Evaluating: Limeade Blast 18,000mAh battery inside an Atari 130XE

[jmccorm]

I'm playing around with putting a Limeade Blast 18,000mAh battery inside of an Atari 130XE's case. The battery itself fits almost perfectly inside of the top center portion of the case, with just the tiniest bit of bulging around the Help and Start buttons.You'd probably have to be looking to notice it.

 

Initially, I would decouple the external power connector of the 130XE to the motherboard. The power connector would go to the power input of the Limeade. The power output of the Limeade would go to the motherboard. Perhaps the power switch would be bypassed so that the charging circuit was always on (which the Limeade was designed to handle and auto-shutoff when charge is complete)?

 

Two apparent challenges:

1] The limeade will sleep if it isn't be used. You normally have to hit a side power button on the Limeade before using it. So the rocker power switch on the Atari would have to be re-purposed (or a new momentary switch added) to perform that function.

2] The limeade will not power the outputs while charging. So it would be ideal to be able to run off of external DC power while the internal battery is charging.

 

I'm pretty sure a simple circuit can solve problem #1. Problem #2 has me a bit concerned on how it would play out in the real world. I'd need a bigger input power supply going to the 130XE, that's probably solvable. But I wonder if I can get the 130XE to remain online while the power cord is pulled and have the Limeade Blast take over in time to keep the OS running.

 

My ideal design would make the magic under the hood totally transparent to the user so they can plug or unplug the external power connector as they wish and the power switch would work like you normally would want an Atari 130XE to operate. But I'm wondering how much trouble that is going to be.

 

Once the battery solution is complete, I'd want to embed an SIO2SD and a mini 2-3" LCD display into the case in order to make the system truly portable.

 

Thoughts?

[jmccorm]

I guess I'll use this forum post to blog my progress.

 

So far, I have temporarily jury-rigged the Limeade Blast to the power connector on the Atari 130XE with a spliced USB cable . (I understand that a 2-pin connection is not a good idea long-term, and I need to use all 6 or 7 pins to reduce resistance. I've seen this issue play out on Pole Position PCBs and it is not pretty.) As far as the load, the 130XE was expanded to 320k and has an SIO2SD attached. So it has a bit more draw than a stock machine.

 

Power from the battery appears to have drooped from 4.9VDC to 4.5VDC under load, which is a little bit of a disappointment (they rate it as capable of 2.1A at 5VDC). The Atari BASIC screen is green and there is a soft interference pattern in the background. A capacitor might filter out the soft background interference pattern. Or is it because I'm not using the shield connection? While the machine clearly isn't running to spec, it is running. Archon II, Stealth, and MULE loaded off of the SIO2SD very well and are playing just fine.

 

Assuming that the Limeade Blast can hold the output steady at 4.5VDC, anyone have any thoughts if the Atari can be adjusted out to work with it? (I thought there was a potentiometer on the motherboard for color correction.) Or should I go for a boost circuit which will cost some battery life but feed it at spec voltages? (If I adjust it out on the motherboard, it would have to be adjusted back when regular Atari power supply voltages are connected, so that doesn't sound like a great solution.)

 

So far I'm 30 minutes into a fully charged battery. The battery isn't generating any additional heat. It still shows all four power bars as full. I'll get a real number, but for now I'm guestimating/hoping that this may be able to run 12-15 hours on battery power. And that'll be awesome for a modern battery with built-in regulation that still fits inside the case.

[Rybags]

My stock 800XL PS only has positive going to 1 pin. Given it only really can peak around 1.5A it's not a problem.

 

Voltage dropping to 4.5 - that doesn't sound good assuming the pack isn't nearly drained. I'm not sure what the lower threshold is but you can almost guarantee that the dodgy video is because the voltage has dropped.

 

The colour circuit I believe uses voltage doubling in an early stage so would be a good indicator if voltage is out of spec. Not having shielding isn't really an issue, plenty of us run without the top half or all of it due to internal additions.

[Faicuai]

This is a GREAT idea, which I always had in mind (especially with mods like Ultimate and Incognito, which once running and replacing OEM ram & rom, *should* yield in overall lower power consumption, as long as your machine's internals have not been butchered with a disparate array of mods).

 

In any case, as soon as I started looking into this possibility, I immediately stumbled across the fact that (like in most DC-based applications), a DC-to-DC power stage/control will be required, either on-board the host machine, or along with the battery itself. As far as I recall from my (say) 800XL schematics, and from what I have seen from a dormant 130XE that I have around here, I could not detect DC-to-DC power regulation on-board the Atari... except a few capacitors here and there... unless I missed something important there.

