+tf_hh Posted March 7, 2018 Share Posted March 7, 2018 I've narrowed the issue to pin 5 of the SIO port (data OUT). At idle on a good working 1200XL, I get get 5 VDC on pins 3,5 and 7 on the SIO port. On this particular board, I get 5 VDC on pins 3 and 7, but an "open" on pin 5. Now to track down that run using the information Jurgen provided for data OUT. DATA_OUT: R128 (100 ohms) and capacitor C91. Also R42 (4.7k pull-up to +5V). O.K., now I'm officially confused. Both R128 and C91 have 5 VDC on both sides (great). Now R42 was not measuring 4.7K ohms (installed) and I got 5 VDC on the side next to the 6502C, but did not get 5 VDC on the other side of the resistor. I verified on a good 1200XL that R42 measures 4.7K ohms (installed). I replaced R42 with a new 4.7K ohm resistor, but when I put the multimeter to it, it shows 276 ohms (installed)??? I'm putting the two leads on both sides of the resistor (installed). I verified with the ohm meter before I installed the new one that it is indeed a 4.7K ohm resistor. I'm an Aerospace guy, not a EE, so my knowledge is limited here. Wouldn't there have to be another resistance in parallel to make the 4.7K ohm resistor read 276 ohms? Also, after I removed the old resistor, I measured it uninstalled and it shows 4.7K ohms. So the original resistor was good??? So, the difference on the bad board is that something is causing R42 to show 276 ohms instead of 4.7K ohms that a good board shows. You shouldn´t measure the value of a in-circuit resistor. At a minimum, lift one side, so that one side is floating not connected to anything. Now you can measure the resistor and should get something around 4.7K. It´s very, very rare that a small signal resistors went bad without having high current and a short-circuit. In that case you should see something burned :-) And yes, because other connections to that signal (at least POKEY itself) are made, the result of a measure having the resistor installed is not accurate. What do you mean with "open" at DATA_OUT? You get ~zero volts or no connection (if you use the diode tester of your multimeter)? Remember, that there´s a 100R resistor in line - most diode testers will respond up to 20R max. So check connection step-by-step from SIO to one end of R128 (100R) and the other to POKEY. The pull-up resistor of 4.7K is "behind" the 100R resistor (facing away from SIO connector), so I would examine the 100R first. Jurgen Quote Link to comment Share on other sites More sharing options...
ACML Posted March 7, 2018 Author Share Posted March 7, 2018 (edited) You shouldn´t measure the value of a in-circuit resistor. At a minimum, lift one side, so that one side is floating not connected to anything. Now you can measure the resistor and should get something around 4.7K. It´s very, very rare that a small signal resistors went bad without having high current and a short-circuit. In that case you should see something burned :-) And yes, because other connections to that signal (at least POKEY itself) are made, the result of a measure having the resistor installed is not accurate. What do you mean with "open" at DATA_OUT? You get ~zero volts or no connection (if you use the diode tester of your multimeter)? Remember, that there´s a 100R resistor in line - most diode testers will respond up to 20R max. So check connection step-by-step from SIO to one end of R128 (100R) and the other to POKEY. The pull-up resistor of 4.7K is "behind" the 100R resistor (facing away from SIO connector), so I would examine the 100R first. Jurgen I measured 5 VDC on both sides of R128 and C91, but pin 5 (data OUT) measured zero voltage. I measured pin 5 on another good 1200XL and it measured 5 VDC idle. What is the path in order from the POKEY? Is it POKEY --> R42 --> C91 --> R128 --> SIO? I measured: 5 VDC -- R42 -- Zero (R42 measured 276 ohms??) 5 VDC -- C91 -- 5 VDC 5 VDC -- R128 -- 5 VDC Pin 5 of SIO -- zero On a known working 1200XL I measured: 5 VDC -- R42 -- 5 VDC (R42 measured 4.7K ohms) 5 VDC -- C91 -- 5 VDC 5 VDC -- R128 -- 5 VDC Pin 5 of SIO -- 5 VDC Edited March 7, 2018 by ACML Quote Link to comment Share on other sites More sharing options...
