Atari STe 1040 trashing UltraSatan SD card

The place to ask all those likely daft questions!
User avatar
exxos
Site Admin
Site Admin
Posts: 23499
Joined: Wed Aug 16, 2017 11:19 pm
Location: UK
Contact:

Re: Atari STe 1040 trashing UltraSatan SD card

Post by exxos »

zChris wrote: Thu May 19, 2022 9:18 pm
exxos wrote: Thu May 19, 2022 2:03 pm Try this program https://www.exxosforum.co.uk/forum/viewt ... ?f=25&t=43
Tried this program and only got * which was good i think. Run it for like 20 minutes.

UltraSatan and card is OK since i can partition it with my H4 board and it reads it correctly on Atari STe. Its when its writing that its a big chance of corruption AND it doesnt work to partition the SD card with ATari STe. Using HDDRIVER 11
DMA and bus resistor fixes are next then.
https://www.exxosforum.co.uk/atari/ All my hardware guides - mods - games - STOS
https://www.exxosforum.co.uk/atari/store2/ - All my hardware mods for sale - Please help support by making a purchase.
viewtopic.php?f=17&t=1585 Have you done the Mandatory Fixes ?
Just because a lot of people agree on something, doesn't make it a fact. ~exxos ~
People should find solutions to problems, not find problems with solutions.
User avatar
rubber_jonnie
Site Admin
Site Admin
Posts: 10472
Joined: Thu Aug 17, 2017 7:40 pm
Location: Essex
Contact:

Re: Atari STe 1040 trashing UltraSatan SD card

Post by rubber_jonnie »

zChris wrote: Thu May 19, 2022 9:19 pm
rubber_jonnie wrote: Thu May 19, 2022 4:53 pm Out of curiosity, what driver are you using?
Using HDDRIVER 11
Good driver, I use it too.
Collector of many retro things!
800XL and 65XE both with Ultimate1MB,VBXL/XE & PokeyMax, SIDE3, SDrive Max, 2x 1010 cassette, 2x 1050 one with Happy mod, 3x 2600 Jr, 7800 and Lynx II
Approx 20 STs, including a 520 STM, 520 STFMs, 3x Mega ST, MSTE & 2x 32 Mhz boosted STEs
Plus the rest, totalling around 50 machines including a QL, 3x BBC Model B, Electron, Spectrums, ZX81 etc...
ijor
Posts: 428
Joined: Fri Nov 30, 2018 8:45 pm

Re: Atari STe 1040 trashing UltraSatan SD card

Post by ijor »

stephen_usher wrote: Thu May 19, 2022 5:00 pm Putting a 2.2K resistor to +5V doesn't shift the voltage at all, not even a centivolt, so whatever is pulling the lines to that value is very strong. Goodness knows how much current the DMA has to sink to pull them down for data transfer.
You are probably measuring at the US end. As Exxos is saying, there is a protection diode that prevents voltage to raise above the threshold. Try measuring voltage at the computer end and it should be higher because there are voltage dividers. Anyway, the diode prevents higher voltages only, it doesn't affect the current when pulling down the signal.
exxos wrote: Thu May 19, 2022 5:03 pm @stephen_usher Maybe they are just zener clamped to 3.3v ? Even so that wouldn't be good for the STE which has driver chips :roll:
There are protection diodes connected to GND according to the schematics. But they are across voltage divider resistors.
http://github.com/ijor/fx68k 68000 cycle exact FPGA core
FX CAST Cycle Accurate Atari ST core
http://pasti.fxatari.com
User avatar
stephen_usher
Posts: 5580
Joined: Mon Nov 13, 2017 7:19 pm
Location: Oxford, UK.
Contact:

