Decca DM45.

[Till Eulenspiegel]

Good evening Gentlemen,
Many thanks for your inputs. The HT supply remains stable when the width fault appears. The amplitude of the line drive waveform remains stabile at -150 volts P - P. I don't think there is a thermistor in series with the line scanning coils, but it is a good suggestion because the scan coils do have an active part in the generation of the EHT. At the end of the line scan the stored energy in the scan coils contributes to the functions of the line output stage. When the scan coils are disconnected to line output transformer primary functions more like a choke, hence the very low EHT.
Since my previous post the width has come in as much as 35mm each side of the viewable screen.
There's not much left in the line OP stage to check and test.

Till Eulenspiegel.

[PYE625]

Hmmmm....interesting the HT does not fall, indicating perhaps there is no leakage anywhere, possibly wrong of course !!
It would be unlikely there is leakage in the actual scan coils I would think. But as we know, anything is possible.

[Till Eulenspiegel]

It's interesting to note the PL81 line output valve displays no distress when the width fault appears. This set has a Telefunken PL81. The construction technique of the TFK valve is quite different to that of the more common Mullard/Philips PL81. Another PL81 has been tried, results are just the same.

Till Eulenspiegel.

[PYE625]

I presume all the DC conditions in and around the line stage are stable?

Any naughty Wimas lurking still?

Just been looking at the circuit and I would check C58, C59 and R68 as a matter of interest.

[Till Eulenspiegel]

No Wimas in the line OP stage. The coupling capacitor between the line oscillator and the grid of the PL81 has an unusually low value, only 2000pF. A more common value is 0.01mfd. It's a Suflex make capacitor, I reckon it is OK but it wouldn't do any harm trying another capacitor.
Also the PL81 screen grid resistor should checked also.

Till Eulenspiegel.

[Derren]

Remember you said there was a cap dated 1972 a later replacement? Any other later work done or even components of wrong value used maybe?

[Till Eulenspiegel]

Remember you said there was a cap dated 1972 a later replacement? Any other later work done or even components of wrong value used maybe?

Hi Derren,
well the two valves in the line oscillator were possibly replaced late 1972. The ECL80 line discriminator and the ECC82 oscillator have been replaced by Tungsram branded valves, a 6AB8 and a 12AU7. Both valves are date coded September 1972. Tungsram was the wholesaler brand of Mullard. It's even possible these valves were fitted in 1973.
Mazda valves were sold under the Solus brand.

A close up picture of the line output transformer.

Till Eulenspiegel.

[Terrykc]

I note that the height increases as the width decreases - if the height was readjusted at this point it wouldn't be far off correct.

Is this the boost rail increasing or, more likely, the EHT dropping slightly?

[Till Eulenspiegel]

Hi Terry, The EHT voltage falls when the width fault condition occurs thus causing the height to increase. The frame oscillator is a multi-vibrator consisting of the triode section of V6B video amplifier and the triode section of V7B sync separator. The valve types are PCL84 and ECL80 respectively. The anode load resistors of the triodes have high values of 330Kohm and 4Megohms. Both these resistors are supplied from the boost line. It'll be interesting to measure the boost volts before and during the fault condition.

Till Eulenspiegel.

[Till Eulenspiegel]

Line output stage voltage measurements taken just after warn up and the picture displaying full width are: HT at the anode of the boost diode 197V. Boost HT 630V. Screen grid of the PL81 175V.
After 15minutes and the picture width reduced. HT slightly down to 194V. Boost volts down to 530V. PL81 screen grid volts unchanged, 175V. The value of the screen feed resistor remains unchanged hot or cold, 1.5Kohm.
Line drive waveform unchanged.

Till Eulenspiegel.

[PYE625]

Hi Till,

The boost HT falling by 100 volts in 15 minutes...... that's quite a drop.
Question is of course, is it external to the LOPT.

Are C54,55,56 and C59 in particular, definitely ok ?

[Till Eulenspiegel]

Somehow, I think the line output transformer is not faulty. Today I will introduce another test procedure. A 10ohm resistor will be connected between the PL81 cathode and chassis to monitor the cathode current of the valve during the non fault and fault conditions. As mentioned earlier the valve exhibits no stress during the time of the fault. Various capacitors and resistors associated with line output stage have been checked. Those scan coils are suspect.

Till Eulenspiegel.

