Scoreboard Project

[dtrueb]

I am working to restore to operation a 1950's vintage basketball scoreboard. The electro-mechanical system is totally shot and inoperative and will be replaced with something electronic ... CMOS counters, drivers, etc.

There are 7 single digit displays as follows
A. 8 minute count down timer with 3 digits (minutes, 10 sec & seconds).
B. Home score - 2 digits (0 to 99)
C. Visitor score - 2 digits (0 to 00)

Each digital display contains 20 each 7-1/2 watt incandescent 120 volt bulbs in a 4 wide by 7 tall pattern as shown in the first attachment with my way of numbering the 13 segments (1 thru D) alongside.

Either 1 or 2 bulbs are illuminated in each of the 13 segments (a total of 13 segments) to form digits 0 thru 9. Not too unlike today's 7 segment digital displays.

I have arrived at the proper CMOS components and design that will drive the clock and score, but am falling short on the proper way to drive the 13 segment display. Of course we know about the BCD to 7 segment display drivers like the 74C48, but there is no such thing as a BCD to 13 segment driver. So, I guess the solution is just to build one... by encoding the truth table into a EEPROM. That could be done with a very small EEPROM. The project would require 7 of these, one for each display.

My problem is.... I don't know anything about programming EEPROMS and I need someone to lead me by the hand into understanding how to do this step by step or maybe someone to make these for me. Also, I don't want to get involved in purchasing an expensive memory programming machine. Maybe there is a cheap way to do this?

The truth table is shown in attachment 2.

The 120 volt bulbs can be driven via individual solid state relays such as the Sharpe PR22MA11 which can pass 150 ma. Two of the 7.5 watt bulbs would draw about 125 ma. These cost only about 40 cents each and I would need about 91 of them (7 displays times 13 segments per display).

I know that most of you will say... gee, this can easily just be done with a microprocessor. However, my stage of learning is not yet that advanced and I prefer to do it with CMOS technology.

Any suggestions and alternate approaches would be greatly appreciated.

Dave
An Old Timer

 

[mursal]

Not sure if I have this correct, 13 segment display, but you will only need to display 0 to 9? So can you drive for example, the top row as a single line in the display (1,2,2,3) and use a standard 7 segment display driver?

Or am I over simplifying?

 

[Harald Kapp]

You can use a BCD-to-7-segment decoder, e.g. 4511.
Re-label your lamps as follows:
22 = a
44 = b
66 = c
etc.
That leaves you with lamps 1,3, 5,7,9 and B as special cases.
You want lamp 1 to be lit when segment a or segment f is lit, so the boolean equation is: 1 = a OR f
You want lamp 3 to be lit when segment a or segment b is lit, so the boolean equation is 2 = a OR b
And so on fpor lamps 1, 4,7 and 9.
Lamps 5 and B need special consideration since they are adjacent to 3 segments. Therefore the equations are:
5 = b OR c OR g
B = e OR f OR g

It may be possible to simplify these equations as not all combinations of segments are possible when displaying 0...9. I'll leave that as an exercise to you

 

[KrisBlueNZ]

Another possibility is a diode matrix. Convert your 4-bit BCD number to a one-of-ten output, and use diode ORs to generate your thirteen "segment" control signals.

Suitable BCD decoders are:
CD4028B - your best option, with ten active-high outputs
CD4514B sixteen active-high outputs
CD4067B sixteen analogue switches

 

[dtrueb]

Harald: Thanks for that reply. I had already thought about using the 4511 and then adding some other logic to decode the other 6 (13-7) outputs. Seemed however, maybe easier to encode the whole table into a single EEPROM (except for my lack of knowledge about how to do that).

If anybody can illuminate me on the EEPROM thing.... I would appreciate it.

 

[KrisBlueNZ]

I guess you mean Flash ROM. It's technically a type of EEPROM but EEPROMs normally have a serial interface. You could use Flash ROMs but they have data widths of 8 or 16 bits, so you would need either 7 or 14 of them, and you would only use the first ten locations, and waste the other 65,526 (for a 512 kilobit Flash ROM, which I think is the smallest you can find). Also, you really need a programmer. There are companies that will program them for you, but not cheaply.

 

[dtrueb]

Kris: Thanks too for your input. Like in my response to Harald, the 4511 could take care of 7 of the 13 segments and the other 6 could be encoded separately. I think what you are referring to in your comment is constructing a ROM using diodes. Similar to diagram 9-4 in my CMOS Cookbook (Lancaster 2nd Edition) page 447.

I could encode just the 6 segments and use the 4511 for the other 7 segments. That would take 23 diodes and a bunch of soldering.

Again.... my feeble brain says "code the whole 13 segment table into a EEPROM". Much simplier and more straight forward. Fewer parts, etc.

Again. Can anybody point me in the right direction concerning programming EEPROMs?

I will post some more diagrams and truth tables shortly.

D

 

[dtrueb]

Kris: By the time I finished my last comment, you had already added another post. Thanks for that. Not following all you say about EEPROMs, but will attempt to educate myself further.

So... EEPROMs are mostly serial, not parallel? I didn't know that. So what kind of memory chip do I need to generate an output on 13 leads that is generated from a BCD? I would call this a lookup table. Convert the truth table into the memory so that the output leads have 1's and 0's corresponding to the truth table. Then drive the solid state relalys which light up the bulbs.

What kind of money is involved in a programmer?

D

 

[KrisBlueNZ]

Well, EEPROM just means electrically erasable PROM, so Flash ROM is a type of EEPROM, but EEPROM is usually used to describe serially accessed devices in compact packages.