+tf_hh Posted March 7, 2018 Share Posted March 7, 2018 I measured 5 VDC on both sides of R128 and C91, but pin 5 (data OUT) measured zero voltage. I measured pin 5 on another good 1200XL and it measured 5 VDC idle. What is the path in order from the POKEY? Is it POKEY --> R42 --> C91 --> R128 --> SIO? I measured: 5 VDC -- R42 -- Zero (R42 measured 276 ohms??) 5 VDC -- C91 -- 5 VDC 5 VDC -- R128 -- 5 VDC Pin 5 of SIO -- zero On a known working 1200XL I measured: 5 VDC -- R42 -- 5 VDC (R42 measured 4.7K ohms) 5 VDC -- C91 -- 5 VDC 5 VDC -- R128 -- 5 VDC Pin 5 of SIO -- 5 VDC The R42 is measured one side up (not connected to anything) with 276 ohms, really? Then it´s defect and should be replaced. Here the path: 1 Quote Link to comment Share on other sites More sharing options...
ACML Posted March 8, 2018 Author Share Posted March 8, 2018 (edited) The R42 is measured one side up (not connected to anything) with 276 ohms, really? Then it´s defect and should be replaced. Here the path: The 276 ohms was measured installed. That is on the bad board. When I removed R42 from the bad board, it does read 4.7K ohms. I made the same installed measurement on a known good board and it read 4.7K ohms. I now will look at pin 28 of the POKEY (SOD). The good board shows 5 VDC on both sides of R42. On the bad board, only the POKEY pin 28 side of R42 reads 5 VDC. I'm not getting 5 VDC on the R125/R41 side (node) of R42. On that side it reads radom small milivolt values. Edited March 8, 2018 by ACML Quote Link to comment Share on other sites More sharing options...
_The Doctor__ Posted March 8, 2018 Share Posted March 8, 2018 (edited) lift one end of r28 to rule out the cap 91 being shorted or bad on the sio side. something doesn't add up... when measuring voltage... black lead should be to common ground rail... you then move only the red lead to each side of components to get voltages... when doing diode tests you do it across a component one way then the other.. resistors don't require testing both directions be certain to disconnect power and discharge caps before you test them... all of my meters including cheap one I have from harbour freight years back have diode, transistor, and cap testing on them. Not very good or accurate but when an item is truly dead it's good enough to catch it. Edited March 8, 2018 by _The Doctor__ 2 Quote Link to comment Share on other sites More sharing options...
+tf_hh Posted March 8, 2018 Share Posted March 8, 2018 The 276 ohms was measured installed. That is on the bad board. When I removed R42 from the bad board, it does read 4.7K ohms. I made the same installed measurement on a known good board and it read 4.7K ohms. I now will look at pin 28 of the POKEY (SOD). The good board shows 5 VDC on both sides of R42. On the bad board, only the POKEY pin 28 side of R42 reads 5 VDC. I'm not getting 5 VDC on the R125/R41 side (node) of R42. On that side it reads radom small milivolt values. Ok, that´s the point I already wrote some posts earlier. Lift one side of the R42, when it´s 4.7K, then this resistor is fine. Also check connection of +5 volt to one side of R42. After that, remove one side of C91 and check voltages again. At last, check the duct of pin 28 from POKEY to the junction of R42/R128 - not that the problem is just a simple defect at the POKEY´s socket... 1 Quote Link to comment Share on other sites More sharing options...
ACML Posted March 12, 2018 Author Share Posted March 12, 2018 (edited) Sorry it took so long to get back. Crazy busy week at work. I hooked it back up today to do some troubleshooting. At first I was getting the zero volts at pin 5 of the SIO and also measured zero volts from pin 28 of the POKEY. After measuring R42, R128 and C91, it all changed. The POKEY was now showing 5 VDC on pin 28 and so was pin 5 on the SIO. The SIO started working. I can boot to the SIO2SD. It now always loads the SIO2SD version 2.2 menu, but on 15% of the attempts, the screen appears to go into screen saver mode and the menu does not finish loading. So now the hard failure is intermittent, but the SIO issue seems resolved. I did nothing to remedy this. Makes me think there is more to the story. The intermittent behavior implies to me that something else is not right. I tried flexing the board gently (power off) thinking there is a cracked trace, but can't get it to hard fail anymore. If I let the machine run, it does not freeze of lock up. I really don't think its the POKEY as I had already replaced the POKEY and got the same hard failure earlier. P.S. Something unrelated I learned. Here's a tip: Don't hold the business end of the 9VAC tip in your hand when you are also trying to remove the SIO plug. I accidentally touched the power adapter tip to the 1200XL heat sink and ZAP!, killed the the 9VAC adapter (most likely burned the 5 amp fuse). And it was a new in box CO17945. That will teach me. Edited March 12, 2018 by ACML Quote Link to comment Share on other sites More sharing options...
