Power System Lynx 1

[ThatGuyUKnow]

Hello all.

 

I have one I can’t figure out. The system will work if I put 5V into C39. I replaced the Q4/13 transistors, Q11 mosfet, and ZD1 zener diode with no change. Replaced U6, no change. Pulled up the schematic, pulled every component out of circuit and found R78 was bad so I replaced it. Still no change. Advice on what I’m missing here? 

[obschan]

Have you had a look at RetroSix power circuit walk through? It's quite thorough.
https://www.retrosix.wiki/power-circuit-atari-lynx

[ThatGuyUKnow]

I have. I used that schematic to check every component in the power system. I think I’ll look back through forum posts for U6 voltage values and then do some testing with an oscilloscope

[Brek Martin]

A game cartridge.

The Lynx can't power up without it because the supply for U6 is routed through the inserted cartridge,

but if you power the PCB with your own 5V supply, then the board is powered independently of the latching circuit and regulator,

and will work without a game cartridge.

 

Alternatively, the following rant:

 

You've shown you can identify components that concern the latching and regulation, but could have provided a bunch of diagnostic.

If you replace every part of a circuit with correct and working components, then the only place left for error is in your work, or the PCB itself.

Your post offers little to accommodate the people you're asking for help.

 

The following refers to Lynx 1 schema, which is easy for me to read.

 

You can check that U6 is receiving power on it's supply pin when a cart is inserted, and that it's ground pin is common with the ground rail

to the left side of Q11 in the schema (the section of the ground rail that is never switched out).

You can check that Mikey is pulling the base of Q4 (lynx 1) low, shortly after powerup, even when powering the board with an independent 5V supply.

You can check that neither the on nor off switch is shorted across the membrane (with gunk in the physical buttons).

You can check that when powered from 9V, that the U6 NOT gate outputs that form the latch, pins 10 & 12, share a common state.

 

 

 

[ThatGuyUKnow]

With the schematic I said that I removed the resistors and capacitors and tested them out of circuit. I did offer what I have replaced thus far. I’m not a pro at circuit design by any means, still learning there.
 

I appreciate your insight and advice . Mikey pulling Q4 low was something I didn’t know. I’ll also check the common ground on U6 that you noted. I’ll do some testing and report back on all findings.

[Brek Martin]

22 hours ago, ThatGuyUKnow said:

With the schematic I said that I removed the resistors and capacitors and tested them out of circuit. I did offer what I have replaced thus far. I’m not a pro at circuit design by any means, still learning there.
 

I appreciate your insight and advice . Mikey pulling Q4 low was something I didn’t know. I’ll also check the common ground on U6 that you noted. I’ll do some testing and report back on all findings.

One thing I was alluding to, is a pic would help.

My insight has ruined more than one lynx board this year, so we can all make mistakes,

and someone could catch yours (or not) with a good picture.

 

I guess most important to check the computer end of the PCB isn't getting higher than 5V,

or the zener isn't getting hot at this stage while you're actually working on it.

 

 

[ThatGuyUKnow]

On 7/25/2024 at 6:04 PM, Brek Martin said:

One thing I was alluding to, is a pic would help.

My insight has ruined more than one lynx board this year, so we can all make mistakes,

and someone could catch yours (or not) with a good picture.

 

I guess most important to check the computer end of the PCB isn't getting higher than 5V,

or the zener isn't getting hot at this stage while you're actually working on it.

 

 

Q11 is getting 8.9VDC. U6 pin 1 (in 1) is getting 8.8VDC, and pin 2 (put 1) is outputting 8.8VDC. Q4 is measuring an odd base voltage. The collector is measuring 8.9VDC, the emitter is measuring 8.5VDV, but the base is measuring 8.9VDC. My ground reference point is the negative of C38. Not sure if using C38 ground is causing the bad reading. Measuring in diode mode with the red probe on ground the base is OL, collector was about .700, and collector was 1.1.

[Brek Martin]

Well I'm coming up blank.

