Ferguson 354U

I repaired one of these attractive burgundy and cream painted Bakelite sets for a customer a few years ago. I had been on the lookout for one for some time, when I found this one listed on the Wireless Works website.

It was in as-found condition - complete but untested. The only problems noted in the description were the missing metallic trim behind the controls and a few minor chips and scratches in the paint. The asking price was £28, which was fair.

The set arrived very well packed in a large box surrounded by screwed-up newspaper, and was just as described. Also included was a photocopy of the "Radio and Television Servicing" service data - a useful touch. To compliment this I printed a copy of the manufacturers data from one of my Service Data CD-ROMs.

The colour of the set is more red than it appears in the photo (it is a similar colour to EU passport covers). I will try to get a better photo when there is more sun.

Initial Inspection

From a quick examination, it was clear that the set would need a fair amount of work to return it to good reliable working condition. The on/off switch did not operate positively (note that this is separate from the volume control on this set). The mains lead had been replaced in the past by means of a choc-block connector inside the set, and the new lead had been cut off where it enters the set, at some later date. The wax-paper capacitor connected across the mains input had blown itself apart (as usual). Much of the PCB was covered in wax (both top and bottom). Most of this had clearly come from the three wax-paper capacitors, since these seemed to be lacking in wax. They were mounted vertically close to valves, so the heat was probably sufficient to melt the wax. There were also a couple of Hunts capacitors, which were falling to pieces (as they do).


The set is not particularly engineer-friendly. To remove the chassis, first the back must be removed and the wires to the aerial and earth sockets desoldered. The tuning knob must then be released, followed by the glass and bezel. This is held by four tags bent over inside the cabinet - three are easily accessible and one isn't. The tuning pointer must then be removed. The chassis itself can then be released (three screws) to the extent of the speaker leads. These leads need to be unsoldered to completely withdraw the chassis. A set like this can only be dismantled and reassembled a few time before the bent metal tags retaining the glass and bezel break off.

With the chassis out, I decided that it would probably be easier to remove the PCB from the chassis for cleaning and capacitor replacement. About a dozen connections needed to be desoldered, so careful notes were taken. I left the waveband switch and the volume pot attached to the board, so removed the nuts holding them to the chassis before removing the two screws securing the PCB and lifting the whole assembly out.

The PCB is only held to the chassis by two screws and the controls at the front. There is no support at the back, however it is not particularly deep or heavy and has survived over 40 years like this.

PCB Restoration

The area around the rectifier was so filthy that I decided to remove some components from the PCB to allow proper cleaning. This method was so successful that I continued with the whole PCB in the same way.

I cleaned the PCB itself with isopropyl alcohol, using paper tissue and cotton buds. I found that the best way to clean the small components was to unsolder one end, lift them vertically then wipe with a dry tissue while still warm from the desoldering (the warmth softens the wax). Isopropyl and other cleaning solvents tend to remove the markings on the resistors etc. A couple of resistors lost their markings in the process, so were replaced with parts of a similar age. All the resistors were checked with a meter and were found to be within spec.

The wax-paper and Hunts capacitors were replaced with modern PCB mounting types. A few of the smaller capacitors around the mixer-oscillator section had been replaced with RS components in the past. I checked these against the service data and found that some were wrong, so I replaced them with the closest values I had in stock.

The valve-holders were soaked in the isopropyl for a minute then wiped with tissue. One valve-holder broke into two halves during this process. Since it was a clean break and fitted back together positively, I secured it with a black cable tie (the holder is black so a black tie is less visible). I decided against superglue due to the risk of getting it in the contacts. Once back in the board the tie was probably not needed but I left it there anyway. I fitted this holder in the mixer-oscillator position, where things do not get too hot, and it is partly hiddeb behind an IF transformer.

I checked the main smoothing capacitor with my reformer before refitting - it was fine.

The whole process took about four hours, and the result was worth the time taken.

Chassis Restoration

Perhaps the word "chassis" suggests something substantial? In this set, the chassis is an inverted T-shaped metal assembly that fits into the front of the case. On it is mounted the controls, dropper resistor, output transformer, tuning capacitor, lamps, ferrite rod aerial and PCB.

I started with the on/off switch. This was disconnected from the circuit and removed. The switch body is the same as those fitted onto the back of combined switch/pots, but it is operated by a spindle with more limited rotation. On this switch, there was sufficient movement to flick it on, but not quite enough to flick it back to the off position.

