morelenmir Posted June 25, 2015 Share Posted June 25, 2015 I read with interest this section from the material for the SIO2SD: In most cases SIO2SD can work with speed up to 127kb/s (hsindex = 0) without major hardware modifications. The only important thing is to add 4k7 ohm resistor (if you do not have one already) between SIO DATAOUT and +5V (pin 1 and 5 of SIO connector).It should be enough, but more advanced users should remove also at least two capacitors from Atari. Capacitors to be removed connect DATAOUT and DATAIN with ground. Those are small, 100nF capacitors (looks like resistors). In 800XL, 65XE and 130XE computers they have numbers C77 and C78. To show the difference please look below at oscillograms - with and without those capacitors. Has anyone tried this? Does it work? Can it be applied to all SIO devices or is it only relevant when you are strictly using the SIO2SD and would cause problems for a real 1050 or other drive? Quote Link to comment Share on other sites More sharing options...
Bryan Posted June 25, 2015 Share Posted June 25, 2015 It will certainly help things that can talk faster than 19.2Kbps. The 400 and 800 were designed before the relaxed FCC rules, so SIO was bandwidth limited to work with 19.2 peripherals. With a few modifications, it can go much faster without problems 2 Quote Link to comment Share on other sites More sharing options...
Rybags Posted June 25, 2015 Share Posted June 25, 2015 It's relevant to any SIO device capable of the higher speed turbo modes. Though at the higher speeds the SIO driver itself can come into question - IRQ driven SIO hits a wall then you need to go to polled mode. Quote Link to comment Share on other sites More sharing options...
lemiel Posted June 25, 2015 Share Posted June 25, 2015 And you can use Simius SIO2FIFO buffering device, but it was limited run only. Quote Link to comment Share on other sites More sharing options...
morelenmir Posted June 25, 2015 Author Share Posted June 25, 2015 I have a perfectly working 1050 that becomes somewhat unreliable when I attach the Happy mod. I wonder if this would cure the transmission problems? I may try it. Quote Link to comment Share on other sites More sharing options...
+Larry Posted June 26, 2015 Share Posted June 26, 2015 Have you run the Happy diagnostics? There is a high speed transafer test + docs. -Larry Quote Link to comment Share on other sites More sharing options...
Bryan Posted June 26, 2015 Share Posted June 26, 2015 I have a perfectly working 1050 that becomes somewhat unreliable when I attach the Happy mod. I wonder if this would cure the transmission problems? I may try it. Happy drives normally communicate at 19.2K (as all drives must) unless you're running something that supports warp mode. So, is it flaky when doing normal stuff? Quote Link to comment Share on other sites More sharing options...
morelenmir Posted June 26, 2015 Author Share Posted June 26, 2015 Have you run the Happy diagnostics? There is a high speed transafer test + docs. -Larry I have and it passes that one okay. I also run the ROM/RAM test and that works for a few cycles but it tends to lock up and require rebooting around the 16-20th run through. Happy drives normally communicate at 19.2K (as all drives must) unless you're running something that supports warp mode. So, is it flaky when doing normal stuff? The problem I have principally is when using Q-MEG. It was recommended to me the best way to create real floppies from *.ATR images was setup an SIO2PC as D1:, the 1050 as D2: and then use Q-MEG to first copy the D1: to a RAMdisk and then copy from the RAMdisk to D2:. It often fails during this process - giving a handshaking buzz two or three writes sessions in to the copy and then what is I think a time-out error. This is with the drives set to 'F' in the weird Q-MEG interface. Quote Link to comment Share on other sites More sharing options...
morelenmir Posted July 18, 2015 Author Share Posted July 18, 2015 To go back to the resistor and capacitor mod, is it really just two capacitors that need to go? I got the feeling it was four or five. Quote Link to comment Share on other sites More sharing options...
Rybags Posted July 19, 2015 Share Posted July 19, 2015 Data In, Data Out are the important ones. If GND had a component then it'd probably be a good idea to get rid of it too. The Clock line operates at the data rate but virtually no peripheral makes use of it, Pokey can operate such that the input clock sets the data rate but it's not in general use either. Command doesn't really mattter, generally a single command is issued per block based operation. Quote Link to comment Share on other sites More sharing options...
morelenmir Posted July 23, 2015 Author Share Posted July 23, 2015 I finally got around to doing this - and can detect an appreciable difference in the transfer rate even before adding the resistor. On inspecting the reverse of my SIO socket in order to do so I found a 'bodge' component already in place, but not I think a resister. See attached photo. Neither of my other 800XL's have these added, so they were possibly already removed by their previous owners or this one received special attention for some reason. Is it safe to get rid of ot do you think? As a matter of interest how critical is the resister rating? Does it have to be 4700 ohms or is that just what was convenient at the time? Annoyingly I don't have that precise value. Quote Link to comment Share on other sites More sharing options...
Kyle22 Posted July 23, 2015 Share Posted July 23, 2015 That is a capacitor on the audio input line. It won't hurt anything. Quote Link to comment Share on other sites More sharing options...
morelenmir Posted July 23, 2015 Author Share Posted July 23, 2015 That is a capacitor on the audio input line. It won't hurt anything. Ahhh - excellent! Its odd it was removed from the other machines I own. Do you happen to know if any of your machines have them fitted? Do you think a generic multi-kilo-ohm resister would do for the high-speed mod Kyle22? I've got what feels like a whole reel of every other size of resister except 4700!!! Quote Link to comment Share on other sites More sharing options...
