LOW VOLTAGE ALARM

[Burgerman]

Would it work with pins 123 and 24v 12v 0v ???

That would be better as it will read like 13.2 12.9 then both, then repeat?

[Mark]

Well it might, but from the limited specification and instructions (all combined on the reverse of the device) I interpreted it as only being able to accept the higher voltages (as opposed to differential cumulative voltages) on pins one and two I.e. "S1". Worked acceptably until 29V+. However, so cheap that the only way to find out is to buy more and try. Did have a quick look at the current draw at 27 V and it was of the order of 45 mA.
Did in fact spend more on beer tonight - and I arrived late!

[Burgerman]

Did have a quick look at the current draw at 27 V and it was of the order of 45 mA.

Way high. There is a bad omen clue that it was going to fail at that voltage on 2 pins!

[shirley_hkg]

http://item.taobao.com/item.htm?spm=201 ... id=&pm_id=

Found this small combo metre . It has 4 modes ; TIME ; TEMP ; VOLTAGE ; CYCLE through all 4 modes at 5 second interval .

Two versions ; 0 - 33V ( or 200V ) . Price is in RMB , so is 1/6 in $ USD . . . DISPLAY SIZE : L 32mm × W 10mm× Thickness 13mm

[Burgerman]

Does it have the low voltage alarm that you can set for the life batteries? My reason for using one.

[shirley_hkg]

Does it have the low voltage alarm that you can set for the life batteries? My reason for using one.

NO. IT'S A VOLTMETRE .

[Burgerman]

Well I will stick to the 2 that I am using. They monitor all 13 series "sets" of cells. And any that get down to 2.7v beep at me!

Thats really all I need, hidden under the seat. Theres no real need to see the actual voltage for me. Just a method of knowing if I ever get out of range far enough to do damage.

[flagman1776]

I ordered a few of these http://www.ebay.com/itm/2in1-Lipo-Batte ... OC:US:3160 Enough for my battery packs with some spares. Cheap enough. I'm hoping I can plug them in the balance charging connector when the pack is in use. I'll know more when they get here.

[Burgerman]

Those are the same as the ones I have here, by the look of things. Same specs exactly. Thers 2 mounted on double sided tape under the seat on my BM3 that I am working on at the moment. Monitoring 7 cells, + 6 cells = 13 cells total. 2.7v is a bit low, may increase this to the next level, so as to give me some early warning to get charged or home...

[flagman1776]

To order these small parts, I've cut & paste the description that I see here linked to UK Ebay or other sources & then paste into my USA Ebay search. I'm sure it would work the other way as well. But it saves a boatload of work trying to find the correct pieces.

[Burgerman]

Or any hobby shop, quick search for lipo battery monitor etc also finds them easily. Mine have clear shrink wrap. Yours appear to not have any?

[robnnorthaustin]

Sorry for the fuzzy pics, my hands have gotten a bit fuzzy the past few years also LOL. I will setup a tripod when I finish the project and take some better pics if anyone is interested. . These are pics of my new volt/amp meter and lights I added to the Revo. I am going to remove the temporary gooseneck I have the meter mounted on and replace it with an 18" black microphone gooseneck that will clean up the mounting also. I am very happy with the LED lights which were dirt cheap and I ended up mounting both so I didn't use the voltage converter. I added two small toggle switches to the meter housing to be able to turn the lights and meter off/on. Total out of pocket expense so far is around $35.







Before I had the amp meter I had always wanted to know what the amp draw was on the batteries and how my driving affected it. Since it's digital I'm sure I'm missing some peaks but I am very satisfied in what I'm seeing. The results bore out what Bm has said about current draw being lowest when driving flat out. I was surprised to see that my peak draws are mostly below 40amps and my scooter averages between 7.5 and 9 flat out at 4.5 miles an hour on flat. sidewalks or pavement. With the new Gels I'm seeing a very flat voltage and haven't driven it far enough to see any voltage sag. That is very comforting and makes me so glad I added the amp/volt meter.

[Burgerman]

This is why scooters go further than powerchair on the same batteries. They dont use power for turning... And motor amps are a very different thing to battery amps.