The negative end of C38 is an appropriate place to get unregulated ground, but remember once the regulation is working,

it's the ground supply being regulated, so to measure 5V would be done on round at the output side of Q11.

With the regulation working, 9V is being switched on & off very fast to maintain a 5V charge on all of the capacitors, but particularly check across C39.

U6 is not yet a concern. The part was selected to tolerate higher voltage because it acts as the latch for power switch prior to regulation.

(U6 is powered directly from the unregulated supply).

 

I don't find anything wrong in your image, but it wasn't just for me, but anyone looking.

 

Are you shorting the points to try to power on and off from the component side of the board without using buttons?

I mention this because ON and OFF labels are swapped in the Lynx 1 schematic!

 

If the zener gets hot all a sudden through your experimentation,

Q11 is stuck on, and the zener is doing all of the regulation for a second or so before it pops.

A second or so later, both Mikey and Suzy fry.

It could pay to ignore ESD precaution while working on it, and keep a finger on ZD1 as far as possible, ready to pull power out!

 

 

 

[Stephen Moss]

On 7/28/2024 at 10:01 PM, ThatGuyUKnow said:

Q11 is getting 8.9VDC

You don't specify which pin it is on so that may or may not be correct.

On 7/28/2024 at 10:01 PM, ThatGuyUKnow said:

U6 pin 1 (in 1) is getting 8.8VDC, and pin 2 (put 1) is outputting 8.8VDC.

Are you powering it from batteries, a regualted 9V supply or an official power supply as your initial post suggest it was not powerin on as with the Lynx off using the offiacal power unit I would expect 12-14V and 9V with the other options so with out knowing which you are using I don't know if 8.8V DC in pin 1 is to be expected or not, although I would expect it to be high when the Lynx is off.

However if you are saying that you are getting 8.8V on both pins 1 & 2 of U6 you might want to re-test that voltage, if it is still correct look for a solder short/measure for short between pins 1 & 2 of U6, as with U6 being a inverter if pin 1 is high then the output at pin 2 should be 0V.

 

On 7/28/2024 at 10:01 PM, ThatGuyUKnow said:

Not sure if using C38 ground is causing the bad reading

Should be Ok for that part of the circuit. 

[Brek Martin]

14 hours ago, Stephen Moss said:

U6 being a inverter if pin 1 is high then the output at pin 2 should be 0V.

 

Good catch! 

[ThatGuyUKnow]

On 7/29/2024 at 4:55 AM, Stephen Moss said:

You don't specify which pin it is on so that may or may not be correct.

 

Ok so I had to look up the datasheet if I have the correct one I’m getting 8.9V on the drain, the large pad/middle pin. It’s def the drain and connected to L6 which also has 8.9V. I checked R84, no voltage, checked U6 pins 6/8 no voltage. I do have voltage on one side of C77 (8.9V) but not the other. Thought the cap was bad, tested it as good but replaced it anyway, no change. If I put a wire on each side of C77 bypassing it the Lynx will go straight to powering on. Using a DMM in diode mode, red on C38 ground I am getting about .689 however after I bypassed C77 it’s fell to .099 and stayed there until Q11 fully discharged.

Edited July 30 by ThatGuyUKnow

[ThatGuyUKnow]

On 7/29/2024 at 4:55 AM, Stephen Moss said:

9V with the other options so with out knowing which you are using I don't know if 8.8V DC in pin 1 is to be expected or not, although I would expect it to be high when the Lynx is off.

However if you are saying that you are getting 8.8V on both pins 1 & 2 of U6 you might want to re-test that voltage, if it is still correct look for a solder short/measure for short between pins 1 & 2 of U6, as with U6 being a inverter if pin 1 is high then the output at pin 2 should be 0V.

 

I tested pins 13/14. Im used to ICs have the dot/line by pin 1 but apparently this is not the case for the 4069. Pin 1 is 8.9VDC, Pin 2 is 0VDC. Power supply is a regulated 9V.

