810 drive won't stop spinning?

[spacecadet]

I recently purchased two Atari 810's (along with a bunch of stuff) - one works totally fine but the other has an odd problem in that it just never spins down once it gets going. It otherwise works and loads data just like it's supposed to; it's just that the motor never stops. The other drive that works fine does stop, so I know it's not just the programs I've been loading.

 

Has anyone seen this and knows what it is? If it matters, I noticed that the "busy" LED on this drive is also very dim. I'm not sure if that could be related or if it's a separate problem.

[+Nezgar]

I'd start by re-seating all cable connections inside that connect between the boards.

 

Since you conveniently have two drives, I'd start by swapping the side board between the two drives. and see if the symptoms transfer. This would help narrow down if it's the side board, or the mech. (edit: or the power board...)

Edited June 1, 2018 by Nezgar

[spacecadet]

Well it's unfortunately not just a bad connection; I did try reseating everything (cables and chips) and it didn't do anything.

 

Unfortunately the good drive is up for sale right now so I can't really open it up anymore. I don't have a need for two drives at all, so I'm selling the good one to try to pay for some of the other stuff I bought and want to keep from this package. I don't technically even need one (I have an SIO2SD) but I was just hoping to get the second one working and then I might keep it because these drives are just cool. If I can't though, I might just sell this one too (as broken).

 

I noticed that it looks like one of the big caps in the back (on the power board?) might be bulging a bit at the top. Could a bad cap cause these issues? Photo is attached - the angle isn't the best for seeing that cap, but I didn't notice the small bulge until later.

Edited June 1, 2018 by spacecadet

[+Nezgar]

If you have a multimeter, you can verify the voltage going into, and coming out of the 2 voltage regulators.

 

One will have 9VAC going in, 5VDC going out. The other will have 18VAC going in, 12VDC going out.

(Centre pin is ground, careful not to short them, it makes sparks).

[Fox-1 / mnx]

it just never spins down once it gets going.

 

Hmm. I have one that doesn't spin and never get's going. Not that this is of any help...

[7bits]

I have a similar problem, I have both the power and busy light on, and it just spins, no seeking, track 0 reset or anything else. The field service manual says check chip a104 pin 9 to drop from 5v to 0 after 7 seconds, if it's doesn't replace the chip. This is the riot chip on the side board so I ordered one.

 

However I doubt this is it but the chip is only 5 dollars so ordered to test and replace. Says to then replace certain transistors. 

 

I did check the test points on the power board, all good, and the regulators have good voltage.  I have the rev c rom, the data seperator board, last psu and the analog board on top of the drive.

 

Only other thing I noticed is some of the transistors get hot on the side board on top, above the psu board connector.

 

I also pulled all the chips and deoxit and tested all the 74ls chips on the data seperator board

Edited February 18, 2021 by 7bits

[+Nezgar]

32 minutes ago, 7bits said:

I have a similar problem, I have both the power and busy light on, and it just spins, no seeking, track 0 reset or anything else. The field service manual says check chip a104 pin 9 to drop from 5v to 0 after 7 seconds, if it's doesn't replace the chip. This is the riot chip on the side board so I ordered one.

A104 is the 6532 RIOT.... Pin 9 goes to pin 9 on the pin header - which goes to a transistor array I believe controls the ~12V (depending on tachometer feedback) to the drive motor...

 

Personally I'd suspect A103 (The 6810 SRAM) or A102 (The 2316 mask ROM). My rationale is that at least from experience in the 1050, the drive won't 'spin up' unless the CPU actually gets far enough to execute code to turn on the motor...

 

It would be useful to know if swapping any of those 3 chips from a working drive helps... The 6810, 6532, and 6507 are also used in the 1050, so can be swapped interchangeably there too:

• The 6810 SRAM (810's A103) can be tested in the 1050's U8 socket.
• The 6532 RIOT (810's A104) can be tested in the 1050's U7 socket.
• The 6507 CPU (810's A101) can be tested in the 1050's U9 socket.

 

The 1050 will clock these at 1Mhz, vs the 810's 500Khz - so it's a good test of their tolerance too...

