I made this couple of years ago. Since I also fix Spectrums, I modified the code to do 4116 DRAMs too. Because 4116s need -5 and 12V supplies besides the 5V, I made 2 boards, one for 4116s only which has the necessary voltages generated on board, and the other one is like this one, does 4164 and 41256. I thought of adding a display at first but then I decided it would be overkill, so I added a few lines of code to display what's going on during the testing and the result on a serial monitor. Needless to say, you need to connect the boards to a PC for this to work. Both boards are externally powered with a 12V adapter, so there is no powering them over USB option.
I'd love to see a DRAM tester that also measures/tests the speed grade, to see if the chips perform well under full speed in a system. Maybe even combined with an automatic refresh test that lowers the refresh rate and tests when the chips start to actually lose bits
@Mr Guru should be possible with a Pi Pico (there was a project some years ago that tried something like this with a BeagleBoard Black and it's PIO engines, which are similar to the Pi Pico's programmable IOs). Also, FPGAs are not that expensive anymore...
@@scottgilson3247 AFAIK the ATMEGA 328 can't go faster than 20MHz so your fastest cycle time is 50ns but the IO's are not synchronous so you would end up with something around 80ns which is far from the needed Speed to really check the RAM Timings. It's not only RAS to CAS that needs to be checked. Also the Output speed should be checked, but the ATMEGA 328 has around 90ns Input settle time.
Hi, love your channel, I learn a lot from you. I bought an Atari 800XL for $100 with just photos of the original box and it is very water damaged, but the surprising thin is that everything in the box is intact plus never used. I scored a brand new Atari 800 XL. Everything wrapped in plastic and not a smudge on any of it. I just learned that you can't judge a book by it's cover.
Amazing to think that stuff like this is so trivial to make nowadays, versus say 20 years ago when RAM testers were still considered "professional equipment" and not always easy to come by, whereas today, you can buy things like the Retro Chip Tester Pro, or make something like this, depending on your needs... :D
Well ... good that such hobbyist equipment is available (wait a minute ... for approximately "say 20 years" now? Hehehe). But in my opinion, your comparison does lack a little bit "versus say" ... reality? If you were doing anything in this field 30 years ago (even just as a hobbyist) then you probably also had a multi-programmer[1]. These cost around $400 (e.g. the devices in the Galep series[2]) and could not only test TWO components (for the price of $30), but also program tens of thousands of PALs, GALs, (E/E)PROMs types etc. on top of that. [1] Or a friend with one those things like me, hehe ... There wasn't a thinkable scenario without having such a device. Oh ... and there were even another decade earlier real EPROM Programmers and CHIP testers for the C64!!! Not that expensive. [2] You won't believe it ... because I am an idiot, like all n00bs are, I killed that device in the first week. It came back a week later and although it was definitely my fault, it was treated as a warranty case and replaced at no cost to me. This very versatile off-grid device works and is used to THIS day! (You can compare my device with the modern TL866 devices, but they lack the wide programming voltage range of my device ... which isn't needed anymore today) Jan: Good Job and great device!:) Edit, Trivia: You won't believe how much motherboards I have "fixed" and "rescued" from the 286 to the Pentium era. Because of the usual "mishaps" when flashing a new BIOS ... which is in many cases the first-time and last-time experience (at the same time:P) of most "handyman" in their life on this topic, hehehe.
