EEVblog #554 - Sinclair FTV1 TV80 Flat Screen Pocket TV Teardown

[Macbeth]

The schematic shows it directly tied to 1440V DC.

Didn't have a chance to check the schema, just me musing on how on earth it was done? The charge on those plates couldn't be changed so quickly (think big capacitor), so I guess the clever stuff was all done in the Ferranti and hand tuned using all those trimmer pots. It certainly was not as simple as an X/Y CRT like used in an oscilloscope, though not as complex as the driver circuitry, magnets, inductors, and shuffle rings used on a colour CRT.

Thinking of oscilloscopes with vector rather than raster scan, I always wondered how the old Vectrex portable arcade game managed it in such a small form factor. I was amazed to see vector graphics back in the 1980's on such a machine - but it was way too expensive for my family, never mind the cost of the game carts. I loved the old Atari vector graphic arcade machines like Battlezone, Asteroids, etc. I notice Vectrex are still pretty darn expensive as '80s nostalgia on ebay.

Aww, heck, even in the early 1990's I was amazed when I saw Hameg CRT storage scopes with proper vector drawn readouts for timebase and volts. I had only learned on very badly calibrated Telequipment D61's at college back then.

[Terabyte2007]

Awesome! Gotta love the 80's. 

[tecman]

Sony's flat CRT for the "Watchman" TV sets was very unusual, in that it used electrostatic deflection for one axis, and electromagnetic for the other. The pole pieces for the magnetic axis were ferrite bars enclosed inside the envelope, with only the coils external.

Actually the Sony series used magnetic deflection for both axes.  The other interesting feature of the Sony and I assume on the sinclair as well is a trapezoid signal applied to the horizontal deflection, based on the vertical scan.  In operation the bottom of the screen, closest to the gun, required more deflection than the top, since it is a shorter distance.  They modulated the sweep to apply this compensation.

paul

[EEVblog]

I did not know this thing ever existed, amazing! And yes using the horizontal output stage as a power supply is very common in CRT screens.

Indeed. I am quite surprised Dave didn't quite know this.

Why would I?
I have almost zero experience with TV's, monitors, or CRT's in general.
I don't come from the consumer repair or monitor design industry.
Never had to design one, never had to repair one.

[EEVblog]

I agree. I was thinking of the exact same thing all along. Surely the repeller had to have a swept EHT charge linked to the flyback?

No, it's a fixed voltage field.

[N2IXK]

Actually the Sony series used magnetic deflection for both axes.  The other interesting feature of the Sony and I assume on the sinclair as well is a trapezoid signal applied to the horizontal deflection, based on the vertical scan.  In operation the bottom of the screen, closest to the gun, required more deflection than the top, since it is a shorter distance.  They modulated the sweep to apply this compensation.

paul

The tube I was referring to was the original design, Sony designation ED15. Electrostatic in vertical, electromagnetic in horizontal.  The internal ferrite bars were used as both the horizontal magnetic pole pieces, but also as the vertical electrostatic plates.  Covered in Pete Keller's CRT book, along with the Sinclair tube, and dozens of other obscure and downright bizarre CRT types developed over the decades.

http://www.amazon.com/The-Cathode-Ray-Tube-Technology-Applications/dp/0963155903

[N2IXK]

Why would I?
I have almost zero experience with TV's, monitors, or CRT's in general.
I don't come from the consumer repair or monitor design industry.
Never had to design one, never had to repair one.

This is the whole reason that the horizontal output transformer (Line Output or LOPT) in a CRT set is commonly referred to as the "flyback" transformer.

The horizontal sweep waveform is a sawtooth. At the end of each scan line, the horizontal output transistor/tube gets suddenly cut off, and the collapsing magnetic field in the yoke winding and transformer core creates a high voltage pulse when the beam "flies back" to the left side of the screen. This pulse is then rectified (originally with a simple diode tube like a 1B3 or 3A3, later with a Cockroft-Walton multiplier housed in a potted "tripler" module, and eventually a diode string integrated into the transformer itself)), and used to provide the second anode or "ultor" voltage for the CRT.

My first real job in electronics (during high school) was part-time bench tech at a local TV shop. Learned how to fix the things as a source of extra income as a kid. I was always scrounging dead sets from the curb to strip for parts for building stuff, and eventually read enough and played around with enough different sets that I could fix quite a few of them for little to no cost, and sell them at garage sales/flea markets. Of course, this was in the days that there was actually a market for used TV sets. Got out of consumer electronics just as the first wave of "disposable sets" (single PC board and all plastic cabinet) started to hit from Japan and Taiwan. The writing was on the wall back in the mid-late 1980s, and the consumer electronics repair industry is pretty much dead today. Glad I headed for greener pastures (broadcast/industrial video, then scientific instrumentation) when I did, that's for sure.

Still like to play around with vintage video as a hobby, though.  Am currently restoring a 1948 RCA Victor 10" TV....

[EEVblog]

My first real job in electronics (during high school) was part-time bench tech at a local TV shop.

My first full time job at 17 was as a repair and testing tech on security gear, so mostly digital and processor stuff, but that included video, and slow scan video too. So I knew all about video signals, but I've never touched a TV CRT, and never had any real interest in them personally.

