Another black blank screen issue in Atari Lynx 2

[Rollinpizza]

Hi! First of all, I'm very sorry posting this post because I am searching info about the problems that I have with my Atari Lynx 2 during months so that I can solve them, but without success. I arrived some conclusions but before confirming it, I would like to assure with your knownledge and your help.

 

But, before, I begin from the beginning:

 

This Atari Lynx had problems to power on. I've recapped it but the problem persisted. I've detected there was a problem on the power stage with the "headphone trick" (the screen woke up with "Insert" title). So, I've replaced all critical components of the power supply stage (Mosfet, Q8, Q7, D13). With this, I got it power on the console but a black blank screen. The speaker sometimes sounds noise and after silence, or a very distorted sound and after silence, or a very distorted sound.

 

I've checked the following:

 

• All the chips has his Vcc voltage (4'9V more or less).
• Some pins of the 74H164 chip. Pin 1, 14, 9 with 4'9V, but pin 2 & 8 has 0-0'2V.
• Some pins of the 74HC4040 chip. Pin 16 with 4'9V, but pin 11 & 10 has 0-0'2V.
• TP20 => 0-0'2V.
• The temperature of the chips. All fine except U1 that is very slightly warm. I have to check other chips first so I can differentiate the temp.

 

All the measures I've did with power supply AC. Not batteries.

 

With these checks I think screen is not receiving data because 74H164 and 74HC4040 is not sending these data. These chips only works if they receive a reset signal from U1 (Mickey/Hayato) but I don't have an oscilloscope to check if this chip is generating the reset signal at the beginning. Is there a way to check if U1 is dead ? (Maybe for this reason I have this distorted sound too). Are my conclusions correct? Or maybe I have a chance doing another test?

 

Thanks a lot for all the help!!

 

[oski_98]

I recommend spending some time carefully searching through and reading previous posts on this forum. There's a lot of information that should be helpful but it's not all in one place.

 

Does your Lynx have a Mikey or Hayato chip? I have some rather long-winded posts here and here on the topic of testing Mikey, that should also apply to Hayato. TLDR you'll need an oscilloscope or at least a logic probe (cheaper alternative).

 

The other ICs you mentioned are cheap and readily available, it's probably easiest to just replace them if you suspect they are defective.

 

It pains me to say it, but you may have actually inflicted more damage with the "headphone trick". It's a horrible hack that subjects the ICs to a voltage that is much higher than they are built to withstand. I really wish there was a way to delete all posts that recommend doing this.

 

Anyway, here's an educated guess given the information you've provided, a "blank" screen is often (not always) a sign of an issue with Suzy. Best case scenario, there's a connectivity issue between Mikey and Suzy. Worst case Suzy needs to be replaced. I recommend testing continuity to all Suzy pins and the components they should be connected to and/or reflowing all the solder joints.

 

Another idea is that I believe you can short pins 31 & 33 (please search other posts to verify this) of the game cart to force the "insert game" message. If you're not able to get that, I would bet that the "headphone trick" that worked for a bit also maybe inflicted further damage.

[Rollinpizza]

Hi Oski_98!! Thanks a lot for the information and your posts.

 

I answer by parts  

• I've looked the revision of the motherboard and I have a Hayato chip.
• Looking one your posts, I've seen it happens me the same. In TP1 there is 0'4V. So I think that the problem is in Suzy or Hayato. But I don't have an oscilloscope to test it fine and I can't afford one. I've tested the continuity to all Suzy pins and the components and it's ok.

Yes, I had shorted pins 31 & 33 and, yes, before it showed the "insert game" message (with the headphone trick) but after fix the supply stage, I get the black blank screen. The headphone trick I've used twice and not more. Anyway it's a pity to know now that this trick is dangerous  and I understand you perfectly.

 

If Suzy or Hayato is broken, my main problem is where to buy a new replacement because I'm from Spain and to buy in Best Electronics is a pain due to the taxes and shipment (almost 40$ plus customs). I don't know more places... maybe I'm thinking to make a lynx pi if Suzy replacement is more expensive than build a lynx pi.

