Back in the Game!

[Ralph]

Finally got a chance to sit down and start to work on the electronics of the televisor. The attached image shows the video/regulator board and the LED cluster sub-assembly.

The LED driver circuit is Klaas's single transistor circuit using a BUZ11 MOSFET. The BLK SET potentiometer sets the black threshold while the peak video drive sets the white level. The TIP29A NPN power transistor is a basic series voltage regulator that provides a regulated 20V from the 25V power supply. The 7812 regulator provides +12V from the 20V bus, powering the audio subsystem board (not shown). The BUZ11, TIP29A, and 7812 will all be equipped with TO-220 heat sinks which have yet to be installed.

The LED cluster board is a 3 x 4 matrix of 12 very-high brightness white LEDs with six parallel string of series-connected LED pairs. The diffuser sheets have not yet been added. With a peak drive level of 100 mA the cluster is very bright. The cluster flickers when NBTV video is applied to the input, so I assume everything is operating properly.

Tomorrow night it will be the sound sub-system that gets built and tested.

Ralph

[Metallica Man X]

Ooooooh! Looks snazzy

[Ralph]

Given that the Club CD's have audio along with the images and the Nipkow televisor is intended to be used for demonstrations at meetings and conventions, an audio channel was needed! Many of my projects over the years employed a National LM380 chip which operates from a basic 12V supply and will generate up to 2 W of audio. At first I thought about integrating the audio circuits with the video/regulator PCB, but later decided to use a stand-alone audio board to provide greater flexibility if I wanted to swap circuits at a later date.

Prior to making up a PC board, I decided to see if there were any kits available that would save the cost of prototyping the audio board. I found an excellent kit (#1690) from NightFire Electronics (https://vakits.com) in Ocala, Florida. The cost was very modest ($10.99 USD plus shipping) and the kit was complete with quality parts and a solder-masked and silk-screened PCB.

Tonight I wired up the kit and ran it through its paces using one of the Club CD's. Audio quality was excellent driving a 6-inch speaker and there was plenty of volume in reserve should I need to use a bigger speaker.

So, two boards (video and audio) done, leaving only the motor speed controller which I will tackle tomorrow evening. After that there will be some packaging and wiring to do, after which I should be watching a functional Nipkow televisor.

For those that are interested in following this mini-saga in more detail, feel free to check into the project web page at: http://taggart.glg.msu.edu/nbtv/de_mech.htm.

Ralph

[Ralph]

Finished the motor control circuit the night before last. Dennis supplied a complete video and motor control board with his DE Televisor replica kit. However, that is hardly a fair test of what it would have been like to use the original DE Televisor back in 1934.The neon would have been hooked into the plate circuit of the final audio stage of the radio receiver. I didn't care to do that, but Klaas's one-trasistor LED driver is a lot more authentic than a handful of integrated circuits.

The original DE motor drive was through the AC mains with a series power resistor and a rheostat to vary the motor speed. I had no no interest in a high-voltage solution and opted for a 12V DC motor. In the spirit of the original, I used a PWM motor control kit that uses a 10K potentiometer to vary the motor speed. I swapped out the standard pot with a 10-turn panel-mounting unit which would mount where the rheostat would have been mounted to the rear tower/pylon. The control module (with the extension leads for the multi-turn control pot) is shown below. The "box" will be mounted with Velcro (hook and loop) tape between the two pylons. With a black cover, the module is virtually invisible and motor speed gets controlled with a potentiometer just where the rheostat would have been on the original.

I made up strobe disc and already had obtained a KeyStrobe Duo 50/60 Hz strobe module from the UK. The controller rapidly brought the disc up to 750 rpm and then came the training to see if I could get it to hold speed without drifting. There is a certain technique that you develop to get around the hysteresis and, by the end of last evening, the speed could be coaxed into holding steady (hands off) for 15-20 seconds. Nothing on the speed controller gets even warm and things seem to work from a cold start. I turned off the power with the pot set at 750 rpm. When I turned it on again tonight, only a very slight tweaking was needed to get it back to speed. I am quite confident that this "old school" approach will actually work and that enthusiastic and patient DE user could actually have watched programming back in 1934!

Now its time to integrate all the finished boards and watch some NBTV myself!

