# GAL Voltages

• posted

I'm about to attempt building my own basic GAL programmer after doing much reading on the web. But first, I've got me a few questions!

1.) For starters, why do GALs use all different programming voltages? Is it based on the chip model? Is it based on the company who made it? Both? Seems like it would have made more sense to just design them for a standard voltage, much like how apparently 12v is the standard to read their contents. Though maybe that'd have made them more expensive, I dunno. Or maybe they did it on purpose so you'd buy their programmers/software?

2.) Does going above the programming voltage by the fractions of volts by which a programmer can increment actually kill the chip? Cause they seem to be rather small increments.

3.) Why are the programming voltage and timings stored in the PES, when from what I understand the PES is erased when you do a chip- erase? I guess this means you have to make sure you read the PES out first and save it, then write it back in after writing in your new fuse settings, and hope nothing goes wrong in the meantime to lose that configuration info.

4a.) Since I don't have a bench power supply, I'm going to have to try to make do with something else. Would it be possible to use something like a 7805 with a potentiometer on its ground, powered by maybe 18-20v (whatever AC adapter I can find), to generate the programming voltages?

4b.) Is it even safe to run an '05 much higher than its standard output?

4c.) How does that affect its current output? And relatedly, am I right in the assumption that the higher the output voltage (getting closer to the input voltage), the cooler the '05 would be?

4d.) Would I need a resistor temporarily across the output of the 7805 to properly test its output voltage with a multimeter? Since I assume the voltage might differ when connecting the meter directly compared to when the '05 is connected to the chip and erasing. If it's as sensitive to the voltage range as I've assumed, then I'd have to make sure the voltage I generated was very accurate and all.

Forgive my latter questions if they sound dumb, but there's still much that confuses me when it comes to "analog" electronics, as opposed to digital circuits. But I'd appreciate any help anyone can offer!

• posted

Use an LM317 or equivalent as your variable regulator. Dirt simple to setup; see the data sheet. The voltage divider that sets the output voltage also provides it a minimum load.

```--
Rich Webb     Norfolk, VA```
• posted

Hiya, thanks for the response. Did ya get buried in the snow there where you are in VA? Northeast TN here just ended up with a big pile of rain.

7805 in particular for a couple of reasons. First, because that's all I have in parts at the moment without having to order anything. But second, because I saw it being done. The schematic for an older Willem programmer uses a 7805 and 7812, and both can have their outputs changed based on diodes in series on their ground pin. Jumpers can be set to bypass certain diode(s) to change the output voltage. The jumpers on the 7812 for example can be configured to output 12.5, 15, 21, and 25v. I thought this might be possible using a pot or something instead of diodes. Though the way they're doing it in series with the ground is what prompted my questions about load, too, since I'm not sure how it affects the output to do it the way they are.
• posted

Just enough to cover the ground, this time.

A quick look at the LM78xx datasheets shows that most of the adjustable examples use a unity gain opamp to isolate the ground (which becomes the ADJ pin) from the voltage divider that sets the reference voltage. E.g.

Nothing wrong -- and probably much to be gained -- from trying both ways and comparing the operating characteristics.

```--
Rich Webb     Norfolk, VA```
• posted

Yes, you can use the 78xx series of regulators to produce higher output voltages. Use the same circuit as for the LM317, but if you are using a 7805, use 5 volts where the LM317 formula uses 1.2 volts - the LM317 is really a 1.2 volt regulator optimised for this "lifted ground" use.

When using the 78xx series this way, you should check on the current through the ground pin, as it will affect the output voltage if it varies significantly with load current.

```--
Peter Bennett, VE7CEI
peterbb4 (at) interchange.ubc.ca  ```
• posted

Not in my experience. The 78xx family of regulators do have a minimum load current spec, below which the device is not guaranteed to meet its specs. However, I have always found that, with no load other than my meter, the output voltage is close enough to the spec for me.

```--
Peter Bennett, VE7CEI
peterbb4 (at) interchange.ubc.ca  ```
• posted

(I speculate): The main design goal is to get the logic array as fast as possible, so they would design the chip towards that goal, and then characterize the programming algorithm to give the best results for speed and storage reliabilty. Toss in the use of different semicondutor processes as time goes on, and thing get very variable.

PLD manufacters outsourced the manufacturer of device programmers (and the customer support costs that go along with that).

My guess is that the numbers came from what gave the best programming yield and aren't that critial.

As I remember, the PES erase was seperate, but it's been over 15 years since I was hacking around with this stuff.