[Stephen Moss]

On 7/28/2024 at 10:01 PM, ThatGuyUKnow said:

Q4 is measuring an odd base voltage. The collector is measuring 8.9VDC, the emitter is measuring 8.5VDV, but the base is measuring 8.9VDC.

You did not actually specify in your original post what he issue was but as you mentioned the system work when powered from 5 connected to C39 I assumed it was not powering on, thus the reading for Q4 base does make sense because it is connected to the floating ground on the Drain side of Q11 and so with the system off (I have not checked it against my Lynx) but I would therefore expect it to be at the supply voltage and then drop to about 4V when measured relative to your C38 ground point.

 

On 7/30/2024 at 10:20 PM, ThatGuyUKnow said:

I do have voltage on one side of C77 (8.9V) but not the other. Thought the cap was bad, tested it as good but replaced it anyway, no change. If I put a wire on each side of C77 bypassing it the Lynx will go straight to powering on.

It would turn on if you short C77 as you are pulling the Gate of Q11 high turning it on permanently (which it is not meant to be) thus that is not something I would generally recommend anyone doing as a test because...

1.  The power circuit is meant to be a switching circuit and it wont be which could cause some components to over heat and 
2.  Pins 6 & 8 of U6 are trying to pull the Q11 Gate voltage low and could be damaged in their effort to do that as they try to sink more current than they can handle in the attempt.

 

On 7/24/2024 at 10:22 PM, ThatGuyUKnow said:

I have. I used that schematic to check every component in the power system. I think I’ll look back through forum posts for U6 voltage values and then do some testing with an oscilloscope

While the RetroSix link posted to looks quite good from a quick scan though, you should be able to find a lot on posts on U6 voltages and testing it (many from myself), and they will worth reading as they should help.

However, if after reading all those you are still drawing a blank start by splitting the circuit onto two parts (left and right of Q11) and start on the left side, in that does the On/Off power latch formed by pins 10, 11, 12 & 13 of U6 work as expected (pin to switch between high an low and stay there) as that would help indicate in which half of the circuit the problem is likely to be.

[Brek Martin]

On 8/1/2024 at 6:24 PM, Stephen Moss said:

 

It would turn on if you short C77 as you are pulling the Gate of Q11 high turning it on permanently (which it is not meant to be) thus that is not something I would generally recommend anyone doing as a test because...

Game over because unregulated supply voltage delivered to the computer.

[Stephen Moss]

6 hours ago, Brek Martin said:

Game over because unregulated supply voltage delivered to the computer.

Maybe, but if the IC's use CMOS technology that traditionally has a 5 to 18V supply range, so the higher voltage in and of itself does not necessarily mean instant death, but should be avoided wherever possible.

What the higher voltage does do is increase the power dissipation (heat) in the device. In this case nearly doublng it, that will stress/kill the IC's over time, so it may well survive if operating breifly at a higher voltage but certainly long time operation at a higher voltage will not do it any good.  

[Brek Martin]

On 8/5/2024 at 5:54 PM, Stephen Moss said:

Maybe, but if the IC's use CMOS technology that traditionally has a 5 to 18V supply range, so the higher voltage in and of itself does not necessarily mean instant death, but should be avoided wherever possible.

What the higher voltage does do is increase the power dissipation (heat) in the device. In this case nearly doublng it, that will stress/kill the IC's over time, so it may well survive if operating breifly at a higher voltage but certainly long time operation at a higher voltage will not do it any good.  

That's not the actual occurrence the two recent times I experienced it.

You can take a CMOS pic microcontroller and power it with a 12V supply all day long.

If you clock that pic with it's own internal RC clock that will heat up, and in turn, overclock the device and it will still be fine.

 

In the Lynx, the zener will take on all of the regulation if the FET fails short, and quickly fail itself,

then one or more of the custom chips stop functioning within seconds, and doesn't function after cooling.

At least those two times I saw. I can't see any particular potential for CMOS latchup occurring.

