USB power mod - switch ideas

[xhul]

Welcome.

 

First of all, here's the mod i performed: https://atarigamer.com/pages/installing-the-5v-usb-mod-into-an-atari-lynx-model-2
I use a wall adapter, and a 3 meters long cable, that has a 90° angle on the micro USB side.
Currently, i have to unplug the cable to reset the lynx, which of course isn't ideal, especially when using the gamedrive.
So i'd like to have a switch somewhere, preferably as close to the lynx as possible.

 

Changing the cable doesn't exactly seduce me, i may lose the length, as well as the cool 90° angle i have on my current one.

 

I guess i could add a tiny mechanical switch near the micro USB port myself, though that may be a challenge.

 

As another option, i was wondering, do you think it's possible to actually reroute the original ON & OFF buttons?
If yes, any idea how it could be done?
In the process, disabling the original power port entirely would also add an additional security layer for when the MOSFET|zener fail.

 

Thanks in advance for your time =]

[+DarkLord]

Why not just buy a long cable with a built in switch? 

 

https://tinyurl.com/bdz57c8n

 

 

[xhul]

12 hours ago, DarkLord said:

Why not just buy a long cable with a built in switch? 

 

https://tinyurl.com/bdz57c8n

Thanks a lot for the link, DarkLord =]

Yeah, i've seen those already, but 1.5M is too low for me, and i'd also lose that very comfortable 90° angle near the micro-USB side.

 

Right now i'm really looking for the necessary knowledge to reroute the ON|OFF buttons.

If that's not even possible, or if nobody knows how to do that, i may eventually try to install a switch where the second strap loop is, at the back of the lynx.

[+DarkLord]

Er...that's 150 centimeters. 59.x inches. Almost 5 feet.

 

It's the same one I bought (brought it up from my past

orders on Amazon) and it's right at 5 *feet*.  

 

PS And the switch is on the end closest to the Lynx.

 

[xhul]

12 hours ago, DarkLord said:

Er...that's 150 centimeters. 59.x inches. Almost 5 feet.

 

It's the same one I bought (brought it up from my past

orders on Amazon) and it's right at 5 *feet*.  

 

PS And the switch is on the end closest to the Lynx.

 

Regarding the cable length, i'm probably nitpicking, yes, though i still believe the longer the better, far outlets aren't that uncommon.

And the more i think about it, the more i like the idea of turning on|off USB power from the lynx itself.

[+karri]

There is a more complicated solution. The newer USB standards can negotiate to provide higher voltages that can be fed to the front of the power chain of the Lynx. Then  you can use on/off as intended.

 

Or you could add a buck converter to increase the 5V USB to 9V and feed that to the Lynx power jack.

[xhul]

1 hour ago, karri said:

There is a more complicated solution. The newer USB standards can negotiate to provide higher voltages that can be fed to the front of the power chain of the Lynx. Then  you can use on/off as intended.

 

Or you could add a buck converter to increase the 5V USB to 9V and feed that to the Lynx power jack.

Hello karri, thanks a lot for the input.

Do you think there could be a possible trick to do that while staying at 5V?

I mean, 9V would probably kill the safety purpose of USB power.

I'm obviously lacking the knowledge here, but i was thinking about replacing the MOSFET or something, so that the motherboard expects 5V all the way.

[+Stephen Moss]

4 hours ago, karri said:

There is a more complicated solution. The newer USB standards can negotiate to provide higher voltages that can be fed to the front of the power chain of the Lynx. Then  you can use on/off as intended.

I don't think that would work here, would it not required both the power unit to be an intelligent USB host and the Lynx power module to be an intelligent USB slave (both of which are doubtful) in order for them to negotiate for and then supply that higher voltage?

 

You could try something like this crudely drawn circuit...

 

The ON/Off power circuit of U6 should run from a 5V supply, so...

1. Remove R73, this disconnects to On/Off signal from pin10 of U6 to the power oscillator (Pin 1 of U6).
2. Connect the On/Off signal from pin 10 of U6 to the gate of a P channel MOSFET (preferably one designed for a 5V logic drive and without an internal diode across the Source & Drain), using the relevant vacant pad of R73 (easier than soldering to U6) .
3. Tie pin 1 of U6 to the supply 5V voltage (use relevant pad of R73) to prevent the power oscillator running and turning original power MOSFET Q12 On and Off.
4. Connect the USB 5V to across either D11 or C39 rather than C41 and the USB source GND to both U6 GND and C41 GND as shown in the schematic.