 

 

[oski_98]

I would check a few more things before assuming both Hayato and Suzy are fried. It's possible you fried some other cheaper component of the power circuit before any permanent damage was inflicted to the more expensive ICs.

 

I would first try bypassing the power circuit completely. You can do this by supplying 5V (e.g. from a spliced USB cable) directly to C41 (the largest/470uF capacitor). Make sure you get the polarity right with the ground (usually black) wire to the - cap terminal and VCC (usually red) wire to the + terminal. With cart pin 31 & 33 shorted see if you get "Insert Game". The reason why I recommend this, is because it is both safe (5V should not cause any further damage) and it completely bypasses the power circuit (so if it works we know the issue is just a power circuit issue and not the expensive ICs).

 

Also, do you have a camera that can take pictures with good detail? Could you upload some pictures of your board? Specifically, pictures of the power components you replaced, the power circuit in general (U6, Q12/11 MOSFET, Q8, Q7, D13, etc.) I'm happy to take a look and let you know if I spot anything wrong.

 

[+Stephen Moss]

13 hours ago, Rollinpizza said:

In TP1 there is 0'4V. So I think that the problem is in Suzy or Hayato.

According the the schematic I have TP1 is the Reset pin, soassuming you are measuring correctly a voltage of 0.4 volts would keep Mikey/Hayato/Suzy in Reset and so if that is the case the system would not do anything.

 

Check that you have the 5V supply running, i.e. measuring across the GNC & VCC pins of Mikey/Hayato. If that is correct then measure the voltage across R1 which is what pulls the Reset signal high (or D1), if you measure 5V that would suggest the capacitor C3 is short circuit, if you measure 0V then either R1 is open circuit (power off and measure resistance) or is open circuit at one end, try checking continuity to VCC pin of Mikey/Hayato and to C3.

Although, if the problem is a bad R1 solder joint then the pressure from the probes when continuity checking could temporarily remake the joint causing it to look Ok so maybe just try adding a little extra solder to the joints first and see if that cures it, then check continuity.  

[Rollinpizza]

@oski_98

 

Maybe you can surprise but I don't have an extra usb cable to prepare a splice. I moved one year ago and I threw away all the extra cables that I had. When I go to parents home, I'll take a look.

 

I've attached some photos.

 

 

 

 

@Stephen Moss

 

I've measured TP1 with the COM probe in TP16. Anyway I've measured the vcc pins in Hayato (COM in Pin61) and all the pins is ok (4'9V) except pin 6 (VDDI) that measures 0V. I'm searching reasons looking the schematic :

 

 

But without success. I've tried:

• Reflow the Hayato pins (from 1 to 6).
• I stared that pin 6 is connected to C83 but I don't see this in the schematic. I've reflow the C83 joints.

 

I can't find the Hayato Datasheet... so I don't know the usage of pin6 but it's not normal.

 

Tomorrow I'll test C83 out of the circuit to see if it's default or not. I hope my multimeter could do it

Edited March 4, 2022 by Rollinpizza
Organise photos. Thinking tests

[oski_98]

WRT TP1, I'm not sure if TP16 is the right ground point but I'm pretty sure you can just use the negative terminal of C3 as your ground point to measure the voltage on the reset pin. R1, D1, and C3 form a power-on-reset circuit that delays when the Hayato RESET pin goes high long enough for other components to "get ready". When C3 is charged the pin will go high, so you can just measure the voltage across C3. R1 provides resistance to delay charging of C3 and D1 provides a path for discharge on power-off.

 

Actually, if you look at the board carefully there are many ground points (small vias onto the ground plane) for using an oscilloscope probe with a ground clip. There's one to the right of R11 and one immediately to the right of pin 1 of Hayato. You could use either off those as grounds as well.

 

+1 to what @Stephen Moss said, if you can confirm that C3 is a short then the RESET pin will never go high and this Hayato will never "turn on". You should be able to do this with a continuity tester. You should get a short beep across C3 but not a solid tone.

