29-08-2012, 12:59 PM
(This post was last modified: 29-08-2012, 05:00 PM by Robert Darwent.)
I halted the completion of this thread due to the project appearing as an article in an upcoming issue of the BVWS Bulletin. Since that issue has recently been sent out, and the 'cat is out of the bag' as it were, I am pleased to continue the project here too for the benefit of non-BVWS members.
Chassis restoration
Continuing, I renewed all the wiring between the major components mounted on the circular chassis and the tag-board;
and then fully rebuilt the chassis;
Fortunately, the original screening braid which carries the wiring between the tag-board and volume control was in good condition and could be reused, it was just a case of carefully threading the new wiring along its hollow centre.
Regarding a replacement for the missing loudspeaker. I was fortunate to already have a 6.5 inch diameter, 3 ohm speaker amongst my previously mentioned spares. Although not from an A22, it originally came from a related Ekco model, it was exactly the type required.
The bronze speaker ring was more problematical. I didn't have one and a number of possible solutions were contemplated during the course of the several months this project took to complete. None proved really satisfactory, then quite unexpectedly an original bronze A22 ring was listed and purchased via eBay. The final missing piece of the jigsaw had been found!
Modifying the mains transformer
A standard A22 has three tapping points on its mains transformer to select different voltage ranges, the A22T being an export set has an additional three. The body and size of the two types of transformer are identical, the A22T just has a much larger voltage selector panel fitted.
Obviously I couldn't replicate the additional three lower voltage tappings, but I could replicate the appearance by making an appropriate reproduction A22T voltage selector fascia and fitting the existing A22 selector panel to its rear.
This would involve extending the existing wires between the transformer winding and the selector panel and rotating the selector itself from the horizontal and refitting vertically in the centre of the new fascia.
The reproduction voltage selector consisted of two parts, the existing paxolin panel which the tap selector was attached to and a graphic of the A22T selector fascia, printed and laminated several times to achieve an appropriate thickness, fitted in front.
The 'U' shaped metal bracket that forms the front of the transformer, mounting the selector panel with two rivets, had to be fitted the opposite way around in order to provide the clearance required from the transformer windings.
The rivets were carefully released using a modellers drill and a burr tool. This enabled them to be reused to secure the new selector panel. Observant readers will note that they have been refitted a little lower and should actually be in-line with the middle voltage taps. This was a necessary compromise due to the 'U' bracket used being approximately a quarter of an inch shorter in height than the one used in an original A22T.
Although not necessary in terms of making the set work, I felt the cosmetic modifications worthwhile. The voltage selector panel along with the dial are the only two things that visually differentiate the A22T from the A22, so I thought it important to try and reproduce them both for the sake of authenticity.
Initial testing
The time had finally arrived to switch the set on! After double-checking everything I connected a temporary aerial and applied power via a lamp limiter with the set switched to the medium wave range. No indications of anything amiss from the lamp and twenty seconds or so later I'd got a definite background hiss from the loudspeaker. Switching out the limiter and tuning around brought in several stations loud and clear.
I'd expected this of course, or rather hoped. The cores of the medium wave coils and the two IF transformers had been untouched, so providing my component and wiring renewal was carried out correctly the set should have worked on this band. But what of the two 'new' short wave ranges.
Trying the S2 range first produced little until I reached the 10 o'clock position on the dial, when suddenly a foreign language station came booming out of the loudspeaker! Further careful tuning found several fainter stations as well. Switching to the S1 range produced very similar results. Clearly both ranges were working, which was something of a success in itself. Whether they were in the correct band of frequencies was unclear at this point, but I was very encouraged by what had already been achieved.
Alignment
Adopting the alignment procedure outlined by Ekco in their service data for the A22 as a guide, I began with the I.F. alignment. Using a signal generator coupled to the grid (top cap) of V1 (ECH35) via a 0.1 uF capacitor, I injected a modulated signal at 465 kHz. As instructed, the set was switched to the medium wave range and the tuning capacitor was fully meshed.
I don't have an audio output meter, so the adjustments were simply gauged by ear. Working backwards from the second I.F. transformer, cores L17, L16, L9 and L8 in that order, were each peaked for maximum audio output. As expected, the I.F. alignment was pretty much perfect already and no significant adjustments needed to be made.
