Jump to content
IGNORED

The "perfect" Lynx....


Level42

Recommended Posts

Well, the big problem is what the draw at idle on the powerboost will be. Most likely it will be around 5ma, so just sitting there with the lynx "off" it would be sucking out 5ma, so after a week of just idling it would probably drain the battery pretty significantly.

 

Ideally it would be nice to use the soft power switch somehow without the draw.

REmove the Lynx cart and the battery will not drain at all because the power rail goes through the card connector. It's what I do with my (unmodded) Lynx 1 since I store fresh batteries inside the Lynx. But if you plan on keeping the flash cart in there permanently, that would be an issue I think.

  • Like 1
Link to comment
Share on other sites

Or, I just figured out an alternative! Gut the Lynx's power and start button circuitry.

 

Tie both the On and Off button in parallel to a LTC2950 pushbutton controller chip, which is tied to the enable pin on the powerboost. The pushbutton controller IC uses 20uA of power and when you push the On button, it would send the enable to the regulator, firing it up. Then if you hold the On or Off button down for a second, it powers the whole thing off. Now you've saved your buttons, and power too.

 

http://www.linear.com/product/LTC2950

 

I may have to try this!

 

Yeah, I thought about a simple transistor set-up to reach the same effect but that would mean yet another hack ;)

So would this be...but it looks promising. However, I'm not clear what the EN output does, will it provide a voltage ? Because the Powerboost1000c Enable input wants a short to ground AFAIK.

Link to comment
Share on other sites

Well I just ordered this battery:

 

http://www.aliexpress.com/snapshot/7983567168.html?orderId=77521877040311

 

The sizes are 3,8x73x125 mm. This should well fit even with the original screen.

 

No idea about the quality but somehow the well-known brand battery makers simply don't seem to bother with Lipo "tablet" like batteries ? At least I couldn't find any....so we'll see how it goed. At least this one has a protection PCB built-in.

 

I figured I could go with something crazy like a 10000mAH battery but who needs a Lynx that can run over 24 hours without charging ?
OK it would be fun but there would be some disadvantages:

 

- weight. I saw one of the 10000mAh batteries weighed 200 grams. This would be quite a bit more than what a set of 6 AA Nimh batteries would weigh (6x26 grams= 156 grams) and remember that my speaker is already a lot heavier, plus there will also be the Powerboost board, although that is not much....but still.

- charging time. The bigger the battery, the longer the charging time....simple. Of course you don't have to do full charges/decharges on a lipo, but still...

- size. Even if it would fit JUST right...why bother, why not keep some nice space for air to move a bit.....

 

So I figured 5500mAh would be plenty.

It will take a couple of weeks to arrive with Aliexpress as I have experienced with all my orders so I still have plenty of time to get the Powerboost 1000C and a nice fat wall-charger with micro USB connector. And some time to figure out how I will fit he Powerboost inside. Would be nice if I can use the micro-USB port directly.

 

It would be cool to add a multi-colored LED somewhere to replace the original LEDs on the Powerboost board, so that the status of that one (power-on/ power low/ charging etc.) is visible at all times. Don't think I'd put it on the front of the Lynx though but f.i. at the top.

 

About the brightness/contrast pot: remember I'm planning to use that one (or at least the spot and wheel) for a pot to control the clock frequency mod :D

 

Pic of battery:

UT8rBxDXNBaXXagOFbXi.jpg

Edited by Level42
Link to comment
Share on other sites

Nice! Can't wait to see it all come together. If I really had some time, I'd just design my own PCB/circuit that has the boost controller, charger, and button controller all on the same thing. I was thinking about doing things cheaper and just chaining together a charger and a booster, but the big issue is dealing with using the battery while charging. I imagine if you don't design things right, it could be very bad for the battery. Which is why I figure the powerboost circuit is tried and true, and not too bad..a little expensive, but doable.

  • Like 1
Link to comment
Share on other sites

Yep it's a very unique concept and I agree it's rather expensive but exactly what is needed. I'm a bit surprised they actually put the schematics up for download which leaves it very open for copies from the Far East....

 

But I guess this is the open-source mentality. There's even a drawing of the PCB lay out.

Link to comment
Share on other sites

......However, I'm not clear what the EN output does, will it provide a voltage ? Because the Powerboost1000c Enable input wants a short to ground AFAIK.

Just checked the data sheet of the LTC2950 and according to the schematic the enable output is a MOSFET that switches to ground so yes that has to work for sure.

