A new 90-line Mirror Screw

[Panrock]

Ah... 'Skytrain'... for me, redolent of urban transport in Asia. I don't know if that was what it was called, but I remember something like it in Kuala Lumpur... very impressive... you remember, when I was visiting a factory in Penang in 2011 and we almost met.

Back to the project: something that will land in my lap soon and about which I have not a clue: the suitability or otherwise of the 470K log 'volume control' pot to operate as the motor speed control (with two transistors lending a hand of course). I can imagine having to string various parallel and series resistors to adjust the 'law'. Any thoughts? My method would be to trial-and-error it until it works reasonably well.

Steve O

[Steve Anderson]

Even with the two trannies either in a Darlington or Sziklai pair arrangemant [1], I would think 470k is way too high a value, I would guess 10k is more reasonable. But it does depend on how the motor responds to the load imposed (inertial mass, bearing and air-mass drag). It is a case of 'cut and try'. I can't recall how this was done on the 120-line version.

[1] Here is where I might have considered a MOSFET.

Steve A

[Panrock]

Thanks Steve. I'm not conversant with the Sziklai pair. I'll do some googling.

The 120-line version, in comparison, was a real beast. It used an 1/8 hp American 60 Hz AC mains washing machine motor, with governor. This, run from 50 Hz and at a reduced voltage of about 160v, drove a belt and pulleys of appropriate ratio so that the governed screw speed would 'click in' at 25 Hz. A small Variac was used to run the assembly up and set the supply at 160v.

Unfortunately, when I set up Karen's unit for the 90-line corrections, I shall have to lose the 120-line ones. This means the 120-line screw probably won't be used again. Maybe it would make a nice door stop at the Science Museum... or perhaps a bacon slicer?

Steve O

[Steve Anderson]

Thanks Steve. I'm not conversant with the Sziklai pair. I'll do some googling.

Here's a start...

https://en.wikipedia.org/wiki/Sziklai_pair

...also this is worth reading...

https://sound-au.com/articles/cmpd-vs-darl.htm

The 120-line version, in comparison, was a real beast. It used an 1/8 hp American 60 Hz AC mains washing machine motor, with governor.

Unusual mix? 220/240V and 60Hz. But in the US for more power-hungry devices they can get a domestic two-phase 120V supply, sort of 120-0-120V. In the Philippines the standard wall-socket outlet is 220V/60Hz, i,e, no 110/120V.

Unfortunately, when I set up Karen's unit for the 90-line corrections, I shall have to lose the 120-line ones. This means the 120-line screw probably won't be used again. Steve O

As long as the 120-line version is well documented/photographed the donor parts could be replicated in the future if it's desired to bring it back to life.

Steve A.

[Panrock]

Thanks. I shall have to take a serious look at those two articles...

https://sound-au.com/articles/cmpd-vs-darl.htm

...for I was going to omit having an 'R1' in my own simple Darlington arrangement.

As long as the 120-line version is well documented/photographed the donor parts could be replicated in the future if it's desired to bring it back to life.

No need for too much concern. The mechanical parts and the screw itself won't be touched. All that will go are the unique delay settings for the 120-line screw in Karen's converter - to be replaced with new settings for the 90-line.

Steve O

[Steve Anderson]

The mechanical parts and the screw itself won't be touched. Steve O

That's what I was hoping. To replicate (if needed) the electronic bits is comparitively easy than a re-build of the 120-line screw itself.

It would be an interesting exercise to compare the 90-line and 120-line versions side-by-side.

Though the logistics for the 2026 convention may be a headache! You might need a van rather than a car for that! Plus it would be quite a complex task even just for the electronics. I really don't think it's a viable proposition...it's an awful lot of work.

Steve A.

'R1' in Rod Elliots' article serves two functions....
1) It bypasses any leakage from the input transistor.
2) It also 'swamps' the B-E capacitance of the output transistor. Not really required in this very low speed application, but at high audio frequencies or more it can/does make a difference.

If you look at the internal diagram of most Darlington devices they have internal resistors that perform those functions. e.g. the TIP142/7 and the like....

[Panrock]

Here is the sub-unit, ready to be mounted on the floor panel of the mirror screw, and the circuit. Make of this what you will. The values are "wot works" and give good control of the speed.

The mirror screw itself will need its black areas tidying and the next job is to see how close to the ideal I can set the slat angles - before resorting to Karen's unit.

Steve O

9912.jpg (151.04 KiB) Viewed 259 times

cct.png

(8.83 KiB) Not downloaded yet

[Panrock]

I have now mounted the powering assembly.

This morning I have made a simple U-shaped slat setting tool, with a 22.7 mm gap between its jaws. This distance had been derived from a session on the computer.

