams partial construction. At last tested PLL's lock after debuging firmware for burning controllers. Repeated experience with PLL's lock gave a positive and identical result.

    All stages of design maded by my self , from idea, include program writing and burning controllers , loop filter calculation, pcb routing, parts ordering and soldering.

    Total box size 100x56x24mm.

    Now partial explane about block diagram.

    Main synth lock VCO at 1660,5 MHz, second PLL do it at 949 MHz. PLL1 serve downconvertion operation for product of 711,5 MHz. By the way this frequency will change final downconvert to any HF band. Mixer connected in those manner not given so mutch on 2609 MHz, also LPF 720 MHz disable it, especially IF max range is 1500 MHz.

    Next converter is 1660,5 MHz mixed with 711,5 MHz. This point serve RX or TX mode. If mixer M2 is "shorted", no product of up conversion can run trouth this path - there M1 receive signal from LNB and mixed product going to the owner TRX.

    In TX operation mixer M2 is up converter (IF port at input) with 1660,5 MHz and product of M2 - 711,5 MHz. Last converter based M1 mix transmitter carrier from TRX and product of M2. Here i'm set power attenuator from TRX to prevent damage of mixer M1. Maximum saturated power to RF port is 0 dBm.

    Input-output path's signals distribution diveded by diplexer - LPF for LNB in and HPF for PA's driver.

    Abouth harmonics or products from mixed signals.

    949 x 2 = 1899.

    1899 + 711,5 = 2609,5.

    949 x 3 = 2847.

    1660,5 x 2 = 3321.

    1660,5 + 711,5 = 2372.

    711,5 x 2 = 1423.

    1423 + 949 = 2372.

    711,5 x 3 = 2134,5.

    There is no product of combinations above is close to 2400 MHz.

    About x-harmonics of 28 MHz not discuse.

    Even i'm not shutt down PA and driver in RX mode, nothing to amplify bacause mixer M2 is "shorted". In any case, port's short technics used in frequency doublers based mixer.

    Final stage will have BPF for 2400 MHz, but i not support any PA, also not support BPF/LPF for the range.

    Both VCO's choosed with low RF output power, to prevent high level radiated power into system. For PLL' f-in port low power arround -3 ~0 dBm found in working windows range. Actualy real RF power of VCO is lower than datasheet value.

    Hello people!

    I have low noise FET's. Design not nessecary, because today in the net found a lot of applications. Only 2 questions :

    If somebody here build dual stage FRONT-END amplifier?

    if it possible to use FRONT-END stages from standart LNB, to connect RF OUT after 2-nd stage, were internal controller on default open bias voltages?

    i have GERBERS for FRONT-END, but FR-4 it's will not working like i expect. In China can order few PCB for very low price in perfect quality, but ROGERS's cost very high, and not every manufacturer can do it.

    In any case in my stock FET's ATF-36077, NE3210S01-T1B .

    Few pictures of hands-on standart LNB to use it as FRONT-END only, BUT can't examine for gain level after upgrade, because AKD-12000 mixer not tested yet.