Renesas/Hitachi H8/36049 EPROM signals to pins mapping

[gashtaan]

Okay.

I'm just skeptical, different order of these major boot mode events are suspicious. Renesas has too many different bootloaders in their MCUs, not sure why they reinvent the wheel for every family. Anyway, thanks for the SH7407 bootloader, I'll look on it. How did you obtain it? It's directly accessible by ISP or from the program in flash?

[Postal2]

.... It's directly accessible by ISP or from the program in flash?

It's accessible from RAM.

[Postal2]

I got the chips but only see one purchase from the seller. This doesn't surprise me, but some people have asked why I paint the world in dark colors.

[gashtaan]

I got the chips but only see one purchase from the seller. This doesn't surprise me, but some people have asked why I paint the world in dark colors.

What is the problem? Do those ICs at least look like genuine? I didn't order them, it seemed shady that aliexpress did not show me that seller when I searched for those ICs.

[pcprogrammer]

What is the problem? Do those ICs at least look like genuine? I didn't order them, it seemed shady that aliexpress did not show me that seller when I searched for those ICs.

That might be a cookie related thing. The Aliexpress site has a very poor search engine that at times needs the cookies removed to produce better results.

[Postal2]

The chips are normal, new, in the factory grid.

[gashtaan]

That might be a cookie related thing. The Aliexpress site has a very poor search engine that at times needs the cookies removed to produce better results.

I tried that before, all the cookies and the cache wiped out and that seller didn't came out in the search - it could be related to the country I've set up in my account.

[gashtaan]

The chips are normal, new, in the factory grid.

I ordered those MCUs and today I received 1pcs completely different IC... it was a scam.

[Postal2]

I sometimes get the same, last time I got some small chip from a cell phone instead of AM3352. I got all 5 chips, I can take a photo. If you were the first to order - it would be the other way around. Check the price of the chip you received and return the difference.

[gashtaan]

If you were the first to order - it would be the other way around. Check the price of the chip you received and return the difference.

Maybe, to me the seller seems shady from the first sight. You got lucky. I received some DSP for CD players instead, useless. Anyway, AliExpress approves my request for full refund so I guess I got lucky too.

[amyk]

Here's another pinout for QFP80 adapter programming Hitachi/Renesas MCUs: https://www.mqp.com/ad633.htm

Looking at the pinouts for all the adapters I could find, it seems /RES is always Vpp and TEST is pulled high if present, while other I/O port pins are used for the rest.

It might be worth asking Renesas itself, since it clearly published this information for other MCUs in similar families around the same time, so the omission of this one could be accidental instead of deliberate.

[gashtaan]

It might be worth asking Renesas itself, since it clearly published this information for other MCUs in similar families around the same time, so the omission of this one could be accidental instead of deliberate.

I'm sure it's not accidental - Renesas began to delete this information from the datasheets after they acquired the Hitachi and they didn't publish such info on their later designs at all. Often I have to search for older revisions to get this info:

[gashtaan]

Here's another pinout for QFP80 adapter programming Hitachi/Renesas MCUs: https://www.mqp.com/ad633.htm

Looking at the pinouts for all the adapters I could find, it seems /RES is always Vpp and TEST is pulled high if present, while other I/O port pins are used for the rest.

Thanks. Yes, I know about that one. So far, I've found out socket mapping for 25 different MCUs and actually the H8/38024 is the closest one to H8/36049. Unfortunately some pins doesn't exist on H8/36049.

VPP and TEST are not a problem, they are only known variables in this equation. For me, the problem are I/O pins that must be pulled high (PGM and some others).

[amyk]

There is a bit more (Russian) discussion here which may or may not be relevant: https://monitor.espec.ws/section7/topic88228p40.html

I did a bit more digging and found mentions that this family actually doesn't map directly to any EEPROM but uses a protocol called "FZTAT64V5" or "FZTAT128V5" (probably depends on the capacity) - and from there found other Hitachi/Renesas CPUs that use this type, such as the H8S/2138 and H8S/2134 whose hardware manual documents the full protocol. It appears to be a self-timed write (needs an oscillator connected) and notably does NOT use a high Vpp.

[gashtaan]

There is a bit more (Russian) discussion here which may or may not be relevant: https://monitor.espec.ws/section7/topic88228p40.html

I did a bit more digging and found mentions that this family actually doesn't map directly to any EEPROM but uses a protocol called "FZTAT64V5" or "FZTAT128V5" (probably depends on the capacity) - and from there found other Hitachi/Renesas CPUs that use this type, such as the H8S/2138 and H8S/2134 whose hardware manual documents the full protocol. It appears to be a self-timed write (needs an oscillator connected) and notably does NOT use a high Vpp.