The idea is to push the tool in so it engages on the away-facing edges of each adjacent pair of slats. If the angle between the two slats is too great, the tool will push the slats together by just the right amount. If the angle is too small, the adjusting action will have to wait until I am working 180 degrees round on the other side of the screw.

Hopefully, this will get the settings not too far from where they should be.

Steve O

[Panrock]

The bench and floor areas now fill up with the ancillary units that will be needed (very much a combined club effort, these were) and the triple RGB lines-of-light. I shall be re-using the 120-line gear.

Steve O

[Panrock]

This is a real "warts 'n all" blog on the project. Visible to all is the mess and chaos in which I work. I don't have anyone to tidy up after me, which is actually a real blessing!

Some good news and some bad. The motor and its control board are fit for the task... I had nice frame pulses rolling in every 40 milliseconds yesterday as seen on the scope. That's 25 frames per second in old money. The screw seemed to rotate in reasonably relaxed fashion at this speed. It helps that it's a lot smaller and lighter than the old 120-line screw. That one had me take cover across the workshop whenever I ran it up to full speed.

But, where's the video signal? In its place is just instability, signified by spurious readings on the level meters when I plug in the signal lead. Take note Steve A, these are DIN connectors (for which I admit full responsibility). Looks like I am now getting my comeuppance! The interconnection lead has been tested (it's OK) and I've also tried tightly coiling it. The next thing then, will be to be to open up the Control Box.

Steve O

[Steve Anderson]

Take note Steve A, these are DIN connectors (for which I admit full responsibility). Looks like I am now getting my comeuppance! Steve O

DIN connectors, yep, nice idea. shame about the implementation. Replace the DINs with D-types with retaining thumb-screws. I'm sure an overblown DIN commitee of accountants designed the DIN series....and probably took years to do it. Likewise with SCART connectors.

Visible to all is the mess and chaos in which I work. I don't have anyone to tidy up after me, which is actually a real blessing!

Agree entirely!! I know where everything is, the rest of the world wouldn't have a clue.

Steve A.

Actually I shouldn't be too harsh on the DIN organisation, the DIN41612 series of PCB edge-connectors have yet to be beaten, IMHO.

[Panrock]

Anyway, I had the LED Driver Box running last night 'with the hood off' (on which the controls are mounted). It's stable like this, but I shall have to get to the bottom of the problem.

I attach a 'publicity' photo of the box in question, and your circuit, repeated for each of the colour channels. Has it really been over 12 years? !! The problem in this case will likely be a layout issue. Funny though, 'cos it didn't do it before.

Actually I shouldn't be too harsh on the DIN organisation, the DIN41612 series of PCB edge-connectors have yet to be beaten, IMHO.

Yes, and plenty of use of these was made in the signal chains of the 405-line transmitter I showed here earlier. However the RF connections to the cards had to use SMA connectors - a bit of a fiddle to undo when removing the cards.

Steve O

(Later) I am still not quite satisfied with the state of polish of the slat edges. Although they are 'mirror finish', the underlying surface still has some tiny dimples, visible only in LED light coming from the Line-O-Light. I'm going to try to improve this, at least. It means dismantling the screw and starting again. This time I also have my 22.7 mm setting tool available of course, and it will be easier to use this when I'm laying the slats down one-at-a-time.

The saga continues!

[Steve Anderson]

Really 12 years? I recall drawing up the circuit, but only just! Looking at it today I might be tempted to replace TR101-3 with an op-amp, say a TL081/2/4, but it does the job (I presume) so why bother? But if there are improvements/changes required, just let me know....the only thing I might suggest now is adding some form of safety current limiter for each LED chain, but that may be thought as going a bit overboard. If the build is done in a workmanlike manner, it should be OK as is.

As usual decoupling/bypass caps are required/implied for the circuit, as is a 'smallish' heatsink for TR104.

Steve A.

[Panrock]

The 90 slats have been removed. The spots of 'goo' I originally applied to them have now been smeared over their faces. Although they can still slide over one another, this will mean they will tend to maintain their angular positions, once set.

Now to do the job again - properly this time.

The thermometer is accurate and that's with the heating on! This ain't Bangkok. Can anyone say what make of car is on the security camera? I've had it 22 years.

Steve O

[Steve Anderson]

The thermometer is accurate and that's with the heating on! This ain't Bangkok. Steve O

It's cool/cold here too, an overnight low of 17 Celcius, daytime high today of 32 degrees, currently 20 outside, 24 inside the house...we have no heating in the house. Time here is 0714.

The car? No idea, I'm simply not a car guy. Cars are as interesting to me as a potato peeler, that's why I have a Toyota. It starts, it runs, that's all I want, and that's after 7 years and some 270,000 km. (174.000 miles). That too has no heater, but it does have air-con.

Steve A.