 

BTW if you can't afford an oscilloscope, you can still troubleshoot a lot with a logic probe, and that only costs $10. If you do get a logic probe I can help suggest what pins to probe for activity.

 

I'll spend more time this evening looking at your pictures and let you know if I see anything out of place.

[oski_98]

I think you're counting the pins wrong because pin 6 does not connect to C83, that's actually pin 15. For reference, pin 1 is labeled by the "1" that's by the bottom middle of the chip. It's next to the right side of D1.

 

Also WRT to Hayato pin 6, it seems like the traces on the board don't match the schematic exactly. The schematic shows pin 6 connected to a 5V source and and an inductor. But if you look at the traces on the board that pin connects to two "104" SMD resistors. This is unlike some of the other pins connected to a power source. For example, if you look at pins 18, 29, 49 they connect to an SMD inductor which matches the schematic. I guess what I'm trying to say is that I wouldn't trust the schematic implicitly, there were multiple board revisions and does anyone really know the source of these schematics? Instead I would try to visually follow the traces and use the continuity tester on your multimeter to figure out what's connecting to that pin. But do make sure you're testing the right pins

 

 

[oski_98]

A few observations from staring at the pictures.

 

* Nothing looks obviously wrong with Q12, Q8, Q7, and D13 though I can't quite tell the part number for Q12.

* What happened to C58? It's gone and there's just two wires, what are they connected to?

* It looks like the electrolytic capacitors were replaced? If this is true, I would double check the polarity and specs for every one.

 

Finally what kind of multimeter are you using to test voltage and on what setting? Make sure you're not using a DMM in any auto-sensing low-Z mode as that will not provide accurate measurements for this application.

[Rollinpizza]

19 hours ago, oski_98 said:

I think you're counting the pins wrong because pin 6 does not connect to C83, that's actually pin 15. For reference, pin 1 is labeled by the "1" that's by the bottom middle of the chip. It's next to the right side of D1.

 

Also WRT to Hayato pin 6, it seems like the traces on the board don't match the schematic exactly. The schematic shows pin 6 connected to a 5V source and and an inductor. But if you look at the traces on the board that pin connects to two "104" SMD resistors. This is unlike some of the other pins connected to a power source. For example, if you look at pins 18, 29, 49 they connect to an SMD inductor which matches the schematic. I guess what I'm trying to say is that I wouldn't trust the schematic implicitly, there were multiple board revisions and does anyone really know the source of these schematics? Instead I would try to visually follow the traces and use the continuity tester on your multimeter to figure out what's connecting to that pin. But do make sure you're testing the right pins

 

 

Uff!! What a mistake... I didn't realize... I thought this number was related with other thing of the board. Ok! I've measured the "true" pin 6 and measures 4'9V like others. And, yes, I really understand you and you're right. I can't get the schematic as an absolute reference.

 

I've measured againt TP1 with COM in negative terminal of C3 and measures 4'9V. I've also test it with COM in Hayato pin 61 or the ground point as you suggested. The measure is the same. So I think the C3 is working properly. I measured R1 without powering on the console and the value is between 98K-99K. What it means? Maybe Hayato is sending always reset?

 

Quote

* What happened to C58? It's gone and there's just two wires, what are they connected to?

* It looks like the electrolytic capacitors were replaced? If this is true, I would double check the polarity and specs for every one.

 

Finally what kind of multimeter are you using to test voltage and on what setting? Make sure you're not using a DMM in any auto-sensing low-Z mode as that will not provide accurate measurements for this application.

About the questions. C58 is in these two wires. I've put this wires because in the board I didn't have enough space. The electrolytic capacitors I've replaced myself and I am sure it's ok. Anyway I'm gonna check it just in case  . My multimeter is a Velleman DVM890 (https://www.velleman.eu/products/view/?id=18686). I didn't find any auto-sensing low-Z mode.