 

The 810's 2316 ROM in A102 can be dumped by plugging it into a socketed brown-shell Atari cartridge, ie BASIC and using DOS 2.0 to save a chunk of memory to disk. If uploaded here i could analyze that dump to ensure it verifies against a known good Rev C 810 ROM dump:

 

 

[7bits]

Thanks, I have a usb rom burner, will see if I can read these mask rom chips. If they are the same as the c64 kernel/basic/os ones I think I can dump them. 

 

Here is the flow diagram I was following from they guide: to be attached later

[7bits]

[7bits]

17 hours ago, Nezgar said:

A104 is the 6532 RIOT.... Pin 9 goes to pin 9 on the pin header - which goes to a transistor array I believe controls the ~12V (depending on tachometer feedback) to the drive motor...

 

Personally I'd suspect A103 (The 6810 SRAM) or A102 (The 2316 mask ROM). My rationale is that at least from experience in the 1050, the drive won't 'spin up' unless the CPU actually gets far enough to execute code to turn on the motor...

 

It would be useful to know if swapping any of those 3 chips from a working drive helps... The 6810, 6532, and 6507 are also used in the 1050, so can be swapped interchangeably there too:

• The 6810 SRAM (810's A103) can be tested in the 1050's U8 socket.
• The 6532 RIOT (810's A104) can be tested in the 1050's U7 socket.
• The 6507 CPU (810's A101) can be tested in the 1050's U9 socket.

 

The 1050 will clock these at 1Mhz, vs the 810's 500Khz - so it's a good test of their tolerance too...

 

The 810's 2316 ROM in A102 can be dumped by plugging it into a socketed brown-shell Atari cartridge, ie BASIC and using DOS 2.0 to save a chunk of memory to disk. If uploaded here i could analyze that dump to ensure it verifies against a known good Rev C 810 ROM dump:

 

 

So I swapped the following in the 1050 I have seemed to work ok i.e. booting DOS 2.5:

• The 6532 RIOT (810's A104) can be tested in the 1050's U7 socket.
• The 6507 CPU (810's A101) can be tested in the 1050's U9 socket.

I didn't do this, as I had never opened the 1050 since it just worked and I have had it for around 5 years now. Last time I opened a 1050 was in 1985 to install a Happy Upgrade

• The 6810 SRAM (810's A103) can be tested in the 1050's U8 socket.

Reason I did not swap the A103 is because I saw this in my 1050:

 

 

Don't want to get too far off topic  I am guessing I have a us doubler type mod? Never knew I had it. I assume this doubles the SRAM via some weird chip select method and I have a custom ROM thats been flashed... Looks like a poor piggy back mod and I think I may break some solder connections and I don't need a 2nd broken drive right now!

 

Is this likely just a bootleg US Doubler?

 

Anyways back to the 810. I was not able to dump the ROM off the Side Board, I need to find a spare LS logic chip to float some pins low on a bread board to get that chip to read ok on my burner. Believe it or not I have zero 8 bit carts!

 

I don't dare put the good chips in the 810 for fear that the 810 has some other over voltage situation I am missing.

 

So what I have noticed, the 1050 hooked to my 800XL with no disk and powered on I get the following when turning on the 8 bit host:

Boot error, repeated, which I think is normal

 

When I hook the 800XL to the 810 powered on with no disk, it just goes straight to the Ready prompt.

 

So I could get an SRAM and Mask ROM chip for giggles, it they are a few bucks to try, but I wonder, should I be looking a the power board again? The diagnostic flow chart says replace 104, 103 and 102 first, so RIOT, SRAM, ROM. 

 

The 810 below is what I have for reference. I am putting in a lot of info for the next googler with the endless spinning drive

 

If anyone has the SRAM an ROM chips that are known good for sale, let me know please!

 

Edited February 18, 2021 by 7bits

[_The Doctor__]

US doubler upgrade.. double check the solder on the stacked chips and seating.... they may have not done a great job or used acid core solder, so per usual clean and re seat make sure chips are oriented by the silk screening... white outlines and writing on the pcb

 

Edited February 18, 2021 by _The Doctor__

[7bits]

3 minutes ago, _The Doctor__ said:

US doubler upgrade.. double check the solder on the stacked chips and seating.... they may have not done a great job or used acid core solder, so per usual clean and re seat make sure chips are oriented by the silk screening... white outlines and writing on the pcb

 

Thanks! Once I get the 810 up and running, will clean that up

[TGB1718]

11 hours ago, 7bits said:

Is this likely just a bootleg US Doubler?