[ThatGuyUKnow]

On 8/1/2024 at 4:24 AM, Stephen Moss said:

You did not actually specify in your original post what he issue was but as you mentioned the system work when powered from 5 connected to C39 I assumed it was not powering on, thus the reading for Q4 base does make sense because it is connected to the floating ground on the Drain side of Q11 and so with the system off (I have not checked it against my Lynx) but I would therefore expect it to be at the supply voltage and then drop to about 4V when measured relative to your C38 ground point.

 

It would turn on if you short C77 as you are pulling the Gate of Q11 high turning it on permanently (which it is not meant to be) thus that is not something I would generally recommend anyone doing as a test because...

1.  The power circuit is meant to be a switching circuit and it wont be which could cause some components to over heat and 
2.  Pins 6 & 8 of U6 are trying to pull the Q11 Gate voltage low and could be damaged in their effort to do that as they try to sink more current than they can handle in the attempt.

 

While the RetroSix link posted to looks quite good from a quick scan though, you should be able to find a lot on posts on U6 voltages and testing it (many from myself), and they will worth reading as they should help.

However, if after reading all those you are still drawing a blank start by splitting the circuit onto two parts (left and right of Q11) and start on the left side, in that does the On/Off power latch formed by pins 10, 11, 12 & 13 of U6 work as expected (pin to switch between high an low and stay there) as that would help indicate in which half of the circuit the problem is likely to be.

Yeah the initial issue was no power on but working via C39.

 

U6, 10 is 9V, 11 is 0V, 12 is 0V, and 13 is 9V which is correct. I confirmed shorting pins 11/14 momentarily that the console boots and works correctly.

 

I thought maybe the issue was with the ON button itself. I am measuring 100k between the points and when the button is pressed it goes to 0 ohms so its properly conducting.

 

I wanted to confirm the traces were intact around U6. Pin 13 on U6 connects to R70 which measured correctly out of circuit, and the other side of R70 connects to pin 10 on U6. I confirmed pin 11 on U6 connects to R85 which measures correctly out of circuit and then it connects to pin 12 on U6. R85 does connect to C63 which also measures correctly out of circuit. All in line with the schematic.

 

Thoughts on where to go from here?

[Stephen Moss]

8 hours ago, ThatGuyUKnow said:

U6, 10 is 9V, 11 is 0V, 12 is 0V, and 13 is 9V which is correct. I confirmed shorting pins 11/14 momentarily that the console boots and works correctly.

If it stays on after the temporary short which is what you appear to be saying then the on/off latch and oscillator formed by U6 and associated components would appear to be working so I think you can eliminate them as being the problem.

 

8 hours ago, ThatGuyUKnow said:

Thoughts on where to go from here?

If you can find TP17 and TP18 try temporarily shorting them (simulating what the power on button does), using TP17 & TP8 would be best but if you cannot find them the alternative would be U6 pin 13 to the non ground end of C34 or R2 (try both as on may not work indication which could be the issue).

 

Using TP17 & 18. if the Lynx turns on, stays on and works as expected then the most likely issue is...

The power on button and/or a connection between it and TP17 (open circuit track on the flex circuit, flex circuit not correctly in connector, open circuit between connector and TP17)

 

However, of the Lynx remains off, that suggests to me that pin 13 is not being pulled low when the power on button is pressed, reasons for that would be...

1. There is no connection between both the power button and TP17 to C34 and/or R2 (C37/R68 respectively on the Lynx 2)
2. There is no connection between C34 and/or R2 to ground, 
3. Everything is connected as it should be C34 has gone open circuit
4. Everything is connected as it should be but R2 has gone high resistance/open circuit 

Personally, from the symptoms I think that if the connections are fine then it is more likely to be a C34 issue then an R2 issue but you should check both as I seem to recall diagnosing a power on/off issue to either R2 or C34 (or their Lynx 2 equivalents) before.

