Programming (non-JTAG) MAX7000 devices

[abyrvalg]

migry, I've tried to disassemble the AMAX70.EXE a bit and have identified the erase function. I can see the 10 42 08 21 84 10 42 08 21 84 pattern matching your log, the ReadID function sends it and receives 5x16 bits back from the chip. Those bits should contain an "ALTERA92" (for EPM7032) or "ALTERA93" (for EPM7032V) in some form. I could dig further if you share LPT pins to programmer regs mapping (don't want to dig into programmer schematics and PAL contents since you did it already).
I can see some sequences looking like prommer regs accesses, i.e. write reg addresses to LPT DATA port, pulse INIT pin, write reg value to DATA port, pulse AUTOLF, but it would be much faster to reuse your info (i.e. reg addresses, who are all those E0, E1,E3 etc I'm looking at).
Or you just follow https://stackoverflow.com/questions/6947413/how-to-open-read-and-write-from-serial-port-in-c to access serial port from DOSBox and implement that bridging. Beware, there are two IC-specific delays (called Tbe and Tpg in AMAX70.EXE, looks like it gets them straight from the ID data), both affect some pulse lengths sent via reg E2. Not sure how critical they are.

[fisafisa]

I am sitting on quite some of these (for now unusable) devices...

Just an idea:
Run dosbox-x in a PI and use the PI IO/pins to interface to the device to be programmed.
Some voltage level shifting and possibly some IO expansion might be necessary but modifying dosbox so that every attempt to write/read from the HILO board is trapped has already been done successfully right ?

Would it be possible to  share the modified dosbox-x code...

Ciao

[migry]

It's been quite some time since I last looked at this project, but it has not been forgotten.

I have about 50 new EPM7032s, but no way to program them.

Just recently I completed the following break out PCB. This will allow me to tidy up my breadboarded circuit.



I am happy to share the DOSBOX code, although I need to check that it is still working now I have transferred all files to a new PC. If this board allows messaging, please contact me via that method.

[Beta_vulgaris]

Altera Classic devices, PROGRAMMING Pin-outs for reference.

[Zapotec]

Good morning !

I need to program EPM7032SLC44 with files made by a project writte in Quartus 13.0.1. I need to use all pins, JTAG pins too
The nice thing is that I have an (old) Elnec LabProg48LV programmer, that support this device. BUT.... its fail to program leaving the chip blank, after no error during programming, but (obviously) fails the verify.

OK I bought an USBBlaster and made a off circuit only programming board.

The problem is that in Quartus, when I try to program device, it told me that JTAG is disabled in project, closing programming window.
I start from Menu Tools->Programmer, scan JTAG, found device EPM7032S, load .JAM file, program. All fine, device works .. but !  jtags pins still enabled (so don't make my project funtions).
If I try to load .POF.... message : JTAG is disabled in file, can't load

If I put +12V (by resistor) on pin44, Quartus program fails also to find (jtag scan) the (blank) 7032S device in home made programmer

Have some idea ?
And wy my Elenec (but I try also with a newer one) falis to program the device (also fail Chip ID .. but is a known problem reported by Elnec).

One Thousand dollars question
thanks

[GTT95]

Hello,

I'm facing quite same situation.

In my EPM7128S based design I need all 84 IO pins for the expected function.

However, I need the ISP feature to be available so that I can reflash CPLD's EEPROM while it is already mounted on the target circuit.

So my question is: once flashed and JTAG pins assigned to functionnal IO, is there a way to take back control over those pins so that the chip can be reflashed? Or is a "special" (non JTAG) programmer required to "factory reset" the chip and make the JTAG specific pins usable again for reprogramming?

Thanks a lot.

[marcopolo]

That's not possible, you need a LP6 ISA board + PL-MPU+ PLM7000-84 Adapter

This is the problem that many people who buy Altera CPLDs on Ebay encounter.

Perhaps you can use a FLEX10K FPGA.

Marc

[GTT95]

So only by keeping the 4 JTAG pins as dedicated to the ISP function I can continue in-circuit reflashing?

