Moderators: Dave Moll, Andrew Davie, Steve Anderson
smeezekitty wrote:So around the time I made the first revision mechanical monitor, I also made a mechanical camera.
Needless to say, it didn't really work.
I'm currently in the process of reworking it to hopefully make it functional. I'm replacing the bearing with a VHS head and 3d printed pullies like I did for the monitor.
In my original design, I used a large CDS cell for a pickup behind the disc. But I think they are too slow for NBTV so I am probably going to need to switch to a phototransistor or PV cell.
My design was not a flying spot but rather but the sensor behind the disc. Has anyone else done it this way? I am concerned about whether enough light will pass through the holes to create a usable signal. Also, what type of lens do I need to use?
I also need to think about whether I need a sensor preamp (maybe an opamp circuit?) and how I am going to generate sync pulses.
Try and try again was what i did its a pity Gary is no longer on the forum but he did leave a lot of advice on the forum as it was something of interest to him .
This is good i think that is a good move ...but the main problem i found was you need to make the device in a light proof box apart from light focused onto the nikpow disk scanning area ...any light getting in from sides such will show up as noise .
I used a cd or dvd nipkow using the tiny dc motor they use...... when rotating it looks ok but there is a very slight wobble to the disk fine for the video dreadful for making mechanical sync !
You could just mechanically use the same shaft for the monitor and camera and film the results but i like to make my own video to record and play back as its not that hard and the long run a lot easier really .
Ok i tried photo sensitive resistors for flying spot scanning pick up and photo transistors diodes ...here the problem is lack of surface area ..think of a matrix of leds in the nipkow monitor idea you would sort do this in reverse or has others have done main way a lens to focus it down to the one photo diode bit of a pain really when you can use something else with a larger surface area for light sensitivity of the whole scanning area and do away with that ! a pmt or solar cell ...
smeezekitty wrote:Yeah it is. I still wonder what happened. It might motivate me to tackle working on a viewer program again at least.
I'd much rather keep them separate units. It feels like cheating to have them on the same shaft. Besides, that is a lot more difficult mechanically and mechanical design is my weakest.
I am surprised you got this much bandwidth the the photoresistor. My tests showed a very poor response. Surface area isn't a problem though. When I say a large CDS cell, I mean a LARGE CDS cell.
I read all the linked threads completely. What a rabbit hole! Looks like I'll have a lot of pain with synchronization and noise. I'm not nearly as picky as Gary was though.
I was honestly tempted just to free run the motor around the right speed and hopefully create sync pulses that are good enough for a stable picture. My disc is large enough and is decoupled from the motor by a belt so that there is little speed variation over the course of one frame. My mechanical monitor only syncs once per frame and it works well.
Apparently my disc is quite thick for a camera but I don't think there is much I can do about that.
It does seem like flying spot cameras are easier to build but they have their own uniqe challenges. Like finding a bright enough light source that won't overheat the disc. Also, it's a bit annoying that they require total darkness. If it doesn't work out with a sensor behind the disc, I might try it.
[/quote]Lots of good amp circuits in the threads linked. BC109's are not easy to get in America. But I think I can substitute in 2N5088's. They have similar Hfe and noise figures. Somewhat lower current but I shouldn't need to run a lot of current through them anyway.
smeezekitty wrote:I thought I had some 2N5088 in the parts box but I couldn't find them.
In the mean time, I added a reed switch and magnet (like I used on the monitor) to generate a frame pulse. I recorded the frame pulse on the right channel and the video on the left channel.
Inspired by Gary's work, I knocked together a timebase correction program that would correct the video based on the frame pulses. It actually works surprisingly well and yields a stable picture in TBP
Here I was holding a piece of plastic with an "X" written on it between the light and camera lens:
https://www.youtube.com/watch?v=SdiIMNKJ8oo
Steve Anderson wrote:Yes, the 2N5088's are similar in terms of noise to BC109's, but the 2N5089 is 1db better on noise factor, and slightly higher gain, assuming you can get them.
Don't forget to use metal-film resistors, not anything based on Carbon. Clean noise-free power supplies are essential, be wary of the usual 78xx & 79xx series regulators, they can be quite noisy, follow them with a series resistor at the output then a decoupling/bypass cap to mop up most of the rubbish.
Screen all as best you can to eliminate hum pickup.
The above generally only applies to low-level circuits (pre-amps etc.), but is good practice generally.
Steve A.
The ultimate in low noise from that era is the LM194/394. Two super-beta transistors only in an 8-pin DIL package. Almost impossible to get now and hideously expensive. An alternative is the SSM22xx series, similar idea. (SSM2210, SSM2212 & SSM2220). Note the SSM2220 is two PNP devices, the rest are all NPN.
Yes lots of head amp circuits lots of trany versions i like the ne5534 the trany circuits can be touchy to get right but fun to experiment and see what works for you .
ON what i see is a problem is lack of surface area of the light sensitive resistor thus the light gaps top and bottom of scan might help a larger light in front or focusing behind the disk scan to the LDR's smaller area ,you also have a noise problem from the motor from the brushes ...perhaps try some caps from motor positive to ground ,the motor might be a bit worn ...Keep going doing well its good experimenting .
Once you get to the reflective light off an object is where the fun starts ...simple test cards ,toys make good stooky bills ; )
Harry Dalek wrote:Pity about the motor noise brushes might be to warn out to much ...which there's only one fix and that's a replacement pain as it is ...that's if you can be bothered it doesn't stop it displaying ...perhaps some thing for the next camera .
The wav file plays well on Big picture but will not load on big picture v2 unless you use the chris long setting then its only half a frame ...always had trouble with V2 displaying correctly at times have not tried it on Doms software yet ...
some circuits below ..the sstv one used a photo transistor notice the light sensitive area is small i had to follow the scan down the scopes crt screen as it was painting the image the photo transistors pick up is very narrow with out optics....There was a transparent test card over the scopes screen.
smeezekitty wrote:Harry Dalek wrote:
I was thinking; since we use an LED array behind a defuser for mechanical displays, why can't one do the opposite for a mechanical camera. That is, have an array of phototransistors in parallel behind a defuser?
Users browsing this forum: No registered users and 48 guests