Here are my hand drawn reversing notes. good luck. They are about 85 percent complete. I have not been able to figure out how it decided to use single or double conversion. The switch bank wafers are allmost innaccessible and I cannot decode them. This switching is determined by a switchbank at the rear of the chassis underneath the 2nd if module. The switch bank is driven by a rack and pinion drive from the main bandswitch turret.  There is enough information here to debug most A618 faults.





second mixer module at rear of chassis using ECH81 or CV2128 with 1995khz crystal osc



part of second mixer, under chassis wiring, mixer and first if transformer



main if amp bandwidth selection wafers This bit I dont understand. The switch bank is driven by rack and pinions from the front panel if bandwidth control. There are 2 stages of transformer coil determined bandwidth, one position inserting the quartz resonator in narrow mode and in "very narrow mode" a 800Hz parallel resonant LC filter is inserted in the audio path.




bandwidth selection wafers, very narrow mode LC audio filter is inserted here



post dectector audio amplifier and compressor. see the inferred schematic below.




detector, agc, delayed agc, noise limiter, manual IF gain control, S meter drive



audio power amp, filter CW in very narrow mode



under chassis central tag board, detector,agc,agc control switch



schedule of valves used in the A618 and base diagrams



detector and noise limiter and under chassis tagboad RHS



2nd RF amp, vfo regulator, mute relay



first rf amp




turret contacts, first mixer vfo








turret biscuits example









post detector af amplifier and audio compressor.This contains my hamification. I could not find a replacement 6AV6 triode  diode so I have rewired it to use a 6AM6 triode connected as the audio agc driver and a silicon diode as audio agc detector. This has worked well. I choose the 6AM6 because I have many and it required the least amount of socket rewiring in a hard to get to location.


My unit is powered from a high voltage dc supply. It wants about 250 volts DC, 6.3VAC filament supply and a negtive DC voltage source to supply negative bias for the agc system. I have found that negative 12 volts  is sufficient but I believe that it officially should be negative 40 volts to get the full agc range.   The power connector has many more conductors than I have yet mapped. The remote mute relay is controlled through the power harness and a few other things that I havent got around to understanding yet.  They dont impact on simple normal operations.  

I have owned this unit, serial number 0006, for 20 years now. It has given little trouble over this time however in the last couple of years there has been rapid deterioration.  The unit is constructed with one watt carbon composition resistors and nearly all resistors greater than 10Kohms that are disipating any significant power have had to be replaced as most have doubled or tripled their nominal resistance.  The biggest surprise was finding that audio interstage coupling capacitors used were silver micro 0.01uF.  Silver micas should have NO leakage...right....nope your wrong. After 50 years these devices have at least a mego-ohm leakage, some had 100K worth of leakage.  This completely deranges grid bias of suceeding stages. Replacing these , and, all paper caps with polyesters  restored the audio compressor to life and the audio once again sounded clean.

Keep the turret contacts polished , dont use abrasives, bond paper is sufficiently abrasive to get and nice shiny metal surface on the turret bump contacts (silver metal!) .  The original mil  spec valves with the CV number should still be serviceable....after fifty years!  mine are! Will YOU be serviceable in 50 years ?  The only valve that may cause some greif is the converter valves, CV2128 also known by the commercial equivalent of ECH81. They are very rare, some may still exist in some hams bottle bank.  The others have common equivalents like 6BA6 et al.  I could not find a 6AV6 to fix the compressor so I have used silicon diodes and a 6AM6 pentode wired up as a triode as a fucntional alternative. I think that when the time comes to replace an ECH81 I will replace it with a TV pentode/triode 9 pin valve. It should even have lower noise than the ECH81  pentagrid.

The dial cord was orginally steel standed fine piano wire,  I have restrung this several times with heavy nylon fishing line. Its a complete pain to restring but nylon will last 7 years even in the hot interior.


Still to do:
First mixer and first if  bandswitching. This is hard due to innaccesiblity in the chassis
2nd if bandwidth selection. easier said than done. This complex wafer switch taxes my tiny mind too hard. It works and I hope it stays working!
Fix band 3, for some reason it has become deaf. No good for NDB dxing. Suspect a silver mica has gone bad, however these bakelite units were not marked in indelible ink, the markings have evaporated.
Hamifications to think about.  The radio does not have a product detector and BFO energy leaks into the AGC. This makes SSB reception rather trying.  The radios performance on its highest band could be better, perhaps a shock..horror  solid state front end, mosfet, to replace the first RF amp.

And the existential question? Do I "restore" the radio hence no hamifications, or do I bend the orginal design to my will and improve on some notable design deficiencies especially the lack of product detector?