Next came the I.F. filter consisting of L1 and C1. Injecting a 465 kHz into the aerial socket, the core of L1 was adjusted for minimum audio output. Again the alignment was close to ideal already, as L1 had also not been disturbed.
Then the adjustments for calibrating the three tuning ranges. To begin with, the physical position of the dial needs adjusting so that when the tuning capacitor is fully meshed the shadow of the pointer on the illuminated tuning arm coincides with the calibration spot just under 'M' on the lower right of the dial. I had already ensured this was the case when fitting the dial earlier.
S1 range (8.5-23.0 MHz) - Dial tuned to and signal injected at 22.0 MHz, adjustment of oscillator trimmer TC4, followed by aerial trimmer TC1 - Dial tuned to and signal injected at 9.0 MHz, adjustment of the core of oscillator coil L10/11, followed by the core of aerial coil L2/3.
S2 range (3.0-8.5 MHz) - Dial tuned to and signal injected at 8.0 MHz, adjustment of oscillator trimmer TC5, followed by aerial trimmer TC2 - Dial tuned to and signal injected at 3.5 MHz, adjustment of the core of oscillator coil L12/13, followed by the core of aerial coil L4/5.
MW range (550-1550 kHz) - Dial tuned to and signal injected at 1500 kHz, adjustment of oscillator trimmer TC6, followed by aerial trimmer TC3 - Dial tuned to and signal injected at 600 kHz, adjustment of the core of oscillator coil L14/15, followed by the core of aerial coil L6/7.
All the adjustments were made for maximum audio output and repeated until no further discernible improvements could be made. The medium wave range needed little adjustment because it had not been disturbed, and surprisingly the two short wave ranges were close to their optimum settings as well.
I attributed this to the fact that I had carefully adjusted the coil inductances and trimmer capacitances as accurately as I could with a digital LC meter during the rebuilding of the coil-pack board, setting them to their calculated values.
Using the signal generator to inject various frequencies on both short wave ranges confirmed that they were accurately operating within the required band of frequencies. Tracking was as good as I had hoped for, with all frequency points matching the dial markings to within a couple of widths of the tuning pointer shadow.
Regards
Chassis restoration
Continuing, I renewed all the wiring between the major components mounted on the circular chassis and the tag-board;
and then fully rebuilt the chassis;
Fortunately, the original screening braid which carries the wiring between the tag-board and volume control was in good condition and could be reused, it was just a case of carefully threading the new wiring along its hollow centre.
Regarding a replacement for the missing loudspeaker. I was fortunate to already have a 6.5 inch diameter, 3 ohm speaker amongst my previously mentioned spares. Although not from an A22, it originally came from a related Ekco model, it was exactly the type required.
The bronze speaker ring was more problematical. I didn't have one and a number of possible solutions were contemplated during the course of the several months this project took to complete. None proved really satisfactory, then quite unexpectedly an original bronze A22 ring was listed and purchased via eBay. The final missing piece of the jigsaw had been found!
Modifying the mains transformer
A standard A22 has three tapping points on its mains transformer to select different voltage ranges, the A22T being an export set has an additional three. The body and size of the two types of transformer are identical, the A22T just has a much larger voltage selector panel fitted.
Obviously I couldn't replicate the additional three lower voltage tappings, but I could replicate the appearance by making an appropriate reproduction A22T voltage selector fascia and fitting the existing A22 selector panel to its rear.
This would involve extending the existing wires between the transformer winding and the selector panel and rotating the selector itself from the horizontal and refitting vertically in the centre of the new fascia.
The reproduction voltage selector consisted of two parts, the existing paxolin panel which the tap selector was attached to and a graphic of the A22T selector fascia, printed and laminated several times to achieve an appropriate thickness, fitted in front.
The 'U' shaped metal bracket that forms the front of the transformer, mounting the selector panel with two rivets, had to be fitted the opposite way around in order to provide the clearance required from the transformer windings.
