Siemens

[BJoe]

Pred casem tu probehla diskuse na tema prilohy.
Myslim, ze prilohy jsou z mnoha duvodu prakticka vec:
1. Na rozdil od uschovny zustavaji v archivu neomezene dlouho.
2. Clovek si muze postu stahnout, odpojit se a pak vsechno
prohlednout, zniz by se musel znovu pripojovat (pokud
si dobre vspominam, tak delka stahovani byla hlavnim
argumentem s ohledem na modemisty). Nekolik sekund
stahovani urcite prijde levneji, nez se nekolikrat znovu
pripojovat.
3. Protoze mam pevnou linku (kabelovku), tak me 2 zas
tak moc nezajima, ale je docela prijemne, kdyz si clovek
muze postu i vsechny souvisejici materialy prohlednot
treba na notebooku pri ceste vlakem.
4. Vetsina mailovych klientu treba obrazky rovnou zobrazuje
za textem.

Jedinym problemem je delka priloh/souboru v uschovne,
coz je ale spise lenost autoru. Obecne staci po oskenovani
jeste aplikovat par filtru, zmensit a ulozit jako GIF. Vysledkem
je cisty obrazek, ktery zabira par kB. V pripade potreby
lze doprovodny text prevest s pomoci OCR do TXT a je hotovo.

Ze to takhle opravdu jde, lze dokazat na prilozenych souborech, jenz
jsem ziskal z puvodne 329 kB dlouheho JPG, ktery byl v uschovne.
Obrazky byly upraveny v PaintShopu (asi za 5 minut), text byl
ziskan s pomoci OCR Readiris a nasledne mirne korigovan.
Prevod trval asi 3 minuty (mam WinChip 200 MHz),
korekce asi 4 min.

Spise nez protestovani proti priloham by se melo dbat na lepsi
kulturu zprav, jako zbytecne necitovat cele zpravy (zejmena
treba texty jako "Odchozi zprava neobsahuje viry..."), nebo
odpovedi stylu "Ano, souhlasim", jenz zadne nove informace
neobsahuji a neposilat do konference soukrome zpravy ve stylu
"Vojta zdravi Frantu", ktere opravdu nikoho jineho nezajimaji...

Pavel Hudecek
-------------- next part --------------
Figure 2.. Circuit diagram.

nique so that the Y component of the recorded signal will have a better bandwidth.
The circuit block diagram is shown in FIgure 1. The C part of the signal is cleanly filltered out from the composite signal using an RLC high pass filter circuit while the Y part uses an RLC high or bandstop filter to remove the C information. Transistors T1 and T2 act as buffers to produce low impedance Y and C outputs.
The circuit diagram shown in FIgure 2 closely follows the block diagram layout. R2, C1 and L2 form a high pass filter to recover the C signal. The filter output is coupled via C3 to the base of transistor T2. This transistor is a common-emitter configuration and buffers the filter while providing a low impedance output via C4 to the S-VHS connector.
The bandstop filter used in the Y path is made up of R1, L1 and C9. L1 is again a standard 47 uH fixed inductor while C9 is an adjustable trimmer, this allows some optimisation of the picture. The Y signal contains frequency components going down to 50 Hz so the signal coupling capacitors C2 and C7 need to have a larger value than in the C signal path, otherwise low frequency parts of the signal would be attenuated. The ac-coupled low impedance signals C and Y are output on pins 4 and 3 respectively of the S-VHS connector K2. Two supply decoupling capacitors C5 and C6 complete the circuit.
The input and output impedance of this circuit deviate from the optimum value but were chosen to ensure that the output signal levels are correct. At 75 Ohm the signal levels should be 0.3 Vpp (burst-amplitude) for C and 1.0 Vpp for Y.
The layout of the PCB is shown in FIgure 3. As in all RF circults it is important to keep the component leads as short as possible when fitting them to the PCB. An earth plane surrounds the tracks so a little extra care is required when soldering to ensure that no solder bridges are formed. In place of a genuine Hosiden S-VHS socket we have specified a standard mini DIN socket for PCB mounting. It will happily accept a Hosiden plug as long as the small plastic stop pins are first removed from the plug.
During set-up, trimmer C9 should be adjusted to produce minimum Moire (colour patterning) effect on the picture. The circuit was successfully tested on many different S-VHS inputs. Generally the bandstop filter in the Y path was necessary but with a WinTV card it did not make any noticeable difference if the filter was in or out. In this case you could simply bypass the Y circuitry and just use the composite video signal as the Y signal.
Finally a reality check, as mentioned before, the quality of the S-VHS compatible output signal generated hare cannot be any better than the composite video input signal. To get real S-VHS picture quality, you do of course need a 'genuine' S-VHS signal source connected directly to the Y/C input of the monitor, TV or recorder.

COMPONENTS LIST

Resistors:
R1,R2 = 2k2
R3,R4,R6 = 120R
R5,R7 = 47k

Capacitors:
C1 = 33pF
C2 = 100uF 1OV radial
C3,C4,C6,C8 = 100nF ceramic
C5 = 4u7 35V tantalum bead
C7 = 470uF 10V radial
C9 = 25pF trimmer capacitor (3-25 p)

Inductors:
L1,L2 = 47uH

Semiconductors: TI,T2 = BC550B

Miscellaneous:
KI = Cinch socket, PCB mount, e.g., T-709G (Monacor/Monarch)
K2 = 6-way Mlni-DIN socket, PCB mount, angled pins
-------------- next part --------------
A non-text attachment was scrubbed...
Name: S-VHS_3.gif
Type: image/gif
Size: 16357 bytes
Desc: not available
Url : http://nethouse.cz/pipermail/hw-list/attachments/20011120/5cb5ca31/S-VHS_3.gif
-------------- next part --------------
A non-text attachment was scrubbed...
Name: S-VHS_2.gif
Type: image/gif
Size: 10083 bytes
Desc: not available
Url : http://nethouse.cz/pipermail/hw-list/attachments/20011120/5cb5ca31/S-VHS_2.gif