EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: 1slickvdc on January 12, 2016, 11:09:57 pm
-
I want to build a computer into my car, but I'm broke, so instead of buying a bunch of new specialty hardware, inverters, DC-DC converters, etc. I was trying to figure out if I could power laptop hardware with a variable 11.8-14.5V supply (basically my car depending on operating conditions, at the wiring I'm using). Already tried feeding that into the charging port, and it must be below some specific minimum voltage, as it won't POST. I've disassembled an old fried battery from this machine to get at the controller. Two questions to bounce out:
1) If the battery controller is fed up to 14.5V and reports accordingly to the laptop, is it likely to go into a failsafe shutdown mode?
2) If the battery controller is fed a bogus reading more appropriate to the computer to bypass said failsafe mode, but the actual voltage going through the terminals to the machine still approaches 14.5V, am I likely to start things on fire, or are there internal regulators (switching or otherwise) that could handle it, if given proper thermal consideration?
Donor laptop models: Dell Latitude E6x00 series
Stock battery rating: 11.1V
Combined cell voltage outside of battery housing: 11.3V (with the on-battery indicator reading 20% charge)
Stock charger rating: 19.5V
-
I think you could run the laptop of the car with 12V.
I have seen peopel run old ipads that had a dead battery with the 12 volt plug(ofc you would need to step it down to the 7.4v or 3.5v)
and integreate them as a hole unit in the dash.As for the question asked i would
guess they have since over charging a lipo-li-ion would just go in a huge :-BROKE.
-
Yeah, but my target voltage is so close to the lowest operating voltage of the car that even a crude linear regulator with something like a pass transistor (which I have parts for laying around) would still have too much drop and will likely be too low to operate on.
-
The charger circuit is inside the laptop. It expects a voltage near 19.5V, not 12V.
-
Yeah, and I figured as much after some testing, which is why I've moved on to putting voltage directly through the battery contacts, which is much closer to my available supply.
-
When the car is running the voltage will be 15-16V with spikes typically up to about 60V. Even when the engine is off the voltage can be up to 13V.
Use a car laptop charger to get 19V or you'll fry the laptop.
-
Practicality out the window on this since these aren't even complete laptops (neither have keyboards, one's missing the palmrest) they're parts machines. Sh*ts and giggles. Just want to see if it's possible.
Also, I've never seen mine go over 14.4-14.5 at 3500+ RPM. At idle it's about 12.5V. Under load at idle it's almost exactly 12V. Yeah, there may be transients in there, I haven't exactly stuck a scope to it, but I've never seen an operating voltage in a car that high. Also, do you have a link or something to these high voltage transients, because now I'm curious.
Back to it though: given a collection of spare parts (for the laptops) and a pile of car audio amplifier repair parts (including transistors and all other sorts of bits), could I make a simple (or almost simple) device or circuit to avoid those spikes and/or regulate the voltage? I'm going for something "Junkyard Wars" style here, work with only what I've got.
If I blow something up, I learned something. If it works, I learned something. Either way I learned something and that's the biggest point of it all. Don't take this as me fighting back though, I'm feisty but I learn by experimentation and hands-on, and I seriously do appreciate all the feedback.
-
These transients (or some subset at least) are called "load dumps". Here's a thingee on them (http://www.ti.com/lit/an/snva681a/snva681a.pdf).
Also, I've never seen mine go over 14.4-14.5 at 3500+ RPM.
That's the thing about transients (i.e., brief events) -- your multimeter may be too slow to catch them, but that won't stop them destroying your laptop.
-
I think that quote may have been taken out of context, but my craving for information was not. ;D
The quote was referring to Ian's reference that running voltage is in the 15-16V range.
Now knowing about the issue of these spikes, what do sensitive pieces of equipment already in the car (like your ECU, sat nav, etc.) have for protection/suppression?
-
get a cheap DC-DC converter of ebay if you want the cheapest way possible, last i checked it was about $7 posted for something that could push 3A at 20V, cars are horrible, noisy and transient filled beasts, e.g. you flash your high beams and most cars will have a short 40V spike when the relay contacts are opening from inductance in the wiring, but way to fast for the ~5hz low pass filter most multimeters have on there DC input,
In regards to the 1slicks question, most automotive rated electronics have quite beefy input conditioning, before it even reaches the automotive grade regulator,
-
Also, I've never seen mine go over 14.4-14.5 at 3500+ RPM. At idle it's about 12.5V. Under load at idle it's almost exactly 12V. Yeah, there may be transients in there, I haven't exactly stuck a scope to it, but I've never seen an operating voltage in a car that high.