Re: Atari STe 1040 trashing UltraSatan SD card

Post by stephen_usher »

ijor wrote: Fri May 20, 2022 12:52 pm
stephen_usher wrote: Thu May 19, 2022 5:00 pm Putting a 2.2K resistor to +5V doesn't shift the voltage at all, not even a centivolt, so whatever is pulling the lines to that value is very strong. Goodness knows how much current the DMA has to sink to pull them down for data transfer.
You are probably measuring at the US end. As Exxos is saying, there is a protection diode that prevents voltage to raise above the threshold. Try measuring voltage at the computer end and it should be higher because there are voltage dividers. Anyway, the diode prevents higher voltages only, it doesn't affect the current when pulling down the signal.
Nope this was inside the ST when I was trying to diagnose DMA issues.
Intro retro computers since before they were retro...
ZX81->Spectrum->Memotech MTX->Sinclair QL->520STM->BBC Micro->TT030->PCs & Sun Workstations.
Added code to the MiNT kernel (still there the last time I checked) + put together MiNTOS.
Collection now with added Macs, Amigas, Suns and Acorns.
User avatar
exxos
Site Admin
Site Admin
Posts: 23499
Joined: Wed Aug 16, 2017 11:19 pm
Location: UK
Contact:

Re: Atari STe 1040 trashing UltraSatan SD card

Post by exxos »

The gigafile did odd things. Sometimes it was 3.3V and other times 5V. I assume when it was driving the bus it was driving 3V by itself and not relying upon the pullups. I know Wolfgang had a lot of problems with the thing overall. Possibly he was driving the bus because the pullups were just inadequate to drive a logic high. And by adding the pullups solved a great many problems. So it would seem logical in driving 3V from gigafile in that respect. But things are slightly more complicated on the STE.

d2.png
d2.png (7.14 KiB) Viewed 1161 times

All here, but note that is all very old research and invalid these days. Likely that page will be taken down in the future anyway.

https://www.exxosforum.co.uk/atari/last/ ... /index.htm


In any case, I do have the original black box ultrasatan is the one I generally use. I do have the updated one in the grey metal box but I have not used that one much because the case seems to be "live" :roll:

With so many clones of hard drives out there these days it is becoming increasingly difficult to work out what is wrong with the things. But realistically, the people developing the hardware should figure out the problems with it all.

Capture.PNG
Capture.PNG (68.28 KiB) Viewed 1161 times

Indeed its clamped to 3.3V. But via a 220R. With pullups (DMA FIX) it would be like 10K clamped to 3.3V. But with driver chips like on the STE, 5v driven via 220R is like 22mA.. 114mW. Might be lower as I think the STE had 33R in series as well.. To lazy to check.

Not sure I follow why the 100R are in there. If its hard clamped to 3.3V, then you only need to drive low anyway. Maybe just noise immunity. No point in a voltage divider if your hard clamping to 3.3V anyway. I am assuming the chip is running on 3.3V. Even so, if the chip is driving high, then the 100R may make more sense as the chip voltage might be higher, say 3.5V depending on regulator tolerances. But anyway, getting off topic a bit.

So as @stephen_usher mentions, 10K DMA pullups will have no effect on the STE in terms of voltages on the ASCI bus. But the DMA pullups are for the DMA IC side of things, like the floppy drive also. 10K on the STFM still won't effect the ASCI port as its still clamped to 3.3V. But the fix should still be done for the floppy drive anyway. If ultrasatan is driving a logic high, then it would probably work more reliable on a STFM as it will be basically circumventing the 10K DMA pullups.

As I was testing my ultrasatan originally on my STFM, I will make the assumption that UltraSatan is not driving a logic high but relying on the pullups. If it was driving high then I probably would not have came conclusion that the 10K DMA pullups were needed to fix the problems with it. Of course I am reeling this off the top of my head I have not looked into all this for a long time.

I don't know what Lothrek did changes wise over the original ultrasatan design ? I think I might have taken mine apart and took a image of the PCB at some point, cannot remember. But I think we have all had this discussion before somewhere.

TL;DR:

People should still do the 10K DMA pullup fix. Its a known point of failure, so just don't risk it.

