Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Using memory without a microprocessor/microcontroller

Status
Not open for further replies.

carbonzit

Active Member
Reading another thread here, in which someone who may or may not be a troll (jury still seems to be out there) asked about using an EPROM they have without a microprocessor. Of course, their proposal was hopelessly naive, not well thought-through, etc., etc.

So I'm asking, seriously and out of curiosity, whether it is possible to do this, to use EPROM (or some other flavor of non-volatile memory) in a circuit without the usual support of a µc/µp.

Now, I know that this is possible, given enough supporting circuitry; what I'm really asking is whether it's at all practical.

Even if it is not, which I suspect is probably the case, I am curious to know just what kind of support circuitry would be required to read and write such memory without a proper micro. I do have a bunch of electronics textbooks, but none of them deal specifically with memory.

I mean, it would be nice to be able to squirrel away a few bits and bytes of information in NVRAM without having to go through all the programming and such ...
 
hi CZ,
The parallel NOVARAMs ie: 8*16Kb,32Kb etc are fairly easy to do.

Basically it requires a binary counter IC and one or two extra gates.
Which have you got of this type of NRAM.???
 
Last edited:
The parallel NOVARAMs ie: 8*16,32 etc are fairly easy to do.

Basically it requires a binary counter IC and one or two extra gates.
Which have you got of this type of NRAM.???

Nothing; I was just responding to that earlier thread. But I am curious just how hard it would be to be able to save a few configuration parameters in a small circuit.

So, hypodermically speaking, what would it take, say, to save say half a dozen 16-bit values? Any ready-made circuits out there? What kind of memory (non-volatile) would be best? What support circuitry would be required? Counters, obviously, plus other gates, I suppose?
 
Nothing; I was just responding to that earlier thread. But I am curious just how hard it would be to be able to save a few configuration parameters in a small circuit.

So, hypodermically speaking, what would it take, say, to save say half a dozen 16-bit values? Any ready-made circuits out there? What kind of memory (non-volatile) would be best? What support circuitry would be required? Counters, obviously, plus other gates, I suppose?

hi,
Say for 16 bytes by 8bits, you would want a HEF4024 [ or equivalent] 7 bit counter to sequentially address the first 0 to 15 addresses of the NRAM.

If you just wanted to select certain bytes from the 16 address, you could use bit switches or parallel load binary counter.

Control lines to a NRAM are CS, R/W and of course the 8bit I/O pins
 
To read a non-volatile memory is generally quite simple. You just apply the binary address of the desired memory location to the address pins and the bits appear on the data pins. I used a ROM that way to perform a conversion of a logarithmic A/D converter output words to generate the linear output from a DAC. The A/D words went to the PROM address which held the corrected linear word for that particular input value which it then output.

Writing is somewhat more complicated and depends upon the write characteristics of the memory, which you get from the data sheet. But basically you again apply the address to the address pins and the desired word you want to write to the memory data pins. The you perform whatever commands the memory requires to write the data.
 
So, no messing around with row & column strobes and such? I never delved deeply into memory operation, and looking at circuits and datasheets always made it seem like a daunting task.

I'd still like to see a simple circuit to illustrate all this ...
 
So I'm asking, seriously and out of curiosity, whether it is possible to do this, to use EPROM (or some other flavor of non-volatile memory) in a circuit without the usual support of a µc/µp.

Now, I know that this is possible, given enough supporting circuitry; what I'm really asking is whether it's at all practical.

Yes -entirely practical, given the right parameters. I've used EPROM for non volatile storage of audio messages - used, in my case, with emergency audio systems - for many years.
When used for the broadcast of evacuation instructions, in conjunction with fire alarm systems for example, there is absolutely no need for a microprocessor or microcontroller to be involved. In fact, for many years (not so much now) the use of any software control of the system was specifically excluded. Deemed to be too unpredictable!
This was primarily due to the tendency for simple microprocessor software systems to either involve the use mechanical 'spinning discs', or more controversially, the use of 'untestable' software routines, that used interrupts.