It varies a bit with the exact type of Lead acid battery, alternator and its condition. Old vehicles that were designed for batteries with removable caps tend to run a higher voltage. Maintenance free batteries have to be run at a lower voltage or they die quickly.
Back to it though: given a collection of spare parts (for the laptops) and a pile of car audio amplifier repair parts (including transistors and all other sorts of bits), could I make a simple (or almost simple) device or circuit to avoid those spikes and/or regulate the voltage? I'm going for something "Junkyard Wars" style here, work with only what I've got.
No. A Latitude E6500 PSU is 3.34A @ 19.5V. (65W) Assuming that's chosen to run a fully kitted out machine flat out with a bit of margin, you are still goiing to need 50W peak at between 10V and 11.1V (assuming the battery is 3x 3.7V LiPO cells in series). You aren't going to find parts for a low dropout 5A spike-resistant regulator in your average car audio amp junk pile, and if by some miracle you do, it will dissipate close to 20W of heat so will need a *BIG* heatsink.
GPSes take far far less current so are much easier to protect + if fixed mount generally have DC input circuits designed to take the voltage range and withstand the spikes. The ECU is custom designed to match the vehicle's systems, and only its sensitive digital core needs protection. All load control circuits are simply designed to accept and tolerate the raw vehicle supply.
-
No. A Latitude E6500 PSU is 3.34A @ 19.5V. (65W) Assuming that's chosen to run a fully kitted out machine flat out with a bit of margin, you are still goiing to need 50W peak at between 10V and 11.1V (assuming the battery is 3x 3.7V LiPO cells in series). You aren't going to find parts for a low dropout 5A spike-resistant regulator in your average car audio amp junk pile, and if by some miracle you do, it will dissipate close to 20W of heat so will need a *BIG* heatsink.
It would likely be light to moderate loads, audio/media decoding, SDR, and a program monitoring my ECU. The board I'm hoping to use doesn't have the nVidia chipset either, so aiming for 30-35W. Could over-design for something like 90W though.
I dunno, do 8-10A big ass BJTs, 60A+ FETs, spare heatsinking and active cooling bits, a couple TL494's, old school logic bits, my personal favorite: the new 3D printer, and enough caffeine to kill a horse count as enough parts? >:D
As typically happens with my trains of thought, this may derail into a thread on designing a simple regulated SMPS or boost converter.
When I said "broke", it meant in terms of buying finished product. I'm an EE student at a local tech school and parts are, well, basically free if I have a plan and can integrate it into some coursework.
Hmm... a possible learning experience... :D especially since I've had a hell of a time trying to get a 494-based anything working past the oscillator. Sadly, anything in depth on SMPS is not part of our normal curriculum, so not something I've had good hands on with yet.
-
How high voltage are the MOSFETs, and are there any P channel ones? There's a few spike-stopper circuits that work by turning off a series MOSFET *quickly* when input overvoltage is detected and letting the load ride through the transient off a bulk decoupling cap.
Stick a 3.5A boost converter to 19.5V after that and you'd have a reliable solution. Magnetics for it will be a problem - you may have to wind your own on a salvaged transformer core and bobbin from a PC PSU if you can liberate a reel of magnet wire.
-
How high voltage are the MOSFETs, and are there any P channel ones? There's a few spike-stopper circuits that work by turning off a series MOSFET *quickly* when input overvoltage is detected and letting the load ride through the transient off a bulk decoupling cap.
Stick a 3.5A boost converter to 19.5V after that and you'd have a reliable solution. Magnetics for it will be a problem - you may have to wind your own on a salvaged transformer core and bobbin from a PC PSU if you can liberate a reel of magnet wire.
I've got a whole pile of 55V, 60V, 100V, and 200V N-channel MOSFETs, since these suckers are the most common culprit in failed car audio amps. However, I do have 5x IRF9540 P-channel MOSFETs in my parts drawer. I can get something else from the campus parts drawers if need be.
http://www.irf.com/product-info/datasheets/data/irf9540npbf.pdf (http://www.irf.com/product-info/datasheets/data/irf9540npbf.pdf)
As it happens I also work at a computer repair shop, so there's no shortage of PSU's to nick parts from.
