• Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

2 voltage sources

Status
Not open for further replies.

stellad

New Member
He;llo,
I'm doing a project, where I need to connect 2 or 3 voltage sources in the same time and I should switch between them. Should I connect this voltage sources in parallel?


Thanks in advance,
Stellla
 

dknguyen

Well-Known Member
Most Helpful Member
NOt exactly in paralell since that will cause a short-circuit because you are connecting two different voltages to the same point.

Add a diode in series with each source, then connect the source+diode combination in parallel with each other. THe diodes will make it autoselect between the higher voltage source.
 

stellad

New Member
But what if I want to select the lower voltage source? I want to be able to choose which one to turn on, I'll do the programing part by labview.

Thanks,
Stella
 

dknguyen

Well-Known Member
Most Helpful Member
[Post removed]

EDIT:
Post incorrect...using FET transistors as either high or low-side switches to change between power sources will not work. This is because the MOSFET parasitic diode will allow the higher voltage source to back-charge the lower voltage source when the higher voltage source is allowed to connect to the load.
 
Last edited:

Boncuk

New Member
Diodes are one way to connect different voltage sources, but they drop the supply voltage by 0.7V which is not indicated on your panelmeter in the power supply.

Additionally - as dknguyen already said - paralleling voltage sources that way only the higher voltage will be selected.

A DPDT relay (operated by an SPST switch) acts as you want and is a clean solution.

Boncuk
 

dknguyen

Well-Known Member
Most Helpful Member
I just realized that neither high-side nor low-side FETs will work very simply for your application. FOr high-side, the MOSFET's parasitic diode that is anti-parallel to it won't properly block the voltage for you between two supplies (backcharging will occur between the two supplies).

So yeah, relay.
 
Last edited:

Sceadwian

Banned
I hope this isn't for a battery controlled application. Selecting between various cells independently for discharge of the pack is a good way to destroy the battery pack and cause a nasty chemical leak, or with lithium's fire becomes a strong possibility.
 

omgwtfbyobbq

New Member
I have a similar question, so I figured I'd bump this rather than start another thread. Given battery prices/specs, for an EV I'd like to run two different battery strings with two different controllers since high power batteries are very expensive per kWh, and lower cost batteries offer much greater energy for my buck, but also have much lower peak power levels. The ideal situation would be that I'll have a box connected to the throttle that sends the voltage to one or the other string controller depending on position/power requirements, so that only one string or the other is on at a time. Of course, as a fail safe I'd like to have a diode on each battery string just in case something went wrong, so my question is, on sites like digikey, what kind of diode or combination of diodes would I need to look for with a ~120V/350A string so that, worse case scenario, it get discharged into from the other higher voltage pack.
 
Last edited:

dknguyen

Well-Known Member
Most Helpful Member
THat's over complicating things to me. You're saying that the portion of high power batteries working alone (with the high energy batteries being unused) are able to provide higher peak currents than if you replaced those high power batteries with high energy batteries and parallel them with the pre-existing high energy batteries? I find that somewhat hard to believe.

Suppose you had 50/50 high power and high energy batteries. Even if the peak currents of the high power batteries was double that of the high energy batteries, you would get equivelant performance and lower cost if you just had all high energy batteries and used them together.
 
Last edited:

omgwtfbyobbq

New Member
The problem with the lower power batteries is that the cycle life (~2-3k cycles to 80% dod) is also associated with a lower discharge rate, specifically .3C. I don't think higher discharge rates would benefit cycle life in this case. While the manufacturer does state 10C instantaneous is possible, supposely it's closer to 5C real world, provided it isn't winter time. ;)

Otoh, A123's M1 cells are supposedly capable of 33C rates, and according to A123's spec sheet, tolerate abuse well, such as high discharge rates and so on, so even though they are about three times the price, they still put out nearly twice the power per buck and also seem to last as long if not longer, with really nice aging IIRC, supposedly a ~1+% drop in capacity per year.

Also, this isn't strictly an EV, but a PHEV conversion, with a 9" ADC (maybe smaller) motor mounted in-line after then trans, so I'm trying to manage peak power, range, weight, cost, durability, and pack aging. I could theoretically go with a larger TS pack, but in order to get 50hp consistently (including cold weather), I'd probably have to get a ~17kWh pack (3C rate consistently), which would weigh in at ~400lbs. Toss in the ~200lbs for the motor/inverter/cabling for a total of ~600lbs, and the car only has ~200-300lbs of carrying capacity left. There's also the possibility of the cells not responding well to higher discharge rates, since there isn't as much info on 'em as there is on A123's stuff, and/or aging out, since ~54-70 miles of all electric range is a bit more than I can use.