I drilled out the two rivets holding the Bakelite contact assembly onto the front spindle plate. The contact assembly mechanism seemed to work correctly. I bent the operating tag on the front section to push further to the off position. This means it would push less to the on position, but with the contact assembly back in place, it all operated correctly. I fixed the switch back together with two 6BA screws and nuts.

I removed the remains of the blown-up capacitor and the dropper resistor from the chassis and cleaned it with isopropyl. I then refitted the switch and dropper, and fitted a new Class-X2 capacitor (since it is directly across the mains input). All leads were sleeved with new heat-resisting sleeving.

I replaced the wires that connected the output transformer and tuning capacitor to the PCB since the insulation was melted in places (probably by a soldering iron when those capacitors were replaced in the past). I also replaced the tone correction capacitor mounted on the output transformer.

The dial lamps were both blackened, so I fitted two new 12V 0.1A bulbs.

Two of the wires to the ferrite rod aerial terminate on a tag strip mounted on the aerial. This was originally held in place with tape, but this had lost its hold over the years. I used a thoroughly modern alternative - two cable ties.

I then refitted the PCB and connected everything back up.


With the amount of component replacement and testing that had been done, I expected there to be little chance of a discrete component fault. On the other hand though, a lot of things had been disturbed so there was a risk of human error!

I refitted the valves (the only components that had not been tested), connected a speaker and applied power via the lamp limiter. A meter on the HT showed that it was a bit high. The set just about received a signal, but it was very quiet and distorted even at full volume. A quick meter check on the cathode of the output valve showed a reading of just 1.5V (it should be around 9.2V). This confirmed what I suspected - the UCL83 output valve was very low emission. I fitted a good used valve and tried again. Much better - the HT was about right, as was the output valve cathode voltage, and the set worked fairly well.

After a few minutes, the sound started to become weaker and distorted. A few voltage checks showed that the HT was low. It was not being dragged down by excessive output valve current, since the output valve grid was correctly at 0V. I was about to try another UY85 rectifier, when I noticed that the smoothing capacitor was very warm. Although this seemed OK with the reformer, it obviously did not like the idea of working for a living again. The capacitor was a dual 50uF 275V PCB mount can. The closest I had was a triple 32uF 350V type in the same style of case. I folded the extra pin flat below the case and soldered a wire link between it and the pin for the second smoothing section, so I had 32uF for the reservoir and 64uF for smoothing. I connected it this way because UY85 rectifiers are rated for no more than 50uF on their cathodes. The reduced value of reservoir capacitor could result in increased hum, but none was noticeable through the small speaker in the set.

It still seemed to be struggling to receive strong stations however. As a test, I tried replacement UCH81 and UBF89 valves, which made no significant improvement. Time to check the alignment. With the wrong capacitors that had been fitted it was quite possible that this had been fiddled with too.

I connected a meter to the AGC line and carefully adjusted the four IF cores for maximum reading with the set tuned to a local station. Three were correct but one was about half a turn off. With this corrected the set sounded better, but I was still not happy.

When looking at the circuit diagram to find where to connect the meter onto the AGC line, I noticed that the decoupling/filter capacitor across that line seemed very low at just 0.005uF (the diagram and set agreed). Comparison with a few other similar circuits confirmed that 0.05uF was a more likely value. I temporarily connected a 0.05uF capacitor, which improved the sound quality by removing the slight off-tune sound. By connecting the 'scope to the line it was clear that with 0.005uF some low frequency audio remained on the line, which would have caused the gain to be varied in sympathy. With 0.05uF there was no audio visible on the line, so I fitted this value. This made a significant improvement, and is worth remembering for the future!

The capacitor in question is circled on this circuit diagram section.

After some more testing, the set started to crackle intermittently. Some gentle tapping with the screwdriver handle narrowed it down to the right dial lamp. I cleaned and retensioned the contacts in both lamp holders, which fixed this problem.

After another half-hour the set suddenly went off. The dial lamps went out and the sound faded away, so it was clearly a problem with the heater chain. I tipped the chassis forward to investigate, and it came back on! I left it and a few minutes later it went off again. This time I was able to carry out a few voltage checks, which showed that the UY85 rectifier valve filament was open-circuit. When the valve was tapped it started working again. I fitted a good used replacement.

No further problems were found during soak-testing.


The cabinet was in good clean condition and needed very little work to make it look good. The most visible problem was the chips in the cream paint on the front.