ACML Posted March 12, 2018 Author Share Posted March 12, 2018 UPDATE: I let the machine run for 20 minutes and tried it again. The hard failure is back. This one is pegging the fun meter, I may just harvest the parts and toss this board. I won't sell a machine to someone with an intermittent problem like this. Quote Link to comment Share on other sites More sharing options...
ACML Posted March 14, 2018 Author Share Posted March 14, 2018 (edited) Well I almost gave up hope, but decided to try one last thing. I blew compressed air under all the sockets and Eureka!, it started to work again. Sometimes it's just something that simple. My guess is that when using a solder sucker to clean out through holes for the OS modification, a super thin solder flake from the solder sucker must have lodged itself under one of the sockets (likely the POKEY). By picking up the board here and there, the solder flake would make intermittent contact and short two traces (my theory). I've been running it on and off for several hours now and its is working good. I'll wait a few more days before I declare victory. Thanks to Jurgen, The Doctor and DrVenkman for the assistance. Edited March 14, 2018 by ACML 4 Quote Link to comment Share on other sites More sharing options...
+DrVenkman Posted March 14, 2018 Share Posted March 14, 2018 Well I almost gave up hope, but decided to try one last thing. I blew compressed air under all the sockets and Eureka!, it started to work again. Sometimes it's just something that simple. My guess is that when using a solder sucker to clean out through holes for the OS modification, a super thin solder flake from the solder sucker must have lodged itself under one of the sockets (likely the POKEY). By picking up the board here and there, the solder flake would make intermittent contact and short two traces (my theory). I've been running it on and off for several hours now and its is working good. I'll wait a few more days before I declare victory. Thanks to Jurgen, The Doctor and DrVenkman for the assistance. Glad it was something simple! I ran into a similar issue with my much worked-on “Ugly Duckling” 1200XL a couple months ago. It suddenly started locking up at random. When I disassembled it, I decided to flush under all the sockets with iso. I found a single large-ish solder flake under the CPU socket I had replaced. Problem solved. Quote Link to comment Share on other sites More sharing options...
ACML Posted March 16, 2018 Author Share Posted March 16, 2018 (edited) Its been a few days and I tested it off and on several times. I think it's actually cured. That is a relief. In my 30+ years working with Atari 8-bits (took a break between 1997 and 2010), I found the 1200XL to be the most reliable. I have owned many in my time, keeping a few and selling most, but everyone has always been serviceable to include the keyboards. It's great that Best is selling replacement mylars, but I've yet to encounter a 1200XL that actually needed one. That being, everyone I've had was easily remedied using BARE conductive paint and some patients. I have never had to swap any of the VLSI chips or a capacitor for that matter. I've had issues with 400's and 800's, but most were related to having seven boards not always making good contact with each other. The 800XL was easy to work on, but I've only had a half dozen of those. Not enough to draw any conclusions. After 30+ years, my opinion remains, that there is no finer 8-bit than a fixed 1200XL. Edited March 16, 2018 by ACML 3 Quote Link to comment Share on other sites More sharing options...
