Atari XF551 floppy driver / power switch

[macsonny]

Hi All,

 

I'm currently finishing off building the @Dropcheck XF551_B v 6.0 board. I've got all components placed except the power switch which I think I can get from Best but being in Australia the shipping is staggering and ridiculous.

 

I was wondering if anyone had an alternative for the XF551 power switch - even if it's just functionally the same but not actually cosmetically the same switch. I plan on putting the board and floppy drive unit in an external PC case so it doesn't need to be a perf match - switch just needs to turn the right power rails on and off.

 

Any ideas?

 

Thanks 

 

macsonny

[sup8pdct]

Will this be ok? Sorry for rubbish pic. Found on mouser australia. Not sure what jaycar have.

[macsonny]

3 minutes ago, sup8pdct said:

Will this be ok? Sorry for rubbish pic. Found on mouser australia. Not sure what jaycar have.

Could work. Do you have a part number?

[sup8pdct]

A105MD9AB04

 

Use search function on mouser, toggle switch, pcb mount 90deg etc. heaps of options and sizes.

 

[macsonny]

I don't have the spec for the XF551 switch so have no idea what the equivalent is from the MOUSER list. Does @Dropcheck or anyone else know the spec for the switch?

[Dropcheck]

On 2/5/2024 at 4:22 AM, macsonny said:

I don't have the spec for the XF551 switch so have no idea what the equivalent is from the MOUSER list. Does @Dropcheck or anyone else know the spec for the switch?

 

 It's a close fit, but this should work.  C&K 7201J1AQE2

[macsonny]

9 hours ago, Dropcheck said:

 

 It's a close fit, but this should work.  C&K 7201J1AQE2

That looks like a 3 legged switch when I think the one in the XF551 is a 6 legged switch. Are you sure this switch will work in the XF551?

 

https://au.mouser.com/ProductDetail/CK/7201J1AQE2?qs=HrqAcaC9xK5GxVU6mL50Vg%3D%3D

[Dropcheck]

40 minutes ago, macsonny said:

That looks like a 3 legged switch when I think the one in the XF551 is a 6 legged switch. Are you sure this switch will work in the XF551?

 

https://au.mouser.com/ProductDetail/CK/7201J1AQE2?qs=HrqAcaC9xK5GxVU6mL50Vg%3D%3D

Mouser has a bad habit of putting the wrong pic on a part #.  This a more accurate pic of the part.  https://www.digikey.com/en/products/detail/c-k/7201J1AQE2/67006

 

You might see if they have the J50 lever model change out.  I think this rocker lever is a tad to big.  ie see if you can get 7201J50AQE2 or 7201J50ABE2

Edited February 7 by Dropcheck

[macsonny]

2 hours ago, Dropcheck said:

Mouser has a bad habit of putting the wrong pic on a part #.  This a more accurate pic of the part.  https://www.digikey.com/en/products/detail/c-k/7201J1AQE2/67006

 

You might see if they have the J50 lever model change out.  I think this rocker lever is a tad to big.  ie see if you can get 7201J50AQE2 or 7201J50ABE2

Aaaahhh, that looks more like it. Have ordered a few to see how I go. Will report back 🙂

 

[macsonny]

On 2/7/2024 at 2:52 PM, Dropcheck said:

Mouser has a bad habit of putting the wrong pic on a part #.  This a more accurate pic of the part.  https://www.digikey.com/en/products/detail/c-k/7201J1AQE2/67006

 

You might see if they have the J50 lever model change out.  I think this rocker lever is a tad to big.  ie see if you can get 7201J50AQE2 or 7201J50ABE2

Ok. Board complete - finally.

 

Dumb question but is the power into the board the same as the 1050 9VAC?

 

[Dropcheck]

Yes.  Very same.

[macsonny]

On 2/24/2024 at 3:13 PM, Dropcheck said:

Yes.  Very same.

Great.

 

So built up first board. Haven't bothered putting in all the IC's for the track display just yet and have kept all the main IC's off the board until I'm sure it's ok.

 

Apples 9 VAC to the board and no LED's lit.

 

Tested that I'm getting 9VAC in ok. Then tested the output of the bridge rectifier and only getting 0.7 VDC from the top of D8 and bottom of D10.

 

Any obvious ideas why the voltage drop would be happening and hence not getting a decent DC voltage to the two 5vdc and 12vdc regulators?

 

I've attached a picture of the populated board if that helps.

 

[Dropcheck]

I'd check that all four diodes and power switch have sufficient solder flow and that at the same time no bridges.  It looks like a couple diodes are pretty dry on the component side of the board.  Also from ancient prototype experience when I was part of the 1088XLD testing group, we had issues with needing to clean the legs of those very same diodes because of dirt/oxidation that had happened at the supplier.  A good scrub with IPA solved that problem, but it needs to be done prior to installation. 