As a result, EPROM stored recordings, simply 'read' by binary address counters, could provide a non-degradable, non mechanical, and reliable, audio message replay system.

On the analogue side, there was quite a lot of audio 'compansion' involved, to extract reasonable audio quality from 8 bit recordings, but the control of the actual 'memory' media did not require any 'microprocessing' power at all.

Bit of a special case perhaps, but definitely viable!
 
I built a wind speed indicator of a counter, a timebase and a pair of eproms of display values. The timebase would set cadence of the counter, whose output would be the addresses of the eproms. Its' output is the segment data for the displays. Mine is calibrated to the Heathkit anemometer from their weather station series units.

If only as example: a storage object isn't necessarily bound to a processor, BUT it Does take creative thinking to apply it without.

P.S. tables on request, schematic negotiable. G.H. <<<)))

Memory/data storage is more a concept than a hardware circuit / software consideration.... it's your design...
 
Last edited:
So it's starting to look as if the person who started that other thread (link given above) really got a raw deal there. It's true that they came barging in, as so many Indian posters here seem to do, demanding answers and using very bad English. However, he really got the bum's rush. I thought that this answer from Nigel Goodwin was particularly dismissive, especially given the answers I've seen here:

RITESH KAKKAR said:
OK, i was thinning they use ic's counter which automatically store the data.
Nigel Goodwin said:
If you want a portable radio that costs hundreds of pounds, is huge, and requires a strong man to carry it, then you 'could' do it in hardware - but it would be a completely pointless exercise.

Nigel's unspoken assertion being that you have to use a PIC or some other kind of microprocessor here, or else use literally pounds and pounds of hardware, which is clearly not the case ... oh, well, I won't say anything more.
 
Last edited:
I have used EPROMs such as 2716 and 2732 as a form of logic.
Use the address lines as the inputs, the data lines as the outputs and program the EPROM with data which sets the outputs that you want for any given input.

I have also used an EPROM to store a morse code message for a radio beacon.
The address lines of the EPROM are driven by a counter and one of the data lines drives the radio keying circuit.
Write data to the EPROM to give the required message, easy.

Neither of these applications uses a microprocessor of any kind.

However, in the general context of using an EPROM or any other kind of electronic memory to store data, you need some kind of microprocessor to read and write the data.
Relating this to the question posed by Ritesh Kakkar, yes you need a microprocessor
to use the EPROM.
No he did not get a raw deal, in common with a number of persons who have visited this forum over the years, he keeps asking questions from position of complete ignorance of the subject, lacking the underlying knowledge required to understand the answer.
When he is given a reasonable answer he not only does not understand the reply, he refuses to understand the reply.
Having visited India on business on several occasions, this is a trait exhibited by many of his fellow countrymen.

Many of us here work with electronic and computery things for a living, and have done from several decades. We also like to do electronic and computery things for a hobby and are happy to assist others to grow their knowledge.
However in case like this, my patience ran out a long time ago and I can do without nonsense like this.

JimB
 
I have used EPROMs such as 2716 and 2732 as a form of logic.
Use the address lines as the inputs, the data lines as the outputs and program the EPROM with data which sets the outputs that you want for any given input.

I have also used an EPROM to store a morse code message for a radio beacon.
The address lines of the EPROM are driven by a counter and one of the data lines drives the radio keying circuit.
Write data to the EPROM to give the required message, easy.

Neither of these applications uses a microprocessor of any kind.

However, in the general context of using an EPROM or any other kind of electronic memory to store data, you need some kind of microprocessor to read and write the data.
Relating this to the question posed by Ritesh Kakkar, yes you need a microprocessor
to use the EPROM.
No he did not get a raw deal, in common with a number of persons who have visited this forum over the years, he keeps asking questions from position of complete ignorance of the subject, lacking the underlying knowledge required to understand the answer.
When he is given a reasonable answer he not only does not understand the reply, he refuses to understand the reply.
Having visited India on business on several occasions, this is a trait exhibited by many of his fellow countrymen.