 

Last but not least, internal space is not really an asset on the XL and XE families... there is very little rom in there, especially once you start adding piggy-back / drop-in mods in several places... even clearing the existing keyboard may be a problem in some cases (XL).

 

The 800, on the other hand, has ALL the space in the world, and a dedicated power-board that could be adapted for such purpose. Much better machine / platform for your invention (I have them all, BTW).

[jmccorm]

Thanks, Rybags. I'll worry a bit less about the connection. And thanks for explaining the connection with the shifted color palette.

 

I started with a full battery, and after 3 hours, it is still holding very steady at 4.5VDC. The Limeade has just gone town to 3 out of 4 bars (admittedly, that doesn't necessarily mean anything). I pulled out my multimeter and tested the draw from the modified Atari 130XE. It was a little surprising. At 4.5VDC, it was 730-760 mA, with the variation based on system activity. The SIO2SD appears to only account for 30 mA of the draw. I've got to watch it to ensure that it regulates the output voltage consistently as the battery fades.

 

I'm looking over at another project that was done in 2007, the Atari 1500XL portable by Beetle. The case was absolutely incredible and was a mashup of a 600XL, two 800XLs, a 1050, 1027, and a 1010. Absolutely crazy. He has 3300mAh NiMH cells in front, and managed to hit 2h 40m of uptime. (He had a custom power supply circuit to handle the batteries.) That huge screen had to have drawn a lot of power. He says the whole package was pulling 1.14A.

 

I really need to get myself a 3D printer. It would look less authentic, but if I get anywhere with this, it could make my project reproducible. But I don't think there is anything affordable out there right now with a build bed that can handle a 13.5" wide object and I'm not crazy about the idea of piecing together two smaller builds. Oh well, that's further down the road, if it happens.

[jmccorm]

In any case, as soon as I started looking into this possibility, I immediately stumbled across the fact that (like in most DC-based applications), a DC-to-DC power stage/control will be required, either on-board the host machine, or along with the battery itself. As far as I recall from my (say) 800XL schematics, and from what I have seen from a dormant 130XE that I have around here, I could not detect DC-to-DC power regulation on-board the Atari... except a few capacitors here and there... unless I missed something important there.

It looks like the output from the Limeade should be regulated at 5VDC. It is designed to charge cell phones and iPads over a USB cable. Either I'm doing something wrong, or it is a first generation bug with the output regulation. I'll work with their people to find out what is going on. If it can be made to work correctly, buying this battery off the shelf solves a lot of charging and output regulation issues. And that means I have something that is very easy for others to build for themselves.

The 800, on the other hand, has ALL the space in the world, and a dedicated power-board that could be adapted for such purpose. Much better machine / platform for your invention (I have them all, BTW).

I've got a couple of 800s. My main complaint is that they're not very portable. The 130XE was small and light (a reflection of Jack's cost cutting) and can easily be carried around or fit in a backpack. I enjoy reading the eBay listings for the Atari 130XE and the seller stating they they only have the "keyboard" and not the main computer. Of course, a new case that can hold the guts of an 800 without all the RF shielding could solve some of that. Edited June 10, 2013 by jmccorm

[BillC]

I enjoy reading the eBay listings for the Atari 130XE and the seller stating they they only have the "keyboard" and not the main computer. Of course, a new case that can hold the guts of an 800 without all the RF shielding could solve some of that.

I believe those are myatari/B&C listings for just a replacement 130XE case & keyboard. They state that it doesn't include the computer so people won't make the mistake of thinking it's a complete system, which is what it appears to be in the images.

[Rybags]

Problem if you do this with 400/800 is that internally they need +12V for the Ram unless you use some modern day replacement upgrade there. Also overall power draw would be somewhat more - the XL and subsequently XE have drastically reduced chip count which lowers the power requirements.

 

I'm no electronics guru but possibly what you'd want is to have an intermediate circuit that does a voltage double then regulates back to +5V.

 

Generally modern devices are tolerant of 4.5V as they often regulate the 5V down to about 3.3-3.5 anyway.

[jmccorm]

Problem if you do this with 400/800 is that internally they need +12V for the Ram unless you use some modern day replacement upgrade there. Also overall power draw would be somewhat more - the XL and subsequently XE have drastically reduced chip count which lowers the power requirements.

