Project [Sticky] 1969 Philips G22K511 Missing I/F & LOPT, Conversion

Ok, I’ll carry on and add the next part which will be the sandcastle pulse generator. This will be adapted from the Luke Theodossiou Colour project in July 1981 PT mag, it uses the TDA9503 which was also employed in the TX9. Component count isn’t too bad.

Here you go. Hope this is not getting too confusing with the duplication of components numbers. If you have any queries on components such as R1 let me know which stage we are talking about eg IF. Note that C8 in the sandcastle stage is a critical component.

Thanks, John 👍 

Can I ask is there anything more we need to design in after that? For example is the Field blanking circuit needing anything? 

Appreciate you're doing this as and when time permits, and releasing to me, so I have something to get on with. I'm happy to wait. I like the modular way you're showing the circuit as It really helps me transpose that onto a matrix board for point-point wiring. However, the reason I ask is I need to plan out on the board and know what available space I have or more importantly, don't have. I'm reluctant to start fully building and populating, only to find out later I've run out of space as I needed to add more circuitry.

Can I also ask about the contrast, brightness and colour circuits as shown? I take it we're using the existing G6 user controls and all we're doing is modifying the associated passives?

Just so you're aware I've yet not made any of the required mods (component removal and the added link) to the decoder, that's on the to-do list.

Once this is all built I'll be shocked if it works, not because I doubt your design, I doubt my ability to have interpreted it all and transposed it correctly onto the boards. I suspect I'll be fault-finding man made errors until Christmas. 🙄

That’s all the main items covered now I think. Getting the required pulses from and interfacing syncs to the existing G6 circuitry will require a lot of experimentation but I hope your existing component stock will provide the required items, nothing I believe that isn’t off the shelf.

Regarding C8 in the sandcastle stage, have a look at the style C158 in your TX9 If my memory is correct it was a small round ERO type but wasn’t particularly unreliable. If any of your stock came from a Thorn dealer and you have a TDA9503 you may have one or two caps in stock.

I’m not sure if the values of the customer pots in the G6 are suitable, I’ll have to check that.

As I said in an earlier post, do some testing with 12v power applied along the way and you certainly should see some meaningful waveforms especially around the TDA9503.

Most important job is to get the G6 to generate timebases and EHT before we can even attempt at interfacing the finished IF/decoder.

Posted by: @jayceebee

If any of your stock came from a Thorn dealer and you have a TDA9503 you may have one or two caps in stock.

I do have a few TDA9503 in stock and yes a lot of my stock did come from a TV shop that handled Thorn products.

I can't visualise the cap you mention, I may have it, that leads on to another query.

Posted by: @jayceebee

have a look at the style C158 in your TX9 If my memory is correct it was a small round ERO type but wasn’t particularly unreliable.

Can you explain, so I can understand, why does the cap has to be so specifically constructed in polystyrene. I can understand the value being critical, the voltage required being a minimum of 63V, but not its construction. My suspicion being it's to do with line pulse high frequency? But if it that is true like a line tuning cap, surely polypropylene is just as, if not more suitable than polystyrene.

Posted by: @jayceebee

I’m not sure if the values of the customer pots in the G6 are suitable, I’ll have to check that.

Had a horrible suspicion that might be the case. The reason that's a pain in the butt is its yet more of the G6 disappearing. You can guarantee if I have to replace the pots the G6 knobs won't fit. Then starts another exercise in how to make them fit, and I'm not replacing them with something else.

Anyway I'm likely to go quiet on this now for a few weeks whilst I build and test. I might also start another parallel project on the go to get a break from it as I'm itching to tackle one of my continentals or that little 719 chassis set.

For C8 it’s all about precision and temperature stability, it’s the timing capacitor for the 15,625Hz oscillator in the IC. Their properties are akin I believe to silver mica but with much higher values.

I've not looked in my TX9, but checking my stock I believe these are the exact ones you describe.

Wrong value I’m afraid and 10% tolerance, ones in the TX9 were cyclindrical radials.

<face-Palm> Sorry John, I misread the circuit, for some reason I thought 10%.

Well, I've all manner of ERO caps but none are 1% jeez!!! That's tight. Finding those will be a nightmare. Closest I've got is a polystyrene 2%

I must be having a grumpy day today as I'm stating to get well miffed/irritated by this project, must be due to turning 58. I best put it to one side until the cloud moves on.

I suspect that will be fine.

On google I’ve just seen a bargain of 50 @ 10nF 500v 1% mica reduced from over £700 to £420 🤣 🤣 🤣 🤣 

Yes indeed, looking online I stumbled upon the Audiophool sites too.

Chris, Farnell to the rescue.

https://uk.farnell.com/lcr-components/fscex-10000pf-1-63v/cap-0-01-f-63v-1-ps-through-hole/dp/9520414

£1.95 plus VAT each and they're in stock.

The G6 LOPTs arrived today, to say they're well past their first flush of youth, is an understatement. Still I'm a "Glass half full" man rather than a "Glass half empty" as maybe now I've a fighting chance now to get the G6 time bases spluttering to life. I'm very grateful for the kind donation to the cause of getting this G6 up and running in some shape or form.