 

Edit:

 

I've checked the polarity and specs and it's ok. 

Edited March 6, 2022 by Rollinpizza
Add more info. Capacitors Check.

[oski_98]

Hayato doesn't "send" reset. Other circuits can use the reset pin to turn processing on/off. Setting that pin low turns Hayato off setting it high (e.g. 4.9V) turns it on. R1 and C3 just delay when Hayato turns on after you press the Lyn on button.

 

In other words, it seems like there's nothing wrong with the reset circuit or the power to Hayato.

 

You should also check the voltage of the Suzy VCC pins as well.

 

If there aren't any obvious signs of a short such as warm chips or VCC pins pulled low. I think you'll have to at least buy a logic probe to test whether the ICs are working properly. Like I said earlier, they are cheap $10-30.

Edited March 6, 2022 by oski_98
typo

[Rollinpizza]

Hi! I'll check Vcc pins from Suzy and is well supplied (4'9V).

 

At the beginning, Hayato is slightly warm but after is ok. I've measured Data and Addresses Pins from Hayato (with Com in one of the pins with Gnd) and at least I can see values changing. I've made the same exercise in Suzy and memories with the same results. So, at least, there is a signal. I would like to check the xtal but with my multimeter I know is difficult. Anyway, if there is signal and values changing, I assume that Xtal is working.

 

I'm looking a logic probes in amazon. @oski_98 do you know a good brand or a good model?

 

Thanks a lot for all the help!!

[oski_98]

You don't need anything fancy. The Elenco LP-560 is cheap and plentiful. There's some posts here on how to use one.

 

Generally what I look for is activity on the RAS/CAS of the memory chips. I've found that to be an indicator that both Mikey and Suzy are working.

 

I would then check the LCD header. I have a post on this forums with a table of the measurements I took on my board.

 

Not sure what to conclude from Hayato being slightly warm at startup I've never noticed any warmth at all from Mikey at startup.

 

A few other thoughts.

1) It's easy enough to check the VCC of the RAM and other ICs on the board. I would do that as well. If the data and address buses are active maybe it's possible that the DRAM chips are faulty and not responding appropriately. This is just a theory, I've never experienced this, nor am I knowledgeable enough of those aspects of the circuit to say so with any certainty.

2) Where did you get the capacitor replacement kit? Are you sure the voltage rating as well as the capacitance for each capacitor are suitable?

 

[Rollinpizza]

Thanks Oski!

 

I've not found the elenco model here in Spain but I've found this one (https://www.amazon.es/Laser-5263-Dispositivo-diagnóstico-reparación/dp/B006IRIP8I/ref=sr_1_9?__mk_es_ES=ÅMÅŽÕÑ&crid=DYNXHV18GYHC&keywords=sonda+logica&qid=1646681208&sprefix=sonda+logica%2Caps%2C91&sr=8-9#customerReviews) that is very similar in features. I've found this one too (https://www.cetronic.es/sqlcommerce/disenos/plantilla1/seccion/producto/DetalleProducto.jsp?idIdioma=&idTienda=93&codProducto=343067005&cPath=455) but I think is too much so that I need it. For the moment, I choose the amazon option.

 

About your thoughs, yes I'm gonna check all your comment and see its behaviour. The capacitors replacement, the voltage rating is ok, don't worry. That's not the first recap I do . I'll post the results soon.

[oski_98]

WRT the logic probe, I think the cheaper one you linked should work fine.

[Rollinpizza]

Hi!! Tomorrow arrives the logic probe. Meanwhile, I've tested all de transistors and it's all OK. Also I've tested all the ICs (UX) and I've found something strange here:

 

 

There is a short between Pin 2 (VCC) and Pin 4 (GND). So I think that the chip (U12) is broken because is not powered and maybe the transistors of the chip are shorted.

 

That explains that randomly there was noise or not. The outputs measure is almost 0v or 1V.