Many of us here work with electronic and computery things for a living, and have done from several decades. We also like to do electronic and computery things for a hobby and are happy to assist others to grow their knowledge.
However in case like this, my patience ran out a long time ago and I can do without nonsense like this.

JimB

Need I say.............Booom!

No.
 
However, in the general context of using an EPROM or any other kind of electronic memory to store data, you need some kind of microprocessor to read and write the data.
Relating this to the question posed by Ritesh Kakkar, yes you need a microprocessor
to use the EPROM.

Not so, according to several other replies above (for instance, #7 and #8). Or are you saying that these statements are not correct? They claim they built devices that use EPROM without a microprocessor. I'm just sayin' ...
 
It can be done but the circuitry is more complex. I remember about 30 years ago a friend and I built a light control board for a band to take on the road. It was used for stage lighting. It had 16 output channels for colored stage lights. What made it advanced for that time was that you could program it with up to 16 different combinations of the stage lights in the memory, and then during the concert select the one you wanted with a keypad. It also had a "real time" mode where the lights were controlled individually with 16 toggle switches. Anyway we used some type of parallel input/output eeprom, they were 4 bit by 16 locations so we needed 4 of them to control 16 lights. So there were a lot of connections.
 
Last edited:
So it's starting to look as if the person who started that other thread (link given above) really got a raw deal there. It's true that they came barging in, as so many Indian posters here seem to do, demanding answers and using very bad English. However, he really got the bum's rush. I thought that this answer from Nigel Goodwin was particularly dismissive, especially given the answers I've seen here:



Nigel's unspoken assertion being that you have to use a PIC or some other kind of microprocessor here, or else use literally pounds and pounds of hardware, which is clearly not the case ... oh, well, I won't say anything more.

hi CZ,
Note there is a significant difference between a 'parallel' and 'serial' type NRam addressing and data read/write.
The serial [ bit wise types] that Ritesh wanted to use does require a lot of logic support compared to the parallel [ byte wise types].
 
So it's starting to look as if the person who started that other thread (link given above) really got a raw deal there. It's true that they came barging in, as so many Indian posters here seem to do, demanding answers and using very bad English. However, he really got the bum's rush. I thought that this answer from Nigel Goodwin was particularly dismissive, especially given the answers I've seen here:



Nigel's unspoken assertion being that you have to use a PIC or some other kind of microprocessor here, or else use literally pounds and pounds of hardware, which is clearly not the case ... oh, well, I won't say anything more.

Best not to as you're talking rubbish :D

He was asking about I2C EEPROM's, all the answers here (which still require a goodly number of chips, LOT'S more complicated than using a processor) are refering to bus addressed EEPROM chips, much larger chips, and much easier to use without a processor.
 
Best not to as you're talking rubbish :D

He was asking about I2C EEPROM's, all the answers here (which still require a goodly number of chips, LOT'S more complicated than using a processor) are refering to bus addressed EEPROM chips, much larger chips, and much easier to use without a processor.

Fair enough, I suppose, although I wasn't talking rubbish, I was speaking from ignorance, which is something different.

A lot of the rancor in that other thread could have been avoided had you written there what you just wrote here, instead of jumping all over that person. Just because they were rude shouldn't be an invitation for everyone else to be too. There's always the choice to STFU, you know ... which I'm going to take now.
 
Last edited:
Something I should clarify: while the end/target device uses an EPROM; I don't know if you Could program/set its' patterns without a micros' environment or Very similar. As mentioned, it's just a simple flavor of a FPGA. Can be an elegant answer pending state machine complexity. G.H. <<<)))
 
It is difficult to generate I2C and SPI commands using dedicated hardware, but I think it could be done using the parallel port of a computer.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top