My understanding is that devices from ebay may not be blank leading to the unability to in-circuit reflash through JTAG

The FLEX10K might be a good alternative. However the ratio between max user IO ant total pins is lower than MAX7000S series (only 66 IO for 100pins devices). Other drawback is a serial configuration Flash is required which adds an extra component to the design.

[Beta_vulgaris]

So only by keeping the 4 JTAG pins as dedicated to the ISP function I can continue in-circuit reflashing?

My understanding is that devices from ebay may not be blank leading to the unability to in-circuit reflash through JTAG

The FLEX10K might be a good alternative. However the ratio between max user IO ant total pins is lower than MAX7000S series (only 66 IO for 100pins devices). Other drawback is a serial configuration Flash is required which adds an extra component to the design.

You can use a parallel PROM, although more pins, it can be multiplexed after configuration done as user I/Os (serial PROMs still the same, but not so ubiquitous as parallel ones).

[GTT95]

Are you refering to my last remark about configuring the FPGA? Was not aware it can be configured from parallel Flash. Anyway, there will still be limitation to 66 pins IO which will force selecting a higher FPGA  with more total pins.

[marcopolo]

Anyway, there will still be limitation to 66 pins IO which will force selecting a higher FPGA  with more total pins.

And what is the problem?
Could you not use a 144-pin TQFP EPF10K10 with 102 I/O?

http://marc.retronik.fr/tmp-mj/AN116_Configuring_Flex6000_et_Flex10K.pdf

[GTT95]

A 144 TQFP means more than twice that much area and I am very restricted in PCB size.

[marcopolo]

To stay under 100x100 mm for a cheap PCB?

[GTT95]

Could be but actually reason is it is constrained by credit card size PCB.

[marcopolo]

Why use old 5V circuits?

[GTT95]

Because the design has to interface with a system delivering 5V TTL levels. Using a 3,3V device would require level translators through resistor based voltage dividers or other kind of ICs.

[marcopolo]

A 144 TQFP means more than twice that much area and I am very restricted in PCB size.

PLCC-84  30x30 mm (without socket)
TQFP-144  23x23mm

TQFP-144 area is 60% of the PLCC-84 area.
it leaves room for a serial configuration memory.

[YetAnotherTechie]

Can't you retarget to a ATF1508AS? as described in Atmel’s Application Note 0916, or http://avitech.com.au/?page_id=3195

[GTT95]

PLCC-84  30x30 mm (without socket)
TQFP-144  23x23mm

Agree, but the comparison was against the TQFP-100 version of EPM7128S. PLCC-84 version doesn't provide 84 user IO.

Can't you retarget to a ATF1508AS? as described in Atmel’s Application Note 0916

Was not aware of this Atmel equivalent. What benefit is there from using the Atmel CPLD? There are still 84 max IO pins. My guess is that I will experience same limitation while using the JTAG pins for ISP, preventing their reuse for functionnal IO.

[YetAnotherTechie]

The biggest benefit is that it's still in production, so you don't have to depend on the used market. It's also less power, more erase cycles and its jtag can be re-enabled by putting 12 volts on OE pin. https://www.hackup.net/2020/01/erasing-and-programming-the-atf1504-cpld/. There might be details that don't fit your use case, they are listed on the original atmel apnote.

[GTT95]

That sounds good. And also better documented about how to perform reprogramming. Let's know if it can be synthesized from Quartus II 9.0.

[Beta_vulgaris]

That sounds good. And also better documented about how to perform reprogramming. Let's know if it can be synthesized from Quartus II 9.0.

Yes, MAX 7000S/AE/3000A, FLEX 10K/KA/KE are supported up to Quartus II 13.0.

[GTT95]

OK, but what about ATF1508AS? Seems not to be available as target device.

[Beta_vulgaris]

OK, but what about ATF1508AS? Seems not to be available as target device.

ATF1508AS can be converted from EPM7128S/AE by Microchip POF2JED.

[GTT95]

OK thanks, good to know synthesizing ATF1508AS from Quartus is only a matter of file format.