Jump to content
Shift838

Suitable SRAM Replacement

Recommended Posts

I've been cross referencing chips for the IDE card.  Of course the 512k SRAM is no longer available.  I believe I have a found a suitable replacement but I would like a couple of sets of extra eyeballs to look at it to make sure I am correct.

 

The M68Z512W.pdf is the datasheet for the original.

 

I found IS62WWV5128BLL , CY62148GN and CY62148EV30LL

 

These 3 appear to be the same to me, but the last (CY62148EV30LL) may be a bit light on the voltage side.  

M68Z512W.pdf IS62WWV5128BLL.pdf CY62148GN.pdf CY62148EV30LL.pdf

  • Like 1

Share this post


Link to post
Share on other sites
Posted (edited)

deleted, since the forum won't let me check "notify on replies" if I forgot the first time.

Edited by brain
forum issue
  • Like 1

Share this post


Link to post
Share on other sites

They all look like good options to me (I assume the IDE card has a 3V3 rail for this IC.  If so, the last items should be fine.

 

Jim

  • Like 1

Share this post


Link to post
Share on other sites

The ISSI part comes in two TSOP-32 pinouts, you need the Type-2 pinout, which is designated by the full part number (shown below).

 

The CY62148GN comes in a 5V and 3.3V version (again, full part numbers specify the voltage and package), and the 3.3V version is not available.  The original part is 3.3V, so I'll assume you need 3.3V.  So the CY62148GN is not an option.

 

The CY62148EV30LL is available but costs more.  Also, the Cypress parts are 45ns, so you are paying for speed you don't need.  The ISSI part is 70ns and cheaper.  Octopart is your friend!

 

Also, the original part is available from four vendors, but you will have to make an offer and probably buy the complete stock.  But three of the four have between 30 and 80 pieces, so if that is enough for your build, you can probably get them for a pretty good price.  Offer $2/piece and see what they say.

 

https://octopart.com/search?q=M68Z512W&currency=USD&specs=0

 

Original:
Full Part Number      Package        Vcc  Speed  Availability
M68Z512W-70NC1      , TSOP-32,       3.3V, 70ns, Make offer for complete stock

Replacements:
Full Part Number      Package        Vcc  Speed  Cost   DigiKey  Mouser
IS62WV5128BLL-55T2LI, TSOP-32 type2, 3.3V, 70ns, $3.84,   622,    1067
CY62148GN-45ZSXI    , TSOP-32,       5.0V, 45ns, $4.92,   220,     222
CY62148GN30-45ZSXI  , TSOP-32,       3.3V, 45ns,          not available
CY62148EV30LL-45ZSXI, TSOP-32,       3.3V, 45ns, $4.51,   1771,    268

When you have an equal choice, let availability and cost be your deciding factors.  My rules for using a part:

 

1. Must be in active status with the manufacturer.

2. Must be available from both DigiKey and Mouser.

3. Must be regularly stocked.

4. Must have decent documentation / datasheet.

 

But in this case, I would probably make offers to the three vendors who have new-old-stock, since this is a limited buy for a specific purpose.

  • Like 2

Share this post


Link to post
Share on other sites
8 hours ago, matthew180 said:

 

Original:
Full Part Number      Package        Vcc  Speed  Availability
M68Z512W-70NC1      , TSOP-32,       3.3V, 70ns, Make offer for complete stock

Replacements:
Full Part Number      Package        Vcc  Speed  Cost   DigiKey  Mouser
IS62WV5128BLL-55T2LI, TSOP-32 type2, 3.3V, 70ns, $3.84,   622,    1067
CY62148GN-45ZSXI    , TSOP-32,       5.0V, 45ns, $4.92,   220,     222
CY62148GN30-45ZSXI  , TSOP-32,       3.3V, 45ns,          not available
CY62148EV30LL-45ZSXI, TSOP-32,       3.3V, 45ns, $4.51,   1771,    268

 

Thanks Matt!  I have been using octopart.  I found it a few weeks ago when I was staring to learn Fusion 360 while I was looking for 3d models of components.  

 

I have already made offers to the various vendors that show stock on octopart and have not heard back yet.  That's why I was looking for a suitable replacement.  So what I am interpreting from your post is that the one that will work although it's a bit faster is 

IS62WV5128BLL-55T2LI

Share this post


Link to post
Share on other sites

 

14 hours ago, Shift838 said:

when I was staring to learn Fusion 360 while I was looking for 3d models of components.

Personally I try to avoid the commercial software with hobby projects, even if there are "free" versions (there is always a catch).  My preferred tools for PCB work are KiCAD and FreeCAD.  I know the FreeCAD learning curve is a brick-wall, but it is actually not too far from SolidWorks in workflow and capability, and worth learning IMO.  Check out Joko Engineering on YouTube, example:

 

There is also a plug-in to FreeCAD for importing KiCAD boards, and to help get STEP files aligned with KiCAD footprints.  Importing the PCB lets you do things like validate the mechanical fit, assuming you have your enclosure modeled in FreeCAD.