Here's a Flash ROM that would work: the Macronix MX29F200CxMx at http://www.digikey.com/product-detail/en/MX29F200CTMI-70G/1092-1079-ND

It's a 44-pin SOIC, i.e. it's surface-mounted, so you can't easily mount it in a prototyping system such as a breadboard or a piece of stripboard. You can get adapter boards to convert it to through-hole. I don't know of any through-hole Flash ROMs available new nowadays. It requires a 5V supply. Most modern Flash ROMs run at 3.3V or lower. So that one may not be around much longer.

It has 2 megabits of storage and can be configured (via a pin) as either 8 or 16 bits wide. You would need 16 bits. At 16 bits wide, it contains 131,072 locations, and you will use 10 of them.

You're right, what you want is called a lookup table. Using memory, you have four bits of input data, which would connect to the bottom four address lines, and thirteen bits of output data, which would come from thirteen of the sixteen available data pins.

Programmers start at around USD 100 but that is not a common device so make sure any programmer you consider specifically says it can program it.

I guess you're getting the picture that it's not really practical! You would be much better off using a microcontroller. Here are a couple of suitable ones:

Microchip "PIC" architecture: PIC16F527: http://www.digikey.com/product-detail/en/PIC16F527-I/P/PIC16F527-I/P-ND/3872386

Atmel "AVR" architecture: ATtiny2313/4313 http://www.digikey.com/product-detail/en/ATTINY4313-PU/ATTINY4313-PU-ND/2271083

These are both available in 20-pin DIP (through-hole) packages. They have 17 and 18 general purpose I/O pins available, respectively. They are both under USD 2.00 each.

I would be happy to write the code for either of these. It's less than one printed page. The assemblers for both devices are free. Programmers are around USD 50 but I can program them for you if you want to post them to New Zealand and include return postage.

The logic for a microcontroller to do this conversion is very simple. The only issue is programming. Microcontroller programming is great fun, and that would be a good introduction for you, if you're interested. If not, it's still probably the cheapest way to go.

 

[dtrueb]

Kris: Great response. Understand totally. Wonderful offer on your part.

My understanding is...
1. The table lookup can easily be done by the PIC
2. Put in a BCD and receive out the proper 1s and 0s on the 13 output pins

Rest of the design is un-effected by this component.

I'm not the speediest person on the planet and this is not a red hot gotta get it done fast project. I'm going on 75 and am a retired EE who was trained shortly after the transistor was invented. Hard to keep up with all the modern and latest EE stuff, but I keep trying. My training was mostly on motors, generators and power equipment. My career was telecom so I experienced all of the telecom merged with computers, etc.

Anyway.... maybe we should continue this off line via email.

I would propose sending you via email, several documents including an Excel file which shows the 13 segment lamps and the truth tables involved.

I could simply order the chips (order 10 with only 7 being required) and have them shipped directly to you in NZ. You can program them and send them back to my home address.

I would want to compensate you something for your time and talent. My benefit is that I get to see the code and how all that is done, plus get the finished product for my project.

FYI: My son-in-law owns and operates a local Pizza Resturant called the ScoreBoard and he acquired this old board and naturally asked me if I could make it operational. Plan is to put it up on the wall in the interior in it's natural state with the control boards so that patrons can operate it for fun. Naturally, I said YES and quickly found out that the board electro-mechanical was totally fried. Took quit some time to reverse engineer it and figure out how it was supposed to operate. SIL doesn't want to restore the board to original condition, but leave it like it is.... chipping paint and rusty. Probably spent the last 40 years in somebody's barn. It was full of bird nests, bird poop and mouse nests.

It's been an interesting project and a good reason for this old man to try and get modern and learn all about TTL, CMOS and microProcessor stuff. I got nothing but TIME on my hands.

I can be reached by email at [email protected] if you care to continue this to it's conclusion. Your offer is too hard to refuse and I appreciate it very much.

D

 

[KrisBlueNZ]

Yep, I'll take this to email.

 

[Harald Kapp]

Harald: Thanks for that reply. I had already thought about using the 4511 and then adding some other logic to decode the other 6 (13-7) outputs. Seemed however, maybe easier to encode the whole table into a single EEPROM (except for my lack of knowledge about how to do that).

Come to think about it, you don't need to encode anything extra. Take for example lamp 1 which is lit for segments a OR f. You can just use 2 solid state relays, both of which would drive lamp 1. One relay is driven by the decoder output for segment a, the other by the output for segment f. And if both outputs are active, both relays are on without any harm.

 

[gorgon]

Why don't you get a bunch of old CMOS EPROMs? Those are easy to program and you can get all the codes you want from them. Since they have 8bit outputs, you'll need 2 EPROMs for each digit, You can use any size of EPROM, but I think the 27c64 8kb, or 27c256 (32kb)is the easiest to get.

I think I have a bunch of 2732s (4kb)you could have, but I don't know if they are CMOS, if that is very important. I could even program them for you, if we agreed on your design.

 

[dtrueb]

Gorgon:
Thanks for the tips and comments. I will get back to you after I get a chance to study your comments.

thx

D

 

[KrisBlueNZ]

gorgon, it looks like we're going ahead with seven 20-pin PICs in narrow DIP packages. EPROMs would have worked, but fourteen 24-pin or 28-pin wide DIPs would be much larger, and anyway, Dave is keen to get involved with microcontrollers. Thanks for your offer though.