Its also why running flat out is the most economical.
Going slower uses the same amps but does it for longer to get to the same destination.
So traveling at half speed would need to use less than half the amps. To be more economical. But thats not what happens. So go full speed.

http://www.wheelchairdriver.com/gopro/motoramps.mp4

[robnnorthaustin]

I'm measuring current between the battery and controller, so would that be motor amps?

[woodygb]

battery Watts in = Watts consumed by motor

Amps x volts = watts

Battery Amps x Battery voltage = Motor Amps X PWM Voltage of controller ( or divided by % of PWM Cycle )

Battery Amps x Battery voltage/ PWM Voltage = Motor Amps

[Burgerman]

Exactly what woody said. But to make it simple to understand, lets say your controller is 100 amps max.

As you set off hard, or stall it against a wall your motor might pull 200 amps (or more) if it was allowed to see the full 24v. The controller cant do 200 Amps, so it throttles back the pulsewidth to 50 percent (or 12v equivelent) to the motor. (actually ON/OFF/ON/OFF/ON/OFF etc at high frequency)

That equates to 50 battery amps, and 100 motor amps (controller is then maxed out - it can do no more). Almost. Some gets wasted as heat.

If you fitted a 48v battery then battery amps will fall by half. Because now, 25 percent pulsewidth is 100 motor amps. Thats ON/OFF/OFF/ON/OFF/OFF/ON/OFF/OFF/ON At around 16k times a second. Dont try it there will be smoke.

Of course as speed increases, as you accelerate hard up a steep hill, the motor speed increases to the point where it CAN accept 24v and will take 100 amps only. So now the controller can use 100 percent pulsewidth. So then battery amps = motor amps.

[robnnorthaustin]

This is so interesting. It also explains a comment I read a few weeks ago. I'm wondering what the relationship is between the switching rate and the current reading I'm seeing between the battery and controller, if there is one? I'm still glad I installed meter though as this is becoming a hobby of sorts. Is the data, schematic published for a PG 45a controller? I have also been looking for the min and max supply voltage the controller would handle if I choose to try a 24v to 30v set of lithium's without changing the controller.

[Burgerman]

Most PG stuff errors at 35v. Some like Dynamic at 32v.
Low end is usually caused by voltage dropping UNDER LOAD due to cable resistance, or old/discharged batteries. So rather than stopping working they reduce amps (pulsewidth) at about 18 volts to maintain function and control.

LiFePO4 = 8 cells 28.80v fully charged, and actually the charge voltage. Thats 3.60v per cell. Drops to 3.4 volts per cell in the first 10 feet. Then almost doesent change more than 1/10th of a volt until its about 80 percent used up. Then drops fast. 2.7v is the 100 percent point. You dont want to go less. Theres no point. Nothing left.

LiPo = 7 cells at 4.20v per cell or 29.40 volts charged. And it stays there. Drops steadily and predictably to around 3.7v then drops off a clif. Dont go below 3.5v sensibly.

Li (cobalt) as used in laptops, same as above but 4.10v per cell charged.

[woodygb]

Is the data, schematic published for a PG 45a controller?

You mean this for the S Drive?

s-drive wiring.jpg (72.38 KiB) Viewed 7690 times

Voltage specs on the final page of this.
http://www.pgdt.com/pdf/s-drive.pdf

[shirley_hkg]

HERE THEY ARE , HUNDRED OF THEM : http://s.taobao.com/search?spm=a230r.1. ... c#J_Filter
Currency is in RMB ; that's 1/6 of USD.

[g7pvs]

Hi,
I don't know how helpful this will be for others, but I've designed a low voltage alarm for my LiFePo4 batteries (8s).
It uses TC54VN3002ECB713 voltage supervision chips (3.0V) - they are also available at different voltages, but obviously cannot be adjusted.
The circuit diagram is as follows:

voltage_monitor3a.jpg (136.7 KiB) Viewed 7525 times

I've measured the voltage the alarm switches on for each of the cells and for most of the cells it switches on at 2.90V and stays latched at 2.95V, but the latch drops out at a voltage of 3.05V. I didn't measure the exact voltage because I only wanted to check it worked, so used potential dividers (with an extra resistor in parallel for the input being tested) to give a rough idea of the switching voltage. (I may do a proper test and measure the exect switching voltages another day).