If you think there's something that can be built with the stuff I've got, or with stuff I can get, what sort of design are we talking here?
-
It expects 19.5 at the charger. 11.1 at the battery. I was trying to figure out the best interface and options.
Also, as mentioned, I have no shortage of the boards and parts for these laptops, so if I try a few during experimentation, I'm not going to lose any sleep.
Having screwed with the battery interface a little, I feel it would be easier to do a step-up and interface it on the charger connection instead.
-
See: http://www.ti.com/lit/an/snva717/snva717.pdf (http://www.ti.com/lit/an/snva717/snva717.pdf)
N.B. load dump transients can occasionally go well over 100V, so use a 200V P channel MOSFET.
-
Since that TI note shows what looks like something that would cut off during spikes, is it fast enough that a large amount of capacitance at the end of the SMPS could keep the computer going during the protection power off phase?
-
Maybe, and for low current applications that's the normal way to cope, but it would be better to have a battery that is at least somewhat usable fitted to the laptop so you only have to supply the average power used and so it doesn't matter if the PSU drops out during occasional longer spikes
-
That would be the practical solution, what about the heat involved when stuffing the guts under a seat? Lithium batteries creep me out a bit in environments that can get real hot like that.
One of my goals in this project to have the minimal amount of components required actually in the car, so caseless aside from a custom mount and airflow piece I'll print up, nothing that's really not part of the motherboard that's not required, etc.
-
That depends on how hot it can get in contact with your car's floorpan on a hot day - which depends on your summer climate. If its high summer, and you've got an unwashed black car parked on asphalt that been in the sun all morning in Death Valley, you are FUBARed.
-
Being Wisconsin we don't quite hit Death Valley level temps, given a beige interior it's a little nicer, and the thing will be under the seat. However, the car is a giant fishbowl in terms of glass coverage, and I've observed temps in the 130's in there before.
Anyway, what's a good approach to designing the step-up converter/good place to start? Again, this would be my first experience in any sort of switch mode power system, so a little 'training wheels' on the thread for a little while would be greatly appreciated.
-
If you put a relatively large capacitor on the end of a dozen feet of zip cord you should be good for killing the load dump transients. Now what might happen putting 15vdc into the battery port is some 15vdc tantalum caps might explode. A number of laptops I've disassembled are simply 3 buck converters in series. 19vdc to battery.. battery to 5, 3.3, 5vdc turns into 2.5 with a flying capacitor divide by two, processor is powered from a buck converter run off the battery. So over voting the battery will overheat the proc buck inductors on a hot day.
I suspect you could run a 3 cell laptop off 25vdc if no capacitors explode. Usually the feta are 30vdc which means the avalanche at about 32 to 35. Another problem is if the switching transients start dumping energy into the fets. This is of course dependant on many things
Sent from my SM-G530T using Tapatalk
-
If you put a relatively large capacitor on the end of a dozen feet of zip cord you should be good for killing the load dump transients. Now what might happen putting 15vdc into the battery port is some 15vdc tantalum caps might explode. A number of laptops I've disassembled are simply 3 buck converters in series. 19vdc to battery.. battery to 5, 3.3, 5vdc turns into 2.5 with a flying capacitor divide by two, processor is powered from a buck converter run off the battery. So over voting the battery will overheat the proc buck inductors on a hot day.
I suspect you could run a 3 cell laptop off 25vdc if no capacitors explode. Usually the feta are 30vdc which means the avalanche at about 32 to 35. Another problem is if the switching transients start dumping energy into the fets. This is of course dependant on many things
My original idea had no battery in use, and was just debating which connector to use and at what voltage. I figured due to how hardy the regulation is on a board anyway that it could be fairly resilient, but not sure the best approach. If what you say is true, then simply a sort of surge/transient suppression device and feed the battery terminals the car's main voltage. Sort of what I was thinking earlier, and worth a shot (again, if I fry a board dicking around, and least I learned something).
Still wouldn't mind building some sort of SMPS type of thing though.
