Failed to boot from floppy

Hi there,

Your motherboard in an Apple II+ (not an Apple //e).

Can you share pictures of the boot screen without any card  in slot ?

The disk controller should be in slot 6. A language card with the cable to this white socket should be in slot 0.

BIT wrote:

Then I removed a chip from the card (ET 4116N-3 / M5) and put it in RAM socket C58 and the system entered the monitor; I immediately turned off because the chip was too hot (maybe it burned). 

No, no, no, no... What exactly have you done? Did you plug a RAM chip into the white socket? Nooo... This socket is not for a chip. It's for the ribbon connector from the language card  - your gray ribbon cable with the blue plug. That's why this socket is white - and not black like all the other sockets to be used for chips.

If you really removed a single RAM chip and plugged it into this socket, then at least this chip will be fried. Maybe more. Wish you luck! ;-)

Don't be too brave before testing such things. Better ask first...

Btw, C58 is the label of the "capacitor 58" - which indeed is next to the white socket for the ribbon cable. Oh dear!

MacFly wrote:

This socket is not for a chip...

 Actually, what BIT did was perfectly legitimate. That socket should contain a RAM chip if you want to restore the motherboard to 48K. They made the socket white in later revisions to help people identify which chip needed to be removed and replaced with the jumper cable.

If the chip overheated it was either inserted incorrectly or was already bad. You can contine to run the computer as is - with the RAM chip missing - but it will only recognize 32K which will prevent some programs from running. Eventually you will want to get a working language card installed so you can bring the computer back to 64K of RAM.

Hi Fabio, 

Please ignore the PM I just sent you.  I see this is an Apple II+ and you are asking the forum for help to troubleshoot this.   What disk are you trying to boot with?  If you have a DOS 3.3 disk, it should boot without the 16K card in the unit.  

Also - Can you describe what happens during the boot process?  Does the motor continue to run and light stay on, or does it stop?   

Thanks,

Jay 

BIT wrote:

Hi everyone,I tried to put this card, not original, which should be a memory expansion, in slot 0.The boot tried to read disk drive 2.Then I removed a chip from the card (ET 4116N-3 / M5) and put it in RAM socket C58 and the system entered the monitor; I immediately turned off because the chip was too hot (maybe it burned).If I leave the RAM socket C58 free, and remove the RAM e

You might have inserted the ribbon cable (16 pin DIP) backwards, and damaged all of the RAM on the card. These cards are thankfully cheap and plentiful, last I knew.

Please circle, on that photo, the exact chip that you pulled and tried to use. The 4116Ns are the actual RAM chips. 

MacFly wrote:

BIT wrote:

Then I removed a chip from the card (ET 4116N-3 / M5) and put it in RAM socket C58 and the system entered the monitor; I immediately turned off because the chip was too hot (maybe it burned). 

No, no, no, no... What exactly have you done? Did you plug a RAM chip into the white socket? Nooo... This socket is not for a chip. It's

Ehhhh? I've seen systems with a RAM chip in the language card RAM-Exp. What is this, now?

In fact, I have pulled that chip to insert language+16K RAM cards.

BIT wrote:

replay :)Thanks to everyone for the answers: you are very kind.Sorry for my english.1- my Apple II is "Europlus" (thanks amauget)2-the flat connector of the expansion card was mounted correctly; the 4116 chip was also installed correctly3- (to MacFly) I remembered that the white socket could be used to complete the memory bank if the expansion card was not installed.4-no

6. This could either be because (1) You lack sufficient RAM (48K is a general minimum), (2) the diskettes are bad, or 93) the speed of the disk motor is wrong. DOS 3.3 is easy to load as it contains no special copy protection schemes, but games usually do, and disk speed issues will cause havoc with them. 

2. That board has eight 4116 chips. Perhaps tey each one, and be absolutely sure that it is oriented properly and that its pins are all seated. 

In BASIC, use the command FRE and tell us how much RAM it reports, please:

PRINT FRE(0)

BIT wrote:

Hi Timelord,

 

without boards, the FRE (0) command returns me:30717

 

with only the board expansion RAM card, it returns me-18435

with expansion RAM and controller Floppy Disk, it return me the error:\0105-    A = 5C   X = 0B   Y = 00   P = 36   S = FE*

 

Thank

This tells me that your system seeks 32K. The last bit is a drop to monitor. 

Try each of the 4116 chips on the board, and try the same FRE command if any of them allow you to boot to BASIC. In that event, you should see a 48K system. 