Comparing the caps of the various cans of car spray paint I had, suggested that Ford Sierra Beige should be a good match. I sprayed a small amount into the can lid and applied it to the chips with an artists paintbrush. It was a perfect match! Three applications were needed to get the chips and scratches fully filled.

Looking closely at the front, I am undecided whether it is the original paint or if it has been resprayed at some time. I think it probably is resprayed - some of the edges where it would have been masked are perhaps not quite as sharp as the manufacturer would have done, and the retouching of the FERGUSON letters is not quite up to manufacturer standard either. The exact match of the car paint is another clue too. If it is resprayed it is a very good job!

The scratches in the burgundy paint were a bit more of a problem. As with the front, I only wanted to touch up the damage, not repaint the whole cabinet. I had no suitable coloured car paint in my stock, and I could not find anything dark enough in Halfords. However, I did have a selection of small tins of enamel model paint, so I decided to try to mix a small amount of the correct colour. After some trial-and-error, I managed to get a reasonably good match with a combination of black, red and a couple of mauves. I mixed it in a small screw-top spice jar, so that the mixed paint could be kept liquid while small samples dried (it dried slightly less red than it looked wet). It took three attempts to get the close colour I achieved. By this point the mixed paint was beginning to thicken slightly so I decided to settle with what I had rather than continuing to try improving the match.

The main damage was to the area between the volume control and the waveband switch. I painted the whole area between the controls to avoid visible edges. In the other places, I filled in the scratches with a thin artists paintbrush. Once the paint was dry (24 hours), I smoothed the paint with Brasso wadding. In the areas where I had filled scratches, I used the wadding to carefully remove the paint from the surface while leaving it in the scratches. The result is not perfect, but from a normal viewing distance on a shelf, it looks fine.

Brass Trim

The brass trim from around the tuning scale was badly tarnished. This responded well to Brasso used on a scouring pad. I also removed the tuning pointers from their carrier and the trim from the tuning knob, and polished these with Brasso Wadding. All these parts had originally been lacquered, but most of this was now thin or missing and was removed by the polishing.

The area of the cabinet behind the controls would originally have had a piece of trim fitted - the glue that held it remained. I had no idea whether this should have been brass, aluminium or something else. I decided that brass was the most likely because the set was too old for brushed aluminium and the remaining trim was brass.

I used some brass shim material, about 0.25mm thick. I had obtained this from a previous employer when they ceased trading, but it is readily available from engineering suppliers (companies who supply nuts, bolts, tools, etc.).

The gap was conveniently exactly the same width as my narrow 12" steel rule, which is 3/4" (19mm). I cut a suitable strip with sharp scissors, then spent some time trimming it exactly to length and rounding the corners as necessary.

By placing the strip in place with the ruler on top it could be held flat enough to allow the cabinet to be turned around and the control cut-outs to be marked from behind with a fine permanent marker pen. These were then carefully cut out with the scissors, remembering that I had to cut to the outside of the line. Again, some fine trimming was required to get it exact - sharp curved nail scissors were the ideal tool.

All the brass pieces were then cleaned with isopropyl before being carefully fixed to a piece of newspaper on the bench in the garage with loops of masking tape, then sprayed with clear polyurethane varnish. They were left for 24 hours to dry before being disturbed.

Before sticking the trim strip in place, I protected the adjoining edges of the case with masking tape. This was to prevent me getting glue on the case as I inserted the strip. I sprayed the back of the strip with spray-adhesive then carefully placed it into position and pressed it down. I then removed the masking tape. A row of cassettes in cases, stood on end, provided enough weight to hold the strip down while the glue set.

I refitted the pointers to their carrier with Evostik, and supported these with a roll of tinned-copper wire. I left everything for 24 hours while the glue tried.

These waits of 24 hours do not cause as much time wasting as may be imagined, since the jobs were done on successive weekday evenings after work.


As with most restorations, the final reassembly was a quick and uneventful procedure. By this stage almost all the work has been done, and there isn't much to write about!

As mentioned before, there was no obvious means of securing the mains cable. I secured it to one of the speaker mounting clips with a cable tie.

The radio performs well for this class of set, and does not run excessively hot. I think it is rather attractive too!

This website, including all text and images not otherwise credited, is copyright © 1997 - 2006 Paul Stenning.
No part of this website may be reproduced in any form without prior written permission from Paul Stenning.
All details are believed to be accurate, but no liability can be accepted for any errors.
The types of equipment discussed on this website may contain high voltages and/or operate at high temperatures.
Appropriate precautions must always be taken to minimise the risk of accidents.

Last updated 14th April 2006.