Project 1970 Philips G22K511 Missing I/F & LOPT, Conversion
I was wondering about hand winding a new primary for the G6? - I remember having modest success with one, but unfortunately, didn't have any new dielectric tape, and what we recovered from the old LOPT was a bit jaded to say the least.
It wasn't actually very difficult, just a bit time consuming to unwind the old transformer, counting the turns, and set about rewinding anew. The biggest issue was the old tape, which failed and led to an unfortunate conflagration. A bit of a bugger really, because a subsequent phone call to CHS resulted in two BNIB LOPTs - The last two they had, and wouldn't/couldn't be replaced. A bit like the Crown Jewels!
Of course, this was somewhere around 1979/80.
Granddad just said "Should'a made that phone call first. Would'a saved a lot'a time an' effort" - Sage wisdom, if ever I heard it!
The 1967 Mullard book "PAL COLOUR TV..." describes the PAL system and the circuits of a dual-standard colour TV receiver.
Also described is the construction of the line output transformer for the receiver. The core is the Mullard/Philips FX2935 and FX2936. Are those components still available? The main winding 1 -9 has 440 turns of 27 s.w.g. (0417mm) Sims QSE, wound at 48 turns/layer at 32, 48, 95 145and 340 turns. Winding 10 - 13, 68 turns 30 s.w.g. (0.315mm) Sims QSE, wound at 33 turns/layer The "25KV" winding consists of 1680 turns 40 s.w.g. (0.122mm) Sims QSE, wound at 30 turns/layer.
The latter can be replaced with a winding to produce 8kv, essentially a 2KV winding.
it's easy to see that many TV receivers were developed using the Mullard application notes. An example is the GEC 2028.
@till Brilliant piece of work!
I couldn't have quoted the number of turns, but I know it's feasible, and even my effort of 40-odd years ago did actually work. It just failed after a few hours of running. A more carefully wound primary coil isn't beyond the realms of possibility. I'm pretty sure that the LOPT yokes and insulating bobbins would be more or less universally interchangeable. There are some additional components that might be interesting to make, or work around (Harmonic tuning coils etc), and some mounting hardware would have to be fashioned.
It has to be worth a consideration?
Hi Marion, I'm sure it's the properties of the Ferroxcube core that is the key component of the design of the transformer. The real surprise is the few number of turns required to construct the windings. Assuming the actual voltage produced by the EHT winding is 19KV the volts-per-turn is 11. Of course the operating frequencies of 10.125KHz and 15.625KHz helps.
Consider the number of turns required for a 5KV EHT transformer in a pre-war TV set is something like 12000 turns. Again, it's the properties of the core that determines the turn-per-volt figure.
Of course the operating frequencies of 10.125KHz and 15.625KHz helps.
Another helpful factor is that this is also a single standard model, which simplifies the construction of the LOPT unit as a whole.
The low turns count of the primary winding is the one thing that drove me to experiment with rewinding one in the first place. I'd have happily considered a second attempt had it not been for the chance of the new ones from CHS, and that CHS' depot was literally only about 12 miles away (Pocklington).
I think the EHT overwind is the same across all the makes and models that employed GY501/PD500 rectification and regulation. If so, then it would be likely that there will be an overwind out there somewhere, looking for a home.
More information about making a line output transformer for a colour TV set. It's most likely the Ferroxcube transformer core in the G6 is the FX2935 and FX2936.
Sorry I've not added much to this thread of late, it's not that I've been ignoring you all, I've just been melting my head trying to fathom the possibility of using this spare Decca CTV25 LOPT.
Decca CTV25 LOPT, Overwind removed, and converted to tripler, includes thick film Focus unit
It's already converted to tripler which is what I want and was working on the decca TV prior to a NOS lopt being fitted. In addition to that I believe I have a complete set of boards for a decca CTV25, so I should be able to harvest any specific line output coils, caps, inductors required and transfer them into the TV replacing the G6 one's so the decca lopt will work. I think the primary has the right number of taps, but again not completely sure.
I'm struggling to make progress, as I've not only got to get my head around the two circuits as they were but try and imagine them both without the PD&GY, EHT replaced by tripler. Furthermore, I believe the Decca lopt is a direct copy of Hugh Cocks. Anyway, I'm not giving up, yet. I might get there eventually as I think this line of investigation is worthy of pursuit even if it ultimately ends up a no go.
Checked the primary on the scope and at least it looks healthy from the "ring".
Circuit diagram of the line timebase of the single standard Decca CS2550 shows the 2KV winding between the anode of the PL509 and tripler input.
Honestly, I don't how Mr. Cocks go away with supplying the tripler with the flyback pulse from the anode of the line output valve. The fact he got 25KV implies the peak voltage at the PL509 anode was >8KV!
