Continue to Site

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.

  • 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.

Automated Load Tester

Status
Not open for further replies.
ronv

I like the idea
Simple is great!
I wasn't aware that 4 1 ohm resistors in parallel would result in .25
That's a part of ohm's law that I did not know.
Is the current caring capacity of the four resistors shared?
So if I figured 7 volts (to be on the safe side) x 25 amps or 175 watts can that be further divided by 4 to 43.75 watts.
Seems to me a lot of the high wattage resistors are wire wound.
Is there one type of construction that is better or worse?

In an earlier post there was concern about resistor value changing as it heated up.
What do you think?
 
For what it may be worth years ago I did a small demo load bank project before we built a much larger unit.

We could simply use 4 1 ohm resistors in parallel and get .25. I like this idea. Do you?

That is what I would consider doing. What remains of the prototype can be seen here. I used low resistance Ohmite brand "stackohm" series resistors just like these from DigiKey. The obvious downside is the cost at about $17 USD per unit. When the final load bank was built I opted for using RES ALUM HOUSED WW 1.0 OHM 50W - 850F1R0E which while not real cheap came in at less than half the cost of those used in the prototype unit. The prototype used manual switches to switch the loads in and out but the final unit used logic level mosfets. The actual final unit also used some pots to trim the measured current which was only because the loads were critical and monitored.

Something you may want to watch is the wire gauge to and from the load resistors. The aluminum housed resistors were easily connected in parallel configurations with AWG 12 as bare AWG 12 fits through the resistor terminal lugs. I used AWG 8 for to and from the loads in the interest of reducing the I*R drops as much as possible. If you figure about 6 volts at 25 amps you have about 150 Watts, thus using 4 EA 1 Ohm 50 watt resistors in parallel would be a good way to go. I would mount them to a small piece of 1/8" thick sheet aluminum and you may need a small cooling fan.

While you aren't monitoring current it wouldn't hurt on initial testing to beg, borrow a clamp on ammeter just to make sure the current is about where you want it at 6 volts. That would be my best guess anyway.

Just My Take
Ron
 
Load Box

Take a look at this one. We can make you a better schematic and if your not in a hurry a board layout. If you know for sure the voltages you want to use for the drop out of the load we can use some fixed resistors instead of the adjustable. Up to you.
I like Ron's metal resistors, but I get them from Mouser (1/3 cost):D
And yes they will share the current, so 4 50 watt ones should be okay if you mount them on a plate as Ron suggested.
 
Love your old site!

Some day I need to get it there and clean it up. Completely blow away all content and rebuild the thing. That site is over 10 years old and was originally named after Bear a large golden retriever (mixed mutt) we had. There are also a dozen or so folders out there not available to public view I store more crap on. However, every now and then it comes in handy. Like the little load bank demo.

Thanks
Ron
 
ronv
I may have said this before but I'm not an electrical engineer. I am an electrical that works for Texas Parks and Wildlife and as a part of my job I maintain several solar installations. I can build all most any project when it get down to the workbench level but when it is in the design stage I can get lost rapidly.
The circuit you posted with the title of Load. Is it a constant load or a complete LVD and load?
I read up on the LM339 but it seems to me it could preform in either function. I know it is all basic to some one like your self but a bit of explanation would really be great.

You made mention of a fix verses adjustable resistance. I am assuming this is in reference to the cut off voltage of the LVD.

Although this project design is not effected by Peukert Law it is directly effected by temperature.
A batteries' capacity is directly effect by temperature. So the test unit either has to be calibrated before each test for the LVD voltage or I will need to do a bit of math at the end of a test to make a time adjustment. I think the prudent way to go would be with fixed voltage cut off setting.
6 volt battery complete discharge 5.25 volts
50% discharge = 6.10 volts
80% discharge = 5.99 volts
These are all based on the 80 degree mark so at the end of the test a + or - value would have to be added to the time.
If we were to go the other route a + or - figure would have to be calibrated into the LVD before each test. Voltage is also effected by temperature.
I have a chart that will suffice in lieu of an actual formula for the capacity variations due to temperature. I wish I had an actual formula but I have never be able to find one.
See attachment
I would be great if I could select between any of the three test values.
Finally take a look at the timer I attached.
Will it work with your design?
As always thanks! Without your assistance and the input of other on this forum a project like this would be impossible for a person like myself.
 
The circuit is both. Here is how it works:
The circuit at the top is the load. When the switch labled push to start is pushed the relay will turn on and the 4 1 ohm resistors in parallel are hooked across the battery as the load.