@@dieSpinnt 30 years ago I was 7 years old, so my pocket money wouldn't stretch to a $400 programmer, during my time in professional PC repair, between 2004 and 2010 (still did repairs afterwards, but that tailed off because modern hardware sucks!), motherboards you said you repaired were back then considered worthless, and the time it would take to do chip-level repair on them from a commercial sense was uneconomical, plenty stuff passed through my hands that sadly went in the bin which today, when repaired, would have been ridiculously valuable, and at the time DIY tools for doing such repairs at home were expensive and prohibitively so if you were just doing hobby stuff, even things like C64s and Amigas, you could buy a box of those for a tenner (I got my two Amiga 600s for £20, in a bucket of other tat that they threw in just to get rid) at a car boot sale, because they were old stuff nobody wanted, of course today retro hardware is treated like a goldmine by unscrupulous sorts, but again, something so dirt-cheap to acquire back then, easier to just parts-swap than to shell out money on test equipment that cost more than a month's wages, unlike today where open-source hardware being created by the repair-oriented community that does so much being sold so cheaply to do diagnostics, testing, recovery and preservation, seriously, it's amazing!! The stuff I could have done back in my PC repair days to fix "old crap" with the tools available now, I'd have been a miracle worker I'm sure!!! Even something like the Greasewazel, £23, what would an equivalent of that 20 or 30 years ago cost? One of those questions where it's probably replied to with "If you have to ask, you can't afford it!", because they were closed-source, niche-market products aimed at professional markets, not us home-gamers, and even in a professional environment were just not worth the investment...
I also made this and 4164 and 41256 work fine, but I don't understand why HM50256 and the MB81256 don't work. I also use them in retro computers for replacement.
Hi Jan, you did a video repairing an Amiga 4000. It ended with the Amiga not posting due to suspected problems in the data lines around battery damaged traces. Any plans to revisit this machine?
It's on my list! I often get so frustrated by projects that don't work out that I have to set them aside for quite a while before having another go. I'm going to try again eventually when I build up the courage and patience.
Yes, what Darkstar says. There's a strange spacing between the rows of pins on one side of the Arduino boards. The pins themselves are spaced correctly, it's just the gap that is odd.
@@JanBeta the odd gap/pinspacing on one side is intentional and is there as a key(like a keyed connector) so you can't put in an arduino addon board(shield) the wrong way around potentionally sending voltage where it shouldnt go. the arduino formfactor was made as an educational tool made to work with addon boards called shields.
Yes, that’s the way it was intended by the designer. I just find it easier to remember where pin 1 is putting it the way I did. Electrically the same both ways, of course, as long as you put the RAM in the correct orientation.
Can you please recommend a person who can make a circuit board with the push button contact pads that a little plastic keypad pushes through little plastic fingers that are above the pc board ? I have a DTMF keypad that is no longer available It has the plastic piece with the plastic buttons and the PCB that has a film strip with the teaces that slip into a pressure connector I don't know how difficult it is to make this type of board. It has 16 buttons that are pressure metal circles that when pushed makes contact through the circuits I have one that I can send so it can be duplicated. Thanks Bill KC2OVX
Finally built mine today :) The serial output is handy for capturing test results, hoping seller for dram will replace the faulty "new" chips based on the failure log.
Should be doable, the test procedure would be very similar. The main problem is that those older chips require +12V and -5V as well as the +5V supply voltage so it would need some extra voltage regulation to make it work.
detecting those is difficult because there are so many different pinouts and so many ways you can potentially damage the chip. If you want to simply test 74xx chips there are some very nice projects that do just that, like the RetroChipTester Pro (RCT) for example
That requires a much more sophisticated circuit and code. This is just a simple, "quick & dirty" DRAM tester that can do 4164 and 41256 DRAM ICs, and that's it.
@@Darkstar2342 That's what I mean (74xx). I got a few bootleg games for example with scratched ICs. Knowing what they are must be simple if you have a table for input injection/output measurement, so coding that must not be so complicated.
@Mr Guru it can, but only a small subset of all 74xx chips, and it explicitly warns you that it can damage chips that don't conform to the standard pinout (which some 74xx chips don't)
I'm writing this comment because a mistake happened, I wanted to give you a like and I accidentally clicked on the dislike next to it, I immediately corrected it and this time I really gave a like, but when I accidentally clicked on the dislike, it said that feedback was shared with the user, so I am writing this comment to assure you that it was a mistake. (Google Translator used)
I can’t support your content anymore because of how you fed into the drama to rip the retro engineering community apart when you went against the Commander X16 and the 8 bit guy by siding with the Z80 project which misses the point that it is providing a 6502 powerhouse to the community
Look, if David sends me a Commander X16 to review, I’d gladly do it. It just didn’t happen (yet). I am not taking sides for anybody. Every single one of the recent retro inspired projects is awesome in their own right. Especially the fact that they exist at all at this point in time.
