Help repairing a 20-year-old driver board

[acertain]

Hi all!

I'm trying to repair an old Daktronics scoreboard driver board, and I could use some help. I have some basic electronics skills and knowledge, but some of the components here are beyond me. If I can't get it working, we're no worse off than we are today (we'll have to buy a new scoreboard), but if I can save our club $4000 with a little bit of work, that would be great.

Background: I'm a volunteer with Ballard Junior Football, a non-profit organization that fields football teams for 7-to-14-year-olds. The scoreboard was purchased 20 years ago, and died towards the end of this season. There's an internal 12v battery, a hookup for an external 12v battery, plus a 120v AC plug. It was originally designed such that you would charge the internal battery with the 120v AC input and then run the scoreboard off the internal battery during regular operation. The internal battery probably gave out a long time ago, as nobody with the club even knew that it had one, and we were running the scoreboard directly of the 120v AC input until it died.

When it died, another volunteer and I with some basic electronics understanding started pulling the thing apart. That's when we discovered the internal battery. We discovered that the 120v AC was stepped down to 16v AC and then rectified with a KBL02 rectifier. We did some basic voltage measurements with it plugged in, and it seemed that the voltage was low (this was a while ago, so I can't remember exactly what we did). When we discovered that it also had these 12v battery inputs, we hooked a 12v supply to each of the inputs to see if they would work.

When we hooked the supply to either of the inputs, the console (which takes 12v input) booted, but we got a burning smell from the driver board. We removed the power input and couldn't identify for sure which component the smell was coming from. Our assumption is that some component in the AC-to-DC conversion circuit (e.g. a power transistor or diode) has failed and that current is back-feeding from the DC inputs through that circuit at high current. Our hope is to pull the failed components so that we can run the scoreboard on an external battery and just remove all the AC-to-DC components.

So that's the general situation. I'm trying to trace the circuits now, and I'm a little confused. My specific question right now is the following. There's an LM317 voltage regulator that outputs 5v to operate a TMS70C02NL microcontroller (and probably the UCN5815A latches for the digits), but there's also a LT1074 that (from my Internet reading) also seems to be doing voltage conversion as a buck converter.

I'm trying to figure out why there seem to be two different ways that they are stepping down the voltage. The LM317 is clearly setup to output 5v based on the resistors in the circuit as explained by the datasheet, but I couldn't figure out from the LT1074 data sheet how exactly the output voltage is set. It's possible that part of the system runs on 12v, which comes out of the LT1074 and the rest runs on 5v, which comes out of the LM317, but I also don't understand why they would use one method of voltage regulation for part of the circuit and a different method for a different part.

I'm happy to post pictures, do experiments, etc., if anybody has any suggestions, but understanding these two voltage-regulation components is my current goal.

Any help you could offer would be much appreciated.

Thanks in advance.

Andrew

 

[KrisBlueNZ]

A few photos would be pretty helpful. Most of us have probably never seen inside a unit like this, and some photos would give us a general idea of what you're talking about.

I think you're right that the LT1074 is used to regulate the 12V (or thereabouts) rail. The 110VAC goes through a transformer and is stepped down to 16V AC which is then rectified. This will produce around 20V DC, which must be converted down to 12V.

Most of the load will be the display, and this is probably powered from the 12V rail. So they used a switching regulator to supply that rail because of the high current and the corresponding need for efficiency (manageable heat dissipation in the regulator).

I agree that the 5V rail will be for the logic circuitry, which generally doesn't draw much current, so it can be supplied by a simple three-terminal linear regulator, probably powered from the 12V rail.

If the voltage out of the rectifier was low, this could be because of an overload. For example, if the LT1074 has popped its clogs, or perhaps another component in the buck converter (such as the catch diode), the converter would be drawing a lot of current, pulling the DC voltage down, and getting very hot.

Depending on how the battery and the external battery connection are hooked into the circuit, it's possible that a damaged buck converter could also overheat with a battery plugged into the unit.

The LT1074's output voltage is set by two resistors which connect to the FB pin of the device. The LT1074 is available in 5-pin and 7-pin versions. Pins are numbered from left to right if you're looking directly at the part number on the package with the tab at the top and the pins emerging downwards. For the 5-pin version, the FB pin is pin 1, the leftmost pin. For the 7-pin version, the FB pin is pin 5.

If you can find two resistors connected to the FB pin and tell us their values (or just the colour bands) we may be able to tell you the circuit's output voltage. Or you can calculate it. Call the higher value resistor RA, and the lower value resistor RB. The output voltage is 2.21V multiplied by ((RA/RB)+1). This assumes that the designer used the conventional circuit, which I think is likely.

This is all somewhat speculative. Some photos would really help.

 

[acertain]

Here's a folder with some photos of the board. https://drive.google.com/folderview?id=0B3q3JM138i9CT0pYN2xIXzk2aWs&usp=sharing

The first photo is an overview of the whole board, followed by three close-ups of each third, with some annotations. And then a close-up of the LT1074's resistors. The main connector is numbered starting at bottom left (the one with black marks on either side), 1 through 3 along the bottom row and then 4 through 6, etc. The connections are 1,2 -> console signal; 3 -> ground; 4,5 -> 12v out; 6,9 -> AC 16v in; 11,12 -> 12v Horn.

The LM317's resistors are 330 and 1000 ohms, which according to the datasheet gives an output of 1.25*(1+1000/300) = 5v. My assumption is that this gives the power to the logic system. I also assume that the TIP120 banks convert the 5v logic to 12v for signalling to the digits. The digits are a little unusual. They aren't lights, but "glow cubes." They are sort of like the displays in old train stations. They have a black side and a florescent side, so the controller only has to supply enough power to flip the cube to the proper side.