Looks  like both of my 1050 USD's  

[+Nezgar]

On 2/18/2021 at 4:56 PM, 7bits said:

I did not swap the A103 is because I saw this in my 1050

Yes the US doubler needs 256 bytes for the double density sector buffer, instead of the 128 bytes from a single 6810 SRAM. (On top of the 128 bytes of RAM in the 6532 RIOT used for the stack)

 

You could still test the 810's 6810 chip, but you'd have to program that EPROM or a new one with the stock 1050 ROM. You would use Rev "L" from here: https://atariage.com/forums/topic/156462-1050-roms/?do=findComment&comment=3580891

 

You can "upgrade" your 1050 US Doubler ROM with a patch I made that increases the stepper speed, and improves the standard sector interleaves for ED & DD.

https://atariage.com/forums/topic/158768-atari-1050-roms/?do=findComment&comment=4594602

Hah, apparently it's been featured on igi's blog- I Believe there's a typo in the sector order in that post I have to review again - and I intend to make a similarly patched ROM for the WST/Hong kong mechs too at some point:

http://blog.3b2.sk/igi/post/Atari-1050-ICD-US-doubler-Part-9-Nezgar-firmware-for-TANDON.aspx

 

On 2/18/2021 at 4:56 PM, 7bits said:

When I hook the 800XL to the 810 powered on with no disk, it just goes straight to the Ready prompt.

This means as far as the computer is concerned, there is no disk drive attached - so it is not responding at all.

 

On 2/18/2021 at 4:56 PM, 7bits said:

I was not able to dump the ROM off the Side Board, I need to find a spare LS logic chip to float some pins low on a bread board to get that chip to read ok on my burner. Believe it or not I have zero 8 bit carts!

Yeah, the 810's 2K ROM has been a challenge for others to read in EPROM programmers, it does in fact use a unique CS enable combination. Another idea is you could remove the BASIC ROM from your 800XL and replace it with the 810's ROM. That should also work to boot to DOS 2.X and "Save Memory".

 

On 2/18/2021 at 4:56 PM, 7bits said:

If anyone has the SRAM an ROM chips that are known good for sale, let me know please!

The ROM's are a bit harder to come by, but if yours is indeed bad, there are detailed posts on the forums of how to rewire your 810 to use a 2732 EPROM instead I can dig up if needed. If you're going to do that, you might as well upgrade it to an 810 Archiver ROM   6810 SRAM's can be found on eBay, I've had good luck so far for my US Doubler PCB's I've been making until my last batch of 40 fakes from china. ? But this seller from Switzerland sells legit ones: https://www.ebay.com/itm//201459455417 (and ships fast - I think it was less than a week to Canada at least!)

[7bits]

Thanks Nezgar! Will def delve more into your advice and links. Good thought on the 800XL to dump the rom.

 

Since I am stuck a little and the hot transitors got me thinking a short to ground somewhere, I decided to pull the 810 power board and take a closer look. Do some isolated testing by powering the power board up without the side board.

 

While this drive has been opened before, I only see evidence of someone looking around, not replacement or repair. All the boards look factory soldered, no bodge work.

 

I checked the bottom of the power board and while the joints don't look cold, they are arguably short on solder! So I went ahead and reflowed, inhaled 40 year old flux, and added fresh solder.

 

Before:

 

After:

 

I also went ahead and re-checked the voltages. I see about 13 in 5 out and 26 in and 12 out respectively, like before. Now I realize I am a dummy, and never saw nor checked the 3rd regulator

 

I see it is about -12v in and -1.1v out. Keep in mind this is just the power board only with the pin 1+2 jumpered on the header pins for the side board.

 

I assume this is not normal? Would explain why the LEDs look slight low. Looking at this chart is should be -5v I think. Anyone know what part number I need for this regulator on the negative rail?

 

 

 

 

Edited February 19, 2021 by 7bits

[7bits]

Looking at the rev2 service manual, it looks like Q100 is a large transistor with a heat sink, not a voltage regulator transistor or my electronic knowledge sucks. I guess where is the -5v regulator supposed to be?