Calibus Posted March 17, 2018 Share Posted March 17, 2018 Regarding all the points you´ve already checked, IMHO the only reason could be a broken trace OR defect discrete part in line of COMMAND, DATA_IN or DATA_OUT. Do you have a scope? Then it should be really easy to find out the issue. COMMAND line should toggle sometimes to low during "detection SIO sound" after powering on. Also the DATA_OUT line should have some transistions, the O.S. sent "Status" to Drive D1: to test if there´s any. When the display of your SIO2SD doesn´t change during the "buzzing SIO sound", then COMMAND or DATA_OUT is mailfunctioned. Here are the parts to check: COMMAND-Line: R79 (100 ohms) and capacitor C82. DATA_IN: R127 (100 ohms) and capacitor C90. Also R41 (4.7k pull-up to +5V). DATA_OUT: R128 (100 ohms) and capacitor C91. Also R42 (4.7k pull-up to +5V). The caps mentioned are connected to ground. So if the signal is constantly at low level, the corresponding cap should be exchanged (1nF) or removed for test - works also without. If you don´t have a scope, check voltages in idle state. All three lines should have something over +4.5 volts. If not, check all parts listed above. And also examine the traces. Good luck, Jurgen Apologies in advance if I should start my own topic - and I will if asked. I also have a 1200XL that I have been refurbishing. I have already performed the keyboard mylar fix after watching FJC's informative videos, and then installed a UAV and U1MB. The upgrades all seem to be working properly - except for the fact that I can't upgrade the U1MB from the ancient original firmware due to an SIO issue... History of the SIO issue: I have created a simple 3 wire SIO2PC cable and tried accessing an ATR via RespeQt - but was unable to access the disk. A directory command from SDX always results in a 138 error. After triple checking the wiring, I suspected the SIO port might be the factor, so I found an AtariMax SIO2PC USB adapter I tucked into a drawer once I found RespeQt. I figured this commercial adapter would be a good acid test of my SIO port - and either prove it was working or not. Not wanting to spend money on APE - I installed the shareware version - plugged the adapter into the 1200XL - tried to get a directory listing - and same 138 error. Interestingly, the TX and RX LEDs on the adapter never blinked. Should I start my diagnosis with the steps I quoted - or should I start someplace earlier in the chain? I read a comment about replacing R63 - is its presence really a symptom of what I am seeing? I do have a 'scope and a fairly good idea how to use it - I just need to know what to probe. I am using a brand new SIO cable to connect the AtariMax SIO2PC-USB to the 1200XL - so I'm sure it is not the cable. As far as chip swapping is concerned - I do have what I believe is a functional 800 at my disposal, I'd try its SIO port - but I'm not sure how to use it with APE or RespeQt since I don't have a Sparta cart for it. (I'm super new to the World of Atari and I'm trying to break in via the 800 and 1200XL) Again - if I should make my own topic for this issue - let me know and I'll do so. Thanks Jim Quote Link to comment Share on other sites More sharing options...
ACML Posted March 17, 2018 Author Share Posted March 17, 2018 (edited) Apologies in advance if I should start my own topic - and I will if asked. I also have a 1200XL that I have been refurbishing. I have already performed the keyboard mylar fix after watching FJC's informative videos, and then installed a UAV and U1MB. The upgrades all seem to be working properly - except for the fact that I can't upgrade the U1MB from the ancient original firmware due to an SIO issue... History of the SIO issue: I have created a simple 3 wire SIO2PC cable and tried accessing an ATR via RespeQt - but was unable to access the disk. A directory command from SDX always results in a 138 error. After triple checking the wiring, I suspected the SIO port might be the factor, so I found an AtariMax SIO2PC USB adapter I tucked into a drawer once I found RespeQt. I figured this commercial adapter would be a good acid test of my SIO port - and either prove it was working or not. Not wanting to spend money on APE - I installed the shareware version - plugged the adapter into the 1200XL - tried to get a directory listing - and same 138 error. Interestingly, the TX and RX LEDs on the adapter never blinked. Should I start my diagnosis with the steps I quoted - or should I start someplace earlier in the chain? I read a comment about replacing R63 - is its presence really a symptom of what I am seeing? I do have a 'scope and a fairly good idea how to use it - I just need to know what to probe. I am using a brand new SIO cable to connect the AtariMax SIO2PC-USB to the 1200XL - so I'm sure it is not the cable. As far as chip swapping is concerned - I do have what I believe is a functional 800 at my disposal, I'd try its SIO port - but I'm not sure how to use it with APE or RespeQt since I don't have a Sparta cart for it. (I'm super new to the World of Atari and I'm trying to break in via the 800 and 1200XL) Again - if I should make my own topic for this issue - let me know and I'll do so. Thanks Jim Jim, If R63 is still stock (unmodified), that could easily be your issue. I think the stock value of R63 is 1K ohm and it acts as a current limiting resistor. You just need to replace it with a jumper wire or 0 ohm resistor. That will restore +5VDC to pin 10 of the SIO. You can quickly check to see if you have 5VDC on pin 10 of the SIO. If not, you likely still have the stock resistor blocking 5VDC to the SIO. That 5VDC is required to run the SIO2SD, P: R: Connection, etc. Edited March 17, 2018 by ACML Quote Link to comment Share on other sites More sharing options...