The resistors connected to the LT1074 I think are 13700 and 2800 ohms, though that would yield 13v, not 12v. The picture is pretty blurry, but let me know what you think.

My hope would be to yank out all the power-regulation components and just get something like a picoPSU to supply the 12v and 5v lines from a 12v battery. Does this seem possible? Although it seems that the picoPSU requires a regulated 12v input, so I'm not sure that a battery will work.

Thanks for any advice you have.

Andrew

 

[KrisBlueNZ]

Here's a folder with some photos of the board.

It's a bit hard to tell much from them. Try taking pictures outside on an overcast day, with no direct sunlight and no flash. This gives good all-round illumination.

The LM317's resistors are 330 and 1000 ohms, which according to the datasheet gives an output of 1.25*(1+1000/300) = 5v. My assumption is that this gives the power to the logic system. I also assume that the TIP120 banks convert the 5v logic to 12v for signalling to the digits.

Sounds likely to me too.

The digits are a little unusual. They aren't lights, but "glow cubes." They are sort of like the displays in old train stations. They have a black side and a florescent side, so the controller only has to supply enough power to flip the cube to the proper side.

I've seen those. They're a clever idea. We have them to indicate route numbers and destinations on some buses here. You can make a large, easily readable sign with a relatively low power requirement. They don't glow, of course, but for daytime use they're great.

The resistors connected to the LT1074 I think are 13700 and 2800 ohms, though that would yield 13v, not 12v. The picture is pretty blurry, but let me know what you think.

I can't tell from the pictures. You can see them in real life, and you're probably right. It's quite possible that it's a 13V power supply. There may be a diode and/or resistor between the power supply output and the battery, which will drop around 1V.

My hope would be to yank out all the power-regulation components and just get something like a picoPSU to supply the 12v and 5v lines from a 12v battery. Does this seem possible? Although it seems that the picoPSU requires a regulated 12v input, so I'm not sure that a battery will work.

Do you think the 5V regulator is damaged? I'd consider it innocent until proven guilty. If it's OK, or if you can replace it, then you only need to supply 12V.

The picoPSU seems to be designed to run a computer from a 12V source such as a car battery. It provides all the ATX voltages, and you only need 12V and 5V. So I don't think it's a very good fit.

If you have a 12V DC battery, and the 5V regulator is OK (or you replace it), you shouldn't need anything else.

 

[KJ6EAD]

Of course, you also need to be suspicious of the condition of the electrolytic capacitors.

 

[acertain]

It's a bit hard to tell much from them. Try taking pictures outside on an overcast day, with no direct sunlight and no flash. This gives good all-round illumination.

Overcast days are pretty easy to come in Seattle this time of year. Daylight hours are a little more restrictive. I'll see if I can get some made.

If you have a 12V DC battery, and the 5V regulator is OK (or you replace it), you shouldn't need anything else.

I was thinking that a 12v battery might not produce a pure enough voltage, but that's probably silly, eh?

Also, what do you suppose the UDN2998W is there for? From the datasheet, it would seem that it would be there to drive the solenoids that flip the glow cubes, but it's in the wrong part of the circuit for that. It seems to be part of the voltage-regulation circuit. It's hooked into the end of the chain of the four big capacitors, and has a massive heat sink on it (and it seems to be the component from which the burning smell emanates, though it's a little hard to tell for sure).

Of course, you also need to be suspicious of the condition of the electrolytic capacitors.

For sure. I'm assuming they are part of the AC-to-DC rectification and step-down circuit, so I'm hoping I can just cut them all out

Thanks for all your help.

Andrew

 

[KrisBlueNZ]

I was thinking that a 12v battery might not produce a pure enough voltage, but that's probably silly, eh?

Yeah, probably silly The voltage out of a battery is very clean.

Also, what do you suppose the UDN2998W is there for? From the datasheet, it would seem that it would be there to drive the solenoids that flip the glow cubes, but it's in the wrong part of the circuit for that. It seems to be part of the voltage-regulation circuit. It's hooked into the end of the chain of the four big capacitors, and has a massive heat sink on it (and it seems to be the component from which the burning smell emanates, though it's a little hard to tell for sure).

It's a dual H-bridge. I doubt that it drives the cubes, because there are lots of them, and the H-bridge's outputs can't be multiplexed in any clean or efficient way.

I can't suggest what it might be used for, because I don't know much about the unit in general. Use the data sheet to find which pins are the outputs, and see where the tracks go.

 

[acertain]

I'll see if I can trace it out exactly and post a diagram, but I'm concinved that the UDN2998W is participating in the voltage regulation somehow. The I-out pins are connected to either side of the big capacitor bank that seems almost certainly voltage regulation. I'm getting a little confused tracing the outputs, but it seems to me that the output of the LT1074 goes to the battery charging (I found a UC2906 battery charging control chip) as well as to this huge capacitor bank + the UDN2998W. Is it possible that the output of the LT1074 is clean enough to charge the battery (or even maybe that's why they're outputting 13v), but then it gets cleaned up/regulated down with the capacitor bank and UDN2998W to the 12v for the logic driving? Has anybody ever heard of a driver chip like the UDN2998W used for this purpose?

Thanks.

Andrew

 

[acertain]

Reading a little more on the Internet, it seems that this type of circuit is commonly used to build an inverter, but that seems the opposite of what I would expect...

Andrew

 

[(*steve*)]

Reading a little more on the Internet, it seems that this type of circuit is commonly used to build an inverter, but that seems the opposite of what I would expect...

Maybe those digits (or something else) require more than 12V to operate.

 

[acertain]

Maybe, but the lines leading to the device are sized for serious current, plus it's got a major heat sink on it, and I can't imagine pulsing the cubes to flip from black to day-glo takes much current. I'll see if I can draw out the circuit.

Andrew