Kyle22 Posted July 23, 2015 Share Posted July 23, 2015 I don't have any 800XLs anymore. Just 1200XLs and original 800s. a 5K would probably work. If you don't care what it looks like, you could use combinations of resistors in series and/or parallel as required to obtain the desired value. Quote Link to comment Share on other sites More sharing options...
morelenmir Posted July 23, 2015 Author Share Posted July 23, 2015 I don't have any 800XLs anymore. Just 1200XLs and original 800s. a 5K would probably work. If you don't care what it looks like, you could use combinations of resistors in series and/or parallel as required to obtain the desired value. True enough - good idea! It/they will attach to the back of the motherboard, hiding any sins. Quote Link to comment Share on other sites More sharing options...
+DrVenkman Posted July 23, 2015 Share Posted July 23, 2015 I have two 800XL's - I can't speak to the second, as it's currently packed away , but the one I've had open recently has nothing like that on the board anywhere. That is a capacitor on the audio input line. It won't hurt anything I think there were cassette reliability mods but simply adding a capacitor doesn't seem to be a part of them. Here's one example of an early one published by Antic: http://www.atarimagazines.com/v2n1/tapetopics.html Quote Link to comment Share on other sites More sharing options...
morelenmir Posted July 24, 2015 Author Share Posted July 24, 2015 Well... I removed the SIODATAIN and SIODATAOUT capacitors, removed that weird bodge capacitor, ordered a same-day delivery of five high wattage 4700 ohm resisters for £1.25 and attached one in between SIODATAOUT and +5V once they arrived. The machine was then re-buttoned and replaced at my desk, before I connected my SIO2PC up and started "RespeQT" on the windows machine. Turning it up to eleven, I set the transfer rate to 'custom' and 'pokey divisor' to '0'. Switching on the 800XL I see 'RespeQT' is showing the transfer speed of '125,000' which is slightly less than the 127,000 promised but still no mean feat! The final step in the setup procedure is to mount a DOS 2.5 master disk *.ATR as 'D1:' from within 'RespeQT'. Rather than launch a specialist sector copier I just test the works by copying all the contents from 'D1:' to a folder on my SDX hard drive.... And does it ever move! The data beeps sound like a GAU-8 going full bore... Just amazing!!! All appears to be transferred properly. So I think I can say the hardware SIO mod works, as seemingly does FJC's 'global hi-speed' patch from the "Ultimate1MB" BIOS. One area I have encountered problems with is the copying functionality of 'Q-Meg', but I will start a specific thread for this later. Quote Link to comment Share on other sites More sharing options...
Kyle22 Posted July 24, 2015 Share Posted July 24, 2015 High wattage resistors? Why? 1/4 or 1/8 watt should be plenty large enough. Quote Link to comment Share on other sites More sharing options...
morelenmir Posted July 24, 2015 Author Share Posted July 24, 2015 High wattage resistors? Why? 1/4 or 1/8 watt should be plenty large enough. They were cheap and available delivered same-day for free! If the smaller ones had been there I would bought them instead. Quote Link to comment Share on other sites More sharing options...
GoodByteXL Posted July 27, 2015 Share Posted July 27, 2015 And you can use Simius SIO2FIFO buffering device, but it was limited run only. At the abbuc forum there is thread collecting interest to get a new run done. 50 pieces are the goal. http://www.abbuc.de/community/forum/viewtopic.php?f=15&t=8566 Quote Link to comment Share on other sites More sharing options...
morelenmir Posted July 27, 2015 Author Share Posted July 27, 2015 At the abbuc forum there is thread collecting interest to get a new run done. 50 pieces are the goal. http://www.abbuc.de/community/forum/viewtopic.php?f=15&t=8566 Very interesting! I think Simius said they were around €10 a piece? Quote Link to comment Share on other sites More sharing options...
+David_P Posted July 27, 2015 Share Posted July 27, 2015 The discussion thread suggests 20 euros each, shipped. Quote Link to comment Share on other sites More sharing options...
morelenmir Posted July 27, 2015 Author Share Posted July 27, 2015 The discussion thread suggests 20 euros each, shipped. Right you are! So - around £13-15 depending on the current exchange rate. An excellent price for a very useful little device. Quote Link to comment Share on other sites More sharing options...
+mytek Posted July 27, 2015 Share Posted July 27, 2015 On inspecting the reverse of my SIO socket in order to do so I found a 'bodge' component already in place, but not I think a resister. See attached photo. That looks more like a diode to me, but I could be mistaken (notice what looks like a diode symbol). If it is a diode, wonder why? -Michael Quote Link to comment Share on other sites More sharing options...
morelenmir Posted July 27, 2015 Author Share Posted July 27, 2015 That looks more like a diode to me, but I could be mistaken (notice what looks like a diode symbol). If it is a diode, wonder why? -Michael You are absolutely right! My eye-sight is so poor I never noticed that even on the blown-up photo! I wonder what it was stuck between the 'AUDIOIN' and 'GND' for? This particular machine is one that, long ago used to belong to the leader of an Atari cracking group down in Portsmouth. I can only assume he must have thought it would help in some arcane way - perhaps to improve 'CLOAD' reliability? Whatever the case it is gone now and SIO runs like a very fast machine-gun between the "SIO2SD" and the machine! Quote Link to comment Share on other sites More sharing options...
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.