Thanks. You reminded me of the other reason why I didn't want to do this with a 400, 800. The 600XL/800XL/65XE/130XE/XEGS takes a 5VDC input. I wanted the simplest power solution I could use, and that meant working through the Atari's existing power input. (Aside from size/weight, that also ruled out the 1200XL because it also had a 9VAC power input, even though I want to think that it only used 5VDC on the motherboard.)

 

The Limeade battery pack went down to 1 of 4 bars in the past few minutes. I'm at the 12h 30m mark.

[jmccorm]

I'm no electronics guru but possibly what you'd want is to have an intermediate circuit that does a voltage double then regulates back to +5V.

 

The Limeade Blast battery pack already regulates its own charging (via USB) its own output (via USB). I hate to add a second regulator on top of an existing regulator because of the added inefficiency.

 

So, the Limeade Blast blinked the last of the four battery level indicators, powered down its 5VDC USB output, and turned off. The warmth, which was extremely mild, was in the top section of the battery pack where the output voltage is regulated. The system lasted 14 hours and 40 minutes on a single charge. Pretty good. Assuming the battery was up to spec, we didn't come close to the the theoretic 24 hour mark (18000mAh divided by 760mA) for two reasons. The first is that the 5V regulators inside the Limeade aren't 100% efficient. The second reason, I am assuming, is that the Limeade holds back a safe margin of power so that the battery is not completely depleted.

 

I guess I've got to figure out which way I'm going to take this project.

 

I could split open the Limeade Blast and adjust the voltage regulation circuity, but then it no longer meets my design goal of being a simple and easy project. I want to believe there is a potentiometer that I can tune inside the 130XE that will do color correction, but if I do that, then the 130XE would only be good for running on 4.5VDC. To run on outlet power, I might have to step down the voltage of the input power supply to bring it to 4.5VDC as well.

 

Inside the case with the Limeade Blast sitting on top of the expansion slot.

 

Case closed with the Limeade Blast sitting inside of the 130XE.

 

Back and top view of the battery. (The image actually shows the 13000mAh model.) USB input/outputs.

 

Front and side view of the battery.

 

The battery can be shifted to the left or to the right (up to the screw mounts in the Atari's plastic case cover) as needed. My original idea was to embed an SIO2SD in the right side of the case, and there is still plenty of room to do on the right hand side. I was also wanting to put a mini 2-3" LCD display on the top of the case. It looks like I'll have to embed a speaker (or two) in the case, too.

Edited June 10, 2013 by jmccorm

[+Larry]

It looks like the output from the Limeade should be regulated at 5VDC. It is designed to charge cell phones and iPads over a USB cable. Either I'm doing something wrong, or it is a first generation bug with the output regulation. I'll work with their people to find out what is going on. If it can be made to work correctly, buying this battery off the shelf solves a lot of charging and output regulation issues. And that means I have something that is very easy for others to build for themselves.

 

Neat device! Were you able to get any useful info from Limade on the voltage drop issue?

 

-Larry

[jmccorm]

Neat device! Were you able to get any useful info from Limade on the voltage drop issue?

I've submitted the question just now. I don't expect an immediate response because they've just fulfilled a lot of orders. I'll let you know.

[HiassofT]

Power from the battery appears to have drooped from 4.9VDC to 4.5VDC under load, which is a little bit of a disappointment (they rate it as capable of 2.1A at 5VDC).

The culprit might be the USB cable. A lot of people are having problems charging their tablets using third-party USB cables (see for example here).

 

Most USB cables are rated for max. 500mA and use quite thin wires, the cables shipped with tablets use thicker wires for the power supply connection and thus have lower resistance (and therefore a lower voltage drop / higher maximum current).

 

You could try butchering your USB cable (completely remove the original cable) and solder thicker wires directly to the USB A connector. Also check the voltage directly at the USB connector (the second USB connector on the Limeade might be separated, though, so measure at the butchered USB A connector) and compare it with the voltage in your Atari - the difference is the voltage drop of the cable.

 

so long,

 

Hias

[jmccorm]

Most USB cables are rated for max. 500mA and use quite thin wires, the cables shipped with tablets use thicker wires for the power supply connection and thus have lower resistance (and therefore a lower voltage drop / higher maximum current).

Sounds very plausable. I reduced my USB cable from 4' to 4" and alligator clipped the outputs from there. The screen looked to be standard Atari blue. Voltage was measured at 4.95VDC. You nailed it! The Atari 130XE is now running well off of the regulated output Limeade Blast.