I made this couple of years ago. Since I also fix Spectrums, I modified the code to do 4116 DRAMs too. Because 4116s need -5 and 12V supplies besides the 5V, I made 2 boards, one for 4116s only which has the necessary voltages generated on board, and the other one is like this one, does 4164 and 41256. I thought of adding a display at first but then I decided it would be overkill, so I added a few lines of code to display what's going on during the testing and the result on a serial monitor. Needless to say, you need to connect the boards to a PC for this to work. Both boards are externally powered with a 12V adapter, so there is no powering them over USB option.
Any day with a Jan Beta upload is a good day.
Thanks! Hope you enjoyed the video. :)
I'd love to see a DRAM tester that also measures/tests the speed grade, to see if the chips perform well under full speed in a system. Maybe even combined with an automatic refresh test that lowers the refresh rate and tests when the chips start to actually lose bits
@Mr Guru should be possible with a Pi Pico (there was a project some years ago that tried something like this with a BeagleBoard Black and it's PIO engines, which are similar to the Pi Pico's programmable IOs). Also, FPGAs are not that expensive anymore...
If you can live with 25ns resolution, an Arduino could do it. There are per-clock-cycle instructions to control I/Os....
@@scottgilson3247 AFAIK the ATMEGA 328 can't go faster than 20MHz so your fastest cycle time is 50ns but the IO's are not synchronous so you would end up with something around 80ns which is far from the needed Speed to really check the RAM Timings. It's not only RAS to CAS that needs to be checked. Also the Output speed should be checked, but the ATMEGA 328 has around 90ns Input settle time.
Hi, love your channel, I learn a lot from you. I bought an Atari 800XL for $100 with just photos of the original box and it is very water damaged, but the surprising thin is that everything in the box is intact plus never used. I scored a brand new Atari 800 XL. Everything wrapped in plastic and not a smudge on any of it. I just learned that you can't judge a book by it's cover.
Cool, das es so schnell geklappt hat...... gruess Bjoern
Amazing to think that stuff like this is so trivial to make nowadays, versus say 20 years ago when RAM testers were still considered "professional equipment" and not always easy to come by, whereas today, you can buy things like the Retro Chip Tester Pro, or make something like this, depending on your needs... :D
Well ... good that such hobbyist equipment is available (wait a minute ... for approximately "say 20 years" now? Hehehe). But in my opinion, your comparison does lack a little bit "versus say" ... reality?
If you were doing anything in this field 30 years ago (even just as a hobbyist) then you probably also had a multi-programmer[1]. These cost around $400 (e.g. the devices in the Galep series[2]) and could not only test TWO components (for the price of $30), but also program tens of thousands of PALs, GALs, (E/E)PROMs types etc. on top of that.
[1] Or a friend with one those things like me, hehe ... There wasn't a thinkable scenario without having such a device. Oh ... and there were even another decade earlier real EPROM Programmers and CHIP testers for the C64!!! Not that expensive.
[2] You won't believe it ... because I am an idiot, like all n00bs are, I killed that device in the first week. It came back a week later and although it was definitely my fault, it was treated as a warranty case and replaced at no cost to me. This very versatile off-grid device works and is used to THIS day! (You can compare my device with the modern TL866 devices, but they lack the wide programming voltage range of my device ... which isn't needed anymore today)
Jan: Good Job and great device!:)
Edit, Trivia: You won't believe how much motherboards I have "fixed" and "rescued" from the 286 to the Pentium era. Because of the usual "mishaps" when flashing a new BIOS ... which is in many cases the first-time and last-time experience (at the same time:P) of most "handyman" in their life on this topic, hehehe.