-
I want to build a computer into my car, but I'm broke, so instead of buying a bunch of new specialty hardware, inverters, DC-DC converters, etc.
You're wasting your time and energy. This is a problem which has been solved thousands of times many years ago! You have a job and can afford a 3D printer, just buy the right thing and be done with it.
Lynd makes automotive grade power supplies (http://www.amazon.com/s/ref=nb_sb_noss?url=search-alias%3Daps&field-keywords=lynd+power+supplies") available for under $100 which have all the input protection and dc-dc conversion you can shake a stick at. They're small, efficient, and used in pretty much every police car in the world (read: reliable).
Carnetix also makes ideal power supplies, such as their 2140 model (http://www.amazon.com/CarNetix-CNX-P2140-Output-Intelligent-Regulator/dp/B004NDKFVE/ref=sr_1_3?ie=UTF8&qid=1452704715&sr=8-3&keywords=carnetix"). 20V at 150W is no problem, plus it has an extra 5V output so you can power a USB hub. Even has a pulsed output so it hits the power button for you when you turn on your key, and an app that sits in your taskbar and monitors the power supply's voltages. I've installed dozens of these and they work perfectly.
There is no advantage to anybody of you rigging up your own nightmare of a circuit. Just tell yourself that if you absolutely had to, you have the knowledge and parts that you could macgyver it. Then sweep the parts into the bin and buy something that works perfectly and is plug and play, and devote your brain power and time to solving a problem that doesn't have a solution yet.
-
That there is in my list of solutions, but for some reason I like making things difficult and try to turn everything into a learning opportunity. Maybe I'll do both. Build one for the learning, buy one for the safety and reliability. That would be fun. Also, I'm in America, where having a job doesn't necessarily mean I have a decent income. Also have a decent size car repair project coming up too.
"App in the taskbar" assumes I'm going to run Windows. Still debating on that but this is just me nitpicking a post for some reason (caffeine withdrawal).
That 20V supply does sound like the best option, although I haven't followed the link or researched that model yet.
Hmm... Problem with no solution? My Raspberry Pi inside McIntosh car stereo project. Need help reverse engineering the damn unit and no schematics or service manuals are available. That may be something for a different thread.
-
Looking for a power adapter to run my laptop (19v/6.3A) safely from questionable sources (car/air/boat) too.
I don't know enough to trust building one myself yet. Trying to find a trustworthy source is difficult, and the laptop manufacturer doesn't have one (universal travel adapters with cig or aircraft "EmPower"(?) plugs).
Maybe these..?
http://lindelectronics.com/ (http://lindelectronics.com/)
http://www.carnetix.com/power-supplies (http://www.carnetix.com/power-supplies)
IM
-
Bollocks to all this engineering stuff, bodge a couple of wires to the battery terminals of an old laptop with a ciggy lighter plug on the other end and go for a drive!
Then report back
-
Looking for a power adapter to run my laptop (19v/6.3A) safely from questionable sources (car/air/boat) too.
I don't know enough to trust building one myself yet. Trying to find a trustworthy source is difficult, and the laptop manufacturer doesn't have one (universal travel adapters with cig or aircraft "EmPower"(?) plugs).
Maybe these..?
http://lindelectronics.com/ (http://lindelectronics.com/)
http://www.carnetix.com/power-supplies (http://www.carnetix.com/power-supplies)
IM
Why not just get a 12/24VDC to 110/240VAC inverter? I have a small 100W one I use in the car if Im away from home and need to charge things.
-
I've got an inverter. No laptop brick though. Also, that's more stuff taking up more space. I did pick up one of those crappy China boost converters. Works great on the bench. Now I've got to make a housing for that and the laptop bits. Debating mounting techniques.
-
All this crazy talk instilling fear about unconquerable load dumps and spikes is pure bs. If this were true you wouldn't have even one MCU(instead of dozens) or a radio in your car.
All the parts you need to build a 12V to 19.5V battery charger for your laptop can be found for free within a PC ATX power supply, especially the high-power magnetics, first to make an input power filter and secondly to have the core and wire to wind an inductor for use with your 12-18.5V boost converter. In the worst case, you may only have to buy a 18V 1W zener and get a connector to fit the power jack on your laptop.
