Make any changes or corrections to the data fields. When finished, click on the UPDATE button.
Security (anti-spam) Question: How many dits in 72?
Sun, 17 Mar 2013 20:45:36 +0000
Larry Phipps wrote: > Interesting circuit, Diz. I ran a SPICE simulation of it. > Gain is 22 dB and is flat through about 20 MHz, rising to 23.5 dB at 30 MHz. > Max input level appears to be about 50 mV p-p according to the simulation. > The simulation also shows an input R of about 300 ohms, and a resonant peak at about 70 MHz. > gain at the peak is 38 dB, which could pose stability problems. > Do you see that in the actual circuit? > Larry N8LP Craig, AA0ZZ wrote: > 1) 100 mV p-p is about -16 dBm. Adding 20 dB means you must be getting an output signal of > about +3 dBm. Is this about what you were getting before you started getting compression (distortion)? > 2) What is the maximum output you can get when running your version at 12v? > The NE5206 spec sheet shows a maximum of +5 dBm output when running Vcc at 5v and it goes > up to 10 dBm at 8v. (9v is absolute maximum for NE5206AN.) I`m assuming you changed the values > in your version so you can go up to 12v, even though the transistors are getting warm by then. > -Craig, AA0ZZ Steve AA7U wrote: > I built the circuit today and tested it with my bench equipment at 5, 6, 8, 12 volts supply voltage. > Gain is fairly flat at about +22 dB to > about 15 MHz, then begins to rise, reaching a very broad peak at 56 MHz, about +4dB higher, > then rolls off above that. The amp > appears to be stable with no hint of oscillations. I didn`t try to measure in/out impedances. > The circuit was built dead-bug style on a > scrap of PCB groundplane, in about a 1`` square area. All transistors were checked with an M3 > semiconductor tester first. > Standard BNC connectors and coax cables to the test equipment were used. > It appears using 8 volts supply would allow sufficient drive for a standard +7 level DBM before > running too hard into amplifier compression. > 5.0V, 24mA. +16dBm OIP3, +26dBm OIP2, 1dB compression at about +1dBm output. > 6.0V, 31mA. +20dBm OIP3, +33dBm OIP2, 1dB compression at about +2dBm output. > 8.0V, 45mA. +20.5dBm OIP3, +38dBm OIP2, 1dB compression at about +8dBm output. > 12.0V, 72mA. +21.5dBm OIP3, +56dBm OIP2. 1dB compression at about +13dBm output. > Test equipment used: 2-HP8640B generators, -6dB homebrew high-isolation combiner; > HP3585A analyzer; TEK 495P analyzer (for>frequencies above 40MHz). > IMD test frequencies 3,4 MHz with -20dBm for each tone output at the > amplifier`s output to the analyzer. > IMD was observed at 5,2,7,1 MHz. > No resistors were changed from the original circuit for any supply voltages. > Steve AA7U Michael Harnage wrote: >From the datasheet of the NE5205AN: > Q2: eliminate shunt feedback loading on output > Q6: bias to allow direct coupling of R5 to base of Q1 > 73 de w1mt Archie Lamb wrote: > The transistors are used (on chip) as temperature compensation for the bias > points. They will help in discrete form but not as much unless you > ``closely couple`` them to their respective devices. > Archie (KX5X) Cecil KD5NWA wrote: > The hump on the upper part of the frequency range can be fixed by using > higher feedback at the higher frequencies. > Add a 30 pF capacitor across R4 on the LTSpice schematic or R5 on Diz > schematic and the frequency response is flat like a pancake to about > 70MHz then gently does down. Less danger of oscillation because of the > high gain at 60 MHz to 70 MHz. That is at a Vcc of 5V, the higher the > Vcc the higher the hump at the upper end. At 12V a 33pF capacitor is a > little flatter. Hi Larry, Craig, Steve and other respondents I used the NJ 9850 DDS module output through an attenuator to test the circuit. All test results are visual interpretations from my Scope. I pretty much agree with all your findings. I did have an oscillation (near 6 meters) but the 10 pF cap (C2) on Q4/Q5 fixed that. I also noticed a slight gain increase near 25 MHz. Not sure about the 70 MHz peak. Max test freq was 40 MHz. Thanks for all the feedback. 72 and thanks again for all the feedback, Diz, W8DIZ check out the SPICE MODEL here http://kitsandparts.com/potluck/wideband_amp.zip get the LTSPICE program from http://ltspice.linear.com/software/swcadiii.exe
Return to SEARCH
Please use unique filenames that include your callsign. Existing files will be overwritten
The Schematic file must be a GIF or JPG. The Auxfile must be a ZIP
Return to SEARCH