 

Does anyone have a clean copy of the rev2, the ones I have are bad scans and near impossible to read when it comes to the curcuit diagrams.

Edited February 19, 2021 by 7bits

[StickJock]

OT:  Every time I see this thread title, this is what pops into my head:

 

[7bits]

Heh, well that is what I am doing, circles

 

I tracked down all the hot transistors. This is on the side board, its fed 12+ volts from pin 20 on the side board connector off the power-supply board. This is used to drive the stepper motor on connector J2 on the sideboard. All other J2 connector pins tie back from the stepper motor to the bank of transistors that get super hot to the touch. These in turn all connect to the RIOT chip. Probably a red herring...

 

If I remove the J2 plug to the stepper motor its cool as ice. Same problem of course, ordered some RAM and will try to dump the ROM again soon.

 

Transistors I am talking about 4 above the blue wire bundle:

Edited February 19, 2021 by 7bits

[7bits]

I tried placing the 810 ROM in the 800XL, all it would do is boot straight to the Memory Diagnostics and start the test. Would not even try to boot the 1050 drive.

 

I let the ROM socket empty and it would then just go to the self test menu. Not sure if this says anything about that ROM from the 810.

 

Ordering a basic cart off ebay. 

Edited February 21, 2021 by 7bits

[+Nezgar]

2 minutes ago, 7bits said:

I tried placing the 810 ROM in the 800XL, all it would do is boot straight to the Memory Diagnostics and start the test. Would not even try to boot the 1050 drive.

 

I let the ROM socket empty and it would then just go to the self test menu. Not sure if this says anything about that ROM from the 810.

 

Ordering a basic cart of ebay. 

Hmm.... it's normal to boot to self-test since the BASIC ROM is not present (But it should still boot a disk if a drive is attached), but not normal to go immediately to the memory test. Does the memory test actually indicate any red blocks? (should only show 40 blocks though, not 48 in this mode)

 

Also compare with option held down - which should disable the BASIC ROM (or in this case, the 810 ROM)

 

I'll try to give this myself and refresh my memory what's supposed to happen and I'm actually giving you good advice...

[7bits]

14 hours ago, Nezgar said:

Hmm.... it's normal to boot to self-test since the BASIC ROM is not present (But it should still boot a disk if a drive is attached), but not normal to go immediately to the memory test. Does the memory test actually indicate any red blocks? (should only show 40 blocks though, not 48 in this mode)

 

Also compare with option held down - which should disable the BASIC ROM (or in this case, the 810 ROM)

 

I'll try to give this myself and refresh my memory what's supposed to happen and I'm actually giving you good advice...

I did notice, while futzing around, it reported bad blocks of memory, I don't remember how far in, maybe the 5th and 6th block? Either way I did notice that when I put it all back together I ran a memory test to double check it was not wonkiness

[7bits]

So getting closer. You know its always the last thing you check...

 

I got 2 NOS ROM chips, was x fingers there were C rev ROMs. Only paid $15 shipped. Drum roll..... Popped the package, rev C.

 

Plugged it in and no difference. Granted both of these could be bad too. I think these came out of an old repair shop as it had the nice basement smell

 

Waiting on the SRAM ones next.

Edited February 24, 2021 by 7bits

[7bits]

Ok, I picked up a bunch of cheap carts that just arrived today, for a few dollars from an estate sale, don't know what they all are, just a bunch of games and music/writer cart. Will delve into them later/test etc. This lot was cheaper than buying the basic cart alone! My basic cart looks like it is lost in the mail.

 

Anyways will any one of these that are socketed work for dumping the 810 ROM? Anyone better than the other? Just asking before I crack them all open

Edited February 25, 2021 by 7bits

[+kheller2]

why are we trying to dump the 810 rom again?

Just to make sure you have a good working one?

 

[7bits]

6 minutes ago, kheller2 said:

why are we trying to dump the 810 rom again?

Just to make sure you have a good working one?

 

Yep. But I figure at this point, 3 bad ones(1 Original and 2x NOS ones) ? Maybe not. Or the NOS ones I got were bad ones stored back in a package. Pretty sure they were new as they both had Pin Spread so to speak. I just threw it in the Music cart, did EXACTLY the same thing as when I put the 810 ROM in the Basic socket in the 800 XL!

Edited February 25, 2021 by 7bits