Video Splitter from Onvon

Stuff I am working on
Post Reply
Daniel Wee
Site Admin
Posts: 2449
Joined: Wed 25 Feb 25 2009 8:00 pm

Video Splitter from Onvon

Post by Daniel Wee »

I got the video splitter from John today and so I decided to take a peek inside it to see what makes it tick. This is a five-way splitter for one input. First up, some photos:-
IMG_0817a.JPG
IMG_0817a.JPG (352.48 KiB) Viewed 15215 times
IMG_0815a.JPG
IMG_0815a.JPG (336.02 KiB) Viewed 15215 times
IMG_0816b.JPG
IMG_0816b.JPG (338.24 KiB) Viewed 15216 times
As you may be able to tell from the photo, the circuit itself runs on 9V which is provided by the 7809 regulator so the supply should be 10V and above to allow for the voltage drop through the regulator. The input resistance measures exactly 75-ohms and it looks like the individual channels are amplified with a transistorized amplifier. The "1AM" marking is for the MMBT3904L (similar to the 2N3904) NPN bi-polar transistor so we're looking at a very straightforward signal amplification here.

Daniel
Daniel Wee
Site Admin
Posts: 2449
Joined: Wed 25 Feb 25 2009 8:00 pm

Re: Video Splitter from Onvon

Post by Daniel Wee »

As for the performance - the splitter has unity gain when fed into a 75-ohm load. The following trace shows this, where red is the input (DC-coupled) and yellow is the output (AC-coupled through a capacitor so the swing goes negative):-
Video-Splitter-IO.jpg
Video-Splitter-IO.jpg (276.85 KiB) Viewed 15214 times
In short, this thing works - which is no surprise. Would be nice if it would work at a lower voltage which allows it to be powered off 2S though.

Daniel
Daniel Wee
Site Admin
Posts: 2449
Joined: Wed 25 Feb 25 2009 8:00 pm

Re: Video Splitter from Onvon

Post by Daniel Wee »

Okay, I did a little more poking around the circuit and came up with this:-
Video-Amp.jpg
Video-Amp.jpg (89.3 KiB) Viewed 15209 times
The final stage, starting from the 100ohms resistor after the 2N3906 is repeated for each of the output channels. You can also see what looks like a Darlington-pair arrangement at the input. Next, we will try to do some circuit analysis to see how we can modify this thing to run properly on 5V instead of 9V.

We have also measured the running current with one output connected to be around 92mA.

Daniel
Daniel Wee
Site Admin
Posts: 2449
Joined: Wed 25 Feb 25 2009 8:00 pm

Re: Video Splitter from Onvon

Post by Daniel Wee »

So we capture the schematic into SPICE and run the simulation using a standard 9V supply rail and this is what we get:-
VA-9V.jpg
VA-9V.jpg (108.9 KiB) Viewed 15210 times
Now, we turn the supply down to 5V and the result is as follows:-
VA-5V.jpg
VA-5V.jpg (111.62 KiB) Viewed 15210 times
Notice that while the transient response has changed a bit, the voltage level at the output is still 1V p-p. This is quite promising actually since it would suggest that the circuit will probably work, possibly with a little loss in bandwidth. So, now we know that we can get this puppy to run on 5V with a simple swap of the regulator from the 7809 to a 7805, or better yet, something more modern such as the LDO LM2940-5.0 or similar.

Daniel
Daniel Wee
Site Admin
Posts: 2449
Joined: Wed 25 Feb 25 2009 8:00 pm

Re: Video Splitter from Onvon

Post by Daniel Wee »

Okay, we fed the simulator with the correct SPICE models for the transistors (we were using 2N2222 and some generic PNP parts before) and I am now seeing slightly better frequency response:-
VA-5V-model.jpg
VA-5V-model.jpg (123.72 KiB) Viewed 15208 times
The DC-coupled output has a little bit of leg-room to spare as well so I think we should be quite alright with this as long as the input does not exceed about 1.1V p-p. Even then (at 1.2V p-p for example) the output scales well so we're really in good territory here. In fact, looking at the output, I'd say that the circuit actually works better at 5V than it does at 9V as there is less over- and under-shoots. The high-frequency response is still better at 9V though, so this is a bit of a toss-up.

One way in which this splitter could be better would be if they had used a transistor with a faster response time. The 2N3904 is okay but is not particularly fast. For the price and purpose though, this is more than good enough.

Daniel
Post Reply