Ekco AD65.

[Station X]

I have an intermittent SG problem which may well be the subject of a thread in the Test Gear section.

Having borrowed another SG and injected a 110kHz signal, it merely confirmed what I already knew. Three out of the four IF transformer windings can't be peaked.

With two windings, both in the same can, output increases as the trimmer capacitance increases, but the trimmer runs out of travel at max capacitance. So it's not possible to tell whether a peak has been reached or not. I assume that the windings have lost inductance over the years. I'll experiment with adding a small fixed capacitor in parallel with the trimmer.

With the third winding the trimmer has no effect on output at all.

The good winding peaks with the trimmer at about half travel, an ideal situation as far as I'm concerned.

[PJL]

Are you testing the transformers in circuit? Try taking them out and seeing if they peak. If they still have a problem I would suspect the trimmers first.

[Station X]

These transformers have their terminations inside the cans with long wires leading to the points in the circuit where they connect to. I think there is little possibility of anything external to the can affecting the resonant frequency. So far I've been testing them in circuit.

Sweeping the SG indicates that the resonant frequency is too high, so I deduce that more capacitance is needed. On visual inspection the trimmers look fine, with the mica dielectric/insulation intact. I shall probably start with the old dodge of using a pair of twisted wires to give a low parallel capacitance.

[PJL]

I had a problem with the trimmers in a Philips set and had to drill out the rivets, clean, and reassemble with insulated bolts/nuts to get it to work properly. A combination of leakage to the chassis and dirt/oxidation between the plates/mica. I eventually did every trimmer in the set which vastly improved the performance.

[Station X]

I dismantled one of the compression type trimmers. It has 3 plates and two sheets of dielectric. There are no rivets, the plates are extended to form tabs which the lead out wires and coil terminations are soldered to. The trimmers are an integral part of the IF transformer and sit in two recessed oblongs in a Bakelite? moulding.

The parts didn't appear to be dirty, but I cleaned them anyway. Reassembled everything, but no change. I still couldn't get a peak.

Got out my box of silvered mica capacitors and soldered a 3.3pF one in parallel with the trimmer. Still no change. Tried a 6.8 pF one instead. Now get a peak at just short of maximum trimmer capacitance.

I think the trimmers must have a very limited capacitance range. A quick calculation shows that if resonance occurs at 10 pF, then a swing of plus or minus 2 pF should enable the transformer to be tuned from 100 KHz to 123 kHz. I'm not getting anything like that range.

I now have to sort the other two trimmers.

[Station X]

I decided that bending the plates of the trimmers to extract the mica dielectric "shims" for cleaning wasn't a good idea as there was a risk of the "brass plates snapping in two.

So I removed one of the IF transformers from the set and removed the parts of one trimmer from the transformer. This was the trimmer which wouldn't peak at all. One thing I noticed was that there were two mica shims between one of the pairs of plates, which would halve the capacitance between those plates when compared to a trimmer with a single shim. It is of course a three plate trimmer, so the overall capacitance would not be halved, but merely reduced.

Having reassembled the trimmer with a single shim between each pair of plates, I measured its range as 30 to 100 pF. The inductance of the transformer winding was 17.79mH so the calculated tuning range is 119 kHz to 217 kHz.

To achieve 110 kHz I'll need another 17 pF plus a bit more to allow tuning through the peak. This seems like an awful lot of extra capacitance.

I also noticed that the brass plates and mica shims of this cap were covered in a sticky whitish substance. Was this some sort of adhesive holding the shims to the plates and presumably changing the dielectric constant and therefore the capacitance?

I wish I could buy a new pair of IF Transformers at five shillings and sixpence each as advertised in the manufacturer's service sheet.

[Herald1360]

That's about £17 each in real money- not so attractive, perhaps.

[Station X]

I've now dismantled, cleaned and reassembled both trimmers in the IF transformer. I've checked that the trimmers aren't shorting and that I have continuity through the windings.

I haven't reinstalled the transformer in the set yet though. Before doing so I'd like to peak it up out of circuit, so as to avoid yet another remove and reinstall cycle.

Any advice as to how I might do this? I'm thinking SG on one winding and a 'scope on the other. However I know from past experience that connecting probes to IF Transformers can detune them.

[G6Tanuki]

Have you considered whether the coils might have absorbed damp etc. over time?

It might be worth sticking the IFTs in a jam-jar with a few sachets of Silica Gel for a month or so to dry them out.

[Herald1360]

If you connect in and out via 10K resistors, you should be fine at 120kHz. Levels will be meaningless, but it's just the peak you need to see.