 

Also the leds are nice, but always check the chips power pin and the disk drive power cable to verify +5 and +12V.  I've managed to install leds in backwards somehow, (don't ask me how) and spent an hour chasing my tail trying to figure out why they weren't lighting up. 

[macsonny]

On 3/8/2024 at 12:46 AM, Dropcheck said:

I'd check that all four diodes and power switch have sufficient solder flow and that at the same time no bridges.  It looks like a couple diodes are pretty dry on the component side of the board.  Also from ancient prototype experience when I was part of the 1088XLD testing group, we had issues with needing to clean the legs of those very same diodes because of dirt/oxidation that had happened at the supplier.  A good scrub with IPA solved that problem, but it needs to be done prior to installation. 

 

Also the leds are nice, but always check the chips power pin and the disk drive power cable to verify +5 and +12V.  I've managed to install leds in backwards somehow, (don't ask me how) and spent an hour chasing my tail trying to figure out why they weren't lighting up. 

I decided to just rebuild a brand new board as I have 5 unpopulated PCB's.

 

Just a quick question - what is the value of C19 as it's not in the BOM?

 

[macsonny]

Ok, so worked out he value of C19 form the schematic.

 

Have built up a completely new board - same results. No LED and doesn't seem like there is power getting through.

 

It would be good to have a few test points on the board I can get voltages from. Any suggestions there?

 

Checking the regulators there seems to be nil DC voltage getting to the regulators so it must be something earlier in the circuit?

Edited March 9 by macsonny

[Dropcheck]

My bad on the BOM notation for C19.  Fixed now. 

 

At this early point in the power supply circuit test points won't help very much, essentially that's what the leds are for. 

 

The power design is by Mytek and was used in his 1088XLD creation, so I feel pretty confident that unless I screwed up in the translation between his schematics and my board, you should have power.  Let me dig into my stash of earlier schematics for Marlin's 1035 venture and compare it to Mytek's published schematics and see if I missed something.   Give me about an hour from now and I should have an answer. 

[Dropcheck]

Okay I think I've found the problem.  In previous versions of the Mytek powered board I had only used the original XF551 power jack and switch.  Power was working on my test boards.  As I then went in to add a little more flexibility to use comparable parts in the revised schematic, I lost the connection between the bottom of C19 and the negative side of C109.  Since these changes were so simple, I decided not to do a whole new board order, so the lost schematic connection never made it into the final gerbers.  There's a persistence of vision that happens when you stare at something for a long time, even with breaks.  Your mind tricks you into believing that connection really still exists despite the obvious physical evidence that it doesn't.

 

 

It's a simple fix.  A bodge wire from the bottom of C19 to the negative side of C102.  Use a fairly thick wire, but not excessive.  It needs to handle the 9VAC at about 1A.  See the png attached.  You should see the added connection highlighted in red between the two caps.   Let me know if that fixes it.  It should.  As soon as you confirm it does, I'll redo the gerbers and re upload to the website, so no one else gets boinked. 

 

Sorry, this one I have to chalk up to 'fill in the blank' reason.  It shouldn't have happened, but it did.   🤕

[macsonny]

10 hours ago, Dropcheck said:

Okay I think I've found the problem.  In previous versions of the Mytek powered board I had only used the original XF551 power jack and switch.  Power was working on my test boards.  As I then went in to add a little more flexibility to use comparable parts in the revised schematic, I lost the connection between the bottom of C19 and the negative side of C109.  Since these changes were so simple, I decided not to do a whole new board order, so the lost schematic connection never made it into the final gerbers.  There's a persistence of vision that happens when you stare at something for a long time, even with breaks.  Your mind tricks you into believing that connection really still exists despite the obvious physical evidence that it doesn't.

 

 

It's a simple fix.  A bodge wire from the bottom of C19 to the negative side of C102.  Use a fairly thick wire, but not excessive.  It needs to handle the 9VAC at about 1A.  See the png attached.  You should see the added connection highlighted in red between the two caps.   Let me know if that fixes it.  It should.  As soon as you confirm it does, I'll redo the gerbers and re upload to the website, so no one else gets boinked. 

 

Sorry, this one I have to chalk up to 'fill in the blank' reason.  It shouldn't have happened, but it did.   🤕

Thanks for looking through this. Glad it's just a matter of a bodge wire.

 

 

Edited March 10 by macsonny

[macsonny]

Bodge installed - worked it out from schematics.

 

Now getting 12VDC on the 12v rail and 1.77 VDC on the 5v rail?

[Dropcheck]

38 minutes ago, macsonny said:

Bodge installed - worked it out from schematics.