As the output of pin 10 toggles between high and low the new MOSFET should turn Off and On respectively thereby connecting and disconnecting the 5V supply to the parts of the circuit that are traditionally 5V via the connection to C41.

 

Alternatively...

1. Connect your USB 5V source across either D11 or C39 rather than C41.
2. Remove R73 and connect pin 1 of U6 to the supply as described in point 3 above.
3. Disconnect pins 6 & 8 of U6 so they are not driving the gate of Q12.
4. Connect the pin 12 output of U6 (pick up from applicable end of R70) to the Gate pin of Q12.

If the output from pin 12 of U6 is sufficient to turn on Q12, then as Q12 turns On and Off it should tie the floating GND of the 5V circuitry to the main power ground thereby switching the Lynx on and Off. You might not get the full 5V due to a little voltage drop across Q12 and L15 but hopefully not enough to prevent normal operation.

 

Theoretically both options should work without causing any damage, although they are untested and so...

• These are suggestions only, you implement either of them at your own risk and take sole responsibility of any damage that may result. 
• I cannot guarantee either would work, as it is dependant on the specification of the FETs used as in both cases (and particular in the case of the second option) a 5V gate drive may be insufficient to make the FETs conduct as much current as necessary for correct circuit operation, particular where the FET is not specifically designed for use with a Logic level gate drive signal.

 

[xhul]

7 hours ago, Stephen Moss said:

I don't think that would work here, would it not required both the power unit to be an intelligent USB host and the Lynx power module to be an intelligent USB slave (both of which are doubtful) in order for them to negotiate for and then supply that higher voltage?

 

You could try something like this crudely drawn circuit...

 

The ON/Off power circuit of U6 should run from a 5V supply, so...

1. Remove R73, this disconnects to On/Off signal from pin10 of U6 to the power oscillator (Pin 1 of U6).
2. Connect the On/Off signal from pin 10 of U6 to the gate of a P channel MOSFET (preferably one designed for a 5V logic drive and without an internal diode across the Source & Drain), using the relevant vacant pad of R73 (easier than soldering to U6) .
3. Tie pin 1 of U6 to the supply 5V voltage (use relevant pad of R73) to prevent the power oscillator running and turning original power MOSFET Q12 On and Off.
4. Connect the USB 5V to across either D11 or C39 rather than C41 and the USB source GND to both U6 GND and C41 GND as shown in the schematic.

As the output of pin 10 toggles between high and low the new MOSFET should turn Off and On respectively thereby connecting and disconnecting the 5V supply to the parts of the circuit that are traditionally 5V via the connection to C41.

 

Alternatively...

1. Connect your USB 5V source across either D11 or C39 rather than C41.
2. Remove R73 and connect pin 1 of U6 to the supply as described in point 3 above.
3. Disconnect pins 6 & 8 of U6 so they are not driving the gate of Q12.
4. Connect the pin 12 output of U6 (pick up from applicable end of R70) to the Gate pin of Q12.

If the output from pin 12 of U6 is sufficient to turn on Q12, then as Q12 turns On and Off it should tie the floating GND of the 5V circuitry to the main power ground thereby switching the Lynx on and Off. You might not get the full 5V due to a little voltage drop across Q12 and L15 but hopefully not enough to prevent normal operation.

 

Theoretically both options should work without causing any damage, although they are untested and so...

• These are suggestions only, you implement either of them at your own risk and take sole responsibility of any damage that may result. 
• I cannot guarantee either would work, as it is dependant on the specification of the FETs used as in both cases (and particular in the case of the second option) a 5V gate drive may be insufficient to make the FETs conduct as much current as necessary for correct circuit operation, particular where the FET is not specifically designed for use with a Logic level gate drive signal.

 

My man =]

Unfortunately, i guess my current lynx is in a too good condition to risk it directly.

However, i may try if i manage to grab a very cheap one with let's say, a very damaged LCD (so that i can use my original LCD on it without too much fear).

 

I also started checking mechanical switches.

So far, installing one where the second strap loop is looks possible on the paper.