[RupertGo]

The correction waveform needed to create a linear picture on the Sinclair flat tube phosphor was extremely complex. Company legend had it that the the engineers responsible modelled it by ray-tracing the electron beam in software running on one of the company's ZX80 3.5 MHz Z80 computers, but the process was unsurprisingly quite slow. A quick calculation showed that it would take longer to complete the task than was available in the lifetime of the project (in fact, had they known it, the lifetime of the company)... so they ported the code onto a Cray supercomputer they rented time on. Thought to be the most dramatic hardware-based speed-up on record.

Sinclair really wanted to use that flat-screen technology in everything, but he could never make it work in colour (the closest he got was a rather odd red/green device that showed everything in a kind of sepia) or any bigger than the size in the TV. There were experimental 4" tubes, but the trouble there was the very large expanse of flat glass with vacuum on one side and 1 atm pressure on the other. You were advised to wear safety glasses if you came anywhere near them...

[Frantone]

Hey Dave - good for you moving comments here! 

I noticed something familiar in your EEVblog #554 about the Sinclair flat screen - the CRT tube looks as though it uses the same kind of C-clamp type spring clip lead for routing the power to the inside of the tube as that which the LVDC surface mount devices used. 

It was in my part two LVDC teardown that I accidentally discovered that those logic device leads were held by spring tension to deposited metallic contacts on the top and bottom edges of each ceramic wafer rather than soldered there.  I found that the packages could be slid out of the leads.  Who knew?!   The Sinclair tube appears in the video to be using this same kind of thing, it appears that way anyway.  First used for the Saturn V!   

http://www.frantone.com/designwritings/design_writings7.html#LVDC2

[silicon junkie]

Just looked at the vertical and horizontal drive signals on a scope, they are stepped as if they are coming from a D-A.
Perhaps this is how they got the scan correction sorted, they use a look-up table to adjust the waveform shape.
The repeller voltage is a fixed DC voltage with a little bit of line scan ripple.

[EEVblog]

It was in my part two LVDC teardown that I accidentally discovered that those logic device leads were held by spring tension to deposited metallic contacts on the top and bottom edges of each ceramic wafer rather than soldered there.  I found that the packages could be slid out of the leads.  Who knew?!

Presumably the reason is an anti-vibration type mechanism so as not to crack the joints?

[RupertGo]

Yes, I'm (almost - been years) sure that the scan waveforms come from a precomputed LUT in the IC driving DACs. I don't know how big that table is - be really interesting to find out the details.

Frantone, I had no idea those clips (which pop up all over the place once you look for them) were a Saturn V design! I know that a lot of packaging and basic logic ideas came from the Apollo project but I've never seen a comprehensive list.

As for the LOPT transformer also producing the EHT for TV CRTs - I always thought of those as the precursor to SMPSUs, but then as one of those kids who fixed TVs as a summer job during their school days, TV circuits were the first seriously complex electronics I came into any sort of contact with.

When the idea was introduced (I think in the 50s), it was certainly a big step-up (sorry) from the mains transformer + massive overwind that did the job beforehand. As was the adoption of the voltage multiplier, which meant you weren't directly rectifying 25KV (in the case of early colour) and thus not producing unearthly amounts of X-rays. Those early sets were a lot more like particle colliders (with the gun facing your living room!) than is comfortable to think...

[Frantone]

Presumably the reason is an anti-vibration type mechanism so as not to crack the joints?

That's an interesting observation and perhaps true - I really don't know honestly.  I was very surprised when I found this out.  There were hundreds of pages in the system and thousands of logic devices - and tens of thousands of pins.  I would imagine the only reason would be stability - though I would think that some kind of through hole soldered pins in the ceramic base would be preferred, as they did use through hole pins for the IBM system 360 SLT device packages.   But these clip leads obviously do work and were used for some good reason - the Saturn was designed by very very smart people -  and they obviously were the choice for that glass tube! 

[N2IXK]

Pretty sure that that basic contact design predates the Saturn V by a decade or so.

Have seen those type contacts used on early thick film RC networks called "Couplates" or "PECs" (Packaged Electronic Circuits), which were ceramic wafers with several resistors and caps integrated into a single module, used in a lot of 1950s era tube gear for coupling networks, filters, integrators, etc. Nowadays, you still see them on thin film resistors/dividers, such as the rectangular ceramic ones used in DMM input stages.

I think it is a fairly standard way of attaching metal leads to a glass or ceramic substrate. The spring action is likely to compensate for the differences in thermal expansion rate between the metal and the substrate.

[JackOfVA]

I assume the mention of "D. Gabor" a university professor, was an inside joke. 

That would be Dennis Gabor, Nobel Prize in Physics for holograms. He also worked with CBS Laboratories (the US broadcast network) and I believe that relationship is the genesis of the flat display screen technology used in the Sinclair TV.


 

[N2IXK]

The Sinclair tube seems closer to the Aiken design, with the electron gun to the side of the phosphor screen, rather than below it, as in Gabor's design.

http://en.wikipedia.org/wiki/Aiken_tube

Both the Aiken and the Gabor designs are also covered in the Keller book.

[PChi]

Before anyone gets too nostalgic about Clive don't forget that the general public forked out cash for developing Sinclair  products. The paperback 'Sinclair and the 'Sunrise' Technology' by Ian Adamson and Richard Kennedy is a good read.