Chip indentification

2»

Comments

  • I think the schematic points to U2 being a standard pinout RAM, something like a 6116. Pin 20 on most early 24-pin chips of that type was /OE which appears o go off board along with pin 21 which was /WE (write). Likewise you'd expect to find the I/O on pins 9-17 and they go towards the transistors.

    Thank you, very helpful.
  • Cluso99 wrote: »
    Brian, did you read my comments above. It looks like they are all inputs with only pin20 as an output.
    The circuit diagram is wrong as pin 20 enables all of the NAND gates in U3 and U4.
    Perhaps U2 is used to disable the organ in case of errors???

    I did. But I am confused about the diagram. If I look at the photos that Electrodude tweaked, I can convince myself that pins 21 and 20 are connected to tracks on the top side of the board and that they go to the edge connector.

    There is an issue on the visuals, I will post a photo with the chips removed.
  • The previous couple of posts seems to indicate that the board stores a number of presets. Is this correct? If so, then perhaps it is more likely to be a RAM chip. Those answers on the schematics and the tracks under the pcb will help solve what U2 is likely to be.
  • @jonathanl64 I know it is a bit sideways of the current topic, but I'd love to see a picture of this organ and the internals when you get a chance. I did several big theatre pipe organ rebuilds 30+ years ago back when I was an active member of ATOS and I dearly love anthing organ-ish. Wind or wired, it's all good!
  • I'll get some more information up later, I've had a long day in the workshop on something I have no problems with, woodwork. Thank you all so far for the info provided.
  • Immediate answer to two questions, I have taken the chips of the board to trace the tracks, and have attached the photo of the board without the chips. Secondly, the boards are mounted in a rack and not accessible when the organ is in use.

    To clarify, this is the capture system for the pistons, which means it has not stopped the organ playing, it just means that changing stops has to be done by hand. In my role as organist, I tend to do this much more than use the pistons, and the organ of which I am currently Titulaire doesn't have any aids to registration and is entirely mechanical! However, most organists like to use their pistons!

    A final note before I do a thorough recheck of the schematic, having done some research on RAM chips, I think this must be SRAM, which makes sense, although my schematic doesn't quite fit with what is needed for a SRAM.

    3264 x 2448 - 2M
  • ElectrodudeElectrodude Posts: 1,374
    edited 2020-04-06 - 18:18:38
    @Cluso99: I think pin 20 of U2 has to be an input driven by the external connection on the left of the board, because there are several of these boards that are all told to load their presets simultaneously by off-board circuitry.

    @jonathanl64:

    Thanks for taking the picture with the ICs removed. That eliminates a lot of guessing. If you or anybody else wants me to, I can align that image image with my aligned versions other two. Let me know.

    I agree that it must be an SRAM. Looking at the PCB, it's perfectly compatible with the 6116 RAM that @Brian Fairchild suggested it might be. If this is the case, then Pin 20 (!OE) is externally driven low when a preset button is pressed to load a preset, and Pin 21 (!WE) is externally driven low when the set button is pressed to store a preset. Of the 6116's eleven address lines, they all come from eleven pins of the second-from-the-top group of external connections on the left side; the top five go through U1 and then to U2's A0-A4, and the lower six go directly to U2's A5-A10. The five that go through U1 presumably hold other values when U1 isn't being clocked, which seems to imply that they connect to an MCU bus or some other shared bus. My theory is that the five clocked lines specify which bank of presets is active, and that the six traces that go under U2 from below connect to the other six address lines; if not this, perhaps it's the other way around. I wonder if the the six address lines could be one-hot instead of binary? Are there exactly six preset buttons in each group? Perhaps the cancel button is implemented as a special preset that's always left empty.

    Since it's a RAM and not some form of non-volatile memory, does this mean that the organ forgets all of its presets when it loses power? That seems like it'd be really annoying, but I guess it's better than not having any registration aids at all. (Does anybody know if there are pin-compatible EEPROMs that could be used here?)

    That's great that you can do the actual organ work yourself. The terrible state of disuse and disrepair of most organs in Catholic churches in the US nowadays is really sad. With some exceptions, most parishes nowadays don't seem to care about beauty anymore and are strapped for money anyway. Most have just gotten electronic organs, which sound great but don't shake the air in the building in the same way. The organists I know who are interested in repairing their organs themselves tend to not have enough knowledge to actually do the repairs.
  • I am not an electrical genius, like most of the folks in the discussion, but if I had other boards that were working, and even though I tested all other components and IC's, I would certainly desolder U2 off of one of the working boards and solder it to the non-working board, just to make 100% certain that it is actually the culprit.