 

[ThatGuyUKnow]

On 8/19/2024 at 4:18 AM, Stephen Moss said:

If it stays on after the temporary short which is what you appear to be saying then the on/off latch and oscillator formed by U6 and associated components would appear to be working so I think you can eliminate them as being the problem.

 

If you can find TP17 and TP18 try temporarily shorting them (simulating what the power on button does), using TP17 & TP8 would be best but if you cannot find them the alternative would be U6 pin 13 to the non ground end of C34 or R2 (try both as on may not work indication which could be the issue).

 

Using TP17 & 18. if the Lynx turns on, stays on and works as expected then the most likely issue is...

The power on button and/or a connection between it and TP17 (open circuit track on the flex circuit, flex circuit not correctly in connector, open circuit between connector and TP17)

 

However, of the Lynx remains off, that suggests to me that pin 13 is not being pulled low when the power on button is pressed, reasons for that would be...

1. There is no connection between both the power button and TP17 to C34 and/or R2 (C37/R68 respectively on the Lynx 2)
2. There is no connection between C34 and/or R2 to ground, 
3. Everything is connected as it should be C34 has gone open circuit
4. Everything is connected as it should be but R2 has gone high resistance/open circuit 

Personally, from the symptoms I think that if the connections are fine then it is more likely to be a C34 issue then an R2 issue but you should check both as I seem to recall diagnosing a power on/off issue to either R2 or C34 (or their Lynx 2 equivalents) before.

 

Touched TP18 to the non-ground side of R2 and it turned on. I will look to locate TP17 (I haven't been able to find it on the Lynx I'm repairing or a parts board which has most of the components taken off) and the power button traces. Ill let you know what I come up with.

 

Edit: Found TP17 below the ComLynx port.

 

Ok. Did some testing. TP17/18 go through their respective vias and to the top/bottom of the ON power button respectively. TP17 is the top of the on button. TP18 goes to the bottom of the on button and the top of the off button. I am getting .7 ohms of resistance which is normal. I can use a piece of wire or even my DMM probes to operate the on/off buttons and it will work 100% BUT no lynx buttons make the functionality work. I have tried 5 so far, two which are functional in a tested system. I did bypass the vias to see and that changed nothing. I don't get it. Any suggestions because this makes NO sense to me.

Edited September 1 by ThatGuyUKnow

[Stephen Moss]

12 hours ago, ThatGuyUKnow said:

I can use a piece of wire or even my DMM probes to operate the on/off buttons and it will work 100% BUT no lynx buttons make the functionality work

Well to me that suggest that all component related to the power On/Off latch are connected and functioning as they should.

 

So I would be exploring the connections from the Power On/Off latch to the front panel and the panel itself all the way to the On/Off button contacts and any power connections to the panel, it could be as something like the flex circuit not being correctly inserted into the FCC connector or a bad solder joint on that connector.

If you have a working Lynx then the easiest test would be to swap the entire assembled front panel (buttons, case & flex circuit) from the good Lynx with that of the bad Lynx. All things being equal...

1. If the problem lies on the PCB of the bad Lynx then the problem should persist.
2. If the problem moves to the good Lynx then it lies with the buttons and/or flex circuit.
3. If both Lynxes work then the problem could be that the flex circuit was not in the FCC socket correctly or a bad solder joint on the FCC connector.

I don't know how accurate it is but I recall reading that if the Lynx buttons are not inserted the correct way round they do not work, does not sound right to be but it may be something worth trying as a last resort.

 

[ThatGuyUKnow]

3 hours ago, Stephen Moss said:

Well to me that suggest that all component related to the power On/Off latch are connected and functioning as they should.

 

So I would be exploring the connections from the Power On/Off latch to the front panel and the panel itself all the way to the On/Off button contacts and any power connections to the panel, it could be as something like the flex circuit not being correctly inserted into the FCC connector or a bad solder joint on that connector.

If you have a working Lynx then the easiest test would be to swap the entire assembled front panel (buttons, case & flex circuit) from the good Lynx with that of the bad Lynx. All things being equal...