Otoh, I could go with a ~1.5kWh A123 pack that should provide a reliable ~50hp and weigh in at ~40lbs, and I can drop in a smaller ~5kWh TS pack weighing ~100lbs, that should almost certainly last as long as the manufacturer's specs indicate given it would only see use at ~.3-.5C rates, and if it is a dud for whatever reason, I'll only be out ~$1500-2000 instead of ~$5500+. The installed weight would be ~350lbs, leaving me with ~450-550lbs of carrying capacity, a bit more reasonable. I could even shave another ~90lbs off by going with a 6.7" ADC motor and ~1kW of A123 cells for 30hp peak.

Anyway, if you know of a motor/battery combo that can deliver ~2-3k cycles to 80% dod, ~30-50hp peak, weighs in at ~300lbs or less, and uses batteries that are ~$650/kWh, don't keep it to yourself, but so far this is the best compromise I've come up with. It minimizes risk, both in terms of battery performance and pack aging (too much capacity that I can't use could just end up aging out), keeps weight down, and provides a nice boost in terms of instantaneous power w/o compromising the $/kWh ratio.
 

dknguyen

Well-Known Member
Most Helpful Member
Why do you keep talking about C rates? Those are normalized discharge rates and since you (or rather we) are comparing a system that carries both high power and high energy batteries but only uses one of them at any one time versus a system that uses all high energy batteries but uses them all at the same time, you are comparing systems with different energy capacities. So you should be comparing ACTUAL current output rather than normalized (C-rating) current outputs.

FOr example, a 50Ah high power battery discharging at 10C is actually discharging 500A. But 100Ah high energy batteries discharging at 5C woudl also produce 500A. By using all high energy, low power batteries in the car and using them ALL at the same time you should be able to produce the same current output as the low energy, high power batteries. THe combined capacity of all high energy batteries in the car rate to produce the same output current as a lower capacity, higher discharge rate battery.

Basically it comes down to the fact that you can only carry so much battery weight, and assuming energy vs weight is pretty much consistent, you can only carry so much energy on your car. So what you are basically doing is shifting around normalized discharge rating of the batteries against the proportion of high energy and high power batteries in the car. So unless this ratio holds true:

[High Power C-rating]/[High Energy C-rating] >> [Ah of High Energy Batteries]/[Ah of High Power Batteries]

you will not gain appreciable increase in actual peak current for the added complexity by using a dual battery type "one-or-the-other" scheme rather than a single battery type "all at once" scheme.

THe change of C-rating with temperature is a moot point since if it's halved for lower discharge batteries it also halves for high discharge batteries so the current output of both systems is still the same as temperature changes. Also, just in case it slipped your mind, remember that a battery rated for 10C being discharged at 10C is being worked just has hard as a battery rated at 5C being discharged at 5C.
 
Last edited:

omgwtfbyobbq

New Member
Why do you keep talking about C rates? Those are normalized discharge rates and since you (or rather we) are comparing a system that carries both high power and high energy batteries but only uses one of them at any one time versus a system that uses all high energy batteries but uses them all at the same time, you are comparing systems with different energy capacities. So you should be comparing ACTUAL current output rather than normalized (C-rating) current outputs.
I already did that in my last post. Like I mentioned, a 1.5kWh A123 pack should be capable of ~50hp output from ~40lbs of batteries. Otoh, in order to get that reliably from a TS pack, I would have to go w/ ~15+kWh of batteries, which are obviously a lot heavier since the weight per kWh is more or less the same, so we're looking at a ~400lb pack.
By using all high energy, low power batteries in the car and using them ALL at the same time you should be able to produce the same current output as the low energy, high power batteries. The combined capacity of all high energy batteries in the car rate to produce the same output current as a lower capacity, higher discharge rate battery.
That isn't the case due to the ten fold difference in consistent usable output as I mentioned in my last post. In order to get a 50hp pack from the cheaper cells, I would need to have way more lower power batteries than I want, which is prohibitive in terms of weight, aging, risk, and even cost.
Basically it comes down to the fact that you can only carry so much battery weight, and assuming energy vs weight is pretty much consistent, you can only carry so much energy on your car. So what you are basically doing is shifting around normalized discharge rating of the batteries against the proportion of high energy and high power batteries in the car.
Yup. Pretty much... I'm using batteries rated for 33C peak, that, at least according to the manufacturer take higher discharge rates really well in terms of lifespan, for power. And another set of batteries that doesn't seem to perform nearly as well in terms of power output at only ~3C, but are much cheaper, so I can get more energy for low power use.
So unless this ratio holds true:

[High Power C-rating]/[High Energy C-rating] >> [Ah of High Energy Batteries]/[Ah of High Power Batteries]

you will not gain appreciable increase in actual peak current for the added complexity by using a dual battery type "one-or-the-other" scheme rather than a single battery type "all at once" scheme.
I'm not concerned about energy density, but power density, wrt weight. In this case, for roughly the same weight, there's a ~10x difference in peak power available. This means I can reach my power requirements with a ~40lb pack, and my energy requirements with a ~110lb pack, leaving me with ~250lbs less than if I used enough lower power cells to relatively safely get the same amount of power.
THe change of C-rating with temperature is a moot point since if it's halved for lower discharge batteries it also halves for high discharge batteries so the current output of both systems is still the same as temperature changes.
Voltage doesn't change in the same way wrt current, at least based on what I've read. Voltage sag with TS cells is much greater than it is with A123 cells at the same C-rate.
Also, just in case it slipped your mind, remember that a battery rated for 10C being discharged at 10C is being worked just has hard as a battery rated at 5C being discharged at 5C.
That's true, and in that case I would rather put a battery that, according to the manufacturer and anecdotal accounts, can take much higher discharge rates, in a position where it has to take much higher discharge rates, rather than try to use a battery that is rated for whatever lifespan at a much lower rate of discharge in that position, and risk wrecking it, or having it age out.

Anyway... What kind of diode/s (in what arrangement) would I need to have a good safety net provided I had something fail and both controllers active?
 

dknguyen

Well-Known Member
Most Helpful Member
I'm not concerned about energy density, but power density, wrt weight. In this case, for roughly the same weight, there's a ~10x difference in peak power available. This means I can reach my power requirements with a ~40lb pack, and my energy requirements with a ~110lb pack, leaving me with ~250lbs less than if I used enough lower power cells to relatively safely get the same amount of power.
Ah, okay a 10x different in C-ratings might warrant a hybrid system. I've never seen much more than 3x, usually 2x difference. But if you don't care about energy density then why don't you just not have the 3C batteries at all and save that weight? Replace them with empty space, not with the more 33C batteries. You obviously think you have enough energy on the 33C portion of the batteries to travel far enough without the 3C batteries or else you would be worried about energy density. You weren't gonna speed for part way and then drive like a grandma the rest of the way were you?

3C discharge borders on useless in my mind for motor drives of any kind. Accelerate more smoothly unlike an SUV punk or street racer, don't rush towards stoplights only to wait, or tailgate so you constantly have to tap the brakes, and manage the transmission better to increase efficiency and lower the peak currents.

You can't have both worlds unless you have more energy than you can use available. THat's not the case with batteries right now.

Voltage doesn't change in the same way wrt current, at least based on what I've read. Voltage sag with TS cells is much greater than it is with A123 cells at the same C-rate.
More cells in parallel reduces the effective internal resistance. More cells in series will make up for the voltage lost. You're comparing different battery chemistries now and that's not as straightforward so only you can really decide that.

Anyway... What kind of diode/s (in what arrangement) would I need to have a good safety net provided I had something fail and both controllers active?
No diodes- use a fuse for each motor driver. THat's a ridiculous amount of current flowing through those diodes which is a lot of losses. Even if the losses are insignificant to the power in the rest of the system, it's still going to introduce cooling challenges to the diodes themselves. Also remember that diodes can also fail active. Just use a fuse. Much less losses, after all, if something does go wrong to blow the fuse you aren't exactly going to be able to or want to start up the car again without doing a bunch of repairs in which case replacing a fuse will be the least of your repairs.

Also, after power plant, transmission, charging losses, and motor losses EV is about the same overall efficiency as a gasoline engine, except EV has a lot more resources and energy put into producing the batteries.
 