@@dieSpinnt 30 years ago I was 7 years old, so my pocket money wouldn't stretch to a $400 programmer, during my time in professional PC repair, between 2004 and 2010 (still did repairs afterwards, but that tailed off because modern hardware sucks!), motherboards you said you repaired were back then considered worthless, and the time it would take to do chip-level repair on them from a commercial sense was uneconomical, plenty stuff passed through my hands that sadly went in the bin which today, when repaired, would have been ridiculously valuable, and at the time DIY tools for doing such repairs at home were expensive and prohibitively so if you were just doing hobby stuff, even things like C64s and Amigas, you could buy a box of those for a tenner (I got my two Amiga 600s for £20, in a bucket of other tat that they threw in just to get rid) at a car boot sale, because they were old stuff nobody wanted, of course today retro hardware is treated like a goldmine by unscrupulous sorts, but again, something so dirt-cheap to acquire back then, easier to just parts-swap than to shell out money on test equipment that cost more than a month's wages, unlike today where open-source hardware being created by the repair-oriented community that does so much being sold so cheaply to do diagnostics, testing, recovery and preservation, seriously, it's amazing!! The stuff I could have done back in my PC repair days to fix "old crap" with the tools available now, I'd have been a miracle worker I'm sure!!! Even something like the Greasewazel, £23, what would an equivalent of that 20 or 30 years ago cost? One of those questions where it's probably replied to with "If you have to ask, you can't afford it!", because they were closed-source, niche-market products aimed at professional markets, not us home-gamers, and even in a professional environment were just not worth the investment...
Nice improvement to a useful piece of test gear.
Neat project! The Arduino is a neay multitool. I think everyone should have one. I also use it as a programmer for pic attiny and updi.
wow that tool is neat for bending leads
very cool project!
Ahoi, i have planed the ZIF Socket 180 Degrees turn around
I know, I explained why I put it in the other direction in the video. Just helps me remember where pin 1 goes. Works both ways which is great. ;)
Would be nice if this tester can also support the 41464 (64K x 1) chips from the C64 250466 boards.
Great video, Jan Beta! I have faster firmware and will release it hopefully soon.😉
Nice! Thanks for all the work you put into this! It's a super useful device in my little lab (and I suppose it is for many others as well)!
Great video as always😊maybe that pcb can be a little bit reworked with more grounding polygons, just to eliminate unwanted noise… just as a thought👍✌️
The whole backside is a ground plane as far as I can see so it should be pretty well shielded already. But more ground fill wouldn't hurt, obviously!
I also made this and 4164 and 41256 work fine, but I don't understand why HM50256 and the MB81256 don't work. I also use them in retro computers for replacement.
"MT" (ram) is just Micron Technology that still make memory and SSD's to this day
They make more reliable RAM than their old 4264 chips these days, thankfully! :)
Hi Jan, you did a video repairing an Amiga 4000. It ended with the Amiga not posting due to suspected problems in the data lines around battery damaged traces. Any plans to revisit this machine?
It's on my list! I often get so frustrated by projects that don't work out that I have to set them aside for quite a while before having another go. I'm going to try again eventually when I build up the courage and patience.
@@JanBeta Great, I am looking forward to it. I have an Amiga 4000 in a very similar state so would be good to see your troubleshooting. Thanks!
the reason for the arduino pin spacing is so you don't plug the arduino shields in backwards.
no that's not what he means. He's talking about the distance between the rows not being a multiple of the 100mil spacing of common perfboards
Yes, what Darkstar says. There's a strange spacing between the rows of pins on one side of the Arduino boards. The pins themselves are spaced correctly, it's just the gap that is odd.
@@JanBeta the odd gap/pinspacing on one side is intentional and is there as a key(like a keyed connector) so you can't put in an arduino addon board(shield) the wrong way around potentionally sending voltage where it shouldnt go.
the arduino formfactor was made as an educational tool made to work with addon boards called shields.
I need another Jan Beta video.
