5V PAL or GAL with CMOS Outputs

[GonzoTheGreat]

Fellow Tinkerers,

I am restoring some vintage equipment and I am in a dire need of some 5V Programmable Logic Devices (PALs, PLAs, GALs) in DIP packages, that have CMOS levels compatible outputs.  None of that TTL V_OH = 2.4V rubbish !

Does anyone know part numbers of any old PLDs, that have 5V CMOS compatible outputs with V_OH > 3.5V ? 
They can be obsolete devices...

[edavid]

PALCE16V8Z/20V8Z/22V8Z have CMOS output, although it's not very well specified.

[basinstreetdesign]

Perusing a few of the ones on Digi-Keys website and looking at datasheets there seem to be quite a few that will go high within 0.2 or 0.3V of VCC as long as the output current is LE 100 uA.
Otherwise if you need to supply a bit more current, then hanging a pullup resistor on the output will do it.  It should be at least a few hundred Ohms up to a couple K.

The only other option is more silicon in the form of a buffer or a transistor.

[Zero999]

Yes, a pull-up resistor would help it to reach +V but beware it would also mean it will compromise the output's ability to reach 0V.

A 74HCT logic IC (e.g. 74HCT244, 74HCT125, 74HCT14 etc.) can be used as a level shifter to translate from 3V to 5V.

[GonzoTheGreat]

PALCE16V8Z/20V8Z/22V8Z have CMOS output, although it's not very well specified.

How so?
The datasheet clearly states min V_OH  = 2.4V

[GonzoTheGreat]

Perusing a few of the ones on Digi-Keys website and looking at datasheets there seem to be quite a few that will go high within 0.2 or 0.3V of VCC as long as the output current is LE 100 uA.

The outputs of this PLD drive 74HC inputs so the DC current is minimal but the AC current is significant at 20MHz.
Which parts list non-TTL V_OH ?

Otherwise if you need to supply a bit more current, then hanging a pullup resistor on the output will do it.
The only other option is more silicon in the form of a buffer or a transistor.

Adding more components to support a PLD chip kind of defeats the purpose of using a PLD chip.
Anyway, this is a restoration project that has to look "original".  I can't have someone look and see a budge-job with bunch of hot pull-up resistors.

[edavid]

PALCE16V8Z/20V8Z/22V8Z have CMOS output, although it's not very well specified.

How so?
The datasheet clearly states min V_OH  = 2.4V

That is the wrong datasheet.  It's only the Z suffix "zero power" family that has CMOS output.  (The other PALCE parts and GAL parts appear to have NMOS output, although that seems very odd for the GAL Z series, which is supposed to be a replacement for the PALCE Z series.)

Here's the correct datasheet: http://latticesemi.com/~/media/LatticeSemi/Documents/DataSheets/PAL/PALCE22V10PALCE22V10ZDataSheet.PDF?la=en
See page 30 and page 22.

[GonzoTheGreat]

That is the wrong datasheet.  It's only the Z suffix "zero power" family that has CMOS output. 

Thanks for directing my attention to the "Z" suffix !
I think, I will use these devices because my local supplier has them.

Also, someone else has suggested this device:

True CMOS Outputs:
Each TICPAL22V10Z output is designed with a P-channel pull-up transistor and an
N-channel pull-down transistor, a true CMOS output with rail-to-rail output switching. This provides direct
interface to CMOS logic, memory, or ASIC devices without the need for a pull-up resistor. The CMOS output
has 16-mA drive capability, which makes the TICPAL22V10Z an ideal substitute for bipolar PLDs.

(The other PALCE parts and GAL parts appear to have NMOS output, although that seems very odd for the GAL Z series, which is supposed to be a replacement for the PALCE Z series.)

Indeed it is odd

@Others
Any other PALs or GALs out there with true rail-to-rail CMOS outputs?

[danadak]

One possibility is to use a CYPRESS PSOC 5LP, it has a fabric that
is routable, tool has logic primitives (see attached) that are drag
and drop in a GUI tool, and can also be programmed in Verilog.

Only issue no DIPS, you would have to build an adapter board.

Is CMOS R-R I/O, and has programmable pull up/dwn on I/O pins.

Has ARM core if you need it, but you don't have to use it. But could be
used for startup control, and rest of analog stuff (A/D, D/A, OpAmp, Vref,
....) for other tasks.

Regards, Dana.

[GonzoTheGreat]

So I bought several of these AMD PALCE22V10Z chips but now I have a problem with programming them.

I cannot find the programming algorithm to build my own programmer and the Supported Devices List of a commercial programmer, that I have (the Chinese TOP2011), does not list the chips with the "Z" suffix and another programmer, I have (the TL866), does not even have the AMD PALCE chips on its Supported Device List.

Why is the programming algorithm not available on the Internet after so many years?

What to do?

[mikeselectricstuff]

So I bought several of these AMD PALCE22V10Z chips but now I have a problem with programing them.

I cannot find the programming algorithm to build my own programmer and a commercial programmer that I have (the Chinese TOP2011)  does bot have the chips with the "Z" suffix on its Supported Devices List.