The ATX will also supply you with the high current schottky rectifier you will need, maybe even the zener and the filter capacitors for the input filter but you may need to find at least one 25V electrolytic filter cap for the 18-19.5 V output.
Note also that an ATX motherboard has tons of 30V, 30-amp N-Chan MOSFETS for use to make an input transient clamp and to serve as the switching element in a boost converter.
A simple 18V zener diode feeding the base of a power transistor or the gate of a N-Chan MOSFET on a small heatsink (all the heatsinks you will need are free to pluck form the ATX P/S) will absorb any load dump you could ever encounter. The ATX can also supply you with another high current schottky diode to give you reverse input voltage protection.
The whole thing could even be designed to work with using a LM339 chip or other SMPS controller chip that can be used to make boost converter controller and these are always found in all ATX P/S's as well.
I could put one together in an hour on a perfboard!
I have already successfully made one using a 555 timer as the controller, not bothering to use the controller chip free from an ATX P/S.
I have been using it in my SUV for years now to charge my laptop without any problems. BTW, 555 timer chips can often be found in ATX power supplies on a small satellite board used to control fan speed relative to operating temperature for fan noise reduction.
This forum is so often saying it can't be done..and if it could be done it would be better to buy something that could do what you want, anyway!
-
Here's the 555 Car Laptop Battery Charger PCB.
Note: The output schottky diode is mounted (TO-220 case) is mounted on a heatsink). On the PCB below it is shown on the top left.
A 220uF/25V capacitor is not shown but mounted on the top of the PCB and connects to the PCB Vout vias, this is the output where the charging cable to the laptop connects to the PCB after the schottky diode cathodes. The two large vias on the upper right are the +18-20V output connections to the laptop.
The 18V zener diode is shown mounted vertically on the top right so the cathode connection is not clearly shown. The cathode connects to the +20V output. The 12V zener is clearly marked and can be an 1/2W or 1W 20% device.
Two T0-18 case NPN small signal transistors are shown, they were also salvaged from the ATX P/S. They can be BC547 2N3904 etc.
The PCB shows an I555 CMOS timer. A bipolar 555 version can also be used. There is a small yellow flying wire jumper shown across its footprint.
Unless otherwise noted all other resistors can be 1/8W 5%.
L is 30 turns on a 3-in approx toroidal core from an ATX power supply, and also are most of the parts used here are salvaged from an ATX P/S. The winding wire was already on the core. All ATX power supplies have this large toroid core, you can't miss it. It is used as a transformer in the original ATX circuit. The extra winding can be easily unwound and removed.
The N-Chan Power MOSFET(Middle of PCB, far-left) was taken from an ATX motherboard, there are many of them to be found near the CPU socket of an ATX MB. #22 AWG leads were soldered to the MOSFET to allow the part to be mounted through-hole. It should have a small heatsink(also from the ATX P/S). These MOSFETs are tyipically rated at 30V and >30Amp and have a low on resistance RDSon and it barely gets warm in this circuit.
The .,125 ohm 4W resistor connecting to the source of the power N-Chan MOSFET is made from four .5-ohm 1W resistors in parallel.
This laptop charger/power converter is practically indestructible and has served me well in my SUV for years. It can supply 3-4.5 amps to a laptop, just like the mains powered brick you use in the house.
The 2200uf/16V cap (top , left) was also salvaged from the ATX P/S.
On Sw is where the 12V from the car battery is connected. I wire it to a cigarette lighter connector found on the dashboard with a connector I got at a five-dime store to charge your cellphone.
-
For load-dump protection, don't plug this in while someone is jump-starting your battery..this is just common sense.
If you have the room and the desire you can use the second schottky diode found in any ATX power supply to provide reverse power protection and you can salvage the other large toroidal inductor fro the ATX power supply to be inserted in series with the 12V from the car battery. At the Sw +12 connection on the PCB you can also connect another N-Chan power MOSFET to ground with a 18V diode to the gate, 1K to ground Drain to +12V, source to ground and that will quench any load dump surge.
I have been using this circuit the way it is shown for years without a problem. It sits under my front passenger seat waiting to be used. Really love it when I go camping!
-
There was what could be a .1 confusing artifact on the L inductor footprint shown. Revised attached.
I was in a hurry to get this PCB single-sided designed and finished in an hour, so it wasn't exactly meant to conform to ISO 9001