 

KiCAD itself is awesome these days (and getting better all the time) and provides a lot of parts with 3D models in the stock library.  For missing parts it is pretty easy to make your own.  If you are missing 3D STEP models, check the manufacturer's website first, a lot of them provide models for their parts.  If you can find a STEP file online, then you can import it with FreeCAD and get the part aligned to the footprint and exported to STEP and WRL (KiCAD uses both formats for rendering).

 

I make all my own footprints, or verify any existing ones against the datasheets; I don't need any surprises when I get my boards back from OSHpark.  There is a big difference between just making a PCB, and making a PCB that can actually be manufactured.  You have to pay attention to *all* the details, and make sure you know the capabilities of the PCB shop you are using.

 

13 hours ago, Shift838 said:

So what I am interpreting from your post is that the one that will work although it's a bit faster is ...

Ah, there was an error in my table.  The ISSI "*ALL" part is 70ns at 2.5V.  The ISSI "*BLL" part is the 3.3V version with a 55ns access time.

 

When it comes to memory, "speed" refers to access time, so the smaller the number the faster the memory; and you pay more for faster access. The original part is 70ns, the 3.3V ISSI part is 55ns (so slightly faster), and the Cypress parts are 45ns (faster than both original and ISSI).  Typically, all other characteristics being equal, faster memories can be used to replace slower memories.

 

Share this post


Link to post
Share on other sites
On 3/24/2020 at 4:07 PM, brain said:

They all look like good options to me (I assume the IDE card has a 3V3 rail for this IC.  If so, the last items should be fine.

 

Jim

There is no 3.3 volts available on the original IDE card design!!!

  • Like 1

Share this post


Link to post
Share on other sites

Then you will have a problem, even with the original SRAM.  The absolute maximum Vcc is 4.6V, and you do not want to operate at or near that range.  None of the SRAMs have 5V tolerant I/O either.  I just assumed that since the original part was 3.3V that there was compensation (at least a 3.3V reg) built into the design.

 

If it works, which it probably will for a little while, it will not be stable and reliable long term, IMO.  Best to find a 5V SRAM.

Share this post


Link to post
Share on other sites
5 hours ago, F.G. Kaal said:

There is no 3.3 volts available on the original IDE card design!!!

Then the information in the original post must be incorrect. In the original post, the original part was described as M68Z512W and the datasheet for that part was attached. The PDF shows the voltage span for that device as 3V3 (2.7 to 3.3 or so...) So, if that is the part on the original design, something must have supplied it with 3v3.

 

Jim

Share this post


Link to post
Share on other sites

I assume the design is coming from the TI Tech Pages?

 

http://www.nouspikel.com/ti99/titechpages.htm

 

A quick check and I don't see any specific SRAM called out, only the clock IC, and the strange combined clock+SRAM ICs:

 

bq4847 no SRAM

bq4842 with 128K

bq4852 with 512K

 

If you use the bq4847 the page says you have to use an external SRAM, but no part is suggested or specified, so I'm not sure where the use of the M68Z512W 3.3V part came from.  But again, I would highly suggest finding a real 5V SRAM, or you risk very unreliable operation and early failure.

 

Share this post


Link to post
Share on other sites
Posted (edited)
3 hours ago, matthew180 said:

I assume the design is coming from the TI Tech Pages?

 

http://www.nouspikel.com/ti99/titechpages.htm

 

A quick check and I don't see any specific SRAM called out, only the clock IC, and the strange combined clock+SRAM ICs:

 

bq4847 no SRAM

bq4842 with 128K

bq4852 with 512K

 

If you use the bq4847 the page says you have to use an external SRAM, but no part is suggested or specified, so I'm not sure where the use of the M68Z512W 3.3V part came from.  But again, I would highly suggest finding a real 5V SRAM, or you risk very unreliable operation and early failure.

 

the SRAM part number came from the component list off of Thierry's site.

 

HERE

 

Edited by Shift838

Share this post


Link to post
Share on other sites

Ah, right, I missed that table the second time through the page (in too big of a hurry).

 

I suspect Thierry never actually built his board with that SRAM and probably did not realize it was a 3.3V only device when he found it to list as an alternative.  You could email and ask him though.  But either way, the datasheet specifies it is a 3.3V device, and the I/O will only tolerate 3.6V max (i.e. it is *not* 5V TTL tolerant).  Personally I would not use it on this 5V only board, it will not last very long.  Even if you had a 3.3V Vcc for the SRAM, its I/O is still not rated to interface with the other 5V parts.

  • Like 1

Share this post


Link to post
Share on other sites
7 hours ago, matthew180 said:

Ah, right, I missed that table the second time through the page (in too big of a hurry).

 

I suspect Thierry never actually built his board with that SRAM and probably did not realize it was a 3.3V only device when he found it to list as an alternative.  You could email and ask him though.  But either way, the datasheet specifies it is a 3.3V device, and the I/O will only tolerate 3.6V max (i.e. it is *not* 5V TTL tolerant).  Personally I would not use it on this 5V only board, it will not last very long.  Even if you had a 3.3V Vcc for the SRAM, its I/O is still not rated to interface with the other 5V parts.