If you desperately need a RAM board, I probably have spares, but no spare cables. I can test the bugger and ensure it is good here, but that is the end of my guarantees, as I pretty much do not comply with putting a mainboard in a dishwasher as an appropriate cleaning technique, and I cannot rule out any sort of issue with doing that. 

At the worst, it'll be cheap, but at the same time, I cannot promise any sort of fast delivery, as it simply cannot be at the top of my priority chain. I will test spare boards when I have a chance, in the event that you would want this; if not, you can try a known good cmplete language card from someone else. 

In the meanwhile, try the other 4116s. 

BIT wrote:

Hi Timelord,I don't need 4116s more, but a way to check the efficiency of what I have.I'm looking for an inexpensive DRAM TESTER, even with Arduino or Raspberry.Can you give me some advice?

https://hackaday.com/2016/03/18/testing-dram-one-byte-at-a-time/

BIT wrote:

I looked at the circuit diagram, but I think it's dangerous: if any one of the RAM's pins is shorted to ground, the arduino dies: OCan anyone confirm this to me?Any other projects or kits for sale?

https://atariage.com/forums/topic/289391-dramarduino-dram-tester-with-arduino/

I suggest discussing it here, as several of the people have built one, and you need to modify the design to work with 4116 chips anyway to provide +12VDC.

There are methods of hand-testing these chips for shorts, and thos could, or may already, be designed into this. I have yet to build one, and I do not know all of the specific details on the Arduino devices, as there are so many flipping models and variations. You can potentially ohm out the power and earth pins and see if you get a zero reading between any pairs--that'd be bad. 

RAM, particularly 4116s, that overheat that quickly, are likely shorting on the +12VDC pin, depending on the current draw, +5 or -5 could do it, I suppose. 

Some old PROM and EPROM encoders have the ability to test comon RAM chips, too, but I do not know what to advise there. I could use similar advice on what EPROM programmer to get these days, to pop out ERPOMs for the ][ series, that I can easily interface with a Win7 laptop. (I thankfuly still have serial in/out, because this is a military application unit.)

Keep in mind that the different-sized RAM ICs do not share a single, universal testing method, so the fact that this cheap device with a little work on a perf board can act to test at least three common types seems ideal. IDK how much I truest the construction of the base device anyway, but at the least I'd trust the parts that I assemble, which would certainly not be RoHS. :D

Here is some additional content for you:

http://labs.frostbox.net/2020/03/24/4116-d-ram-tester-with-schematics-and-code/

https://www.youtube.com/watch?v=MVZYB54VD2g

https://spectrumcomputing.co.uk/forums/viewtopic.php?t=1426

https://archive.org/stream/radio_electronics_1991-05/Radio_Electronics_May_1991#mode/2up

This issue has a detailed article about making a 4164 tester. You would need to adapt it to work with 4116, but it has very useful information. 

I will again say that one of the safest ways around this issue is to buy some tubes of new stock 4116Ns. It's a good idea to keep tubes of fresh 4116, 4164, and 41256 chips in sock if you work with a lot of vintage micros. At least all of the chips on the ][ and ][+/europlus are socketed. 

BIT wrote:

Hi Timelord

one moment, I don't understand.You say that I have only 32KB; but studing, I found that the result:-18435means: -18435 + 65536 = 47710which are 48KB

are you sure the problem is a defective 4116 chip?

 

Integer rollover here. AppleSoft BASIC uses signed 16 bit ints, so, as soon as the count hits 32767, it rolls over to negative values.

I was quoting this:

without boards, the FRE (0) command returns me:30717

That is 32K. With the RAM board, you should be getting a negaive value of ~ -30K. 64K rolls over, but BASIC/DOS eat up some RAM, so you end up with around -30000 free if you are seeing all RAM on a 64K system.

Likewise:

with only the board expansion RAM card, it returns me-18435

If you saw that value on a 48K system, then, aye, it is fine, but if you had the language/RAM card installed, then one bank of 16 bits * 8 was not registering. 

With the FDC in slot 6, do this:

Type CALL -151 followed by RETURN key.

At the asterisk prompt, type C600L and return.

Send a screenshot of what you get.

What happened? I thought that you sorted the issue with the Disk ][ card and had this sorted, and now you were on to memory size???

BIT wrote:

Should this command be given with 4116 inserted on the white socket?

 Doesn't matter.

I'm confused that earlier you said the disk drive was working (with some disks) but now when you issue the FRE() command with the controller card present it drops you into the monitor. That would seem to indicate some issue with the card (or the slots, since you've already tried another FDC). That's what concerns me.

BTW, the value -18435 is correct for a 48K machine. Applesoft doesn't use or care about the extra 16K on the Language Card.