    I am sure you are very compotent and correct, but just in case.... :) If there was an error during troubleshooting, you may be trying to reverse engineer a pcb when it is not necessary.

    Please forgive my suggestion if you already tried this. :)
  • idbruce wrote: »
    I am not an electrical genius, like most of the folks in the discussion, but if I had other boards that were working, and even though I tested all other components and IC's, I would certainly desolder U2 off of one of the working boards and solder it to the non-working board, just to make 100% certain that it is actually the culprit.

    I am sure you are very compotent and correct, but just in case.... :) If there was an error during troubleshooting, you may be trying to reverse engineer a pcb when it is not necessary.

    Please forgive my suggestion if you already tried this. :)

    All suggestions welcome! We changed U2 over and proved the board worked, and that the problem was with U2, hence the need to replace it!
  • That's great that you can do the actual organ work yourself. The terrible state of disuse and disrepair of most organs in Catholic churches in the US nowadays is really sad. With some exceptions, most parishes nowadays don't seem to care about beauty anymore and are strapped for money anyway. Most have just gotten electronic organs, which sound great but don't shake the air in the building in the same way. The organists I know who are interested in repairing their organs themselves tend to not have enough knowledge to actually do the repairs.

    I spent the first 25 years of my career not building organs, although I gained lots of experience in my youth. Glad I stopped working professionally as a musician to start my own organ building business, and I'm glad I did.

    It is so sad to see pipe organs disappear in favour of electronics. It's short term saving for long term pain, they don't last long! If you have a good organ builder, and listen to their advice when they come in to tune, the cost is much reduced as they spot problems before they get expensive!

  • An old EEPROM would be a 2816.
  • I have done some more work on the schematic, which has corrected a number of errors. The unknown chip now has all its pins connected, and doing the right thing (I think!) Still checking the electrical connections, and I have tidied up the schematic a bit, although I'm sure a professional could do so much better!

    Revised schematic attached.
  • Another small revision to include pin names for the chip, which now make sense. It must be either a SRAM from what I now see, and thanks to everyone for their help so far.

    It was a long day in the workshop before and evening of tidying up the schematic and rectifying the errors (I'm sure there must be some even now.)

    I'm going to get a good night's sleep and look and check again tomorrow.

  • I don't think pin 21 is tied to VCC as you have indicated. I think pin 21 only connects to R3 and the connector. U2's VCC should instead be connected to the other wire above it that goes to the top of C1. U1's pins 1 and 20, which are also the same VCC, are also connected together, which I don't see reflected on your schematic.

    I adjusted your top image with the ICs removed to line up with my other aligned versions of your images. I also transparently overlaid the new top picture with the bottom picture; I don't know if that picture will be useful, because it's not very clear.

    I think you want an EEPROM or an NVSRAM and not an SRAM; otherwise, it will forget all of its presets when you turn the organ off. I don't know where people buy electronics from in Europe, but a quick search on Digikey finds these weird-looking brick things that should fit in U2's place and have the same pinout.
    2974 x 2289 - 5M
    2974 x 2289 - 5M
  • Thanks for posting the pic with U2 removed. This shows that U2-20 is indeed connected externally, so it can be an input to U2-20, and U3 & U4 too. So it is /OE (output enable) and when low U2 will do its' job and when high U3 & U4 will do their job.

    My reasoning that it could not be an EEPROM/RAM/OTP was that there was no input to U2-20 and as such it therefore had to be an output going to U3 & U4 as an enable, and hence my reasoning it had to be a PAL. But now we have proof that U2-20 is being driven, it is an enable select wire, enabling either U2 or U3 & U4. So it can (an almost certainly is) be an EEPROM/RAM/OTP.

    The original top side photo showed the tracks to U2 21, 20 & 18 were cut. I now presume these cuts were done during fault detection.

    So we know U2 is a storage device, we need to know...
    Does the presets get lost when power is removed?
    If so, then U2 is most likely a SRAM. If not, then U2 is an OTP EEPROM or similar.

    As Brian said earlier, a 6116 SRAM would be most likely.

    An EEPROM/OTP/Flash would be less likley, as typically of that era they require higher voltages to program, and this does not appear to be likely on this pcb.

    So, with what we now know, trying a 6116 is not likely to do any damage.

    BTW I would buy some 100nF mono thru hole capacitors and solder to the pins under each of the ICs from Vcc to GND (ie diagonally). Don't forget to put some insulation on the legs to prevent a short. Small heatshrink will work nicely.