There appears to be two complete LOPTs and then a spare overwind and primary. The two complete are 1x single standard 1x and dual standard, with the primary/overwind bundle, also being from a single standard.

I base this supposition on the fact one LOPT having two peaking coils, one for 405 the other 625? Whereas the other LOPT and spares bundle only have the one peaking coil, I'm hoping that makes sense.

So in essence as I'm looking to convert this SS G6 to tripler EHT, I have three primaries at my disposal, hopefully one or all will be OK.

Nope my assumption was wrong ,I should have read the manual before posting.

It looks like I've got one shot at this, despite two and a half lopts in this bundle I've only one set of L5514 and L5515 coils. All three have L5520. This is all tuning the 3rd Harmonic. Unless.......What I'm not sure about is due to me removing the EHT overwind and going for tripler EHT generation instead, whether I need these coils at all. Can I tune the third harmonic with a polyprop cap.

Another way to see it is that what you've got is a lot better than what you had!

It's been 40 years or so since I last did anything with a G6, but I have the distinct feeling that you'll need the harmonic tuning coil, so treat it with kid gloves, those things break off easier than breaking eggs! - John or David will be the ones with the low down modding the LOPT.

Just out of curiosity, have you rung those LOPTs?  It would seem logical to do that first and see what you've actually got on the bench, but if it came down to the very worst case scenario, you now at least have the essential hardware to allow yourself an attempt at rewinding one.

Other than that, you look like you've got the bits you need to make a start at piecing one together. It's a pity that the junction insulation sleeve isn't there, otherwise you could have gone for the full on GY501/PD500 rectifier/regulator set-up. You probably still could, but you'd need to work out how to mount the junction, and some way to insulate against junction to chassis flashover. The missing metal cover is the easy part to replicate.

I've rung just one on the scope and for good measure verified it on the JABCO. The coils L5514 and L5515 had already broken away from the plastic mount and the wires from the paxolin board. I've remounted them and secured with some heat-shrink and a cable tie. I test the others later.

I've long made my mind up the Overwind is going. The reasons, this TV is a complete basket case, the amount of carnage and bodgery on what remains of its panels, is quite extensive. Secondly I'm not working on this set trying to diagnose a multitude of existing panel faults and also trying to fault-find my I/F decoder and RGB panels, with a honking 25kV EHT X-ray generator a few inches away, its just not going to happen. Final reason for going tripler, I want to run this set in my 70s lounge from time to time and the chances of a tripler surviving hours of use far exceed those of these questionable G6 lopt overwinds. The origin of these lopts also suggest they are all likely pulls. Coupled with that, they've all been poorly stored so have broken parts and missing parts. Hopefully one can be made up to work with a tripler, likewise they could all fail on the primaries soon after re introduced to volts.

That looks good, on test. Might it be worth baking that one whilst other areas of the project are being worked on?

I can sympathise with the desire to go for a tripler, but although people oft used to bemoan the reliability of early colour TVs, I used to run my own G6 from tea time to bed time every day of the the week, back in the day, but also, in the summer months, I used to run it with the back cover removed!

I have to say though, I have seen those overwinds with hairline cracks in the resin, and although still working, nor would I trust them. I guess they'd work for me because I was there, but hand it over to someone else, you could guarantee it would fail!

Attempting to make a silk purse from the pigs' ear of LOPTs, to get one decent one as they all have damage. All the primary tag boards with the LOPT leads were broken or with sections missing. Mercifully all the LOPT winding lead-outs are intact, phew! I've epoxy'd the snapped sections on two, then cut up and fashioned a piece of paxolin to repair the solitary primary winding. I've also managed to piece together the broken board that held the 3 peaking coils, three pieces glued together, it ain't pretty but needs must.

I've selected the one which looks the best and have removed the overwind, boy that was not fun. Now I'm just waiting for them to set. Later I'll test the other two primaries then assemble. I think I'm definitely going to take the 6kV of the top cap of the PL509 and build either a quadrupler maybe also try those BY8410's.

The left LOPT has been chosen as the one to run with, the one on the far right is a spare overwind. The middle one will be kept as is, unless its primary is required.

I really need to prove what exists of the G6 can be powered up and see importantly see if the timebases are even possible to get running as this set is a real dog. To that end I also need to get the lopt sorted and that means sorting out the EHT. If none of these function or can be made to function there's little point investing time and energy on further development of the I/F decoder and RGB panels.

Can anyone advise me on my cap diode ladder, is there a calculation to enable me to determine the outcome as I'm a bit unsure. I want to take the 6kV pulse from the TC of the PL509 input that into my diode/cap ladder and end up as near to 22-24kV and no higher.

Another thing I'm unsure about is where I tap off for the focus unit, which should be around 5kV.

The BY182 are 12kV 2.5mA, the caps are .001uF 15kV. The efficiency of these diodes I believe may get me very close to the desired minimum output so I may not need to build a quadrupler.

One LOPT has been fashioned from the best bits.