Link to comment
Share on other sites

LOL, this LTC2950 idea seems to be getting better and better:

 

I've been studying the LTC2950 data sheet a bit more and I think it should be possible to connect the Lynxes' Power on signal to the /KILL input of the LTC2950. (Maybe even directly). The kill input is used to give the microprocessor controlled device that is powered by the whole set-up a way to shut down the power neatly......well there you have exactly what Power On signal does on the Lynx ! :D : D

 

And it should also be possible to hook up the 5V coming in from the micro-USB input on the Powerboost 1000c as as signal to hook up to the AC "connected" input of the Lynx......now doesn't all of that sound sweet.....it would still shut off the Lynx after 4 minutes when no charger is connected or after an hour when there _is_ a charger connected......Id really love it if the Lynx could still control itself like this even with a brand new power supply section :):):)

 

I'm trying if they will ship me 2 samples of the LTC2950 :D

Edited by Level42
Link to comment
Share on other sites

Looks like I'll need to invert the power-on signal before connecting it to the /KILL input:

 

http://atariage.com/forums/topic/235605-atari-lynx-ii-schematics-for-power/?p=3569571

 

I bet I will be able to use one of the inverters that is "left over" on U6 after I've totally butchered the original power supply section, in fact....if I short the emitter and collector of (former) Q7 there is already a path to the hex inverter and I can pick up the inverted power on signal from pin 12 of U6 :D

post-25272-0-78484400-1470936484_thumb.jpeg

Edited by Level42
Link to comment
Share on other sites

I'm a bit in doubt as to which kind of multi-color LED to use for the Powerboost1000c.

 

O by the way, the Powerboost 1000c actually has four LEDs:

 

 

  • Blue indicates the 5V output power state.
  • Red indicates when the battery voltage is below 3.2VDC (Low Battery Output)
  • Yellow indicates when the battery is being charged
  • Green indicates when the battery is done charging (all full)

 

Of course the RGB LED has only three colors, Red Green and Blue...

 

Now I think that the original Blue LED is the least interesting one. After all, I can know there is 5V output OK if I turn on the Lynx and it works...

The other three are much more interesting in daily use.

 

Of course I could make some fancy combinations but I don't think I want weird combined colors as the primary colors are best related to the status.

 

However, instead of Yellow the RGB LED will have Blue color, so it would be blue while charging. Not sure If I like that. I could activate both the green and red pin to get yellow instead but maybe I'll just go with the blue :)

 

Now I'm not sure if should go with a traditional 5 mm LED or a smaller SMD one so I ordered both.

 

5mm RGB LED:

http://nl.aliexpress.com/item/10pcs-5mm-RGB-LED-Common-Cathode-Tri-Color-Emitting-Diodes-f5-RGB-Diffused/32666275072.html?spm=2114.010208.3.1.DOLl4w&ws_ab_test=searchweb201556_0,searchweb201602_5_10057_10056_10055_10037_10049_10033_10059_10032_10058_10017_10060_10061_10062_10064,searchweb201603_3&btsid=658c9c4e-1481-47a6-ada4-d1132d2da178

 

SMD RGB LED on a tiny PCB:

http://nl.aliexpress.com/item/RGB-3-Color-Full-Color-LED-SMD-Module-For-AVR-PIC/32332024290.html?spm=2114.010208.3.301.2RSDe6&ws_ab_test=searchweb201556_0,searchweb201602_5_10057_10056_10055_10037_10049_10033_10059_10032_10058_10017_10060_10061_10062_10064,searchweb201603_3&btsid=658c9c4e-1481-47a6-ada4-d1132d2da178

 

I'll solder the connector on the other side if I'll use the SMD one. Not sure where to put it yet.

Link to comment
Share on other sites

 

EDIT: I noticed in your photo the foam protrudes beyond the speaker plate, common in front mounted speakers. If the foam is pressed against the Lynx case, this will severely impede movement of the cone muffling the sound.

 

 

@Kosmic Stardust: yes I worried about this and I'm pretty sure that was actually the case. So I decided to also cut-away the "final" speaker support on the bottom of the case. Now the speaker sits deep enough for the foam to be able to move. The speaker isn't really fixed now, but it seems to stay happily in it's place because of the little foam, the PCB and the rest of the case.

 

Sound improved a LOT now !!! The bass is simply amazing. The high tones are OK and I fear if I would enlarge the holes the bass would suffer again, because when I remove the top cover of the Lynx, the deep sound is gone right away. Of course this is the way a speaker box works as you explained much better than me. So: I'm happy with the sound, very happy, this was worth the money and effort. It's not exactly a "drop in" replacement but very doable IMHO.