The CALL -151 command was meant to be issued from BASIC ( ] prompt) to get you into the monitor ( * prompt).

From your response, it appears you were already in monitor so it attmpted to list memory. But even this shows something wrong.

00F1 - 01 00 ORA ($ 00, X) = 79   (I assume you translated ORA to HOUR) is the correct response whether the white socket is populated or not.

Unless you typed it incorrectly the first time, the   00F4-FF ???=79 response is not correct (even though you said you typed CALL -150, the response indicates that you correctly typed CALL -151). If you can repeat these results then there definitely is something weird going on with your motherboard. I'd start with swapping the F8 ROM to see if that changes anything.

Might be the 74LS138 that handles ROM slot select, too; or one of the ones that handles slot select.

The slot select chips are in row H, the ROM select is in row F, iirc. ROM select is in the same bank as the ROMs, to their right as I recall. 

Please don't use dishwashers to clean electronics...

There is also a 74Ls08 that works in the ROM select circuit.

P.S. I would not have suspected F8 vecause it does drop to BASIC w/o the Disk ][ card, and it does try to autostart. I suppose it could still be, though?

TBH, this seems like a slot bus IO conflict or general data corruption. 

Wasn't there a testing aid that was available on cassette at one point?  Perhaps coold booting from some of the tape stream images might help?

Where are you located?

BIT wrote:

Hi Timelord,I'm from Verona (Italy).Why don't you like  my dishwasher?Can you tell me only one scientific contraindication, after 8h in a 45 ° C fan oven for drying?

You are ignoring the basic issues with this cleaning method.

Water jets, ion streams, minerals dissolved in the water that are now on all of your contacts. Heat exposure, depending on your washer, damaging components and joints. 

There is a reason that we don't do this with sensitive components, boards, and sockets. The trace elements and salts residues alone could be causing havoc or igniting long-term deterioration.

That said, if it has this issue aforehand, then, it clearly isn't immediately/directly related, but I will gladly pay the price/litre for denatured alcohol (and I do) to never need to worry about this kind of hazard.

You are entirely ignoring the process involved, speificallt the dissolved elements in the water, its application, and its temperature. If you soaked the boad severl times in a bath of distilled water, then, aye, that may be safe, but spraying jets of your municipal water source at it, and potentially hot jets, could easily damage things.

Anyway, I know someone in Italy with ][+ components. If you want me to put the two of you in contact, I will. It may be the best way for you to find ][+ ICs without resorting to buying International. 

BIT wrote:

Hi Timelord,I tried to replace the 74LS138 with another certainly working, but that's not it.I tried to replace the 74LS08 with another certainly working, but it is not this.I didn't understand the other tests to do ... :(

Did ou test all three locations of the 73S138?

if all of those passed, it is on to the F8 ROM swap, as advised, and then the CPU/RAM. I do not know how much time you are prepared to invest, but ][+ maiboards remain rather inexpensive, when availavble. Once ou delve into an issue this complex, the ASGs say to 'replace the mainboard', and other literature has no one specific answer.  

Keep in mind that you cannot socket an EPROM in the ROM sockets without making an adapter socket to go between the mainboard socket and the EPROM.

BIT wrote:

Hi Timelord,I replaced the N. 3 74LS138, but nothing has changed. :(I can't replace the single ROM D8, because the Apple D8 EPROM clone, I don't think it's compatible.Can I replace all 6 ROMs with the 6 EPROM 2716?I can also create an EPROM, if an image of the D8 ROM is available somewhere: I have a beautiful HILO ALL-11 that still works, but I have to prepare a PC with ser

You cannot put an EPROM into those sockets without construction an adatper using two sockets wired up to do pin conversion. 

See:

http://www.willegal.net/appleii/appleii-integer.htm

https://www.classic-computers.org.nz/blog/2010-03-06-problems-with-replacement-appleII-eproms.htm

https://atariage.com/forums/topic/229489-replacing-apple-ii-roms-with-2716-eproms/?do=findComment&comment=3068099

I am not sure what a lot of you are saying in your latest post means. Did you install ONLY a total of 16K, only filling one bank, with nothing in any slot, including nothing in slot zero?

I suggest ordering 16K of new 4116 RAM, stripping all of the existing RAM out, and installing the new ICs. Be absolutely sure to be ESD safe. Most of the components that we have been describing that can fail are quite ESD sensitive. All RAM chips in a single 16K bank should be of the same brand, and must have the same  timing.

If you are mixing RAM in a bank from different manufacturers, or with different timing, then it can result in bit degredation or shifting that will corrupt values. 

The general procedure for intermittent errors is as follows:

Replace, then test each of these in sequence, one at a time:

RAM, in banks, with known good RAM.