However, it seems the tiny ones usually come with a low maximum supported amperage.

My wall adapter supposedly sends 2.4A, anybody knows how much the lynx actually expects to work properly?

[+Stephen Moss]

13 hours ago, xhul said:

My wall adapter supposedly sends 2.4A, anybody knows how much the lynx actually expects to work properly?

The original PSU is rated at 1 amp and so the Lynx will take approximately that otherwise its input voltage from an original PSU would not be approximately 9V.

It does not matter if the PSU can provide more current than is needed as the Lynx will only take what it needs, but there are two caveats to that...

1. The PSU voltage output must be regulated, which it should be on a USB PSU as they are set to 5V to comply with USB standards. Unlike the original Lynx PSU's which are not regulated which is why they measure 12-14V off load and are only 9V when on load, thus using an unregulated PSU with the wrong current value would result in either too high or too low an input voltage. In the case of a higher current capacity the resulting voltage from an unregulated PSU would be higher that it should and thus result in excessive power dissipation.
2. If a fault develops in the Lynx the result is potentially more destructive it could draw the entire 2.4 Amps (12W @5V) whereas at 1A @ 5V = 5W, which is still fairly destructive to most components but has less potential for any secondary damage.  

[xhul]

1 hour ago, Stephen Moss said:

The original PSU is rated at 1 amp and so the Lynx will take approximately that otherwise its input voltage from an original PSU would not be approximately 9V.

It does not matter if the PSU can provide more current than is needed as the Lynx will only take what it needs, but there are two caveats to that...

1. The PSU voltage output must be regulated, which it should be on a USB PSU as they are set to 5V to comply with USB standards. Unlike the original Lynx PSU's which are not regulated which is why they measure 12-14V off load and are only 9V when on load, thus using an unregulated PSU with the wrong current value would result in either too high or too low an input voltage. In the case of a higher current capacity the resulting voltage from an unregulated PSU would be higher that it should and thus result in excessive power dissipation.
2. If a fault develops in the Lynx the result is potentially more destructive it could draw the entire 2.4 Amps (12W @5V) whereas at 1A @ 5V = 5W, which is still fairly destructive to most components but has less potential for any secondary damage.  

Thanks again.

I guess my current USB PSU is fine, at least no visible side effects so far.

So, i'll be looking for a switch that can pass up to at least 1A.

One last question just in case, isn't there a risk for the switch itself if it receives 2.4A but supports lower than that?

[+Stephen Moss]

On 2/22/2023 at 9:45 AM, xhul said:

One last question just in case, isn't there a risk for the switch itself if it receives 2.4A but supports lower than that?

Difficult to say as the switch is probably capable of handling more current than it is rated at it, however for correct and continuous operation for the specified lifetime it must be used at the specified current, how long it would continue to operate correctly for at overcurrent would depend in how big a gap there is between what it is rated at and what it can actually take.

 

Remember the Lynx only requires 1 amp so the switch is unlikely to take much more than that unless there is a serious short circuit on the Lynx in which case the switch is the least of your problems. I don't think there would be much in the way of additional inrush current as the caps used are not very big. But if the switch did fail it could go one of two ways...

Open circuit: The switch contact burn out or get carbon deposits due to arcing when switching, you would have to run a temporary bypass to make the Lynx work until a new switch was installed or decide to no longer se a switch.

Short Circuit: The contacts get welded together by the arcing and so you cannot use it to turn the lynx off which is no different to the state you are in now.  

[xhul]

7 hours ago, Stephen Moss said:

Difficult to say as the switch is probably capable of handling more current than it is rated at it, however for correct and continuous operation for the specified lifetime it must be used at the specified current, how long it would continue to operate correctly for at overcurrent would depend in how big a gap there is between what it is rated at and what it can actually take.

 

Remember the Lynx only requires 1 amp so the switch is unlikely to take much more than that unless there is a serious short circuit on the Lynx in which case the switch is the least of your problems. I don't think there would be much in the way of additional inrush current as the caps used are not very big. But if the switch did fail it could go one of two ways...

Open circuit: The switch contact burn out or get carbon deposits due to arcing when switching, you would have to run a temporary bypass to make the Lynx work until a new switch was installed or decide to no longer se a switch.