  • Like 1
Link to comment
Share on other sites

Next step....I wanted to have a look how the Lynx pots look like so I removed the dial of VR1 (Brightness) as this will be my target to use for adjusting the clock frequency.

 

This pot has a value of 50kOhm which is probably already better than the temporary test-pot I used which is 20k. So I might be able to actually keep the current pot there which would be nice.

Even nicer is that I discovered that the casing of the pot is soldered with two pins on the side of the pot case to the PCB (it looks like they are connected to the +5V plane) but the actual pot connections are separate.

This is great because it means I can unsolder the three pins and bent them up so they're out of the circuit, while I can still keep the 2 solderings that are only there to keep the pot on the PCB :)

 

See this picture:

 

post-25272-0-85057400-1470950000_thumb.jpg

Link to comment
Share on other sites

 

 

@Kosmic Stardust: yes I worried about this and I'm pretty sure that was actually the case. So I decided to also cut-away the "final" speaker support on the bottom of the case. Now the speaker sits deep enough for the foam to be able to move. The speaker isn't really fixed now, but it seems to stay happily in it's place because of the little foam, the PCB and the rest of the case.

 

Sound improved a LOT now !!! The bass is simply amazing. The high tones are OK and I fear if I would enlarge the holes the bass would suffer again, because when I remove the top cover of the Lynx, the deep sound is gone right away. Of course this is the way a speaker box works as you explained much better than me. So: I'm happy with the sound, very happy, this was worth the money and effort. It's not exactly a "drop in" replacement but very doable IMHO.

Glad I could help.

 

Man you are really frankensteining the hell out of that thing. I'm a bit concerned though if the LCD screen mod will tolerate your efforts to overclock the system. I imagine the mod has it's own clock. If not, you might be outputting 90Hz instead of 60Hz!

Link to comment
Share on other sites

Yes love this forum, even though there only a couple of is into hardware hacking there is some great info from you and Cowdog which really helped ! :)

 

Haha, yes maybe I love modding stuff more than actually playing the games ;) ;) ;)

 

Yes I have my doubts about that too. For sure McWill's screen has it's own clock (you can see it on the PCB) and I am not knowledgeable enough to know how the video built-up of the Lynx is. I do know that the actual screen refresh is variable so I have a tiny bit of hope there. But seeing that SainT's SD card also doesn't seem to tolerate any overclocking (well, the menu doesn't start, the games run fine if I first power up and select the game at regular 16Mhz) it might be a slim chance.

Obviously whatever the Lynx puts out natively to the screen has to be processed quite a bit by McWill's LCD, those chips aren't there for nothing.....but we'll see. I'll have to make a choice either way for sure.

 

The good thing is that I can adjust the overclocking so it won't be 20 or 24 Mhz right away so I can slowly turn it up and see how McWill's LCD reacts....

Edited by Level42
Link to comment
Share on other sites

 

 

The good thing is that I can adjust the overclocking so it won't be 20 or 24 Mhz right away so I can slowly turn it up and see how McWill's LCD reacts....

Assuming there's no feedback, you maybe could crank it until the screen artifacts or completely glitches out, then tone it down again until the picture resolves. Assuming you also have the capability of underclock, that could desync things as well.

 

Is the Lynx screen truly variable frame rate, or does the CPU just push stuff to the screen whenever it feels ready to? I believe Game Boy had a pretty solid 60Hz refresh so the SNES Super Game Boy was only off a few clock cycles syncing to SNES refresh rate.

Link to comment
Share on other sites

This is what I know from Wikipedia:

 

  • Variable frame rate (up to 75 frames/second)

 

And found this about programming it with CC65

 

The supported framerates are 50, 60 and 75 frames per second. Actually there is no real reason to use anything else than 75 frames per second.

Link to comment
Share on other sites

This is what I know from Wikipedia:

 

  • Variable frame rate (up to 75 frames/second)

 

And found this about programming it with CC65

 

The supported framerates are 50, 60 and 75 frames per second. Actually there is no real reason to use anything else than 75 frames per second.

 

True.

 

The Lynx uses a dual buffered screen system in games. This means that it continuously displays buffer 1 at 75 frames per second while the software is drawing on buffer 2. Once everything is drawn on buffer 2 the software tells the interrupt routine that it wants to swap the buffers. At the next VBL interrupt the interrupt will make buffer 2 the display buffer and buffer 1 the drawing buffer.