[Till Eulenspiegel]

A 1 ohm resistor is inserted between the chassis and the cathode of the PL81. After warm up and displaying a full width picture the measured voltage present across the resistor is 0.2volt indicating the valve is drawing 200mA. That would seem too much current but in fact the valve is only conductive for slightly less than 50% of the duty cycle. Now I'd say that the time the valve is turned on is not long enough, doesn't normal practice in a line timebase demand that the valve or transistor is conductive for 55 to 60% of the duty cycle and the reclaimed energy contributes 40% of the scan?
I will report back my findings after the fault returns.

Till Eulenspiegel.

[Till Eulenspiegel]

NEWS FLASH! The width has started to come in and the current through the valve has increased considerably, in fact the measured voltage across the resistor is now 0.7volt. Moving the scan coils about causes the voltage to vary, I think that it is pretty certain it is scan coils that are breaking down.

Till Eulenspiegel.

[PYE625]

Its possible there is leakage to the deflector coil screen which is earthed.
Try lifting the ground at B1....But be careful, it looks to be part of the aquadag connection.

[Derren]

Hi "Till", must admit scan coils are things I've hardly ever suspected on anything, and yet this was also your conclusion on the interlace fault on the Royal Star was it not? Just as an update the Royal Star is now working better than it ever has and yet at the time it was almost condemned for its poor interlace. I think it is the best I can expect from it, super picture (I for one don't mind the emphasised line structure!) but linearity not the best, we've a different scenario with the DM45 here though of course. I suppose we also have to remind ourselves sometimes that these sets were never intended to last for over 50 years

[Till Eulenspiegel]

Hi Derren,
I've found a set of 110 degree scanning coils which have been removed from another project set. These coils will be tried out in the Decca so that we'll know for sure if the original scan coils are faulty or not.

Line output valve: The PL81 continued to be used into the early sixties even though the better valve for wide angle scanning is the PL36.
In my early post it was observed that the PL81 was drawing 200mA during 45% of the total scanning time 0f 98microseconds. Simple calculations confirm that the average current is 90mA. Well within the ratings of the valve.

Till Eulenspiegel.

[Till Eulenspiegel]

The alternative scan coils have been fitted. The wires from the chassis to the coils are bit on the short side so that's why the picture is tilted. The results are more picture width and slightly reduced height.
After fitting the scan coils the set was only tested for ten minutes. It'll receive much long test time tomorrow.

[Till Eulenspiegel]

The ersatz scan coils have not cured the fault, but nevertheless there is a worthwhile improvement. The width does still come in after a half an hour or so. The mark space ratio of the line drive waveform needs some attention. at the present time the PL81 is on is for only 45% of the duty cycle, 55% would be better, so some checks in the line oscillator multi-vibrator are called for.

Till Eulenspiegel.

[Derren]

What a shame, and this one seemed like an easy fix on the face of it! Ah well, keep trying I've thought of another connection between the K-B Royal Star and the DM45 and that is poor ventilation which can't help, the coils in the K-B certainly had a 'blackened' look about them.

[Till Eulenspiegel]

There is no other alternative the line output transformer will have to be replaced. The transformer chosen for the job will be a BRC jellypot.
It should be possible to fit the "new" transformer and the tripler unit inside the line output stage screening metalwork, then it will be hidden and the original appearance of the chassis is maintained.
The BRC transformer delivers power to spare in order that a stabilising circuit can be employed. This can be considered for the Decca or not so nice the width and EHT can be adjusted by series resistors in the anode supply to the boost diode. The STC VC1 and very the early production Pye model 11 used this method of width control.
Sorry I can't post up pictures at this time because the old Kodak C613 camera doesn't allow the transfer of pictures to the computer.

Till Eulenspiegel.

[PYE625]

I can't help but keep looking at the photo of the LOPT and thinking along the lines of my modification to the insulation of the Bush TV125 LOPT that I carried out.....

[Till Eulenspiegel]

The style of construction of the DM45 line output transformer is similar to the LOPTx that was fitted in the DR123 dual standard model. The line output transformer failed in that set and was also replaced with a jelly pot.
Last night the replacement transformer was fitted into the DM45 and now the next stage of the conversion is connecting up the new part.
The jellypot had to positioned in the screening box to allow sufficient distance between the tripler and the surrounding metalwork.
Sorry still no pictures.

Till Eulenspiegel.

[Till Eulenspiegel]

NEWS FLASH! Nice healthy spark at the EHT connector of the ex-BRC line output transformer. Now it's time to fit the EHT doubler and find out what the results are like on the screen.

Till Eulenspiegel.

[Till Eulenspiegel]

The EHT doubler is fitted and there is a very bright picture on the screen, albeit a very wide one. The scanning coils in use at the time I'm writing this are not the original Decca ones. I'll reconnect the Decca coils and report back the results.

Till Eulenspiegel.