For Sale RAM Chip Tester for Commodore, Atari, Apple & More

[tops4u]

Hi all,

I just wanted to reach out, since while working on the next Version of the Code, I have discovered a bug. So for those who are using the 18Pin Mode, testing is incomplete (code address lines 4 &5 mapped to the same physical pin). If you want to upgrade, there is a bug fixed Version on Github. Upgrade instructions on how to use an Arduino UNO to upgrade the Firmware will be posted soon in the 1.st Post.

If it is not urgent to you to have a flawless 18 Pin Mode, I suggest to wait a few days longer and wait for the next version that is testing RAM more thoroughly by implementing a Random Data Check, checks static column RAM and will probably offer a display option (currently for the SMD PCBs)

 

[Bharaputra]

Declaring interest.

---

Payment sent.

---

(Marked as sold/completed by tops4u)

 

[tops4u]

Great News

The new Software is almost done. All relevant Features were implemented today. I need to recheck that the retention Times are really spot on, currently the may be a little longer than the specified times, but to be on the sure side if your RAM turns out tested ok, it really should be ok, even when the Refresh cycle takes a few microseconds longer.

Update is recommended to all users, since there are important fixes.

If you want to upgrade but do not intend to use the Display - make sure to comment the OLED Line out in the source code. This will make your tester faster and will not distract you since the OLED shares some lines with the LED.

What is new?
1. Bug fixes. 18Pin Mode Bug is fixed. Minor Bug (that should not have functional impact) in 16 Pin Mode was fixed.
2. Pseudo Random Pattern checking. This is important and increases thoroughness of the Testing.
3. Fixed the missing retention Checks for the last 1-3 Rows at the end of Tests.
4. Probably the most awaited change: OLED Display - for those who don't want to decode the LED Blinks. Just Plug it in or keep using the LED.
5. Static Column RAM Tests! Yes! Works for 514258 and 514002.
6. Due to internal improvements still most tests do not take significantly longer and all Ram are tested in under 12 Secs.

So how does it look like, and what information does it provide? Still I think nobody really wants to know which Row/Column or "address" is currently being tested. So the display shows the type of ram that was autodetected - make sure it matches your Chips identification! Success or failure are clearly indicated with some text of what failed - even in most cases dead Ram is dead Ram.

​

Disclaimer: Upgrading your Tester is your own Risk. Although it is hard to "Brick" it, but probably not impossible. If you kill your Ram Tester or render it inoperable it will be your responsibility.

​

 

[tops4u]

There is now availability for the Version with Display.

 

[Bharaputra]

Tester received and used with 41256, 44256 as also with 514400, 514402 RAM, tool performed well as described and expected. Display as a good feature to avoid counting LED flashes for errors found during test. Speed of test is fine

 

[BuzzBrain]

Declare interest for one unit, with display.
Shipping to Norway.

 

[Fobozz]

Declaring interest.

---

There will be up to 10 new Units available later this week. Check initial Post.

I would like to order one of the units with display.

---

(Marked as sold/completed by tops4u)

 

[tops4u]

NEWS: Upcomming Firmware will support 411000 (1M x 1) DIP Ram. This will be Firmware upgradable for all HW Revisions.



There is a downside as well :
This will not work for ZIP, since the ZIP Pinout is too different from the 541000/544256 ZIP Pinout. And testing is rather slow. I'm still optimizing code and adjusting timing but expect around 30+ 28sec per RAM.

Goodie: While working on the 411000 I found some optimization potential for all other RAM Types, expect up to 20% 40% faster tests.

 

[BuzzBrain]

Unit arrived. Tested and working Good

 

[tops4u]

Firmware 2.4.0 is ready for download on GitHub.

Installation notes: Make sure you do have default Arduino UNO Settings when uploading - no change in compiler options, as it could mess with retention timing and maybe even with code execution order.

Changes:

• Support for 411000 added. Test takes 28sec, due to less than optimal pin assignment. I had to do some address calculations in the code which adds to test times.
• General Speed improvements, see table below for details.
• Important Notice: I’ve modified the load on the output pins of the 18‑pin chips by enabling the pull‑ups, so that the RAM must actively pull the outputs LOW when required. This change was necessary because parasitic capacitances were holding the input levels even after the RAM pin switched to output mode—even when the RAM was unplugged between tests. With pull‑ups enabled, all I/Os are now held at 5 V unless the RAM actively drives them low.
  However, older RAM modules or those produced with earlier processes often have weaker output drivers. While they may work fine in environments with minimal loading, their driver strength can be insufficient for this test setup—potentially leading to false‑negative results.

Ram Type	2.3.x Time	2.4.0 Time	Boost​
4164	5sec	3.5sec	30%​
41256	12sec	9sec	25%​
411000	n/a	28sec	-​
4416	3.7sec	2.6sec	29%​
4464	6.5sec	5.2sec	20%​
514256	5..5sec	3.8sec	31%​
514258	5.2sec	3.8sec	27%​
514000	14.5sec	8.9sec	39%​
514002	12.5sec	8.7sec	30%​

Issued 2.4.1 today with minor improvements:

• The Random Data which is used in the last stage for Retention Testing but also as "Random Data" Test is now inverted every second test. This is due to the nature of random data it will not test the Retention of each and every cell, only around 50% if it is really random. For users who want 100% retention coverage for all cells, this offers the possiblity to just redo the test and the pattern will be inverted so that to other 50% of cells will be tested.
• In the Past after every update the Diag Mode was active. I have now added a Macro which is by default deactivated. People who solder their own PCB can uncomment it and enter Diag Mode to test their soldering - for all others: you should never see it again.

One note on retention testing
Usually the RAMs have robust times and the minimum retention time can be exceeded by far without the RAMs loosing their data. I have one batch of RAM from a undisclosed CN Source that work perfectly well but all fail to meet retention times. This is systematical and all those RAM fail. Most failed RAM I have seen have single or multiple defect cells, defect address decoders or IO Pins. Still it is important that the RAM can store the data for the specified time.

 

[actionthisday76]

Declaring interest.

---

(Marked as sold/completed by tops4u)

 

[amienslaver]

Declaring interest.

---

Interested in 1 unit with display.

---

(Marked as sold/completed by tops4u)

 

[butjer1010]

Declaring interest.
One with display please

---

(Marked as sold/completed by tops4u)

 

[amienslaver]

Received the tester, everything is good, happy with it

 

[crock]

Declaring interest.
For unit with display, postage to Switzerland!!

** Received, thank you. Great service. **

---

(Marked as sold/completed by tops4u)

 

[nOnAm3]

Does the tester support 4116?

 

[tops4u]

Does the tester support 4116?

No it does not, since 4116 needs -5V and +12V this rather a special requirement.

 

[djcollins]

Declaring interest.

---

declaring interest for 1 unit with Display

---

(Marked as sold/completed by tops4u)

 

[djcollins]

Paid now !

 

[martinen]

Declaring interest.

---

(Marked as sold/completed by tops4u)