Chinese 5.5 inch B&W TVs

[FERNSEH]

The problem with driving the line output stage of a 625 only receiver with a 405 waveform will be too much EHT.
The line output transistor acts like a switch, when it is turned on the current in the line OP transformer ramps up to a point where the drive waveform ceases rapidly. The magnetic field in the transformer collapses and so induces energy into the flyback winding. It's that -E = L. di/dt thing, the faster the current change, then the greater will be the induced volts. It follows that on 405 lines the ramp up time will be longer, hence the EHT will be greater.
So the answer might be to wind on a number of turns on the transformer core, short out the leads. I have no idea how many turns will be required, that's something you will have to experiment with.

DFWB.

[neon indicator]

Or a suitable series resistor on +12V HT line to LOPT coil/transistor to reduce current and thus energy stored and thus flyback pulse size.

[murphyv310]

This tends to cause non linearity in the scan, due to the varying current requirements.

[ParcGwyn]

If you are interested I have attached a schematic.

Dave

GW7ONS

www.parcgwyn.talktalk.net

[neon indicator]

Nice ParcGwyn

Should indeed be easy to mod without adding windings or modifying LOPT.

A pot at R405 / Q7 to adjust Width/Drive

Add 50% across C401, C403 at the Horizontal frequency pot

More capacitance across D9, D10 at the LOPT

L8 in series with Scan coils might need changed or may be OK with drive adjusted.

[murphyv310]

No No No.
You cannot reduce the drive in this way, the driver needs sufficient signal to enable the line o/p transistor to perform as a switch. doing this will change the operating conditions and the transistor will fail.
Can you explain your reasoning for adding capacitance across D9 & 10 please.

[FERNSEH]

Yes, Trevor is right, the line O P transistor must be turned on hard to function as a switch. The turn on and off drive waveform must have sharp edges, otherwise the transistor will operate in linear mode if the waveform has soft edges.
A saturated transistor dissipates low power when turned on hard.

DFWB.

[Red to black]

the driver needs sufficient signal to enable the line o/p transistor to perform as a switch.

Trevor is absolutely spot on here, to keep the switching devices total dissipation in W within the spec. of the device, (Ptot).

When the transistor is switched hard on, it is passing maximum current but the voltage across the device is (or should be) very low, V*I = low Ptot.

When the device is fully switched off, it will have maximum voltage across the device but passing very low (next to no) current through it, again V*I = low Ptot.

The maximum Ptot is when the device is at the halfway point, = 1/2 I*1/2V.
It also dissipates to a lesser extent at either side of the halfway point, and anywhere in between the two states

It is important in both Line output stages and switch mode power supplies that the switching device switches between hard on, to hard off, as quickly as possible, ie. to spend as short as time as possible between the two states (on and off) to keep the Ptot down.

One reason for failure of these devices in both applications, is due to 'lazy drive', there are of course many other reasons for failure.

Cheers,
Baz

[murphyv310]

Personally I think Neon is not grasping what we really need here. We Have an excess of energy that we need to dissipate. We know that the LOPTx is designed for 625, we don't have the option of various tapping's to play with, so if the core of the transformer is available to wind 20 or 30 turns of 18 swg wire around it is then a simple matter of shorting the winding out, this of course can be tried on 625 firstly, a substantial drop in EHT and lack of width is what we are looking for. This is precisely what I have done on my old 625 line monitor and it works quite well with the exception of some slight non linearity.

Why try and get a complex result when an easy and simple way to do the job will work and work well?

[neon indicator]

Yes, of course the drive has to switch the transistor quickly and fully on.

Why not try a combination of resistor load on one of the LOPT windings and series resistor on LOPT supply to reduce the on current a bit?

[dominicbeesley]

Would a starting point not be to break the circuit at R6 and bring the LOPT stage up gradually on a variable power supply with the driver chip's oscillators running at 10k (tweak caps and adjust, I'm pretty sure Karen't got a scope)

The main problem I had when I tried this with normal LOPTs was the core saturating (lack of or excess EHT could be twiddled out as could the width) but when the core saturated things got hot and went pop. Core saturation is a function of time * voltage as the on time is now about 50% higher you'll need 75% the volts to avoid saturation?