Disclaimer :

As to if it will specifically solve your hard drive problems or not I really cannot say. There are that many clones and hard drives out there now, some of them have been manufactured bad out of the box. Some people may have bad connections, bad power supplies, bad SD cards, the list is inherently endless. The fixes I came up with solve a lot of known problems. They cannot of course fix " unknown to me" problems. Like I said some years ago, the faults on these machines are developing faster than I had time to investigate them. Hence why I gave up and just developed a new motherboard.
https://www.exxosforum.co.uk/atari/ All my hardware guides - mods - games - STOS
https://www.exxosforum.co.uk/atari/store2/ - All my hardware mods for sale - Please help support by making a purchase.
viewtopic.php?f=17&t=1585 Have you done the Mandatory Fixes ?
Just because a lot of people agree on something, doesn't make it a fact. ~exxos ~
People should find solutions to problems, not find problems with solutions.
ijor
Posts: 428
Joined: Fri Nov 30, 2018 8:45 pm

Re: Atari STe 1040 trashing UltraSatan SD card

Post by ijor »

exxos wrote: Fri May 20, 2022 1:47 pm Indeed its clamped to 3.3V. But via a 220R. With pullups (DMA FIX) it would be like 10K clamped to 3.3V. But with driver chips like on the STE, 5v driven via 220R is like 22mA.. 114mW. Might be lower as I think the STE had 33R in series as well.. To lazy to check.

Not sure I follow why the 100R are in there. If its hard clamped to 3.3V, then you only need to drive low anyway. Maybe just noise immunity. No point in a voltage divider if your hard clamping to 3.3V anyway. I am assuming the chip is running on 3.3V. Even so, if the chip is driving high, then the 100R may make more sense as the chip voltage might be higher, say 3.5V depending on regulator tolerances. But anyway, getting off topic a bit.
I though the zener diode was more for voltage protection (may be also for hot plug-in) than for level shifting. But I assumed this was clamping to 3.3v after the voltage dividers at the device side only. If you are saying this will clamp the ST side as well? Then that sounds like a bad design to me.
As I was testing my ultrasatan originally on my STFM, I will make the assumption that UltraSatan is not driving a logic high but relying on the pullups. If it was driving high then I probably would not have came conclusion that the 10K DMA pullups were needed to fix the problems with it. Of course I am reeling this off the top of my head I have not looked into all this for a long time.
The US is driving high. Surely the output is not open drain, if that's what you mean. If the US output would be open drain, then it wouldn't work at all. Not even with pull-ups because a 10K pull-up is way too weak for this frequency. I think the pull-ups are mostly to reduce ground bounce and noise inside the DMA chip when the bus is idle, that otherwise would be floating.
http://github.com/ijor/fx68k 68000 cycle exact FPGA core
FX CAST Cycle Accurate Atari ST core
http://pasti.fxatari.com
User avatar
exxos
Site Admin
Site Admin
Posts: 23499
Joined: Wed Aug 16, 2017 11:19 pm
Location: UK
Contact:

Re: Atari STe 1040 trashing UltraSatan SD card

Post by exxos »

ijor wrote: Fri May 20, 2022 2:59 pm I though the zener diode was more for voltage protection (may be also for hot plug-in) than for level shifting. But I assumed this was clamping to 3.3v after the voltage dividers at the device side only. If you are saying this will clamp the ST side as well? Then that sounds like a bad design to me.
If the US chip is running from 3.3V and not 5V IO tolerant ( I Haven't checked), then it would have to be clamped to 3.3V else it would otherwise blow the chip up as the ST side would be driving from 5V. If it was 5V tolerant then why bother with the zener anyway.

D0 comes from the ASCI port, so 5V, via 220R clamped to gnd via the 3.3V zener. Then 100R to the blackfin chip which can only be 3.3V or GND.

If the blackfin was driving IO from 5V, then the output would still be clamped to 3.3V via the zener anyway. The 100R probably stops it from shorting 5V to 3.3V level.

But I haven't really looked into that series of chip, so don't know whats its IO capability is. Eitherway, like you suggest, just seems like a "poor mans level shifter" to me. I'm not a fan of clamping 5V to 3.3V even though its still classed as a logic high. By the time you have connections, cables, series resistors, noise etc in the mix, your probably boarding on the thing malfunctioning anyway.