    You can repair the track damage by using Kynar wire.
  • I think we can exclude NVRAM. When that board was made, IIRC, all the NVRAM around was a lump of epoxy housing a standard RAM chip, some logic and a battery. Or the chip socket with the battery built in. I've just found some of the lumps in my chip drawers.

    Standard RAM is still a fit if the volatile nature is acceptable.

    Likewise a window-less EPROM if the fixed nature is acceptable.

    What I have just found in the same drawer are some X2864 chips which are EEPROMs from the right period of time. Mine are datecode 8504. And there was a 2k version the X2816. These are single rail EEPROMs with a standard pinout. And you can still find them on eBay for not much money.

    So, my money is on a 2816. Datasheet attached.
  • Bingo, I've just just had a thought and turned my chips over. Look familiar?


    800 x 666 - 52K
  • I don't think pin 21 is tied to VCC as you have indicated. I think pin 21 only connects to R3 and the connector. U2's VCC should instead be connected to the other wire above it that goes to the top of C1.

    - corrected
    U1's pins 1 and 20, which are also the same VCC, are also connected together, which I don't see reflected on your schematic.

    - corrected
    I adjusted your top image with the ICs removed to line up with my other aligned versions of your images. I also transparently overlaid the new top picture with the bottom picture; I don't know if that picture will be useful, because it's not very clear.

    Everything is useful!
    I think you want an EEPROM or an NVSRAM and not an SRAM; otherwise, it will forget all of its presets when you turn the organ off. I don't know where people buy electronics from in Europe, but a quick search on Digikey finds these weird-looking brick things that should fit in U2's place and have the same pinout.

    I will look these up, a lot of our stuff comes from China, I use Mouser, RS-online or Farnell.

  • Cluso99 wrote: »
    Thanks for posting the pic with U2 removed. This shows that U2-20 is indeed connected externally, so it can be an input to U2-20, and U3 & U4 too. So it is /OE (output enable) and when low U2 will do its' job and when high U3 & U4 will do their job.

    My reasoning that it could not be an EEPROM/RAM/OTP was that there was no input to U2-20 and as such it therefore had to be an output going to U3 & U4 as an enable, and hence my reasoning it had to be a PAL. But now we have proof that U2-20 is being driven, it is an enable select wire, enabling either U2 or U3 & U4. So it can (an almost certainly is) be an EEPROM/RAM/OTP.

    And can see this now and understand the functioning. I have now read up on how the chips function, so that helps with understand my schematic errors.
    The original top side photo showed the tracks to U2 21, 20 & 18 were cut. I now presume these cuts were done during fault detection.
    - Yes
    So we know U2 is a storage device, we need to know...
    Does the presets get lost when power is removed?
    - Yes
    If so, then U2 is most likely a SRAM. If not, then U2 is an OTP EEPROM or similar.

    As Brian said earlier, a 6116 SRAM would be most likely.

    An EEPROM/OTP/Flash would be less likley, as typically of that era they require higher voltages to program, and this does not appear to be likely on this pcb.
    - noted

    [/quote]So, with what we now know, trying a 6116 is not likely to do any damage.

    BTW I would buy some 100nF mono thru hole capacitors and solder to the pins under each of the ICs from Vcc to GND (ie diagonally). Don't forget to put some insulation on the legs to prevent a short. Small heatshrink will work nicely.

    You can repair the track damage by using Kynar wire.[/quote] - all understood.

    Thank you again for your time, and everyone else, I really am dusting off the cobwebs here.

  • Bingo, I've just just had a thought and turned my chips over. Look familiar?


    Indeed they do! Thank you, lots of very helpful material here.
  • Thank you all, lots of useful information so I think we now know what the chip is. Thank you also for your welcome and engagement, I will now start looking through some of the other threads!
  • I'd now be inclined to fit a DIP socket in U2s position to make testing easier.
  • I'd now be inclined to fit a DIP socket in U2s position to make testing easier.

    I agree!
  • frank freedmanfrank freedman Posts: 1,615
    edited 2020-04-16 - 03:11:40
    Just another thought for consideration: if you have an extra board, especially a dead board, it should not be hard to turn it into an extender board thus making the signals in question accessible so that a more educated guess as to function and timing could be made/observed. May take a bit of digging to get hold of a connector of the type on the backplane, but may be worth it. Mount the connector on the edge of the dead board. Cut traces going on the dead board and run green wire to the connector for the board under test. Now you can scope or meter or logic probe the suspect board during normal operation.
Sign In or Register to comment.