So my plan would be work out what voltage we can bare to drive LOPT at 10kHz by slowly increasing LOPT "HT" until transformer makes a racket. Or just go for 7.5 V and hope.

Knock half a volt off for good meaure then rig up a 317 voltage regulator or a few series diodes to drop however many volts between the 10V HT and whatever we want and put them in series with R6.

I can't remember but I seem to remember that when I did this then the EHT wasn't far off spec, though messing the with tuning capacitor made a fair difference.

While doing this it might be a plan to move R406 over to before R6 so the driver tansistor and LOP transistor bias are always working from the full 10V?

The harmonic tuning of the LOPT will need tweaking but is that not just a case of increasing the tuning cap (C411) by 1/f2 == 2.4 times as big?

Or am I missing something as usual!

D

[murphyv310]

Hi Dom.
No you are not missing anything obvious. The harmonic tuning will need to be changed but if you are able to damp some energy from the transformer the tuning will be different anyway.
Why though are we trying to crack a walnut with a sledgehammer?
If the transformers core is exposed (which we don't know as yet) the simple and easy way is to wind a coil of wire around it, shorting it out works or you can damp it with a R or RC network.
My little montor has two resistor changes for the oscillator, a different "S" correction cap, a coil around the TX and a small change in tuning cap value.
I will dig it out of the loft and take a few pictures of the work and the picture.

[dominicbeesley]

Ok, sorry I'd missed that. My inductor knowledge is still shaky but will the core not still saturate? Every time I think I've understood transformers I have to go and sit in a dark room with a cold flannel on my head...anyway, off to play with the ignition coils on my Land Rover...I hope I understand them enough to get it running!

Dom

[Karen O]

Jim,

PM me with your email address and I'll send you the circuit and PIC hex.

Best,
KarenO

[davegsm82]

Just in case anyone wants a look, here is the PDF of the KA2915 IC.

http://www.datasheetcatalog.org/data...ic/mXvvquq.pdf

Looks like audio may be a bit of an issue, since the chip does FM detection, so a separate audio detector may be required.

Obviously a video inverter will be required.

Looks like the line osc should be easy enough to tweak.

I'd have a go with my TV but I don't have any 405 line source.

Dave.

[murphyv310]

Hi.
You don't need either a video inverter nor any changes to the detector if feeding AV in at baseband, which I am sure Karen is doing, please accept my apologies if wrong.

[davegsm82]

Ah, yes, I forgot this was the 'direct' approach.

At the end it would be a nice touch to get the full mod-demod working with the original tuning equipment, that way you could be sure the converter would work with an original 405 set.

Dave.

[terrykc]

I've just been studying the schematic and a couple of oddities stick out.

The first is a 500kHz oscillator, Q13 (top right). Further investigation shows some interesting SAW filters in the sound stages.

(I know that sound isn't the main topic here but the discussion is relevant to the title of the thread, so I hope no one objects to the diversion ...)

If I interpret the schematic correctly, this is a multi-standard receiver for systems D, I and K but, strangely, not B or G ...!

Video output is buffered (Q5) and fed to two SAWs in parallel (U6/U7) which are 6.5MHz/6MHz respectively. The output is combined (Q12) with the 500kHz signal and fed to the main sound IF SAW (U5) which is 6.5MHz.

Thus it appears that system D/K 6.5MHz signals are processed conventionally but system I 6MHz sound is converted to 6.5MHz (by mixing with the 500kHz local oscillator) before being applied to U5.

There is a further 6.5MHz SAW in the feed to the video o/p (Q8) to suppress the system D/K inter-carrier sound beat but system I sound continues unmolested!

I would have expected to see 5.5MHz filters, rather than 6.5MHz, which would have made it a system B/G/I set that one would think would have wider appeal - or was there no demand for these small sets in mainland Western Europe?

Following the same philosophy, choosing 6MHz for the sound IF and adding a third 5.5MHz SAW in parallel with U6/U7 would, of course, have resulted in a system B/D/G/I/K set that would work on any 625 line system in the world - apart from France and South America, of course.

Terry

[Mike Phelan]

LOP operation is as Trevor and David say.
Also this.

[neon indicator]

On re-reading:
My comments on this this thread have mostly been incomprehensible gibberish. Sorry

murphyv310 aka Trevor explaining it properly. ...