The lower circuit is the lvd.
V2 is a 12 volt power source for the relay and the circuit. We have to decide where we will get this voltage.
R5 and U2 create a very accurate voltage reference at 2.5 volts. This voltage goes to the - terminal of the comparator. The comparator compares the - terminal to the + terminal. If the + terminal is more positive than the reference voltage on the - terminal the output will go plus turning on U3 which in turn "picks" the relay. U3 is only there to get enough power to switch the relay. The potentiometer (U4) allows adjustment of the battery voltage at the trip point to match the reference voltage. We can replace this with 3 voltage dividers and a switch to get your 3 ranges; although I think to get the accuracy you will need to set each one once. Do you have a good voltmeter? I bet you do.. When the battery goes just below the trip point the + terminal of the comparator is more negitive that the reference and the output of the comparator will go to zero and the relay will open and disconnect the load and turn off the timer (which I didn't show but uses a second set of relay contacts).
I'll be gone most of the day tomorrow, but some of the other guys may have some other ideas or can help if you have questions.

Seems like the timer should work.
 
Last edited:
Te Model TH-7NM is the model timer you want. I have used the same family of counter/timers under a few brand names. They last about 3 years plus on an internal coin cell battery. Set it up for a dry contact closure to run.

Ron
 
I am really getting excited about proceeding with this project. How do you think I should Iproceed.
ronv Thanks for the explanation. I do have a digital volt meter 6V 1mV ±(0.8% of rdg + 3 digits) I have experience etching a PCB and feel quite confident I can handle the workbench stuff. However I will need your guidance in the design. Are there software programs for laying out a PCB? One of the things that attracted me to the LVD I refer to in an earlier post was it was all laid out so that an individual such as myself could make the PCB purchase the necessary parts and put it together.

The way I envision this thing working and looking is this.
Attach two leads to the battery to be tested.
The unit need to be self contained without the need for an external power source
Optionally it would be nice to have a volt meter or an LED to indicate the battery was sufficiently charged to continue the test. This would eliminate the need for two pieces of equipment.
Select one of three possible tests Full discharge, 80% or 50%
Press a button to start the test cycle
Again optionally it would be nice to have some visual indication (LED) the test was in progress. So that a a glance from a distance it could be seen the test was still in progress.
Finally LED goes out test complete
Record time from timer and do a bit of math

I feel as if I'm imposing asking for so much guidance but I will do all I can to help. I can list out all the parts research a vendor and then ask your approval.
I can make a PCB or have one made and after that the actual assembly is the fun part.

If you and the other on this forum are willing to assist me I would be most appreciative as completed this would really help me in testing a battery bank.
 
Projects that acually do something are the most fun. So lets go for it.
You'r the battery guy so what do you think about powering it off of the battery under test? At 6 volts the circuit will probably draw about .2 amps. Most of that driving the relay. We could, I supose, reduce this to .02 or 03 amps by using a solid state switch. Probably not much more complicated and we could probably get that much from a "dead" battery. Short of that we could do the low power version (solid state switch) and use AA batteries to power it.
We will have spare comparators so a full charge indicator should be easy enough.
Also pretty easy to do a test in process indicator.
It will probably keep you busy figuring out how to package the whole thing.
One other thing I worry about is do we need to protect against hooking the battery up backwards?
If you don't mind we can get the parts from Mouser. They have a bill of material program so we can all see the datasheets and when you are ready to buy you just push the button.
Have you made a pcb before?
 
Last edited:
Yes I have made a PCB before.
I am a true tinker and done a number of electronic projects before.
Years ago I hung with the Ham radio operators and even built a decent receiver but that was years ago when semiconductor and IC circuits were still just dreamed about. When transistor were a big as a bean and radios were rated by how many of them they had. Guess I'm showing my age a bit.

I guess in a perfect situation the tester would not draw any current from the battery being tested.
However having said that testing lead acid batteries is not an exacting thing even in the lab under perfect conditions. Lead acid batteries are fickle. There are variables in the field that can't be easily controlled so a + or - 10 % is acceptable. That is why the compromise with the load was a good suggestion. + or - the manufacture specifications not + or - within the test procedure. The tester needs to be able to reproduce the same test standard this allow for comparisons which establish trends. So if the test loads the battery at a mean value that is reproducible that's OK.
Are the batteries failing, how much of their capacity have they lost are far more difficult questions to answer than are they bad. This can only be determined from a series of repeatable tests over a period of time.

For now I'm not concerned about hooking it up backwards but if I told you I never made a mistake I would be lying. Also if I should let anyone else use the tester it would be nice to have the additional protection if it not too difficult.

Mouser is fine

I got to go for tonight. Tomorrow requires me to be awake so I will check back in then.
 