+DrVenkman Posted March 17, 2018 Share Posted March 17, 2018 Jim, If R63 is still stock (unmodified), that could easily be your issue. I think the stock value of R63 is 1K ohm and it acts as a current limiting resistor. You just need to replace it with a jumper wire or 0 ohm resistor. That will restore +5VDC to pin 10 of the SIO. You can quickly check to see if you have 5VDC on pin 10 of the SIO. If not, you likely still have the stock resistor blocking 5VDC to the SIO. That 5VDC is required to run the SIO2SD, P: R: Connection, etc. I don't believe that matters for an SIO2USB type device, since the USB port from the host computer (or Raspberry Pi, in my case) powers the interface device. I'd plug the Atarimax device into the 800, boot any old DOS image you can find (DOS 2.0, DOS 2.5, SpartaDOS 3.2, MyDOS ...) and see if the device itself is working. That said, four years ago when I decided to get back into 8-bit computers and pulled my stuff out of storage, one of my 1200XL's was fully functional except it wouldn't boot anything, as if the SIO port wasn't connected. Turned out to be a bad POKEY. Quote Link to comment Share on other sites More sharing options...
Calibus Posted March 17, 2018 Share Posted March 17, 2018 I don't believe that matters for an SIO2USB type device, since the USB port from the host computer (or Raspberry Pi, in my case) powers the interface device. I'd plug the Atarimax device into the 800, boot any old DOS image you can find (DOS 2.0, DOS 2.5, SpartaDOS 3.2, MyDOS ...) and see if the device itself is working. That said, four years ago when I decided to get back into 8-bit computers and pulled my stuff out of storage, one of my 1200XL's was fully functional except it wouldn't boot anything, as if the SIO port wasn't connected. Turned out to be a bad POKEY. OK, I found a DOS 2.5 ATR image on atarimania, mounted that via APE and successfully booted off of it with my 800. That tells me the SIO2USB device is working - as well as whatever ASIC in the 800 that drives the SIO port is OK also. Your symptom of a bad SIO port sounds like what I am experiencing. When the 800 powers up - I see activity on the SIO2USB device - TX/RX - and I do NOT see that with the 1200XL - so it looks like the SIO port *IS* dead. Since replacing the POKEY fixed things for you - I'm going to try swapping POKEYs between my two computers and see if that is it. Quote Link to comment Share on other sites More sharing options...
Calibus Posted March 18, 2018 Share Posted March 18, 2018 OK, I disassembled my 800, removed the POKEY and transplanted it into the 1200XL. Same symptoms - so that tells me the POKEY is fine. Note, since the 800 is a bit of a pain to re-assemble, I did not try the 1200XL POKEY in it - but I expect that it would have worked. I'm now going to see if I can find a readable schematic for the 1200 and see what all drives the SIO port. I assume this is where the previously mentioned resistors and capacitors come in: COMMAND-Line: R79 (100 ohms) and capacitor C82. DATA_IN: R127 (100 ohms) and capacitor C90. Also R41 (4.7k pull-up to +5V). DATA_OUT: R128 (100 ohms) and capacitor C91. Also R42 (4.7k pull-up to +5V). Quote Link to comment Share on other sites More sharing options...
ACML Posted March 18, 2018 Author Share Posted March 18, 2018 (edited) OK, I disassembled my 800, removed the POKEY and transplanted it into the 1200XL. Same symptoms - so that tells me the POKEY is fine. Note, since the 800 is a bit of a pain to re-assemble, I did not try the 1200XL POKEY in it - but I expect that it would have worked. I'm now going to see if I can find a readable schematic for the 1200 and see what all drives the SIO port. I assume this is where the previously mentioned resistors and capacitors come in: Something that helped me was that the path from pin 28 (POKEY - Data out) and the others are 5VDC at idle. So when it's not trying to load, you should see 5VDC on most of those paths. Pins 3, 5 and 7 on the SIO are the ones of interest. They should all read 5VDC at idle. Edited March 18, 2018 by ACML Quote Link to comment Share on other sites More sharing options...