[Station X]

Originally Posted by Herald1360

If you connect in and out via 10K resistors, you should be fine at 120kHz. Levels will be meaningless, but it's just the peak you need to see.

That's what I did. It makes a difference which winding the SG and 'scope are connected to, the output being greater one way then the other. This could be due to the IFT being a step up or step down type or perhaps it's because with the IFTs removed from the set the bottom of the IFT is open.

There is some reduction in output if I touch the can with my hands.

With both trimmers fully tightened IFT peaks at 80 kHz.
With both trimmers fully loose IFT peaks at 108 kHz.

That isn't the whole story of course, because the two windings could be peaked at different frequencies. However I can peak the trimmers at frequencies within the range. The peak output voltage is reasonably constant at frequencies within the range.

I can't get anywhere near the required 110 kHz though.

So it looks like I now need less capacitance, whereas before cleaning the trimmers I needed more.

I'm going to make up a dummy bottom cover for the IFT under test and if that doesn't improve matters, I'll try removing the trimmers and substituting some fixed value silver micas and see what that does.

[TrevorG3VLF]

If you want less capacitance, put a capacitor in series with the trimmer, the bigger the capacitance, the less its effect.

[Station X]

Thanks for the suggestion. I don't want to go down that route just yet. The coils and external wiring to the IFTs are all terminated on the trimmer caps. To provide a series cap would involve shifting terminations and I really don't want to have to unsolder the thin wires of the windings again for fear of breaking them which would be a disaster.

Whether of not the can or a can bottom is installed makes no difference to the frequency. I'm assuming that's because there's a very wide spacing twixt can and winding.

The theoretical trimmer value for a frequency of 110 kHz is 117pF. I set both trimmers to that value and checked the frequency. It was about 90 kHz. There's something fundamentally wrong here but I don't know what. Let's not forget that when the IFT was installed in the set more capacitance was needed.

I think I'll do as G6Tanuki suggests and put the IFT in a jar with some silica gel to see if that makes any difference.

In the mean time I'll get on with the cabinet which presents a different type of challenge.

[Station X]

This set and the removed IF Transformer has been sitting in a warm dry place since I lasted posted in this thread nearly four months ago.

You will recollect that I removed one IF Transformer and dismantled it in order to give the trimmer caps a thorough clean. Having done that I tried to tune the transformer to 110 kHz out of circuit, but had no success.

A trawl of the internet suggested that tuning an IF transformer out of circuit isn't easy to do, so today I reinstalled the transformer in the set and both trimmers peaked nicely at 110 kHz.

Brilliant I thought, I'll give the other IF transformer the same treatment. This I did, but I can't get it to peak at all. I can't tune it for min/no response either. I've made sure that the coils aren't open circuit and the trimmers aren't short circuit. No joy though.

The transformer which peaks is the one (electrically) nearest the aerial. the one which won't peak is nearest to the detector.

The only other untoward things I've mentioned before. The "volume" control is only effective for it's last few degrees of rotation (No it isn't OC) and the delayed agc doesn't operate on strong off air signals.

The manufacturer's service sheet for this set suggests that one of the IFT trimmers should be tuned for min rather than max response. How you're supposed to do that I don't know, as there's no null or trough position. The Trader sheet says to peak everything,

I think I'll try injecting a signal directly to the suspect IFT and sweeping it to see if there's any peak response point.

[Station X]

I think I've finally got to the bottom of this problem, but I haven't actually fixed it yet.

The problem is with the trimmers used to tune the IF Transformer which won't peak. They have insufficient capacitance to tune the coil to resonance at 110kHz. This can be compensated for by adding additional parallel capacitance, but there's a snag. The trimmers also have very limited tuning range, which makes the value of the additional capacitor extremely critical. Too high and the IFT will only tune below 110kHz. Too low and it will only tune above 110kHz. Unfortunately my box of silvered mica caps doesn't contain the correct value.

The service data doesn't give a value for the trimmers. Ideally I'd like to change them for new ones, but they are four tag types with the tags fitting through holes in a Bakelite former. The only replacements I can find are two tag types which won't fit the holes and in any case don't have sufficient tags to support the end of the windings and the external connection leads.

I think I'll have another go at dismantling, cleaning and reassembling the trimmers. They are soldered together rather than riveted and slight misalignment of the plates is likely to affect the capacitance value. I'll check the capacitance range before and after doing this.

If that doesn't work I'll consider putting some additional trimmers inside the can. There's plenty of room for them.

[batterymaker1]

You have the tenacity of a bulldog and patience by the bucketful. I salute you!

[Station X]

After a break of more than a year I started work again on this set a few weeks ago.