The Ekco CT104 in the workshop just manages to produce 6KV.
The fact he got 25KV implies the peak voltage at the PL509 anode was >8KV!
The Ekco CT104 in the workshop just manages to produce 6KV.
No wonder the little blighters used to give a bit more than a playful nip, if you got too close! All these years I've been thinking in terms of high-hundreds!
Was there ever, and that includes B&W television, a universal, or general purpose, replacement LOPT that could be adapted to individual applications? - Probably a stupid question, but from the foregoing, it looks like there was a general outline/profile for LOPTs, developed by Mullard, and adopted by the set makers.
Hi Marion, in the Philips 21TG100U line output transformer substitution topic in the VRAT forum a GEC 2040 transformer is being assessed. Sure enough, the transformer works in the set but the PL81 line output valve is operating under unfavourable conditions as the anode current during the conduction part of the scan is in excess of 160mA, which is the current the PL505/9 was designed to operate on. In fact the anode of the PL81 is glowing dull red.
Returning to the high peak voltage at the anode of the colour TV line output valve, that 6KV pulse also packs a punch! There's lot's of power there. At the end of the forward scan the line output valve is turned off rapidly and that equation -V = L . di/dt enters into the operation of the circuit.
So it follows what is needed a universal LOPT for 110* CRT mono TVs and another transformer for 90* CRT colour TVs.
Just to be a bit clearer, I was thinking in terms of universal replacements for say, 90 degree colour, 110 degree colour, then say, 75/90 degree B&W and 110 degree B&W, rather then one transformer trying to be all things to all types of sets.
Hi Marion, even if one considers the 90* colour line output transformer first we'll need one for valve timebases and another for transistor line output stages. It's doubtful if anyone has saved the Ferroxcube cores so alternatives will have to be found.
Has anyone tried a Bush A823 line output transformer in a Philips G8?
Hi Chris, I'm going have get my myself a coil winding machine, and one that does wave winding. I'm going to need one if that "smokers cabinet" radio recreation ever gets started. The Marconi RB12 which has those mysterious spade tuned coils.
Hi Chris, I'm going have get my myself a coil winding machine,
Funny you should say that I was looking at this earlier today, only £439 with free delivery from Czech Republic. Might be some duty to pay now we're outside the EU. Not sure what the threshold is, I bought a Grundig diagnostic tool from Holland for £54, it arrived yesterday with zero customs charges. Although the Czech Republic is in the EU, like us when we were in the club, they have not adopted the Euro.
This project is now back on, I've managed to secure a G6 Lopt. Whilst I await its arrival, I'll continue working on the IF board and building the RGB section. However John, can we revisit the RGB circuit, I'd rather not if possible, use the BF460's and 15K collector resistors. Is there anyway we can use the K35 RGB module? This would be neat as we're already using the K35 decoder and the whole thing will run way cooler, thoughts please?
Before anyone asks or suggests, no there is still little or no chance of and I/F turning up. I've contacted all the major known collectors, none have G6 spares. One even stated when he started collecting G6 over 40 years ago, spares were virtually non-existent even then. However, as I've said numerous time before if one should magically surface, and I can get it, then the set will be reverted to spec.
This project is now back on, I've managed to secure a G6 Lopt.
Amazing, how do you manage it?
Shouldn’t be a problem using the K35 unit, just need to see what supplies it needs. I’ll have a look at the connections later this evening.
Hmm, not quite as simple as I was hoping. Requires 155, 140, 13 and -20v supplies, the latter which I suspect is for switch off spot suppression could be deleted and I'm sure the 13v supply will be happy enough with 12. The tricky item is the 140v supply which is also used for a blanking/clamp pulse also, I wonder if it's that critical or is it Philips being Philips? The other issue is that the main circuit diagram shows an earlier two chip decoder and I wonder if there are any other significant changes I can't see. I wonder if Sideband or anyone else with more K35 exposure than me could answer this question, Is the single chip a drop in replacement for the two chip version?
Hi John, as there's been no response/input to your queries, and we're now unsure of the K35 RGB suitability, I've decided to start building my own RGB board to the specifications you stated earlier in this thread.
I've changed my approach, rather than trying to cram it all on the IF/Decoder project board as an additional stack, I've decided to build the RGB circuit on a separate board. This will help with taking the heat away, and It should fit in the available chassis space. This will also be pluggable so can be easily removed. I've got everything I need bar the 680pF caps, the nearest I had was 560pF & 800pF. I wasn't sure on the criticality of the value, so I've ordered some NOS Suflux & Lemco 680pF
I've not populated the space on the right of the RGB circuit board, that will have the contrast, Brightness and colour control circuits. I'm using a neat piece of software to plan the layout on matrix board