 

There still appears to a semi-random pattern of very quick soft interference lines scattered throughout the video. I suspect the switching in the regulator is causing this and can probably be filtered out. Other than that, this thing is great.

[+Gemintronic]

My thought is to forget about the external battery and just focus on making a USB power cable for that computer. External batteries like the Limeaid are a dime a dozen. I *REALLY* like the idea of powering classic computers via USB, though One could use a pocket projector that is also USB powered making you completely off the grid!

[jmccorm]

My thought is to forget about the external battery and just focus on making a USB power cable for that computer.

This project isn't going to be an external battery... it is going to be an internal battery! (See the pictures above.) The initial external cabling was done as a proof of concept.

 

Along the lines of your suggestion, though, based on the testing I've done, a USB to Atari power DIN connector cable would do the trick, as long as some good gauge wire is used, and your USB port is capable of delivering 0.9A or better (which I think is standard for USB 3.0 and later). Perhaps some filtering, too. It is absolutely doable.

[jmccorm]

My diagramming could be better, but here are the plans for the next step:

 

 

At the power connector (left hand side of the case), there are two connections that need to be cut. This isolates the input power from the on/off switch. Actually, now that I think about this, it might be wiser just to cut the positive end and to leave the ground intact, right?

 

Once cut, the top end goes to the +5VDC input to the Limeade Blast. This will act as a continual charging circuit (which the Limeade is designed to handle). The bottom half goes to the +5VDC output of the Limeade Blast. The Limeade Blast will power the circuit under two conditions:

 

1] The battery is not depleted and it can successfully generate +5VDC

2] The battery detects a load

 

After making the change, the power switch on the Atari will not control the +5VDC going into the battery. Instead, it controls the +5VDC from the Limeade Blast's output to the motherboard. When you want to turn the machine off, you would use the power switch as you regularly would.

 

Once this step is done, you have a very basic battery powered Atari 130XE. The only flaw is that while the Atari is powered on, it is always running exclusively off of battery power. If a machine is continually left on, the Limeade Blast will automatically shut down the output when the battery's capacity nears exhaustion. The power supply input can only be used to charge the battery while the machine is turned off (and cannot be used to power the machine itself).

 

Next steps:

 

1] Obtain a higher capacity power supply that can simultaneously power the Atari 130XE (.76 amps) and the Limeade Blast (variable based on the charger's capacity, maximum rate of 2.1 amps). I'm probably looking at this 2.2 amp Atari XE/XL power supply on eBay.

 

2] Create a relay or other solution where if power supply input is detected, the battery is disconnected from the motherboard and the power supply directly powers the motherboard as well as recharges the battery.

 

3] Test, and if necessary, balance the power supply's current from between the Atari 130XE and the Limeade Blast's charging input to make sure that the Limeade blast doesn't steal so much current that the 130XE cannot operate.

 

4] Test that the Atari 130XE will be glitch-free when switching between battery and power supply input, while the system is running.

[jmccorm]

So far, I've evaluated three LCD displays to embed in the case:

 

Casio 2" TV-980B [6 vdc, 550mA, RF in]

RCA 3" 16-3062 [6 vdc, 630mA, video/audio in]

Sharper Image 3" CO727 [6 vdc, 550mA, video/audio in]

 

Immediately apparent is that the RCA and the Sharper Image have identical cases. The video in the RCA model, however, is superior to the Sharper Image model. Everything is much more readable. Both of these units take a video in and an audio in and have clean sound. The only problem seems to be that the NTSC artifacting isn't present, so if you have a game that uses artifacting to create additional colors, then those areas are going to show up as lines instead.

 

The Casio uses an RF-in. I went straight from the Atari's RF out into the Casio, which probably is causing an impedance mismatch. The video is a bit grainy and hard to read. But the NTSC artifacting comes out correctly.

 

The initial surge from simultaneously powering on the Atari 130XE and the RCA 3" LCD seems to be a bit too much for the Limeade's 2.1A output. It flickers off-and-on under the load. If I start off on wall outlet power and transition to battery power, though, it is able to handle the load. So I'll have to play around with it. The limeade has a 2.1A and a 1.0A output, which are powered simultaneously. I can probably split the load between the two.

 

It is tough to say which way I'll go with it. A 3" screen is preferable to a 2" screen, but unless there is a solution to reintroduce artifacting, the 2" display may be preferable if I can get a good RF image out of the Casio.