Ralph

[Ralph]

The motor and video circuits were all patched in via test leads and the light box was flashing away, so I decided to try for a picture. Juggling the motor control with one hand, positioning the light box with the other, and trying avoid getting sliced by the disc was a new experience. Despite the gymnastics, my old friend bugs bunny slipped in and out of sync. This thing is actually going to work! I will need a better diffuser (suggestions readily accepted), but the transitory images were quite sharp with good gray-scale rendition. The single-transistor driver is a very useful contribution Klaas!

I will need a few days to convert the rat's nest of test leads into permanent wiring, at which point I will photograph the final version. The big challenge will be to actually get reasonable photographs of the display. With wiring done and a better diffuser in pace, I am looking forward to the photographic challenge!

Ralph

[gary]

Over the years I think I have tried every kind of diffuser there is, and I can honestly say I haven't found anything that works significantly better than a strip of frosty Scotch tape over each aperture - give it a go as it is cheap and easy to do and you can easily pull it off again if you don't like it. Good luck with the pictures I look forward to seeing them.

[Ralph]

The packaging philosophy I had when the project started has evolved considerably as things approach completion. The original idea was to have everything - video, audio, and motor control circuits in a satellite control unit tht would plug into the televisor mechanics with a single cable. Later, to preserve the original DE Televisor concept of having the motor speed control on the front tower/pylon, I tucked the motor control electronics between the base of the two pylons.

That left the video and audio circuits for the external control unit. My original idea was to mount the LED cluster pcb in a small project box attached to the top of the support strut for the replica neon tube in the original replica kit. Unfortunately I would have needed additional mounting hardware to raise the light-box to the required level. By using a larger project box I could eliminate the need for any mounting extensions, but now I had a light-box with a lot of unused internal space. At some point, given the compact nature of Klaas's one-transistor LED driver, I started to think about putting the video/regulator pc board inside the light-box! The idea worked out very well and the result is illustrated below! A POWER LED is on the back of the unit with the CONTRAST control at the top. Signal and power leads exit from the bottom of the box. The distracting hardware on the light-box and throughout the unit will be touched-up with flat-black paint once the endless assembly/dis-assembly work of construction is complete.

The only thing left over is the audio board and that is only required when using the Club CD or a similar source delivering both audio and video. The audio board is going inside a small speaker enclosure and will be available when needed.

As the final configuration of the televisor takes shape, all the motor and video electronics are artfully hidden away amongst the display mechanics. There are three inputs along the rear edge of the base: 12V for the motor, 24V for the video circuits, and the video input. One can easily see all the bits and pieces of the mechanics with no "electronics" clutter. Photos to follow....

Ralph

[Ralph]

I just wrote up a report on the completion of the monitor, but it disappeared into the internet abyss. So, short version:

(1) It is done!

(2) It works.

(3) You can, with practice, actually watch programming with manual control of the disc.

Display photos eventually!

Ralph

[Ralph]

Here are some of the first photos of the new televisor display. I'll get better with time, but didn't expect so much hassle getting a full-feature DSLR camera to handle the job!

The diffuser works moderately, but its limitations are evident in the image of my daughter at the far right where the outlines of some of the LEDs can be noted. I have an idea for a better approach which I will try over the next week.

Ralph

[Ralph]

This afternoon I experimented with using a piece of string to act as a gentle "brake" for my free-running disc. I cannot remember if I saw the idea here on the forum or in something in the Newsletter. The point is, it isn't my idea! Imagine a piece of string, anchored on one side of the motor, wrapped once around the drive shaft to the disc and anchored on the other side of the motor. If possible, space the anchor points on either side so the string makes something close to a 30-45-degree angle with the base.

In the case of your basic manual motor control, you would gently manipulate the speed control, moving very gently back and forth at the end, to frame and hold the image. That is tedious enough, but after all of that, the image will probably be out of phase. You then have to let the disc spin a little fast or a little bit more slowly until the image slides into the proper phase (left edge of the image coincides with the left edge of the display "window"). You then have to lock and hold the image with very gentle back and forth adjustments of the speed control. Can be done, but very tedious!

