W27C512 programmer

[bitman]

Hello - I'm posting in beginners because that's what I am and since I found a few other posts here around the same type of topic, I figured it was the better option than bugging those who know better than I.

I got a few W27C512 very cheap and figured I would give it a whirl to limit the number of chips needed for some of my plans. Reading the datasheet (https://media.digikey.com/pdf/Data%20Sheets/Winbond%20PDFs/W27C512.pdf) gave me more gray hair than it should though. I had not realized I needed different voltages to overwrite data. So my first step is to create a reader and then add a writer component to it. I don't think spending $300+ to do that is worth it, and I think this will be plenty of a challenge to help me grow a bit more into this. But I have a few questions.

I have an arduino mini lined up and ready to do this.

1) How do I protect the 5v circuits connected to A9/OE when I apply 12 or 14 volts? Not sure how to design a "buffer" around this.
2) I plan on using two simple 74HC595's for addresses since well the arduino doesn't have enough data lines to do both data and addresses. I don't see anything preventing this from happening. Most of the time limits are minimum, and the max limits seems to all be around erasing which isn't using addressing.
3) Typical newbie question: Is it best to step up voltage or step down? Should I look for a 14 volt supply voltage and step everything down, or step up from 5v? I think stepping up is best as that's the typical operation voltage. Only during erase/write is higher voltages needed.
4) I'm totally lost on section 10 AC Characteristics. What is this telling me? What's the diagram supposed to do - is that what's in the chip, or something that's expected to be in the circuit using the chip?

[james_s]

Would something like a TL866 do it? I don't have mine handy at the moment so I can't check but it handles most of the devices I've tried it on. Cost me less than 40 bucks.

[bitman]

Would something like a TL866 do it? I don't have mine handy at the moment so I can't check but it handles most of the devices I've tried it on. Cost me less than 40 bucks.

Hmmm - that seems to be software control only - and Windows only too, so that's a no go out of the door, even if the chip was supported.  It's a dirt old chip so I would recon basic old stuff would support it. However, I would lose out on a learning opportunity

[edavid]

I got a few W27C512 very cheap and figured I would give it a whirl to limit the number of chips needed for some of my plans. Reading the datasheet (https://media.digikey.com/pdf/Data%20Sheets/Winbond%20PDFs/W27C512.pdf) gave me more gray hair than it should though. I had not realized I needed different voltages to overwrite data. So my first step is to create a reader and then add a writer component to it.

You would really be better off forgetting the W27C512s and buying some 5V flash chips instead, e.g. 29F001.

Or if you really want to use the W27C512s, find a friend with a programmer.

1) How do I protect the 5v circuits connected to A9/OE when I apply 12 or 14 volts? Not sure how to design a "buffer" around this.

74LS07 or CD4504, or discrete equivalent
(but you don't care about A9 since you don't need to read the ID)

3) Typical newbie question: Is it best to step up voltage or step down? Should I look for a 14 volt supply voltage and step everything down, or step up from 5v? I think stepping up is best as that's the typical operation voltage. Only during erase/write is higher voltages needed.

Either way is fine.

4) I'm totally lost on section 10 AC Characteristics. What is this telling me? What's the diagram supposed to do - is that what's in the chip, or something that's expected to be in the circuit using the chip?

They show the test conditions you need to reproduce the values in e.g. section 12.

[james_s]

I don't see Windows only being all that much of a problem. My last device programmer required a real parallel port, I just kept an ancient laptop with Win98 on it around just for using that programmer. Eventually switched to a slightly newer but still ancient XP laptop. You don't have to use your primary PC.

[David Hess]

Besides using an open collector driver or a discrete solution, some TTL logic families have gates with a diode in series with the emitter of the upper transistor which allows the output to be pulled above the positive supply assuming that the gate output is not low.

The 7400, 74L00, 7402, 7404, and 74L04 are examples of gates like this however it is not guaranteed that they will work this way in their specifications so they need to be qualified by the user.

