I had been listening to 75 meter AM and decided it would be nice to have a smaller radio to devote to it. It came within 3-4 days of ordering.
Assembly instructions were great except for a little confusion on Q2. I think they could reword it better to say “The labeling should face to the side of the board where the MOSFET heat sinks are located”. Inserting it backwards seems to be a common mistake.
I deviated from stripping wire and soldering the connecting wires from the display board to the main circuit board. A header was included with the display in its anti-static bag. I cut it to the correct length (number of pins) and installed it on the display board. I then installed another on the main circuit board, and used some 30 cm jumper wires from my arduino/raspberry pi “breadboard” wires. I also installed single header pins for the speaker connections on the circuit board and used a couple more of the jumpers to make the speaker easily moved out of the way if modifications are needed to be done (as it turned out later, some were required).
I always view winding toroids as the worst part of any kit. All 4 were single windings so only required counting the turns and evenly spacing them. After installation and testing I “glued” them down with fingernail polish.
Power on test did not result in any smoke being released. The display brightness was adjusted and then I calibrated the RX to track with the TX. The TX was 200 hz low but the RX was 20 khz high. The RX was quickly adjusted via the calibration setting. There doesn’t appear to be an adjustment for the TX error. 200 hz is nothing to be concerned about for AM use.
Transmit power on 3885 khz was 5 watts with PEP reading around 20-22 watts. At 3605 khz power output was 7 watts and 4 watts on 3995 khz. The RX did not seem very sensitive. On 3605 a 10 uv signal was barely readable. At 3885 khz the sensitivity was around 15 uv. I did not measure sensitivity above 3885 but it was obvious the RX sensitivity dropped off even worse as you went higher in frequency. WWV on 5 mhz was readable but not very strong.
Research into past posts of N75 at groups.io blamed poor sensitivity on either Q2 improperly mounted (turned backwards) or bad solder connections on the NJM2113 audio amp ic. Also there were complaints of the RX sensitivity above 4 mhz dropping due to rolloff of the bandpass filter, similar to what I was seeing.
The manual states the RX signal is mixed in the RF PA AMP mosfet (IRF510) that is pumped with a clock injection frequency of 6 mhz and the IF of 2 mhz is then fed through Q2 to the 2 mhz filters. Some work needs to be done to check the levels of the injection frequency and determine if an issue exists with either too much injection, not enough injection, issues with the bandpass filter or a need for another stage of amplification is required to resolve the RX sensitivity. Some say use a long antenna to make up for it, but what about the person that must use a compromised antenna?
In my opinion, the NJM2113 SMD device appears to be one of the worst choices in packaging for this application. The device used is an extremely small SMD packaged IC. The spec sheet states 260 mw output for the device used and its required to be mounted on the circuit board. Posts in the N75 group state the device requires mounting flush against the copper to help dissipate heat. They also recommend not running the volume control past half way using the internal speaker.
AUDIO IC REPLACEMENT
After some study, I decided to address the audio output issue first. The PDIP package would have been a better choice for the audio amp when using the NJM2113 series. It does not require mounting flush against the copper and is rated for more power output. The data sheet says the wiring is the same and the sample circuit suggests no changes between the different packages. An order was placed to Digikey as they were the only distributor with PDIP packaging in stock at the time of order. Mouser showed delivery in about 4 weeks.
https://www.njr.com/electronic_device/PDF/NJM2113_E.pdf
The SMD NJM2113 was removed. The PDIP version was mounted nearby “dead bug” style. Pin 7 (GND) was bent and soldered directly to the ground foil. To reduce the number of connections to the original SMD pads, I opted to use new 1 uf leaded capacitors to bypass pins 2 and 3 to ground, also soldering the ground leads directly to the ground foil. (5) Kynar 30 AWG jumpers were cut, stripped, and then soldered to the SMD pads and to their corresponding pins on the PDIP.
On power up there was no flashes of light or smoke released and the audio worked as it should. Most important, the PDIP did not exhibit excessive heating as the SMD part did at high volume.
RX Sensitivity
Noticing the power was higher at 3600 khz than at 39.95 khz, I decided to try shorting 1 turn on various coils. After experimentation, I ended up shorting 1 turn on L2 and L3. This improved RX sensitivity in the amateur band and also centered the power output up so max power is middle of the band and dropping off on band edges. Before mods I was getting 7 watts on low end and 4 watts on high. Now its a little over 5 watts on low end and high end and about 5.5 to 6 watts center. Sensitivity before the mod was over 15 uv at 3.885 and got better the lower you went, now I can detect signals down around 7 uv or better over the entire amateur range. I’m not concerned with 4 mhz and above so I will leave testing to others.
Winding 1 turn less on L2 and L3 might have a different affect than shorting one turn, so I will leave that to someone else to try and welcome their input. I would like to hear from you if have any other mods.
You can contact me through yahoo.com using my callsign.
My personal opinion – If the RX sensitivity issue gets resolved, this would be a nice radio for the money. It’s easy to assemble and looks nice. The only addition it could use is some sort of noise blanker or ANL circuit. Audio output with the PDIP upgrade is adequate for someone like me that suffers from hearing loss and doesn’t want to wear headphones.
73,
Bob N4FV
October 2020 Update:
I ordered one of the LNA preamplifiers from Ebay. It was stated as having 21 db gain from .01 to over a ghz. It did not come with any input protection so I added some 1N4148 diodes to clamp the input voltage to 1.2 volts, I tried inserting it in the 2 MHZ IF and that did not work. Inserting it in the antenna line did result in a respectable improvement to the RX sensitivity, at least 1uv. The background noise floor does rise slightly since the device does add noise. A suitable switching arrangement will have to be done using either PIN diodes or a relay. Also a bandpass filter should be installed to limit overload from broadcast stations, especially if nearby. QRP Labs has one that can be easily duplicated unless you want to place an order with them.
More to come as I pursue this.
The N75a group at GROUPS.IO seems to be inactive. I posted there and never got any responses from anyone about the issues I was experiencing. I definitely would not recommend this kit to anyone unless they are an experimenter willing to put some time into making it work properly.
Dec 2020 Update
Bandpass filter is constructed. Final tuning needs to be done and then a small DPDT relay installed to switch the preamp/filter out of the circuit on TX. Hopefully I will get this finished up soon and report back.