Last edited:

omgwtfbyobbq

New Member
Ah, okay a 10x different in C-ratings might warrant a hybrid system. I've never seen much more than 3x, usually 2x difference. But if you don't care about energy density then why don't you just not have the 3C batteries at all and save that weight? Replace them with empty space, not with the more 33C batteries. You obviously think you have enough energy on the 33C portion of the batteries to travel far enough without the 3C batteries or else you would be worried about energy density. You weren't gonna speed for part way and then drive like a grandma the rest of the way were you?
The 33C batteries have ~1-1.5kWh, and since I want ~6kWh I'm probably going to add ~4-5kWh of low power batteries. That should give me ~30-40 miles of range during the low speed operating conditions I'd use EV mode with, keep the total battery costs around $4000, output ~30-50hp, and keep pack weight around 150lbs, and the system weight around ~250-350lbs.
3C discharge borders on useless in my mind for motor drives of any kind. Accelerate more smoothly unlike an SUV punk or street racer, don't rush towards stoplights only to wait, or tailgate so you constantly have to tap the brakes, and manage the transmission better to increase efficiency and lower the peak currents.
It would mostly be for keeping speed. I would use the ~30-50hp from the A123 pack when accelerating, in conjunction with the engine and alone provided I don't need much power in the city. The ~15hp low power/high energy TS pack would be for cruising.
More cells in parallel reduces the effective internal resistance. More cells in series will make up for the voltage lost. You're comparing different battery chemistries now and that's not as straightforward so only you can really decide that.
I'd definitely need to do some testing to determine what kind of voltage drop I'm looking at. Odds are the TS pack will be all series, since power output isn't maximized there. I'll have to fiddle around with parallel/series strings of the smaller A123 cells in order to figure out what kind of voltage/current I'm looking at.
No diodes- use a fuse for each motor driver. THat's a ridiculous amount of current flowing through those diodes which is a lot of losses. Even if the losses are insignificant to the power in the rest of the system, it's still going to introduce cooling challenges to the diodes themselves. Also remember that diodes can also fail active. Just use a fuse. Much less losses, after all, if something does go wrong to blow the fuse you aren't exactly going to be able to or want to start up the car again without doing a bunch of repairs in which case replacing a fuse will be the least of your repairs.
Fuse it is! 400A/144V fuse should do fine, especially with the smaller 6.7" motor.
Also, after power plant, transmission, charging losses, and motor losses EV is about the same overall efficiency as a gasoline engine, except EV has a lot more resources and energy put into producing the batteries.
It really depends on where the oil comes from. Oil in the states for instance, requires a lot of energy, and extraction/transportation/refining is only at ~50% efficiency, so for most cars we're looking at ~10% efficiency give or take. Even with fossil fuel power plants at only ~30%, since we also have ~30% hydro/nukes, electrical energy generation wrt fossil fuel energy consumption is at ~50% on average, although this depends on region. Toss in EV efficiency at ~70%, and we're at ~35%. Otoh, if we got our electricity from only coal power, and our imported oil only required a lot of energy for refining, then conventional cars could be at ~15% and an EV at around 21%, so in that case there isn't a huge difference.

In my case, the appeal isn't efficiency, but cost. W/ federal and state rebates, solar panels are approaching ~6c/kWh levelized cost over 20 years, and ~3-4c/kWh levelized cost over 40 years, with lifespans longer than that, so levelized costs that may be in the neighborhood of ~2-3c/kWh. Provided TS LFPs have a fairly linear capacity loss curve compared to the number of cycles, I could get ~4000 complete (equivalent) cycles by using the pack down to ~40% capacity, which is about 10-12c/kWh, depending on how much I can get the batteries for. Total costs would be ~12-15c/kWh, and battery/electricity costs would be ~2-3c/mile in conservative city driving at ~200Wh/mile. Since maintenance costs along are about that, being able to go all electric for short city trips and during city travel in general, should at least pay for itself. Since I'm looking at ~30mpg city anyway, then I should see a 6-7c/mile savings based on fuel costs, and assuming a system cost of ~$2000-3000, it should pay itself off in ~30k city miles, which isn't bad IMO considering it'll also add ~30-50hp in a 2800lb car that only came with ~50hp from the factory.
 
Status
Not open for further replies.

Latest threads

EE World Online Articles

Loading
Top