Short Circuit: The contacts get welded together by the arcing and so you cannot use it to turn the lynx off which is no different to the state you are in now.  

Thanks for your time.

I was actually assuming that though the lynx will only suck 1A, the switch would potentially receive more than that, maybe for a brief moment, because it's upstream.

That probably shows how electronically ignorant i am.

 

I'm about to purchase a series of mini switches, hoping i find a convenient place to insert one of them somewhere on the case.

If i ever succeed, i'll upload a photo, for the record.

[sage]

On 2/22/2023 at 9:37 AM, Stephen Moss said:

The PSU voltage output must be regulated, which it should be on a USB PSU as they are set to 5V to comply with USB standards.

You trust a cheap USB switch enough to deliver 5V according to USB standard? And, how much over-voltage is allowed by the standard? Is it below the limits the Lynx chips can safely operate?

[+Stephen Moss]

3 hours ago, sage said:

And, how much over-voltage is allowed by the standard? Is it below the limits the Lynx chips can safely operate?

I don't know without looking at the USB specification or knowing the exact technology used in the in Lynx devices although I would think it is some type of CMOS rather than TTL based in which case CMOS can usually run from a supply voltage between 5 and 18V.

 

Due to a combination of lack of time and being very focused I tend to answer the question posed which was initially regarding power switching, but two things occurred to me...

1) A push to break switch would be required to interrupt the power supply, most small push button switches tend to be push to make so you might not find an applicable one.

2) If as in the original post a GameDrive is being used so there is no need to turn of the power in order to swap out cartridges for playing another game then another possible option could be to just reset the Lynx rather than power off. This can be achieved by using a switch to short out C3, thereby pulling the Reset line low, you should test to see if that would work before committing to it as I am not 100% sure if it would produce a full system (power on type) reset or just halt execution of whatever program is currently loaded but is only suitable where a GameDrive or similar cartridge is used.   

[xhul]

2 hours ago, Stephen Moss said:

1) A push to break switch would be required to interrupt the power supply, most small push button switches tend to be push to make so you might not find an applicable one.

I'm currently looking for one of those types:

https://www.electronicshub.org/switches/#Single_Pole_Single_Throw_Switch_SPST

https://www.electronicshub.org/switches/#Single_Pole_Double_Throw_Switch_SPDT

Of course a SPST would be enough, though i may use a SPDT instead (leaving 1 pin alone), because they tend to be significantly smaller.

2 hours ago, Stephen Moss said:

2) If as in the original post a GameDrive is being used so there is no need to turn of the power in order to swap out cartridges for playing another game then another possible option could be to just reset the Lynx rather than power off. This can be achieved by using a switch to short out C3, thereby pulling the Reset line low, you should test to see if that would work before committing to it as I am not 100% sure if it would produce a full system (power on type) reset or just halt execution of whatever program is currently loaded but is only suitable where a GameDrive or similar cartridge is used.

I've actually already considered a reset type switch, which would indeed work for the GD just fine.

But yeah, regarding regular cartridges, i don't exactly like the idea of inserting them while the lynx is already powered.

[xhul]

Tada!

 

Edited March 17 by xhul

[sage]

On 2/28/2023 at 10:13 AM, Stephen Moss said:

so there is no need to turn of the power in order to swap out cartridges for playing another game then another possible option could be to just reset the Lynx rather than power off. This can be achieved by using a switch to short out C3, thereby pulling the Reset line low, you should test to see if that would work before committing to it as I am not 100% sure if it would produce a full system (power on type) reset or just halt execution of whatever program is currently loaded but is only suitable where a GameDrive or similar cartridge is used.   

short the detect pin on the connector :shrug:

soft-reset often goes back to the main screen only, not reload the cartridge. Thus a reset button is a nice thing.

[42bs]

On 2/28/2023 at 10:13 AM, Stephen Moss said:

game then another possible option could be to just reset the Lynx rather than power off. This can be achieved by using a switch to short out C3, thereby pulling the Reset line low, you should test to see if that would work before committing to it as I am not 100% sure if it would produce a full system (power on type) reset or just halt execution of whatever program is currently loaded but is only suitable where a GameDrive or similar cartridge is used.   

With GD you need a power cycle to be able to select a new game. A reset of the Lynx just starts the last game in the Game memory.