REAL radios glow without smoke.
Designed 2014, Steve Jones, VK2YLD.
Material Copyright 2007-2020, VL2YLD
Material on this site is for informational purposes ONLY, whilst all reasonable steps have been taken to ensure accuracy, errors and omissions do happen, but it wasn't me and I didn't do it! ANY ACTIONS taken from this information is soley at YOUR RISK. However, please feel free to copy, print, repost anything you find useful.
The old workhorse - Philips FM-92
The radio used on the node is a 'U' band (450-470Mhz) Philips FM-92, reprogrammed and re-tuned down to the 70cm amateur band. These radios are a bit 'hit & miss' with the VCOs, some will pull down far enough and others will simply refuse! I replaced the VCO in this unit as the first spent more time out-of-lock than working, but the replacement is fine?? (component tolerence etc i guess)
The FM-92 has been around a long time now, but it still performs well as long as the usual faults have been sorted and mods have been done. First thing to do was to rip the covers and shields off and clean everything. Carefully check the soldering on the PA transistor as this often breaks away giving intermittent TX power failure. Use a jeweller's loup or similar to get a good look... Rework the joints if they are suspect in any way. Assuming the original EPROM is still in the radio, fire it up into a test set and check for noisy VCO and correct TX power output. Repair any failures. Check receiver sensitivity and repair as necessary.
Remove the EPROM and replace it with the new one. Re-tune the VCO as per handbook but, as there is only one channel programmed, set VCO to 10v at the test point. If it won't make it get as close as you can, but if the test point is <4v or >13v the radio won't be stable. Try another VCO, radio, or get out the pliers! Usually the VCO adjustment screws go right in and you need more!!, so the only answer is to remove the VCO from the board and open it up (argghhh). A couple of small 5pF chip SMD caps (NPO if you got 'em) soldered across the trimmers will usually extend the range far enough. Rebuild, replace it and try again... (Also a good opportunity to check the PCB soldering to the case!)
Next is the front-end.. Luckily I haven't had one yet that didn't pull down far enough. Systematically tune each slug on the front-end module whilst reducing the sig-gen level to keep the signal slightly noisy as you peak things up. Some of the front-ends I have worked on required an RF sig-gen level about -45dBm to open the mute before tuning. After you peak things up, the sensitivity should be about -120dBm or thereabouts for mute open. As there is only one channel, the front-end can be peaked right up without having to worry about broad-banding (edge to edge), just screw all you can from the receiver. Then adjust the TX power and check the efficiency (DC current). Tweaking the trimmers in the PA may be enough, cap replacements or additions may be required. These radios can get up to 40w output with some fine tuning, but DON'T DO IT!.. Peak it and then reduce to 25-28W else the final transistor may release it's magic smoke and head off to silicon heaven!! According to the manual, the temperature sensor reduces power when the heatsink gets to 100oC!! That's HOT.. 25w with fan cooling works just fine and seems to remain cool enough to touch.
Last thing to do is the mods... If you got this far without blowing it up!
Firstly, look at the control board and see if the .5F supercap is installed by the systems connector. If its there, remove it.. In the vacant space for module U403, install a lithium cell with -ve to ground and +ve to +5vSW rail via a blocking diode. The diode acts as a switch over element to stop the radio from attempting to charge the 3v lithium up to 5v when switched on with the usual result.(boom!).. This holds the brains up when power is off and allows the radio to auto restart when power restored. Don't expect the battery to last for months without power as the drain is too high, but 3-4 hours every so often doesn't hurt too much. Mine has lasted 2 years now and still has a good charge with about 5 hours a week powered off.
Next is the COS line... This I decided to buffer into an open collector switch that extends the signal out the rear DIN connector's centre pin (normally unused). The two transistors were located in vacant solder pads near the tx scrambler slot (not fitted) and point-wired. I used a 2SC711 buffer into a BD139 current switch configured to give active low on carrier detect. All's well?? Goooood... Clean the cover mating surfaces, add a little heatsink compound to the cover rails (helps transfer heat to the lid and base covers) and screw down the lids. (I'll put up some photos when the camera works for me!)