Not sure what are you referring to. The ICP protocol that is being handled be the software in boot ROM? Yes, this is well documented, but this is what I want to avoid. As I'm understand it, FZTATxVy is not a "protocol" it's type of flash that is supposedly stacked on main MCU die.

I believe that also this family maps EEPROM directly to the pins, because manufacturer of one of programmers that supports programming this MCU claims they support direct access to the flash - but I'm not going to spend thousands € to buy programmer and the socket for it.

[peter-h]

I use the H8/323 and it emulates a 27C256 EPROM. Needs high VPP just like the EPROM.

are you sure that the H8/300 series is even capable of supporting external memory in such that it can directly address it from the CPU core?

Of course they can. 64k external address space for code or data. But that's not the question here.

Unfortunately I don't have the data books to hand but it should be in there and nowadays they are all online.

[gashtaan]

Unfortunately I don't have the data books to hand but it should be in there and nowadays they are all online.

Was Renesas still releasing H8 data books after they bought technology from Hitachi? I believe they stopped publishing those and such data is confidential, shared only to major programmers manufacturers.

[amyk]

There is a bit more (Russian) discussion here which may or may not be relevant: https://monitor.espec.ws/section7/topic88228p40.html

I did a bit more digging and found mentions that this family actually doesn't map directly to any EEPROM but uses a protocol called "FZTAT64V5" or "FZTAT128V5" (probably depends on the capacity) - and from there found other Hitachi/Renesas CPUs that use this type, such as the H8S/2138 and H8S/2134 whose hardware manual documents the full protocol. It appears to be a self-timed write (needs an oscillator connected) and notably does NOT use a high Vpp.

Not sure what are you referring to. The ICP protocol that is being handled be the software in boot ROM? Yes, this is well documented, but this is what I want to avoid. As I'm understand it, FZTATxVy is not a "protocol" it's type of flash that is supposedly stacked on main MCU die.

I believe that also this family maps EEPROM directly to the pins, because manufacturer of one of programmers that supports programming this MCU claims they support direct access to the flash - but I'm not going to spend thousands € to buy programmer and the socket for it.

I am saying that FZTATxV5 looks like a parallel EEPROM, but not any standard type I'm aware of - writing 40h followed by 128 bytes initiates a sector write, 20h 20h is chip erase, and 71h 71h enters status read mode.

[gashtaan]

I am saying that FZTATxV5 looks like a parallel EEPROM, but not any standard type I'm aware of - writing 40h followed by 128 bytes initiates a sector write, 20h 20h is chip erase, and 71h 71h enters status read mode.

Oh, now I see what do you mean. I was mislead by that they also call ICP programming over UART in boot-mode as "F-ZTAT On-board programming".

I was focused on H8/38024 which seems to be most alike to H8/36049. The term F-ZTAT is mentioned in its datasheet multiple times, but now I understand the PROM programmer mode details are related strictly to masked ROM versions and details about F-ZTAT are completely omitted there. It would indeed means that VPP is not needed but the MCU has to be externally clocked.

[gashtaan]

The chips are normal, new, in the factory grid.

Tried another seller. Just received the package - the IC is obviously fake, poorly printed marking on sanded surface

[Postal2]

Sometimes I received chips with bad marking, and they worked. I received one with good marking AT91RM9200 - it didn't start. The Hitachis that I received absolutely exactly matched the product photo, I'm sure that they work. But now "the product doesn't exist", apparently you scared the seller with something.

[gashtaan]

But now "the product doesn't exist", apparently you scared the seller with something.

This has nothing to do with me, I noticed the product already did not exist when I received the package. Anyway, I got full refund from AliExpress. Will see how it'll end with my second attempt. I'll try it, although I highly doubt it's genuine.

[Postal2]

Don't doubt it. I always get Motorola mcu with horrible markings, but the pin is just as easy to break off as the original.

[peter-h]

Not sure if any data books existed with "Renesas" on them. I never collected any.

The H8/325 hardware book, page 202, has the EPROM pin connections. I built an adaptor for a Data I/O chiplab from that, many years ago. The info may enable you to work out the pins; I can scan to a PDF if you PM me your email address.