Category: Uncategorized

Microsynkie v2.0

Unfortunately our favourite input signal conditioner IC, the MAX7450 is end-of-life’d and gets hard to find. For the Synkie to work properly, all video input signals must have their black level DC clamped to ground. the MAX did this very nicely, but for future revisions we must find an alternative. So I tried once again to build a decent input clamp circuit on my own. This led to a complete redesign of the Sync-Splitter and Resyncer modules, and finally to a new attempt for the Microsynkie which incorporates both modules on a small PCB.


Input Section

Analog video input signal is limited for protection between the power rails before it is buffered and AC coupled. A LM1881 extracts all interesting timing signals which are fed to an analog switch. During H-Sync and Backporch we pull C15 to ground – effectively removing the negative going sync tips and clamping the backporch to ground. The result is buffered again and ready to be messed around with (Siff signal in Synkie-speak) The remaining switch lets us extract the colorburst neatly centered around ground as well. V-Sync can be used as a simple means for showing that there is a video signal present on the input. 

Output Section


First the potentially very weird signal that is coming back has to be clipped to meaningful video levels, that is 0V-0.7V. The following circuit seems to do the trick, this time without any potentiometers to set the clipping level. Thanks Max!


Another MAX4053 triple analog switch is used to fake a normal video signal. A simple resistor divider creates about -0.3Volts that can be inserted as sync tips during the H-Sync Phase. Colorburst is cut back in during the backporch and the clipped SIFF-Input signal during the active video period.

We tried to filter out some of the resulting transients with a low-pass filter before the output buffer, some nasty spikes can still be seen on the oscilloscope but even our very picky Analog Video To SD Recorder doesn’t seem to bother. 

Color Phase Corrector

Before feeding the colorburst back to the signal, an all pass filter lets us delay the burst phase to correct for the time delay of the messed-with SIFF signal in order to get correct colors again (or any color at all)


Backporch Clamp Notes

Connecting two different power sources through a DC connection is dangerous, and usually forbidden by safety regulations. Consequently, video-equipment makers have a tacit agreement to AC-couple the input of their equipment, and DC-couple the output—requiring the next stage to re-establish the DC component (see EN 50049-1 for PAL/DVB [SCART] and SMPTE 253M section 9.5 for NTSC, which allow a DC output level). Failure to establish such a protocol leads to “double coupling,” where two coupling capacitors appear in series, or to a short where there are no capacitors. The sole exception to this rule is battery-powered equipment, such as camcorders and still cameras, which AC-couple the output to minimize battery drain.



Biblioteq Mdulair is an orchestra made of some 15 analogue function generators for four hands played by polish/swiss/french duet Emma Souharce and Daniel Maszkowicz.

With its oscilloscopes, sinusoidals, and frequency sweeps, Biblioteq Mdulair is a sound installation producing all kind of waveforms, exploring vibrations, tickeling resonances, and creating breathing beats. Those primitive electronics machines bring the soundspace down to a magma of waves for a dizzying acoustactile experience.

Biblioteq Mdulair is mostly a music instrument, but several collaborations with the analogue modular video synthesizer SYNKIE, built by collective [a n y m a], created a total audiovisual symbiosis, namely while performing at CERN, European Center for Nuclear Research, and at Cave12 in Geneva.

More Info:


Some calculations

time to charge a capacitor:
t = -ln(1-Uc/Ucharge)*RC

Uc = 1V
Ucharge = 5V

so we have t = 0.22314355 * RC

horizontal ramp: 64uS -> 0.0000064 = 0.22314355 * RC
RC = 0.00028681088753
47nF, 6.1 kOhms
vertical ramp: 20 ms (1 field) -> 0.02 = 0.22314355 * RC
RC = 0.08962840235449
10uF , 8.9 kOhms -> 10k pot

does not work: ATTENTION the 4053 can have up to 2k5 Ohms ON resistance. we have to discharge the cap in max 9uS – length of h-sync + back porch
0.000009 = 0.22314355 * 2500 *C -> 16nF max

also: the v-ramp we want to charge only on h-sync, so that’s 313 pulses of 10.4 uS, total charge time of 3.255 ms. Discharge in max 0.5ms.
discharge: 0.0005 = 0.22314355 * 2500 *C -> max 890nF
charge: 0.003255 = 0.22314355 * R * 100nF, R = 145.8k

Linear Technologies 1251 and 1228

This monday I just got some samples of the LT1251 – 40MHz Video Fader and DC Gain Controlled Amplifier and of the LT1228 – 100MHz Current Feedback Amplifier with DC Gain Control and I got really excited, because that’s probably two IC’s we’ve been looking for for a while. Although we have our own discrete VCA module that seems to work quite nicely, an integrated video capable VCA could open a lot of new possibilities… Building a 6×4 matrix mixer would take 24 sk10 modules, for example. I would also love a voltage controlled version of the sk30 gamma module, and for VCF’s an integrated VCA seems almost always necessary. We tried our hands on the Rohm BA7655A Video VCAs, they are nice but need to be AC coupled, which defeats the purpose to be able to interchange CV and video signals. Our current 6-channel mixer built on these chips also seems to have some serious phase shifting issues (which could also stem from the really wild pcb layout I did, but anyway…) I really start to dislike the rohm vcas – to really be able to fade to black you need to put both in series, which means another ac-coupling stage, and the whole mute-if-cv-is-1V shit means a lot of control circuitry around it to prevent them to go to white, because when they’re muted they don’t output black but white! who wants something like this, anyway?