 

[oski_98]

That IC is an LM386 audio amplifier. A replacement part is probably less than $1. In fact I would just buy replacements for all of the ICs (except Mikey/Hayato and Suzy which are difficult to source) just in case. Probably won't cost you more than $10.

[Rollinpizza]

22 hours ago, oski_98 said:

That IC is an LM386 audio amplifier. A replacement part is probably less than $1. In fact I would just buy replacements for all of the ICs (except Mikey/Hayato and Suzy which are difficult to source) just in case. Probably won't cost you more than $10.

Yes, maybe it could be a good alternative.

 

Ok, I've measured with the logic probe on CAS and RAS pins from the RAMs and the results are: 

• CAS: HI : Bad feelings because I expected a signal and not a hight state.
• RAS: HI and LO: That seems OK.

I've tracked these signals and I've stated that Suzy and Hayato/Mickey (Mine: C301719-001) are generating in its CAS pins a High State and in its RAS pins a signal (HI&LO). In this scenario (Case 1) the speaker is doing a noise/buzz. Sometimes in this case, after a time, suddenly the speaker is popping. In this period RAS is in HI but after LO and after comes the signal (HI&LO). When the signal comes, the speaker is doing a noise again.

 

Sometimes when I've powered the lynx, I've found randomly RAS pins are in HI  and the CAS pins too. The speaker in this case doesn't have sound (Case 2).

 

With these results, @oski_98 , what dou you think? Maybe is Suzy and Hayato/Mickey broken? O could it be the xtal component? Is there a way to test the xtal?

 

As always, thanks!!

Edited March 13, 2022 by Rollinpizza
Correct a mistake.

[oski_98]

If the RAS is showing HI&LO that's probably a good sign that Hayato is fine. So maybe the issue is Suzy or the DRAM.

 

I would immediately replace the audio amplifier. Reason why: if you hear audio with a game cart inserted that means both chips are working. So I would replace the audio amplifier insert a working game cart and listen. If you hear sound then do a little dance because that means Hayato and Suzy are OK.

 

If in that situation you don't hear audio I would replace the DRAM first, since that is an easy and cheap part to find.

 

Also, what about the LCD header? Have you tried probing the clock and data lines?

 

 

[oski_98]

Also if you're testing with pin 31&33 of the game cart port jumpered you might instead want to test RAS with a game cart inserted. It should be always HI&LO.

[Rollinpizza]

Ok! I've measured some TP points related with LCD  (TP22,TP23) and the LCD header with logic probe. In addition to this, I've measured more point that I think can help to diagnose what happens with my atari lynx. Let's go!

 

TP 22 (LCD_VDD) = 1'61V (Reading from other posts, a correct value would be between 5'6V and 11V)

TP 23 (LCD_VSS) = 0'79V (Reading from other posts, a correct value would be between -3'3V and -6V)

 

Other Related points:

 

TP12 = between -0'13 and 0'13V (Reading from other posts, a correct value would be between 5V and 10'7V)

 

With this values, we can see the LCD is not well supplied. There is not enough voltage.

 

LCD Header measured with logic probe:

 

Pins with HI Value: 1,2,4,5,6,7,11,14,15,21

Pins with LO Value: 3,9,12,13,16,17,18,19,20,25,26

Pins with No Value: 8,10,22,23,24 (Here the logic probe didn't do nothing)

 

I've measured more TP points to track where is the problem and I've found interesting this:

 

TP27 = 5V

TP26 (DBL) = 0V

TP25 =1'9V

TP24 = 1'41V

TP21 = 9V (Here maybe the COM is not the correct (Minus C23)) 

 

The TP26 is 0V and is the pin 63 (DBL) from U1. Without voltage here LCD_VSS and LCD_VDD can't be generated properly.  

 

 

[Rollinpizza]

On 3/14/2022 at 7:11 AM, oski_98 said:

Also if you're testing with pin 31&33 of the game cart port jumpered you might instead want to test RAS with a game cart inserted. It should be always HI&LO.

Yes, with a game cart inserted, there is always HI&LO. 

 

Next week will arrive me the audio amplifiers so that I can replace them.

