Trying to settle an argument. Based on over fifty years experience as a coin-op tech and consultations with other professionals over the years, I have always set the game logic boards at 5.17 volts to eliminates pesky reset problems. (and provide a cushion for less than ideal line voltages at various locations). An esteemed colleague has now said this is too high. Assuming the power supply is ok, anyone have any arguments for or against? Thank you...

 

If you are saying you adjust the power supply output (i.e. trim pot) to 5.17V then you are basically pre-compensating for voltage drop in the cabling and connectors. This is not a bad technique and would typically be done while measuring voltage on the logic board.
If you are saying the you adjust the power supply until voltage on the logic board is 5.17V then technically its not too high as commercial 5V TTL logic (back in the day had a 10% tolerance, i.e. 4.5V to 5.5V) and your 5.17V is within that operating range.
Now, these game logic boards were quite large and depending on the number of GND and power layers you could see voltage drop across the board, so would also be OK to adjust power supply such that the lowest voltage at all four corners was at least 5.0V then that would also be OK.
"Pesky resets" are typically poor design decisions, but certainly possible that running at slightly elevated voltage alleviates.

 

Hello,

As said, pesky resets can happen due to bad decoupling in poor designs.
Decoupling or Bypass Capacitors, Why?

Bertus

 

"pesky resets" can occur at high voltages as well as at low voltages, as well as for other reasons.
Since standard 7400 series logic has a Vcc tolerance range from 4.5 V to 5.5 V, I would not set the voltage based on the reading at the power supply. I would set it at 5.0 V based on the Vcc to GND voltage at the logic board. This takes into account all the power and ground wiring from the power supply and still gives you ±10% tolerance.

 

I have seen some YouTube arcade techs go as high as 5.3V (and that's asking for it when you go that high).

IMHO, it would be better to ensure the 5V / ground wiring from the supply to the board (or board stack) is in good shape and capable of handling the needed amperage. Then find a 5V / ground connection at the far end of the board or board stack and adjust for 5.0+ at that end of the board.

 

Welcome to AAC!

I have always set the game logic boards at 5.17 volts

Standard TTL voltage tolerance is 5%, so 4.75-5.25V. Military spec parts are 10% tolerance.

An esteemed colleague has now said this is too high. Assuming the power supply is ok, anyone have any arguments for or against?

I agree with your colleague. I'd set the voltage closer to 5V at the boards.

 

Standard TTL voltage tolerance is 5%, so 4.75-5.25V. Military spec parts are 10% tolerance.

Been a long time and I could be wrong but I thought the 54XX series of chips are the military grade (5%) and the 74XX are commercial grade (10%). I think you got it backwards. Again, I could be wrong.

 

From National Semi (1994) 74F00 datasheet.


Then there is TI datasheet (1993) for 74F00 that shows 10% for both



But the 7400 TI datasheet shows:



So kind of depends on which "TTL Family" you were using.

 

Been a long time and I could be wrong but I thought the 54XX series of chips are the military grade (5%) and the 74XX are commercial grade (10%). I think you got it backwards. Again, I could be wrong.

From the clip in my post, 74 series are 5% and 54 series (military) are 10%.

 

Thanks for the correction. I was the one who had it backwards.

 

Trying to settle an argument. Based on over fifty years experience as a coin-op tech and consultations with other professionals over the years, I have always set the game logic boards at 5.17 volts to eliminates pesky reset problems. (and provide a cushion for less than ideal line voltages at various locations). An esteemed colleague has now said this is too high. Assuming the power supply is ok, anyone have any arguments for or against? Thank you...

I don't know that you will find a definitive answer because you and our esteemed colleague are probably talking apples and oranges, to some degree. Is your colleague coming from a background of working for decades with real arcade systems (of the type you worked with) operating in the real world? If not, then they are coming from a more theoretical viewpoint and, as such, is having to make some assumptions that may not apply to the situations you are talking about. You, on the other hand, while you may not have been bothering to explore the situation as deeply as would be required to really back up your approach, used an approach that was based on getting things to work in the real world, without any assumptions at play. As we used to say on the flight line -- sometimes ya gotta do whatcha gotta do.

 

I have seen some YouTube arcade techs go as high as 5.3V (and that's asking for it when you go that high).

IMHO, it would be better to ensure the 5V / ground wiring from the supply to the board (or board stack) is in good shape and capable of handling the needed amperage. Then find a 5V / ground connection at the far end of the board or board stack and adjust for 5.0+ at that end of the board.

I would not trust suggestions found on yootoob. They might be valid, or not.

 

I was a hardware engineer at an arcade game manufacturer back in the 80s and the simple answer is: it depends.

Some of the more old school engineers based their designs on "typical" gate delays to get clock timings. This had a tendency to be both temperature, voltage, and batch dependent. You might get some boards out of a run that would be flakey at the extremes. If these were being used in our cabinets, the voltage would be tweaked to keep the board stable.

I designed to max and min specifications and never had temperature/voltage issues (at least that I heard about) so things were solid as long as the voltage at the board was between 4.75 and 5.25.

Just after I started there, we started making JAMMA kits so a policy was implemented of putting a 5.6 volt 5 watt zener directly across the power rails at the connector. This provided protection against ESD and also kept the boards from being fried by the generally crappy power supplies the arcade owners used. An artifact of this was that the 5% tolerance on the diodes meant that feeding more than 5.32 volts to the board risked engaging the zener which would then melt into a short having given its life to save the board.

In general, I recommend running the boards with between 4.75 and 5.25 at the cable side of the board connector as this was what they were designed to. Running at the low end of the range will keep power dissipation down and extend the life, but that's not necessarily reliable in the presence of noise.

On later generations of boards, there might be an onboard power supply to provide the 3.3 volts (or lower) needed by the logic. If that's the case, a higher supply voltage just means more power dissipated in the power supply with no effect on performance.