-
Notifications
You must be signed in to change notification settings - Fork 40
Pre V2 Amplifier Errata
This document pertains to the V1.0 amplifier board only.
<<< Insert pics here >>>>
There are no detailed build instructions. In general, you will have to follow the schematic to build. Place all SMT parts prior to installing any tall parts (toroids, relays, connectors, etc). Perform the Tap Enable update and the Bias update prior to installing tall parts.
Follow the V1.0 schematic for everything except the filters and the bias change. For those you will follow the V1.1 schematic. Do not install U1 or R9 at this time. They will be installed during the bias change below.
At this point, we are not firm on the 17/15 filter values. We may do away with bank6 if we get a 17/15 filter that we think is easy to reproduce. You may need to order some T37-10 toroids, maybe 6-9 of them. They are pretty inexpensive. We may be using them on the 17/15 and 12/10 roofing filter.
Easy modification to P3 where you cut the trace from K1 pin 1 that goes to P3 pin 1. Then run a short jumper from P3 Pin 1 to the ground side of C14 that is right next to P3.
<<< Insert pic here >>>>
The bias on the RD16HFF1 (from regulator U1, LP2951) is too low at 5volts to allow biasing the FET to the recommended 600mA. We will change the LP2951 output 6 volts, instead of the fixed value of 5V.
- Start by cutting the tracks between pins 1 & 2 and 6 & 7 of the LP2951. If you already installed U1, you will have to remove it to cut the 1 & 2 track.
- Scrape the resist from the track which runs along the left side of the PCB from pin 4 of the LP2951 towards the mounting screw, just above pin 1 of the device.
- Solder a 10K and a 39k resistor in series from the scraped track to the positive pad of C10 or pin 1 of U1 [the 10k end goes to the ground track].
- Solder a thin wire from the 'join' to pin 7 of LP2951.
This then forms a feedback circuit which allows the LP2951 regulator to output 6.07V instead of 5V. (Picture, I used 0603 size but 0805 should fit also). 6volts is then plenty to bias the FET to 600mA or even a bit more. (depending on spread of the FET's.)
As a result the bias on the PZT2222A is now a little high. Change R9 from 1k to 1k2.
<<< Insert pic here >>>>
There are 3 options for T1. The first option is described in the wiki for the superband build.
<<< Insert pics here >>>>
The second option uses 2 binocs stacked with the coax that was tested by Stew and noted more consistent output power.
<<< Insert pics here >>>>
Jacinto also had a pic of the implementation recently.
<<< Insert pics here >>>>
Third is one using turns of enameled wire proposed by Claudio. I am going to try that one.
<<< Insert pics here >>>>
On my v2 frontend board, I did the following -
- I plan to not use the T/R relays U1 and U5. I wired it straight from T1 to L1. T/R is handled by my amp.
- Wire the TX reconstruction filter as 100p,330uh,200p,390uH,200p,330uH,100p using AIML-1206 inductors.
- I plan to eventually populate U2,U3 and U4 but for now I have bypassed them, feeding T2 to P3. RX filter not populated.
- P2 will be used for a power tap from the V2 amp board, but now left un-populated.
- I left P5 and P6 un-populated.
- I have a header for CN3 but do not plan to connect it until Steve has his logic implemented for it.
- I put a 2x6 header in CN2 and will use a 12 conductor ribbon cable to connect to the same style connector on the V2 amp board. The pin-outs are compatible.
- I used BN-43-2402 hand wound transformers for T1 and T2. T1 was wound according to the prior frontend. T2 was wound per Steve's note with T2 (8:1 with no centertap, 14 turns on PCI-E side, 5 turns on antenna side). Jim Ahlstrom has been using some commercial transformers. See his notes. I listed some that Steve suggested in the Frontend BOM.
- I used 2 pin molex headers for P1 and P3.
- I placed a leaded 100 ohm resistor across the input side of T1, per the suggestion of Jim Ahlstrom. This will limit the outout of the frontend to around +10dbm max.
<<< Insert pics here >>>>