Does anyone have Thierry's current email address as the one on his page is not valid.

 

 

Share this post


Link to post
Share on other sites
4 hours ago, Shift838 said:

Does anyone have Thierry's current email address as the one on his page is not valid.

 

I am a little surprised his Yahoo address is no longer valid. It was for me four years ago.

 

...lee

  • Like 1

Share this post


Link to post
Share on other sites

Alternatively just believe the datasheet for the part and conclude that a 3.3V device will probably be damaged in a 5V circuit.

Share this post


Link to post
Share on other sites

It looks like he used the alternative part from Mouser, the TC554001AFT-70, which is 5V:  https://html.alldatasheet.com/html-pdf/31696/TOSHIBA/TC554001AFT-70/253/1/TC554001AFT-70.html

 

Just find a replacement for that IC and all is well.

 

If you're more adventurous and have the space to make a through hole board, it looks like it could be made through hole, and Alliance makes a 32 pin DIP 5V 512kB SRAM.

 

Share this post


Link to post
Share on other sites
9 hours ago, matthew180 said:

Alternatively just believe the datasheet for the part and conclude that a 3.3V device will probably be damaged in a 5V circuit.

In Thierry's design the RAM chip M68Z512W get its power from the clock chip BQ4847 chip (pin #1 Vout) . See also schematic in my website http://www.ti99-geek.nl/Projects/idecard/Files/ide_sch2.gif

 

When using the other clock chip BQ4842 you don't need the RAM chip ... the DSR can be stored in the clock cips memory.

If the RAM chip is added anyway then yo have to disable it when using this clock chip.

 

In the data sheet of the RTC-65271 I don't see a voltage output for some external device. If this clock chip is going to be used some batery backup for the RAM chip must be created.

 

Fred ;-

   

Share this post


Link to post
Share on other sites
10 hours ago, F.G. Kaal said:

In Thierry's design the RAM chip M68Z512W get its power from the clock chip BQ4847 chip (pin #1 Vout) . See also schematic in my website http://www.ti99-geek.nl/Projects/idecard/Files/ide_sch2.gif

 

When using the other clock chip BQ4842 you don't need the RAM chip ... the DSR can be stored in the clock cips memory.

If the RAM chip is added anyway then yo have to disable it when using this clock chip.

 

In the data sheet of the RTC-65271 I don't see a voltage output for some external device. If this clock chip is going to be used some batery backup for the RAM chip must be created.

 

Fred ;-

   

Hi Fred, there is supply of BQ4852YMC-85 but Rochester electronics instead of no longer available Texas Instruments. 

 

https://www.digikey.com/product-detail/en/rochester-electronics-llc/BQ4852YMC-85/2156-BQ4852YMC-85-TI-ND/11538020

 

Share this post


Link to post
Share on other sites
11 hours ago, F.G. Kaal said:

In Thierry's design the RAM chip M68Z512W get its power from the clock chip BQ4847 chip (pin #1 Vout) .

Vout is the same as Vcc (5V) during normal operation, and switches to the battery voltage (probably about 3V) when power is lost.  But the normal operation voltage is still 5V, and this does not address the I/O signals that are still going to be coming from 5V TTL ICs.

 

27 minutes ago, Swim said:

... there is supply of BQ4852YMC-85 but Rochester electronics ...

Those do not provide the battery Vout to power the external SRAM during power failure, if that is important.

 

Also, any of these clock chips with internal batteries are going to suffer the same problem, i.e. they will be dead in 10 years, which in retro-computing is not very long.  And in 10 years they will probably be near impossible to get.  Using an external battery design would be a better option, IMO.

 

Share this post


Link to post
Share on other sites
24 minutes ago, matthew180 said:

Vout is the same as Vcc (5V) during normal operation, and switches to the battery voltage (probably about 3V) when power is lost.  But the normal operation voltage is still 5V, and this does not address the I/O signals that are still going to be coming from 5V TTL ICs.

 

Those do not provide the battery Vout to power the external SRAM during power failure, if that is important.

 

Also, any of these clock chips with internal batteries are going to suffer the same problem, i.e. they will be dead in 10 years, which in retro-computing is not very long.  And in 10 years they will probably be near impossible to get.  Using an external battery design would be a better option, IMO.

 

Hi matthew180, the BQ4852YMC-85 carries the 512k SRAM on board so there is no need for a 512k SRAM at location IC23.

Share this post


Link to post
Share on other sites

I'd possibly be willing to redo the schematic and PCB and make it all through hole, but I'd need help...  If I did, I'd probably make the entire design through hole parts, since I don't know that I see any reason it all has to be surface mount.

  • Like 1

Share this post


Link to post
Share on other sites
5 minutes ago, Swim said:

... the BQ4852YMC-85 carries the 512k SRAM on board ...

Ah, yeah, right.  Too many options, too easy to make a mistake.  Even with the internal SRAM, the internal battery is going to be a problem down the road.

  • Like 1

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...