1. If the problem lies on the PCB of the bad Lynx then the problem should persist.
2. If the problem moves to the good Lynx then it lies with the buttons and/or flex circuit.
3. If both Lynxes work then the problem could be that the flex circuit was not in the FCC socket correctly or a bad solder joint on the FCC connector.

I don't know how accurate it is but I recall reading that if the Lynx buttons are not inserted the correct way round they do not work, does not sound right to be but it may be something worth trying as a last resort.

 

It’s a Lynx 1 so no flex or FPC.

 

I did rotate a number of the buttons but that didn’t have an effect.

 

 

[Stephen Moss]

19 hours ago, ThatGuyUKnow said:

It’s a Lynx 1 so no flex or FPC.

Sorry I forgot that the Lynx 1 & 2 were different in that regard.

 

As we appear to have proved that all the Lynx circuitry is operating using methods other than the buttons to turn it on and off reliably then logically that must only leave the conductive button pads, the PCB button contacts or an open circuit between the buttons and pin 13 of U6.

 

Conductive Button Pads:

This can be tested by using something conductive to short the relevant button contacts on the PCB, half of which should be going to pin 13 of U6, the other half to VCC (Off) or C34/R2 (On).

If that turns the Lynx on and off reliably then it is possible that the conductive pads on the button rubber sub strait may be too warn to make a good contact (see my post here),  

 

PCB Contacts:

You could conduct a visual inspection to see if they look worn, if they do then you could try and replenish them using a pen with conductive ink (used for PCB repair) and putting a couple of layers down on top of them, although I don't know how well it would stand up to the ware it would get from the button presses as it was not really designed for that. 

 

Open Circuit:

That should be easy enough to check by doing a continuity using a multi-meter.

 

If non of those resolve the issue then as the power on/off circuit itself appears to be working I am not sure what else it could be.  

[ThatGuyUKnow]

On 9/3/2024 at 4:21 AM, Stephen Moss said:

Sorry I forgot that the Lynx 1 & 2 were different in that regard.

 

As we appear to have proved that all the Lynx circuitry is operating using methods other than the buttons to turn it on and off reliably then logically that must only leave the conductive button pads, the PCB button contacts or an open circuit between the buttons and pin 13 of U6.

 

Conductive Button Pads:

This can be tested by using something conductive to short the relevant button contacts on the PCB, half of which should be going to pin 13 of U6, the other half to VCC (Off) or C34/R2 (On).

If that turns the Lynx on and off reliably then it is possible that the conductive pads on the button rubber sub strait may be too warn to make a good contact (see my post here),  

 

PCB Contacts:

You could conduct a visual inspection to see if they look worn, if they do then you could try and replenish them using a pen with conductive ink (used for PCB repair) and putting a couple of layers down on top of them, although I don't know how well it would stand up to the ware it would get from the button presses as it was not really designed for that. 

 

Open Circuit:

That should be easy enough to check by doing a continuity using a multi-meter.

 

If non of those resolve the issue then as the power on/off circuit itself appears to be working I am not sure what else it could be.  

I can use a wire like an on/off button and it works perfectly.

 

I thought the same thing on the pcb contacts. I tried some deoxit initially but that didn't work so i tinned the pads and wicked away the solder but that didn't work.

 

I have other buttons including some that work in another Lynx and none of them work. 

 

I'm going to brainstorm a bit further, lmk if you have other thoughts.

[Brek Martin]

Am I reading right that you can place a metallic object across the 'On' PCB pads to turn it on,

but if you use the carbon impregnated pad that works in another Lynx, that won't close the circuit and power on the Lynx being repaired?

 

There's very little margin left there for problem solving indeed. There would be some slight changes in pressures applied to the board.

You are turning the board upside down to use the button pads, and the force of gravity is applied in the opposite direction.

Can you turn the board upside down once you successfully power it on? Have you reflowed the entire PCB?

I'd replace every diode that is encapsulated within a glass tube as a very last resort (at least short of replacing the unobtaium).