74LS139 (2/4 Decoder) at mainboard location F2

74LS139 (2/4 Decoder) at mainboard location E2

74LS20 (NAND gate) at mainboard location D2

74ls138 (3/8 Decoder) at mainboard location F12

6502 CPU at mainboard location H7

ROM Select 74LS138s. 

ROM Set; the rightmost ROM is the lowest in the memory map.  F8 is what has been suggested to replace, but as you can get to BASIC w/o crashing, I feel this is wrong. F8 is the lowest address space, and contains the monitor and autostart, as I understand things. I suppose it could still be bad, and you would have some corruption in memory, which could in theory explain a lot of things, so if you want to replace a ROM, start with F8, then F8, then E8, then E0, then D8, then finally, D0. 

You can probably still find a complete set of Apple ][+ AppleSoft ROMs inexpensively. 

When replacing the CPU, be sure to use only parts 6502, and 6502-II, or 6502A if you cannot find the first two. Do not use alternatives such as the 6510, as these are not pin compatible, and I do not advise using a 65C02 in the ][+.  Although the CMOS 65C02 is pin-compatible with the NMOS 6502, I find that I will want NMOS components when possible on the ][ and ][+, but that could merely be my opinion and experience. 

BIT wrote:

Hi,the computer started with only the 16KB desk (one pin was not inserted well).the CAL -151 command always returns the address to me:E009-*I gave the speedyG command "C050 C053 C054 C057 N 265: FF N 266 <265.BFFEM 266 <265.BFFEV 265: 0 N 266 <265.BFFEM 266 <265.BFFEV 34:14 "[[{"fid":"30967","view_mode":"default","fields":{"format":"default","alignment":""},"link

I have no idea what ROMs you are using there. It is hard-enough diagnosing this system without using ROMs out of some sort of clone machine.

You might try to email Brian at ReActiveMicro and ask if he can make a pin-compatible AppleSoft ROMset for you. Perhaps there are EPROMS or PROMS that he can use that won't need adapters. Otherwise, you should buy a full AppleSoft ROMset that is known to be good. TBH though, buying a full spare mainboard is probably the most cost-effective solution, given that it is likely only to be 50% more than the ROMs and it will give you a complete extra chipset Unless there is something very special about that board, it'd be less trouble, although if you have the Europlus model, that can be trickier  

No. This command fills all of RAM with FF's then verifies that it is correct; then does the same with all 00's. While not strictly necessary, it selects the HGR display so you can get a visual indication if there are gross memory errors.

As writtem it assumes 48K. To mofdify for 16K, change to C050 C053 C054 C057 N 265:FF N 266<265.3FFEM 266<265.3FFEV 265:0 N 266<265.3FFEM 266<265.3FFEV 34:14 (enter)

The screen should fill with all white (plus 4 rows of  "?" at the bottom). Then change to all black (with 4 rows of  "@" at the bottom). Any other output to the bottom of the screen would indicate verify errors.

Change all of the original "B"s to "7."

C050 C053 C054 C057 N 265:FF N 266<265.7FFEM 266<265.7FFEV 265:0 N 266<265.7FFEM 266<265.7FFEV 34:14 (enter)

BTW, don't know what the 34:14 at the end is for. You can omit that.

BIT wrote:

The new 16MB bank (32MB total) works fine (thanks jeffmazur for the command)

Now excuse my question, perhaps stupid.I understood that Apple Disk Server can create FDs starting from the image contained in the server, using the FDC in slot 6.My problem is that I have to test RAM, ROM and also FDC.Is there a way that allows me to load directly into RAM the only tool to perform th

To create disks, you need to first hot-load ADT-PRO, as far as I am aware. You should be able to run anything that you steam, but making a boot disk, not as easy.  Otherwise, you can try to hotload DOS 3.3. I take it that your master disk still doesn't boot after getting 32K of RAM working?

You need 64K for ProDOS, so, 3.3 (or 3.4) is your best initial option. 

IDK what the freight would be, but I would sell you a set of master disks at my cost for the media, plus freight, if that is what you need. 

I have:DOS 3.3

DOS 3.4 (community-developed patch of 3.3)

DOS 3.3 Sample Programmes

DOS 3.3 Muffin Loader (for 13-sector disketes)

ProDOS (all versions)ProDOS Users' Disk

ProDOS Workbench Utilities (inspection)

GEOS v2.1 (4 diskettes, double-sided)

Apple II Desktop (all versions)

PASCAL Software Set (7 diskette sides)

... Pretty much all OS disks, and HW diskettes, beyond this list.