I haven't measured the quiescent current, but the datasheet specifies a typical quiscent current (per device) at 0.9uA, so leaving the battery monitor permenantly connected will over a week drain the batteries by 0.9uA*24h*7 = 151uAh, or over a year 151uAh * 52 = 7852uAh = 7.9mAh which in 15Ah batteries is negligible. Also the drain on each battery is identical.

Best Regards
Trev

[g7pvs]

Hi Again,

Oops the circuit I posted wasn't quite correct (I fiddled with it to get the schematic capture to work as I hadn't got the layout of the SMD components quite correct),
The one below should be correct - the values of R9 to R16 are the values calculated and need replacing with the next preferred value lower than what is shown:

voltage_monitor4.jpg (243.5 KiB) Viewed 7525 times

Best Regards

Trev

[g7pvs]

Hi Again,

I used a single sided PCB design as follows:

voltage_monitor2_c.jpg (79.86 KiB) Viewed 7523 times

Best Regards

Trev

[Burgerman]

I think you will find 3 volts too high. Every time you load the batteries it will beep at you. 2.5 is a better point. Maybe add a series diode or 2?

[g7pvs]

Hi,

It started beeping when going up my driveway which is a slope I've measured at about 10 degrees. I'm in the process of recharging the batteries and it is looking like I'll be putting approx 8Ah back into the 15Ah batteries.

Next time I'll leave it until it beeps whilst using it on the flat and see how discharged the batteries are.

If the voltage is too far out I'll make another PCB and use a different version of the voltage supervision ICs.

Best Regards

Trev

[Burgerman]

You will struggle to get any sense from an alarm other than at the very END of charge.

There Is more difference between different cells, than voltage drop due to discharge over around 85 percent of the charge.

Set to say 2.5v it may/may not start going off under load regardless of charge - depending more on load/resistance than state of charge.

1.8 Amps discharge for 7.5 hours here. Looks much the same for the first 15 hours...
Lowest cells are still better than highest cells everywhere!

[g7pvs]

Hi,

I'm not using the Hyperion to discharge the battery.

The alarm only sounded when the batteries were under lots of load i.e. 10 degree slope.

The recharge process had reduced the current to 0.1 Amps when I posted - now at 0 Amps (still 8Ah).

I'll wait until I've discharged the batteries under normal use, then I'll see.

Best Regards

Trev

[Burgerman]

I'm not using the Hyperion to discharge the battery.

It doesn't matter what actually discharges it. The point was to try and explain that the cell voltage is constant. It does not drop as it discharges. All day long. Voltage will not tell you how discharged a pack is. Until it really drops off a cliff right at the end.

It falls in proportion to load (regardless of charge.) But doesn't drop significantly for almost the whole charge.

And it falls INITIALLY in the first few meters.
And it falls fast right at the end as the pack becomes exhausted.

The bit between 99 per cent full, and about 85 per cent used, stays much the same.

In other words the only thing that the voltage (alarm) can tell you is that the pack is completely empty, or that you are going up a hill. It cannot tell you if its 50 per cent, 20 per cent, 75 per cent used up. Because individual cells in a pack vary by MORE than the difference between almost full almost exhausted.

So you will need to set it to around 2.3 to 2.5 volts. Then it will reliably tell you when its empty. It cant tell you anything else with any reliability.

[Fulliautomatix]

Yeah! Mine tells me I'm going up a hill...through a kerb cut...over a step...accelerating hard...
I suppose if it stays on continuously when stopped or light running, the battery is flat?
Set at 2.7V...lowest those AOK ones go.

[Burgerman]

Fit a forward diode. Various types. You lose part of a volt from .3 to .7 depending on type. Then it will only come on when you reach end of charge.

Or connect via a capacitor / resistor that acts to smooth pulses.