 

This means that the software does only update the screen on every 10th frame or so. It depends on the complexity and the quality of the game. The only advantage of 75 Hz is that the VBL interrupts come more frequently. So the waiting time is shorter. But there is no way that the game would advance 75 frames per second. You are lucky to get 12 frames per second. I know that there are exceptions like Zaku. But that is seriously good coding.

Edited by karri
  • Like 1
Link to comment
Share on other sites

Great info, thanks ! But what this will translate to for McWill's LCD will be a matter of trial and error I guess.....

Maybe I should over lock the LCD proportionally to the Lynx's clock ? ;););)

 

On another note....I found some nice info about the LM386 here....

http://www.circuitbasics.com/build-a-great-sounding-audio-amplifier-with-bass-boost-from-the-lm386/

Link to comment
Share on other sites

Comparing a bit how the designers set up the LM386. Clearly they included the bass-boost with the cap and 10k resistor between pin 1 and 5.

 

They didn't put any components between pin 1 and 8 which, according to the datasheet means the gain is set at 20. By adding a pot and cap here the gain becomes adjustable and could go up to 200......might be fun to experiment a bit with that.....my new speaker clearly "eats" more energy from the output because of the much heavier magnet.....plus it's only 4 ohms....

  • Like 1
Link to comment
Share on other sites

Checking the schematics for what voltage the "AC connected" input can handle.

 

Normally when the power supply is connected 9V (or whatever the AC adaptor puts out) is divided by two resistors R69 (100k) and R73 (390k). If I calculate this there is about 7.1 voltage going to the "AC connected" input of MIKEY (pin 14 "color").

 

Isn't this a bit high voltage ? Has anyone measured this with an original Lynx supply connected ? I don't have one around....

Link to comment
Share on other sites

Checking the schematics for what voltage the "AC connected" input can handle.

 

Normally when the power supply is connected 9V (or whatever the AC adaptor puts out) is divided by two resistors R69 (100k) and R73 (390k). If I calculate this there is about 7.1 voltage going to the "AC connected" input of MIKEY (pin 14 "color").

 

Isn't this a bit high voltage ? Has anyone measured this with an original Lynx supply connected ? I don't have one around....

 

Never mind this......I forgot that the Lynx chip-set is CMOS......sorry, since I've done so much work on classic arcade games I am "always in TTL-mode" ;) ;) ;)

Anyway, makes perfect sense now....

Link to comment
Share on other sites

Thanks for posting this! I was just thinking about doing the same things (except for overclocking), so I'm following closely. I've decided to go with a Model 2 though, so I'm wondering if your speaker and battery solution will work for it. Waiting for the Lynx to arrive now, so I will let you know if I make any progress coming up.

Edited by pixelperfect
  • Like 1
Link to comment
Share on other sites

This is more or less a note to myself:

 

I think I mentioned on one of my earlier postings that the [AC connected] should be connected to the INT signal of the LTC2950, but this is obviously wrong. The INT output of the LTC2950 is meant to let a microprocessor controlled device do a controlled (and usually pretty slow) shut-down....like a computer shutting down....that's not what we would ever need on the Lynx. Simply turning off the EN pin is good enough and that's what the LTC2950 already does.

 

Now this should be correct:

 

- The /KILL input of the LTC2950 will need to be connected to the inverted [Power On] output of the Lynx.

(The Lynx will need to turn the [Power On] signal high within 512ms (so about half a second ) or else the LTC2950 will turn the EN off again....I think 512ms is plenty for the Lynx to do that.)

 

- The output labeled "USB" of the Powerboost 1000C has to be connected to the [AC connected] input of the Lynx. There will be +5V on this pin when there is +5V connected to the micro-USB input. So this will work exactly like when the original Power supply was connected on the Lynx 1.

Since there is only +5V instead of +9V coming from the Powerboost 1000C I will copy the "design"of the Lynx II for this, which has the [AC connected] signal permanently connected to +5V through a 100k resistor..but of course now the =5V is only there as long as there is +5V on the micro-USB input connector .......I guess that should work :D

  • Like 1
Link to comment
Share on other sites

Looks like a solid plan. I can't wait to get some more Lynx II's to try my hand at this as well. I have a powerboost 1000c already, and just need to buy a few LTC2950's.

 

What I'm seriously thinking of doing after this is all figured out is just taking the powerboost 1000c design and building a new PCB that has all the components sorted out, less LED's and make a simple to wire up option for others who want to do it.

 

Soon the Lithium Mod will be the ultimate replacement for the power supply woes!

 

Cuz there's just something about it.. so right.. (Couldn't resist)

  • Like 2
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...