Why is the programming algorithm not available on the Internet after so many years?

What to do?

PLD programming specs were generally very hard to come by back in the day - I think because manufacturers were worried that incorrect programming would impact performance and cause support issues. This probably started when PALs used physical fuses, where it was important to get the waveform right.
Flash devices are  much less critical, and I think it's quite likely that programming is the same for different variants of similar devices, so would be worth trying the programmer you have on the Z devices - from memory Z iwas a variant to reduce standby power draw.

[GonzoTheGreat]

PLD programming specs were generally very hard to come by back in the day

But today this should not matter to the manufacturer, since these chips have not been produced by them for years.
I though that Internet does not forget anything.

...from memory Z was a variant to reduce standby power draw.

Most importantly, chips with the "Z" suffix operate on the CMOS logic levels, which means that they require much more than the 2.4V to register a HIGH logic level with a 5V power supply.

[Andy Watson]

But today this should not matter to the manufacturer, since these chips have not been produced by them for years.
I though that Internet does not forget anything.

As Mike said, the manufacturers were (for some reason?) very guarded about releasing the programming information. So it never made it on to the internet.

The datasheet suggests that the programming for the "Z" series is the same as the standard series - you will probably have to take a gamble and try the programming algorithms for a standard 22V10. I have a vague memory that Elektor published a description of the algorithm many, many years ago - I'll see if I can find the magazine - but don't hold your breath!

[GonzoTheGreat]

As Mike said, the manufacturers were (for some reason?) very guarded about releasing the programming information. So it never made it on to the internet.

There is something very evil about that...or unethical at least. 

There have been successful lawsuits about manufacturers not releasing complete VCR programming instructions, so why none for the PLD chips which cannot be even partially used without programming?

[GonzoTheGreat]

The datasheet suggests that the programming for the "Z" series is the same as the standard series - you will probably have to take a gamble and try the programming algorithms for a standard 22V10.

Right after I reverse engineer the command set for that TOP2011 programmer and patch its software to output CMOS logic levels instead of the TTL levels for the non-"Z" PALCE chips.

...or make an external TTL-->CMOS level converter, once I figure out which input pins need this conversion.  It certainly would not be all of them....

[mikeselectricstuff]

As Mike said, the manufacturers were (for some reason?) very guarded about releasing the programming information. So it never made it on to the internet.

There is something very evil about that...or unethical at least. 

There have been successful lawsuits about manufacturers not releasing complete VCR programming instructions, so why none for the PLD chips which cannot be even partially used without programming?

Simple - people were expected to use a commercial programmer. Nothing unethical about that.
Remember in those days, hobbyists weren't using these things to any significant extent.
The design software ( e.g. CUPL) was expensive ( from memory over GBP1K) .
Why would manufacturers go to the hassle of documenting and supporting programming specs when they could just deal with a handful of programmer manufacturers, and change the specs whenever they wanted if it allowed improved performance for a new part.
 
 

[GonzoTheGreat]

There is something very evil about that...or unethical at least. 

There have been successful lawsuits about manufacturers not releasing complete VCR programming instructions, so why none for the PLD chips which cannot be even partially used without programming?

Simple - people were expected to use a commercial programmer. Nothing unethical about that.

Yes there is, because it creates a racket designed to extort money from the end users, while a cheaper solution exists and is purposely withheld.

Would you feel it was ethical if a VCR manufacturer offered to send you a $500/h serviceman to program your VCR ?
...and a VCR can be at least partially used without programming.

[mikeselectricstuff]

There is something very evil about that...or unethical at least. 

There have been successful lawsuits about manufacturers not releasing complete VCR programming instructions, so why none for the PLD chips which cannot be even partially used without programming?

Simple - people were expected to use a commercial programmer. Nothing unethical about that.

Yes there is, because it creates a racket designed to extort money from the end users, while a cheaper solution exists and is purposely withheld.

Would you feel it was ethical if a VCR manufacturer offered to send you a $500/h serviceman to program your VCR ?
...and a VCR can be at least partially used without programming.

The VCR analogy is meaningless as it's a totally different market.
The purpose was to ensure customers had a reliable way to use their parts.
Most of the customers would be commercial places with the budget to pay for it, for whom making their own programmer would have cost more than buying one.
It would make no business sense for them to support random DIYers lashing up their own soluitons. Remember this was a fast moving competitive market at the time.
I have no idea how hard it was at the time to get programming specs - there were certainly a significiant number of programmer manufacturers so couldn't have been that hard once you'd convinced them you knew what you were doing. 

[edavid]

Right after I reverse engineer the command set for that TOP2011 programmer and patch its software to output CMOS logic levels instead of the TTL levels for the non-"Z" PALCE chips.

...or make an external TTL-->CMOS level converter, once I figure out which input pins need this conversion.  It certainly would not be all of them....

It is quite likely that the programmer always puts CMOS logic levels on its output pins during programming.  There's no reason not to.  Hook up your scope and see.

If not, you would just have to add pullup resistors, not level converters.