@8:08 could you have oriented the socket 180 degrees so the lever is completely clear? just mark pin 1 with a sharpie.
Yes, that’s the way it was intended by the designer. I just find it easier to remember where pin 1 is putting it the way I did. Electrically the same both ways, of course, as long as you put the RAM in the correct orientation.
Can you please recommend a person who can make a circuit board with the push button contact pads that a little plastic keypad pushes through little plastic fingers that are above the pc board ?
I have a DTMF keypad that is no longer available
It has the plastic piece with the plastic buttons and the PCB that has a film strip with the teaces that slip into a pressure connector
I don't know how difficult it is to make this type of board.
It has 16 buttons that are pressure metal circles that when pushed makes contact through the circuits
I have one that I can send so it can be duplicated.
Thanks
Bill
KC2OVX
👍
How did you get that fast at soldering, Jan?
😁
Easy! Years of practice and a bit of time traveling! 😅
I noticed in the code comments leaving the jumper off the program will auto detect the memory type.
Finally built mine today :)
The serial output is handy for capturing test results, hoping seller for dram will replace the faulty "new" chips based on the failure log.
3:00 Hey Jan, noticed you burnt you left thumb... was that a solder burn?! Hope you ok! 😮
It was an accident with a circular saw actually. But I'm fine, probably going to leave a visible scar for some time though. :D
Hi, can I test dram 2107 with this tester? Thanks
Hi the other type with the display , is it disponible with a PCB?
Could this be made to test 4116's and 4416's? Maybe by changing the programme on the arduino?
Should be doable, the test procedure would be very similar. The main problem is that those older chips require +12V and -5V as well as the +5V supply voltage so it would need some extra voltage regulation to make it work.
@@JanBeta Maybe this would make an interesting video?
It would be nice to have a device that would test the DRAM's of several old computers.
Where could one find that Perf board I have no clue???
Search for perfboard/matrix board online. Usually readily available from electronics components sellers, PCB manufacturers and even on Amazon/ebay.
Hey Jan, what about also detecting type of logic 74xxxx ICs?
detecting those is difficult because there are so many different pinouts and so many ways you can potentially damage the chip. If you want to simply test 74xx chips there are some very nice projects that do just that, like the RetroChipTester Pro (RCT) for example
That requires a much more sophisticated circuit and code. This is just a simple, "quick & dirty" DRAM tester that can do 4164 and 41256 DRAM ICs, and that's it.
@@Darkstar2342 That's what I mean (74xx). I got a few bootleg games for example with scratched ICs. Knowing what they are must be simple if you have a table for input injection/output measurement, so coding that must not be so complicated.
@Mr Guru it can, but only a small subset of all 74xx chips, and it explicitly warns you that it can damage chips that don't conform to the standard pinout (which some 74xx chips don't)
A lot of those logic chips can be checked/ detected with tl866 or the newer T56 eprom programmer
recommended solder you're using?
I primarily use Stannol 60/40 leaded solder with .7mm diameter but it's a matter of taste, really.
Do you also sell the PCB still?
I never sold the PCB, the project is openly available to have it manufactured by a PCB house though.
There are 666 likes as I type this... I can't do it!
Haha, understandable! :D
I'm writing this comment because a mistake happened, I wanted to give you a like and I accidentally clicked on the dislike next to it, I immediately corrected it and this time I really gave a like, but when I accidentally clicked on the dislike, it said that feedback was shared with the user, so I am writing this comment to assure you that it was a mistake. (Google Translator used)
No problem! Thanks for pointing that out! :)
I can’t support your content anymore because of how you fed into the drama to rip the retro engineering community apart when you went against the Commander X16 and the 8 bit guy by siding with the Z80 project which misses the point that it is providing a 6502 powerhouse to the community
Look, if David sends me a Commander X16 to review, I’d gladly do it. It just didn’t happen (yet). I am not taking sides for anybody. Every single one of the recent retro inspired projects is awesome in their own right. Especially the fact that they exist at all at this point in time.