With the string brake, all is simplicity by comparison. Bring the motor/disc up to speed with the controller and then set the speed so it is just above the proper value - enough to cause the display frames to move upward at the rate of ~ 2 frames/second. The precise offset from nominal is non-critical, but the disc must definitely be running fast.

Forget about the motor speed adjustment and apply a slight downward pressure to the string (left or right side) and you can easily lock the picture in terms of roll. If the phase isn't just right, reduce pressure on the string very slightly and let the picture roll upward until it is properly phase and increase the string pressure to lock the display. Compared to doing everything with the manual speed-control knob, the string provides a relaxing alternative that lets you watch longer "programs" such as a complete Club CD.

Leaving the televisor on, you can go for a cup of coffee, come back, and have the image locked in moments, compared to extended tedious adjustment with the motor speed control.

It seems they used the "string brake" back-in-the-day and it is a very practical trick. Many thanks to whoever wrote it up!

Ralph

[Metallica Man X]

Looking good! I'll have to give this string idea a whack

As for a diffuser, have you tried using plastic bags?

I've had much success using a few layers of clear grocery bag on my crappy creation.

[Ralph]

It is truly quite amazing how easy it is to use the "string brake" compared with fine-tuning the speed controller. You can see an image of the basic string set-up toward the bottom of my DE Televisor web page:

http://taggart.glg.msu.edu/nbtv/de_mech.htm

There are two issues with the first run of images from the new televisor:

(1) The first is a slight mottling of the images caused by "hot-spots" from individual LEDs that can still be identified despite the diffuser. I think there is a better way to approach the issue of even-illumination and I should be able to get to that late this week.

(2) The second problem is strictly photographic - a soft appearance due to a slightly out-of-focus set-up. This time I did two things to get a sharper image. The first was to remove the magnifying lens (which works fine for general viewing) and set the camera up to photograph the active area of the disc directly. The second trick was to put a focus target on the front of the active area of the Nipkow disk. Of course this target has to be removed before you start up the disc. The point is, it is far easier to reach a sharp focus on the target than is the case when trying to focus on the virtual image created by the disc.

A sample image is shown below demonstrating good resolution of the scan-line structure, including the very slight overlap of adjacent scan lines resulting from the diameter of the scanning holes. Simply looks better now!

Ralph

[AncientBrit]

Ralph,

Regarding diffusers I remember someone on this forum (Klaas maybe) recommending using a pair of diffusers in tandem with a small gap between them.

This apparently produced superior results.

Regards,

Graham

[DrZarkov]

Indeed. I've tried a Fresnell-lense and diffusor with a distance of 5 mm with very good results.

[Ralph]

Good suggestions all around! One of my first thought was that the best way to solve the diffuser problem was not to use one! I built a brand-new lightbox with a new LED arrangement to experiment with reflecting the light from an interior white surface. It did provide a very even field of illumination, but light levels were way down!

While the original LED array was on the bench, I did run tests with a variety of diffuser materials. My original diffuser was mounted perhaps 5mm beyond the ends of the LEDs and produced a very uneven field of illumination as you can see in A (below). Most other materials were somewhat comparable at the same distance. Careful examination of my daughter' image in B (below) shows how the uneven brightness results in a very mottled image.

Double-layers of any of the material (with a space between) produced a better result but at the cost of image brightness. I tend to pay a lot of attention to that factor as I an photographing single frames because I am using manual speed adjustment of the televisor. The smallest f/stop I can use when photographing the display is about 5.6, so light is critical for a good exposure!

With all the materials I tested, a greater space between the LEDs and the diffuser was the key to effective performance. I am now using my original diffuser plate, but separated from the LEDs by almost 30 mm instead of the original 5mm spacing. The brightness is marginal (for photography), but there is a major improvement in image quality once the LED "hotspots" have been muted. Image C (below) was made with this new arrangement.

Since I was working on a new lightbox, I decided to devise a better mount than the rather flimsy bracket that was part of the original DE Televisor replica kit. The new arrangement is illustrated below. I will continue to experiment to obtain greater light intensity, but the next major project will be Peter Smith's motor control. I intend to set it up to switch between manual and automatic operation to highlight the difference between a high end televisor and the manual controls that were the staple of the home-built variety.

Ralph