The rivets were carefully released using a modellers drill and a burr tool. This enabled them to be reused to secure the new selector panel. Observant readers will note that they have been refitted a little lower and should actually be in-line with the middle voltage taps. This was a necessary compromise due to the 'U' bracket used being approximately a quarter of an inch shorter in height than the one used in an original A22T.
Although not necessary in terms of making the set work, I felt the cosmetic modifications worthwhile. The voltage selector panel along with the dial are the only two things that visually differentiate the A22T from the A22, so I thought it important to try and reproduce them both for the sake of authenticity.
Initial testing
The time had finally arrived to switch the set on! After double-checking everything I connected a temporary aerial and applied power via a lamp limiter with the set switched to the medium wave range. No indications of anything amiss from the lamp and twenty seconds or so later I'd got a definite background hiss from the loudspeaker. Switching out the limiter and tuning around brought in several stations loud and clear.
I'd expected this of course, or rather hoped. The cores of the medium wave coils and the two IF transformers had been untouched, so providing my component and wiring renewal was carried out correctly the set should have worked on this band. But what of the two 'new' short wave ranges.
Trying the S2 range first produced little until I reached the 10 o'clock position on the dial, when suddenly a foreign language station came booming out of the loudspeaker! Further careful tuning found several fainter stations as well. Switching to the S1 range produced very similar results. Clearly both ranges were working, which was something of a success in itself. Whether they were in the correct band of frequencies was unclear at this point, but I was very encouraged by what had already been achieved.
Alignment
Adopting the alignment procedure outlined by Ekco in their service data for the A22 as a guide, I began with the I.F. alignment. Using a signal generator coupled to the grid (top cap) of V1 (ECH35) via a 0.1 uF capacitor, I injected a modulated signal at 465 kHz. As instructed, the set was switched to the medium wave range and the tuning capacitor was fully meshed.
I don't have an audio output meter, so the adjustments were simply gauged by ear. Working backwards from the second I.F. transformer, cores L17, L16, L9 and L8 in that order, were each peaked for maximum audio output. As expected, the I.F. alignment was pretty much perfect already and no significant adjustments needed to be made.
Next came the I.F. filter consisting of L1 and C1. Injecting a 465 kHz into the aerial socket, the core of L1 was adjusted for minimum audio output. Again the alignment was close to ideal already, as L1 had also not been disturbed.
Then the adjustments for calibrating the three tuning ranges. To begin with, the physical position of the dial needs adjusting so that when the tuning capacitor is fully meshed the shadow of the pointer on the illuminated tuning arm coincides with the calibration spot just under 'M' on the lower right of the dial. I had already ensured this was the case when fitting the dial earlier.
S1 range (8.5-23.0 MHz) - Dial tuned to and signal injected at 22.0 MHz, adjustment of oscillator trimmer TC4, followed by aerial trimmer TC1 - Dial tuned to and signal injected at 9.0 MHz, adjustment of the core of oscillator coil L10/11, followed by the core of aerial coil L2/3.
S2 range (3.0-8.5 MHz) - Dial tuned to and signal injected at 8.0 MHz, adjustment of oscillator trimmer TC5, followed by aerial trimmer TC2 - Dial tuned to and signal injected at 3.5 MHz, adjustment of the core of oscillator coil L12/13, followed by the core of aerial coil L4/5.
MW range (550-1550 kHz) - Dial tuned to and signal injected at 1500 kHz, adjustment of oscillator trimmer TC6, followed by aerial trimmer TC3 - Dial tuned to and signal injected at 600 kHz, adjustment of the core of oscillator coil L14/15, followed by the core of aerial coil L6/7.
All the adjustments were made for maximum audio output and repeated until no further discernible improvements could be made. The medium wave range needed little adjustment because it had not been disturbed, and surprisingly the two short wave ranges were close to their optimum settings as well.
I attributed this to the fact that I had carefully adjusted the coil inductances and trimmer capacitances as accurately as I could with a digital LC meter during the rebuilding of the coil-pack board, setting them to their calculated values.
Using the signal generator to inject various frequencies on both short wave ranges confirmed that they were accurately operating within the required band of frequencies. Tracking was as good as I had hoped for, with all frequency points matching the dial markings to within a couple of widths of the tuning pointer shadow.
Regards