IF ALIGNMENT

I have finally got the IF Transformers to peak at 110 kHz. This involved a lot of experimentation with silvered mica capacitors soldered in parallel with the IFT coils. The values are very critical owing to the limited tuning range of the trimmers.

Having found the correct values of cap I realigned the IF to 110 kHz. The manufacturer's service sheet suggests that one winding should be adjusted for minimum signal. However I couldn't find a trough in the response, only a peak, so I peaked all four windings as per the Trader Sheet.

RF ALIGNMENT

It was now time to do carry out RF alignment. Interestingly the manufacturer's service sheet gives no details of RF alignment, so I had to use the Trader sheet.

Aligning this set is not an easy job, as the trimmer caps can't be adjusted with the chassis in the cabinet and when the chassis is removed from the cabinet there's no tuning scale to set the pointer to. I tried taping the old tuning scale to the chassis, but it was so distorted (by heat?) as to be useless. So I had to remove and replace the chassis from the cabinet several times during alignment.

There's no mark to indicate where the cursor should rest with the vanes of the tuning cap fully meshed. Also there's only a single trimmer to adjust the local oscillator frequency. What you have to do is adjust the trimmer so that when 200 metre and 500 metre signals are tuned in, the error in the cursor reading is the same and in the same sense for both signals. Having done that you can position the cursor correctly and recheck the results. This took many fine adjustments, but I finally got it right. I then peaked the RF trimmers using a weak off air Absolute Radio signal on 1215 kHz.

Even with the LW local oscillator trimmer set at minimum I couldn't get the cursor to sit at 1515 metres when receiving BBC R4 on 198 kHz LW. It was closer to 1490 metres Only changing the trimmer or connecting a fixed capacitor in series with it would have fixed this, so I left things as they were.

REASSEMBLY

Other threads describe construction of a new tuning scale trim:-

https://www.vintage-radio.net/forum/...d.php?p=772092

and repairing a damaged mounting lug:-

https://www.vintage-radio.net/forum/...d.php?t=117813

I had no problems with assembly and thanks to earlier adjustments the pointer moves freely without contacting the tuning scale or speaker surround.

OUTSTANDING ITEMS

I don't have the correct mains connector or knobs.

[Station X]

OVERALL IMPRESSIONS

Putting this set into good working order (it can hardly be called a restoration) hasn't been easy. Most of the bread and butter stuff is listed in post #14. However having done that work it took me a long time to sort out problems with the IFT's.

I'm still not convinced that the AGC and gain control circuitry is working correctly. The set is silent until the gain control is turned. well beyond the half way mark.

The set gives out a lot of heat from the dropper resistors. Normally I have a vintage set stood on the window sill for daily use. This set threatened to set fire to the net curtains though, so I moved it to the sideboard.

The illuminated tuning pointer is pretty useless. In a bright room you can see the scale, but not the pointer. In a dark room you can see the pointer, but not the tuning scale. I have the correct bulb fitted and think the problem may be down to the material used for the reproduction tuning scale.

Performance on MW is good with plenty of volume in reserve. On LW, performance when receiving French and German stations is good. Not so when receiving BBC Radio 4 though. It's much weaker than the continental stations and there's a lot of background noise which I think is coming in via the aerial.

EDIT

Performance on LW is much improved with a 150 foot long wire aerial connected. The usual wire up to the picture rail isn't good enough with this set. Connecting a signal earth made no difference.

[mark pirate]

I certainly admire your patience with this set, it has certainly been a long journey!
To be honest, I have never had a satisfactory conclusion with my AC85. That too is silent until the gain control is two thirds advanced
I may revisit it one day....

The best performing round Ekco is the A22, I have restored two of these now, both of which were in a very derelict state when received.

Mark

[Phil G4SPZ]

Well done! What a marathon.

Originally Posted by Station X

The set gives out a lot of heat from the dropper resistors. Normally I have a vintage set stood on the window sill for daily use. This set threatened to set fire to the net curtains though, so I moved it to the sideboard...

Just a thought - is there an internal heat deflector fitted to the back? I've not worked on an AD65, but I once restored an AD75 which required me to make a complete new back from hardboard, and the original heat deflector (first picture) was missing. A replacement was made from tinplate, second picture. It sits above the dropper, and as well as protecting the Bakelite cabinet above the dropper, it channels the hot air out through the vents in the back. In so doing it also creates a draught by the 'chimney effect' that helps draw cool air in from the lower vents, increasing the airflow rate across the dropper and reducing the temperature of the issuing hot air. It still gets warm rather than hot, but not hot enough to be a fire risk!

Phil