Calibus Posted March 18, 2018 Share Posted March 18, 2018 I took a scope probe to SIO TX this afternoon. I figured that if TX wasn't working, then I needn't bother looking at RX. (I'm sure APE etc isn't seeing anything from the 1200XL - so there shouldn't be anything on RX at this time...) I couldn't figure out how to properly store a trace and extract it from my 'scope - so I held my phone in front of it while I powered on the 1200XL. The trace immediately goes to 5v, and then has a brief surge of traffic (traffic is between 5v and 4v - I assume that is normal?) Pauses, and then another brief surge of traffic. Does this look normal? This is the exact same signal I see coming out of the POKEY itself. https://www.youtube.com/watch?v=xMCdmj4PwVg Quote Link to comment Share on other sites More sharing options...
1050 Posted March 19, 2018 Share Posted March 19, 2018 It says this video is unavailable, maybe you uploaded it as private? Have you looked at resistor R63 yet? First sentence of this thread and you need to be saying the same thing. R63 1200XL delete SIO fix http://atariage.com/forums/topic/268934-help-identifying-1200xl-mods/?do=findComment&comment=3827857 Quote Link to comment Share on other sites More sharing options...
Calibus Posted March 19, 2018 Share Posted March 19, 2018 It was set as private - which I misinterpreted as being viewable if the link was provided. It's now just set to "unlisted" - so hopefully that link works now. No, I haven't touched R63 as per the suggestion by DrVenkman that it shouldn't be a factor. However, I'm willing to do so. Quote Link to comment Share on other sites More sharing options...
+DrVenkman Posted March 19, 2018 Share Posted March 19, 2018 It was set as private - which I misinterpreted as being viewable if the link was provided. It's now just set to "unlisted" - so hopefully that link works now. No, I haven't touched R63 as per the suggestion by DrVenkman that it shouldn't be a factor. However, I'm willing to do so. R63 limits +5V DC. If you’re using a standalone SIO2SD, you need to remove it. If you’re using an SIO2PC-USB device or a real drive, you don’t. The drive has its own power supply, and the USB device gets its power from the USB port on the host computer. Quote Link to comment Share on other sites More sharing options...
Calibus Posted March 19, 2018 Share Posted March 19, 2018 R63 limits +5V DC. If you’re using a standalone SIO2SD, you need to remove it. If you’re using an SIO2PC-USB device or a real drive, you don’t. The drive has its own power supply, and the USB device gets its power from the USB port on the host computer. Since I'm currently using an AtariMax SIO2PC-USB device - R63 doesn't come into play. Bummer. I guess I'm stuck then until I can get a trace of a good working SIO transmission, since I can't tell if there is anything wrong with mine - nothing to compare it to. 1 Quote Link to comment Share on other sites More sharing options...
+kheller2 Posted March 20, 2018 Share Posted March 20, 2018 1200XL SCHEM.PDF This is the horrible officially published1200XL schematic. Quote Link to comment Share on other sites More sharing options...
+DrVenkman Posted March 20, 2018 Share Posted March 20, 2018 Bear in mind too, the third-party 1200XL schematics drawn by Jerzy Sobola are notoriously inaccurate. I wouldn't trust them at all if you run across them. Better to just follow traces by eye, with a multimeter, logic probe or scope, and make notes of your observations and findings. Quote Link to comment Share on other sites More sharing options...
1050 Posted March 20, 2018 Share Posted March 20, 2018 Since I'm currently using an AtariMax SIO2PC-USB device - R63 doesn't come into play. Bummer.Huh? +5 READY line is not used for power if that is what is behind the thinking there, it's only purpose is to tell devices on the SIO chain that the computer has power and is READY. My real drive with R63 intact would not work, so I jumped it and then it did work. Only thing I did too. My advice would be to just jumper R63, putting in the popular SIO fix on your 1200XL, and then give it a try. Could measure it too, before and after. Your video link now working as per normal. Sorry don't have a known good trace to share with you. Quote Link to comment Share on other sites More sharing options...
Like us on Facebook
Follow us on Twitter
AtariAge on Instagram
Watch us on YouTube
Chat on Discord
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.