 

Now getting 12VDC on the 12v rail and 1.77 VDC on the 5v rail?

What is the substitute +5V regulator part #?

 

Where are you getting the 5V rail voltage from?

 

Edited March 10 by Dropcheck

[macsonny]

1 hour ago, Dropcheck said:

What is the substitute +5V regulator part #?

 

Where are you getting the 5V rail voltage from?

 

Using this regulator https://au.element14.com/traco-power/tsrn-1-2450/dc-dc-converter-1a-5v-5w/dp/228024701 Using same (but 12 v version) for the 12VDC regulator and seems to be working fine.

 

I'm testing the 5VD at the floppy drive power cable point and also directly from the regulator. Same output at each point.

 

EDIT: I've used these regulators for 5 and 12 VDC in a 1050 floppy in the past and they seem to work well - zero heat output!

 

Is there a substitute for the 0.1uH inductor which is horrible to try and find?

Edited March 10 by macsonny

[Dropcheck]

When you supposedly have the power switch turned off, are you getting either +5V or +12V at the drive power cable?  Of course you shouldn't have either voltage with the switch off, but the switch can be failing or my footprint is wrong. 

 

I don't see any other issues with the power circuit design itself.  It's now matching Mytek's design, so any problems would be related to soldering and possible part failure or wrong part orientation. 

 

I think you said you had assembled two boards at least to the point of the drive power cable and power leds.  I would recheck the power circuit solder flow on all parts on both the boards you have assembled.  Apply the bodge wire on the other board and see if there is any difference.  Remember that the wire needs to be thick enough to carry not just the voltage but current as well.   If one has +5V at the drive power cable and not the other then you may have a bad regulator on that board.  I would also double check the orientation of the +5V regulator in the board. 

 

If both are still well below +5V, then I would replace the +5V traco regulator with a bog standard 7805 positive TO-220 linear regulator on one board.  You can run it without a heat sink for a little while for testing purposes.  Basically we are testing the ability of the traco converter to handle lower than normal +5V input level.  There's not much margin to work with here.  A max of 18V minus 12V, minus 4/6 diode drop out voltages.  If you hit +5V with the 7805 installed, then the traco just can't step up the voltage enough to compensate for the low input voltage. 

 

The Traco regulators are nice units, but I haven't used them very much.  I don't know how they react to the possible marginal input voltage despite what the datasheet says.   

 

Let me know how it goes.

[macsonny]

1 hour ago, Dropcheck said:

When you supposedly have the power switch turned off, are you getting either +5V or +12V at the drive power cable?  Of course you shouldn't have either voltage with the switch off, but the switch can be failing or my footprint is wrong. 

 

I don't see any other issues with the power circuit design itself.  It's now matching Mytek's design, so any problems would be related to soldering and possible part failure or wrong part orientation. 

 

I think you said you had assembled two boards at least to the point of the drive power cable and power leds.  I would recheck the power circuit solder flow on all parts on both the boards you have assembled.  Apply the bodge wire on the other board and see if there is any difference.  Remember that the wire needs to be thick enough to carry not just the voltage but current as well.   If one has +5V at the drive power cable and not the other then you may have a bad regulator on that board.  I would also double check the orientation of the +5V regulator in the board. 

 

If both are still well below +5V, then I would replace the +5V traco regulator with a bog standard 7805 positive TO-220 linear regulator on one board.  You can run it without a heat sink for a little while for testing purposes.  Basically we are testing the ability of the traco converter to handle lower than normal +5V input level.  There's not much margin to work with here.  A max of 18V minus 12V, minus 4/6 diode drop out voltages.  If you hit +5V with the 7805 installed, then the traco just can't step up the voltage enough to compensate for the low input voltage. 

 

The Traco regulators are nice units, but I haven't used them very much.  I don't know how they react to the possible marginal input voltage despite what the datasheet says.   

 

Let me know how it goes.

Would having the wrong inductors cause the voltage drop from the regulator? I have a sneaking suspicion the 6.8 uH inductor I have isn't working right or doesn't meet spec. Could that be the culprint?

Edited March 10 by macsonny

[Dropcheck]

8 hours ago, macsonny said:

Would having the wrong inductors cause the voltage drop from the regulator? I have a sneaking suspicion the 6.8 uH inductor I have isn't working right or doesn't meet spec. Could that be the culprint?

I doubt it.  If you are actually reading voltage at pin 3 of the Traco regulator and it's way below +5V, the inductor isn't even in the picture yet.  The circuitry after the regulator mainly conditions the shape of the signal, filtering and smoothing out ripples.   You can replace it if you want, but I don't think it will make much difference. 

[Dropcheck]

One other thing you can try is use another 1050 power supply.  These things are going on forty years old now, and while they may have the voltage, they aren't pushing the current they used too.