Okay, Lets see what it will look like. I'm thinking the fully charged value would be like 2.1 volts per cell with no load?
 
One more spec issue before we start.
Will these values work:
Fully charged - 6.351v
80% - 6.094v
50% - 6.000v
80% - 5.256v
If these are close enough we can eliminate the adjustments and just use 0.1% resistors.
 
This is the parameters of the project:

Flooded Lead Acid Battery 2.117 volts per cell
Fully charged 6 volt battery ≥ 6.35 volts (Green go no go LED indicator on)
80% SOC = 5.99 volts (only 20% SOC left in battery)
50% SOC = 6.10 volts
Complete discharge = 5.25 volts

Although SOC measurement are made open circuit these tests are under the prescribed load until the trigger voltage is hit. It is very likely the voltage will recover somewhat after the load is removed. The only measurement that is OCV is the fully charged go no go LED indicator. If the tester itself generates any significant load this may need a bit of calibration.

As to the question of tolerance there is only 1.1 volt difference between a fully charged battery and a completely discharged one. There is .85 volts difference between a 50% SOC and a completely discharged battery.
This is why a reliance solely on a voltage reading to determine SOC is an iffy proposition at best. Many RVs come equipped with SOC indicator that merely monitor battery voltage and many an RVer has been unpleasantly surprised.
For the tester we are proposing to make the tolerance must be pretty close and the ability to recreate the same test standard critical. I envisioned that it would be necessary to place a trimmer pot in each of the circuits to make final calibrations.

You may be interested in who and how the final voltage representing a completely discharged battery was arrived at. It is at that point (5.25 volts) that the curve drops off precipitously and there is no useful capacity left in the battery.

In any earlier post you voiced some concern about the amount of current the relay might draw. Since we have already settled on a mean value for the load I can’t see that the additional draw would be significant.
However I did run across this simple circuit for reducing the current draw of a relay.
It came for someone named Bert (a Ham radio operator) and his design for a LVD.
I have attached both for your perusal.

Tomorrow I have got a 16 battery bank to pull apart and evaluate. Sure could use this tester but for now I will have to get by.
 
I believe back in post #14 ronv stated:

The relay needs to be good for at least the 25 amps.

The concern here is that the relay contacts of the selected relay need to easily handle the 25 amp load current the relay will be switching. I don't think the issue of concern is the relay's actual pull in or drop out voltage or coil current. If taking the relay route for switching the load in and out you may want to consider a beefy automotive type 12 volt coil relay preferably designed to handle about 35 amps. Switching 25 amps DC using a relay requires some pretty beefy relay contacts.

Ron
 
I know this thread has started to get pretty long but back in post 30 ronv stated
At 6 volts the circuit will probably draw about .2 amps. Most of that driving the relay.
This is the issue I was speaking to in my last posting.
Your point about the relays ability to handle the proposed current is a good one.
Selecting the right component will be important.
Thanks for you input!
 
I am not sure what ron has in mind for the relay. Yeah, I missed that little part. There are a few options and I do not see it as a problem but will wait to see what Ron has in mind as he has a really great circuit planned out so I'll stay reserved for now. :) THis will work out fine.

Ron (the other Ron) :)
 
Load Box

The more I looked at high current Relays the less I liked the idea. So here it is with a solid state switch. I've attached a schematic and layout, but If you download Express PCB (Free) I'll attach the files once I look them over a bit. Then you can print out the artwork.

Maybe everyone could take a look and see if there is something wrong with it.

Some concerns I have:
The 339 can't drive much current so turn on will be kind of slow ( 10 usec) but since it only turns on and off every few minutes I guess it's okay.
Is this the correct way to turn on the timer? It has its own power?
Not sure that the resolution on the 1K trimmers will be good enough to hold a few mv on the trip points.
 
Is this the correct way to turn on the timer? It has its own power?

Yes, either a open contact closure or open collector as you drew it. I would just use a 2N3904 switching transistor. Transistor ON count increments up in min. Transistor OFF count stops. Need to allow for a small N/O reset button.

I'll try to look at more later. Traveling for a funeral so I am not at home and sort of harried.

Ron
 
Well I am ready to take this project from planning to the workbench but there are still a few question.
I will address my concerns singularly, one per post so as to avoid and confusion.
PCB
I have some experience with etching a PCB
This particular one is simple and should not present any problems.
When doing some research on the current methods and material I found there are a number of companies that will make the PCB for a relatively low cost.
My question is this:
Does it make economic sense to purchase all the materials and go through the process of making a PCB or is it more prudent to just have it made?
Understand that whatever materials I would purchase probably will not find another use for a long while.
If in fact it is better to farm it out do you all have suggestions as to what company might offer the best service?
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top