When Bright Digital Displays Were Mostly Heat Sink (And Awesome!)

Interesting and really enjoyable. Looks nice Fran.

Damn, that thing is beautiful. Very nice.

Jessica McIntosh

A was going to do that - but the extra time was in my view not worth it - I went for a more 'universal' driver project.

Fran Blanche

Wow Matt - YOU should have done this video!

Fran Blanche

I really enjoy these old kit videos... keep 'em comin' Ms. Fran.

Howard Simons

A fascinating display and indeed a lot of effort but like you said you now have a driver which will drive pretty much any display.

Dr Andy Hill

for directly driving the bulbs off the 4511 you will want to use a preheating resistor, enough to keep the filament slightly warm but not glowing too brightly, this will reduce the peak current into a cold filament and possibly remove the need for the transistors, the sheet for the 4511's recommends around 400R for numitrons but with those you may want 390 - 270 ohms at probably a reasonable power rating. Alternatively use a ULN2003 (or 2803 I think for 8 channels) and a positive power rail on the common, they are designed to drive bulbs or inductive loads and have the exact amount of outputs for a 7 segment display.

William

The lengths she goes to recreate the past!

Dan Swinehart

Fran, these incandescent seven segment displays, as you presumed, were used extensively in avionics since they offer exceptional sunlight readability. And, believe it or not, they are still manufactured today. While never used for new-design, the old avionics equipment from the 70s/80s is still serviced today, decades after their initial design and construction. The companies to who initially made these displays either vanished or changed owners multiple times over the years. Many of them were made by a company called Oppenheimer, and they used blueish fiber segment inserts to change the natural incandescent bulb amber color to a corrected pure white. Oppenheimer decided to stop production of these modules due to the small demand, typically being only for replacements. Finding another vendor was very burdensome. The modules could be ordered with non-color-filtered bulb readouts or another color like amber, and later NVIS green. These were terribly expensive display modules and, believe it or not, they actually could have individual bulbs changed. Or they mounted several bulbs to a tiny circuit board that could be changed. I don’t know about that Tung Sol version. I’ve never seen that brand in any avionics I have taken apart. Believe it or not, the large aluminum block, wasn’t necessary for the modules I’ve seen in avionics. Yes, the bases were usually screwed into a structural component of the overall indicator to dissipate some heat, but they simply made the phenolic, or whatever the black material was, deeper so that the module itself would handle the heat. They were used alone in some cockpit indicators, like distance readouts, but they were also built into extremely complex electro-mechanical instruments. In those cases, they were almost always front-removable from the complex indicator case so the modules could easily be removed and replaced. Having to pull and swap a $60,000 indicator out of the main instrument panel for a burned-out incandescent segment was obviously not practical. There were actually dedicated dimming circuits just for these readouts on aircraft such as the 747-100 to 300 series. (Pre-glass cockpit) Daylight readability during night flight would be very annoying, thus the dimmer pots were placed on the front glareshield just for these readouts. Depending on the criticality of the indication, they had several ways to drive these 7 segment displays. The simplest was to use a SN7447 TTL 4 bit decoder-driver to directly illuminate each segment bulb. But they used the SN5447 higher temperature MIL-Spec IC in a ceramic package for heat dissipation and reliability. Other times they used a chip similar to a ULN2003 between the output of the decoder/driver and the bulbs. They varied the voltage of the common ground pins to allow for dimming. Then they made specialized chips that did the TTL decoding and lamp driving on the same chip which saved space and could handle much more current. There was usually a “lamp test” button someplace that activated all 8 segments, and the decimals, to be sure they would not present a false critical reading because of a missing segment, like an “8” turning into a “6” or something else deceptive. When the reliability of the readout was critical enough, they built in current sensing into the driver chips to flag a readout that had an open bulb, thus drawing no current. The lack of load would then direct the driver IC to flash all the segments in the digit, thus making it obvious to the pilots that there was a bad digit in the readout. There was usually a button to cancel this blinking so the flight could continue. Of course, there were always two identical indicators that used these incandescent displays, to there is never a single point failure. The bulbs used in the indicators, as you stated, were thick-filament and low voltage. Basically, the best they could do in order to ensure long life in a high vibration environment using a DC current to drive the filaments. The problem with using DC power to drive incandescent lamps is that a phenomena called “DC Notching” would pit parts of the filament and lower the life of the bulb. I’ve had 5VDC bulbs die in a matter of months when run by DC, but literally last for years when fed AC voltage. That’s why they used AC power to drive the thousands of tiny 5 volt bulbs in the cockpit used exclusively to backlight the plastic cockpit panels and analog instrument interiors. This allowed for the lettering on the panels to illuminate and they would last for years because of the quality of the bulbs and the AC power supplied to them. Replacing dead bulbs in electromechanical indicators was extremely burdensome since the guts would have to be removed from the case and the multiple bulbs replaced. Usually they were soldered into mini PC-boards. Some manufacturers got smart and designed a removable wedge module at the top of the instrument so that it could be changed out without having to disassemble the indicator. Back in the 60s, and even through the 80s, there was no alternative to incandescent lamps that provided the same brightness. Of course, now, FINALLY, we have sunlight-readable white LEDs, which have been steadily replacing the incandescent bulbs in backlit panels and displays. The LEDs, obviously, offer high brightness in any color, consume very little current and last indefinitely. The only problem is driving the new LEDs with the old AC dimming systems in so many existing aircraft. Since the dimming is not linear on LEDs, and they have a dramatic drop-off at some voltage, ICs were developed to read the dimming input voltage and Pulse-Width-Modulate signals to the LEDs to match the incandescent dimming curve. The data sheet you included in your video depicted one of the Latitude/Longitude modules used in the Litton control/display units that were in almost every jumbo airliner cockpit of the early to late 70s. They were a great success, but after decades of use, inevitably, these displays eventually showed their age. The internal heat, sunlight exposure, aging materials and other physical failures made the faces of them pretty ugly. Still usable, but can look very worn, and sometimes even cracked. No doubt these were an excellent solution for high contrast daylight readability in a world where there was no solid-state alternative. The tiny bulbs used to be made in the US, but like so many other things, even companies like “Chicago Miniature Lamp” were having them made overseas. Kind of deceptive with “Chicago” in the name. Also, the other components that made up avionics in the 60s and 70s were all American companies, with few exceptions. Now avionics are made with parts from all over the world. It’s sad to think we once had the ability to be self-sufficient in high quality electronic component manufacturing during most of the “Cold War” as a matter of national security. The USSR used to reverse-engineer our chips and stole most of their technology from the USA. Now everyone seems to have access to worldwide supplies of electrical components. If China stopped all exports, I hate to think of the parts shortages that would face the world. Even the US Government doesn’t typically levy “nuclear-hardened” requirements for a lot of avionics. It seems that tactical fighters are the only remaining “price is no object” kind of electronics left. I’m really glad you made this this video depicting this very important step in display technology history. It played a critical role in the days before we could even dream of having a white LED, let alone a sunlight readable LEDs. As usual, I wish I wish I had an on-line photo repository of everything I described, but I don’t think there are enough people with this super-narrow niche interest in display technology. You do so well with your examples. I can only elaborate on your already excellent demonstrations.

Matt Wietlispach

That is a fascinating find. I can't imagine how much that assembly cost originally. Obviously intended to be used in critical applications they were probably scheduled to be changed out before any of the bulbs could burn out.

Bill Kerr