[bson]

The TL866 programs W27C512's just fine.  I used those EEPROMs and a TL866 to upgrade the firmware in my Keithley 2001.

[bitman]

You would really be better off forgetting the W27C512s and buying some 5V flash chips instead, e.g. 29F001.

Hmmm - since I already got these I'm going to give it a whirl.

Or if you really want to use the W27C512s, find a friend with a programmer.

I have a PIC programmer but it only has 10 pins or so. Clearly it's meant for serial communication and not this. 

1) How do I protect the 5v circuits connected to A9/OE when I apply 12 or 14 volts? Not sure how to design a "buffer" around this.

74LS07 or CD4504, or discrete equivalent
(but you don't care about A9 since you don't need to read the ID)

Hmmm - seems like a 74640 or similar would be better, since they're bi-directional? One more chip to control. Definitely running out of digital pins on my Nano

Thanks for your reply.

[karoru]

I don't see Windows only being all that much of a problem. My last device programmer required a real parallel port, I just kept an ancient laptop with Win98 on it around just for using that programmer. Eventually switched to a slightly newer but still ancient XP laptop. You don't have to use your primary PC.

Unless I'm getting paid for it I'm not going to keep computing museum at hand, despite the fact I started at DOS era. I've been hearing that "you don't have to use your primary PC" about 15 years ago when ugly "system("@ECHO sth > lpt1");" hack applications started to become a problem. Windows NT is a thing since nearly 20 years. Accessing ports via x86 Assembly was terrible even in Windows 95 era. Visual Basic 6 apps that use DOS commands to write to printer port instead of some sensible API were a poor hack  even in Windows 95 era. These applications for device programmers were terrible even when they were "modern". If we assume that desktop software written by EE's is 10 years behind general-purpose software, then we should be using Windows 7. Probably on 64bit machine;)

[edavid]

1) How do I protect the 5v circuits connected to A9/OE when I apply 12 or 14 volts? Not sure how to design a "buffer" around this.

74LS07 or CD4504, or discrete equivalent
(but you don't care about A9 since you don't need to read the ID)

Hmmm - seems like a 74640 or similar would be better, since they're bi-directional? One more chip to control. Definitely running out of digital pins on my Nano

You don't need a bidirectional buffer for the unidirectional OE line, and the 74640 doesn't help you since it's not HV tolerant.

[james_s]

Well I do use Win7x64 now, that's what I use for my TL866. I've still got plenty of old PCs around because old computers are another of my hobbies but that's beside the point. Win7 is the last viable version of Windows IMO. It's really quite good.

[daybyter]

I am working on a similar project at the moment. Arduino programmer for old eproms. I use a cheap nano clone, so I get data and power via the USB connection. I use 2 74hc595 for the address, although some 16 bit counter might be faster. I use a 5v => 35v step up to create the 25v programming voltage, that I need.

[CJay]

Given that programming times are in the mS per address the 74HC595 should be more than fast enough.

I built a reader with an arduino, a TTL buffer and a binary counter a few weeks ago to examine some old EPROMs (simpler than digging out the PC based programmer I have), I don't think it'd take much to turn it into a programmer, as a base for controlling the high voltages there are chunks of circuitry you could lift from the Willem (I *think* it's still open hardware?) which was IIRC derived from earlier work on PIC programmers, the David Tait one being one of the very first hobbyist ones?

[daybyter]

I used 4 normal transistors bc547 and 1 bc337, that I had here. No mosfets in my box... :-(
But it took me a couple of days and some help from the owner of ltspiceusers.ch to get to those 2us risetime.
I'm more of a software guy and electronics was more digital to me until this project.

[bitman]

I used 4 normal transistors bc547 and 1 bc337, that I had here. No mosfets in my box... :-(
But it took me a couple of days and some help from the owner of ltspiceusers.ch to get to those 2us risetime.
I'm more of a software guy and electronics was more digital to me until this project.

Hah! That is probably the problem I have run into right now. Unless I do a frequency on OE I cannot get even a read to work.