[oski_98]

For reference, here's the measurements I took from a PAG-300 Lynx II (a more rare Lynx II variant with a Mikey instead of Hayato chip). I took these with an oscilloscope, and I also included the mean voltage computed by the oscilloscope. I think the mean voltage should match what you would expect to see with a digital multimeter.

 

The HI&LO on RAS when the game cart is inserted is a good sign that there's some activity on the ICs. However, I'm concerned about your measurements on the LCD pins. Especially that none are reading HI&LO.

• Pins 22-25 are for powering the backlight. Don't test these with the logic probe. The voltages are outside the range of what that device would expect. I don't think it would damage the logic probe, but just to be safe I wouldn't use the logic probe on this. Measure these with a DMM instead.
• Pins 17-20 are the data lines I would definitely expect HI&LO reading on those pins.
• Strange also that 14 is reading HI, I would expect 13 to read LO and 14 to read HI.
• Pins 3-7 are square waves I would expect HI&LO readings there as well.

Unfortunately, I'm not sure what the reading for all the test points should be on a Hayato Lynx II so I can't help much there. Maybe collecting that data would be a good future project. I do know that TP26 (DBL) on the PAG-300 Lynx II should NOT read 0V, it's a 50% duty cycle 3.15kHz square wave with a crest of 4.9V and a trough of 0V - it should read ~2.4V with a DMM.

 

What happens if you try to measure with the logic probe directly from the Hayato pins? See the pin layout for Hayato here, the data lines are pins 2-5 or Hayato, DBL is pin 63, the display clock signals are pins 13, 15, 16. See if you get the same measurements with the logic probe if you take the measurement directly from those pins.

[Rollinpizza]

I've measured de LCD header with DDM and here are the results:

 

• Pins with 4'9V: 1,2,4,5,6,7,11,14,15,21,
• Pin 8: 1'22V
• Pin 10:1'90V
• Pin with 1'60V: 22,23
• Pin 24: 0'78V
• Pin 25: 0'53V
• Pins with 0V: 3,9,12,13,16,17,18,19,20,26

 

Comparing to the measurements of the post, there is a lot of pins that the values are wrong. Especially the power supply where the voltages are too low. I think that the main reason is DBL signal that is 0V and should be a square wave. Then, there is the clk pins that is not a clock signal and the origin os these signals come from Hayato.

 

I've tried to measure with the logic probe directly from Hayato pins and... well, there is pins that generates a signal correctly but others not. Here is the results:

 

• Pins with HI: 6,12,13,16,12,18,22,23,28,29,30,32-38,49,50,52,54,57,62,64,65
• Pins with LO: 1-5,9,11,14,15,19-21,24-27,31,39,48,55,56,59,61,63
• Pins with HILO: 7,8,10,17,51,53,66-68.
• Pins with No Value (DMM Measure): 40-41 (Aprox 1V),42 (2V),43(Aprox 1V),44(2'25V),45(1'27V),46-47(Aprox 1V),58(-0'01V to 0'01V),60(1'94)

 

I've found something strange in Hayato... every time I've power up the console to measure, I find differents behaviours:

 

• Speaker poping: Almost all the pins are in HI or LO.
• Speaker making noise: There is more activity. RAS in HILO.

 

I'll wait fot the arrival of low amplifieres and see results but if the results are the same, maybe I'll consider to save to buy a Hayato CPU. 

 

Edited March 20, 2022 by Rollinpizza
Add more info

[oski_98]

It seems unlikely that Hayato is the issue given the level of activity you are describing. I think it's strange that the chip would fail in such a targeted way rather than fail catastrophically.

 

At this point I'd wait and see what happens after you replace the audio amplifier.

 

The random behavior you are describing could be due to the audio amp chip being faulty or possibly a cold solder joint that is sometimes not properly conducting. You could test continuity between components and leads and their pads or nearby vias. Alternatively just reflow every solder joint. Exhausting work to be sure but it's an option to consider.