I thought the later ultrasatan is used a proper level shifter using mosfets. I think that is what it looked like but I don't think anyone looked into it.

I guess if someone knows Lotherk they can ask him what level shifter system he's using on the current ultrasatan build. I assume the schematic is not published.

ijor wrote: Fri May 20, 2022 2:59 pm I think the pull-ups are mostly to reduce ground bounce and noise inside the DMA chip when the bus is idle, that otherwise would be floating.
It's not just idle though. The pullups fix HDD and floppy drive issues. The bus has to be active in that case.
https://www.exxosforum.co.uk/atari/ All my hardware guides - mods - games - STOS
https://www.exxosforum.co.uk/atari/store2/ - All my hardware mods for sale - Please help support by making a purchase.
viewtopic.php?f=17&t=1585 Have you done the Mandatory Fixes ?
Just because a lot of people agree on something, doesn't make it a fact. ~exxos ~
People should find solutions to problems, not find problems with solutions.
User avatar
stephen_usher
Posts: 5580
Joined: Mon Nov 13, 2017 7:19 pm
Location: Oxford, UK.
Contact:

Re: Atari STe 1040 trashing UltraSatan SD card

Post by stephen_usher »

ACSI doesn't run very fast, slower than the main ST bus, less than a megabit per second, so the rise times aren't necessarily that fast.
Intro retro computers since before they were retro...
ZX81->Spectrum->Memotech MTX->Sinclair QL->520STM->BBC Micro->TT030->PCs & Sun Workstations.
Added code to the MiNT kernel (still there the last time I checked) + put together MiNTOS.
Collection now with added Macs, Amigas, Suns and Acorns.
ijor
Posts: 428
Joined: Fri Nov 30, 2018 8:45 pm

Re: Atari STe 1040 trashing UltraSatan SD card

Post by ijor »

exxos wrote: Fri May 20, 2022 3:28 pm D0 comes from the ASCI port, so 5V, via 220R clamped to gnd via the 3.3V zener. Then 100R to the blackfin chip which can only be 3.3V or GND.

If the blackfin was driving IO from 5V, then the output would still be clamped to 3.3V via the zener anyway. The 100R probably stops it from shorting 5V to 3.3V level.

But I haven't really looked into that series of chip, so don't know whats its IO capability is.
I thought you said it was operating at open drain. I misunderstood what you meant by "high". Sorry. But indeed, the Blackfin is not 5V tolerant.
stephen_usher wrote: Fri May 20, 2022 3:28 pm ACSI doesn't run very fast, slower than the main ST bus, less than a megabit per second, so the rise times aren't necessarily that fast.
Even for 500 KHz, a 10K pull-up is too weak for open drain output. Try running the floppy with 10K pull-ups.
Goodness knows how much current the DMA has to sink to pull them down for data transfer.
...
Nope this was inside the ST when I was trying to diagnose DMA issues.
Yes, I was wrong regarding the clamping. I though it would affect the device side only. But anyway, the diodes are not relevant when the DMA chip is driving low and sinking current.
http://github.com/ijor/fx68k 68000 cycle exact FPGA core
FX CAST Cycle Accurate Atari ST core
http://pasti.fxatari.com
User avatar
stephen_usher
Posts: 5580
Joined: Mon Nov 13, 2017 7:19 pm
Location: Oxford, UK.
Contact:

Re: Atari STe 1040 trashing UltraSatan SD card

Post by stephen_usher »

The problem was that whenever the UltraSATAN was plugged in and switched on the floppy couldn't read or write anything. This had been getting worse the more I'd used the UltraSATAN.

I had to replace the DMA chip to get things to work again.
Intro retro computers since before they were retro...
ZX81->Spectrum->Memotech MTX->Sinclair QL->520STM->BBC Micro->TT030->PCs & Sun Workstations.
Added code to the MiNT kernel (still there the last time I checked) + put together MiNTOS.
Collection now with added Macs, Amigas, Suns and Acorns.
Post Reply

Return to “NOOB QUESTIONS”