Help creating a 200 volt DC discharge circuit

Hi, I am looking for some pointers on how to perform a controlled discharge of a 200-230 VDC battery pack (i know - danger!!). I need to do this to properly balance the pack. I already have a charger, it is just the discharge side I need, as it needs to be very precise to avoid cell damage.

I found this Smart battery discharger/tester - Exhibition - Arduino Forum on discharging small 1.2v lithium cells with a light bulb as load, and I have found a lot of people using the same bulb method manually discharging packs like mine and monitoring with a multimeter, but it is extremely impractical as it takes a couple of days and you must constantly check voltage to know just when to disconnect... time is no big deal, but disconnecting at the wrong voltage is the increasing risk over such a long time.

I could do something similar using a high voltage DC relay, may even potentially need a couple of relays, ie a typical 12-24v relay shield that controls a much higher capacity relay.... but having a hard time finding something that does high DC.

I need to step the battery down in 3 phases, stopping at each level, and lowering discharge current 50% for the next phase, so I could either connect a different load on the relay at each phase, or program a PWM signal to lower it.

Any pointers for particular hardware components are extremely appreciated. This could be extremely useful to others, I bet there are many e-bike users searching for something like this.

"...discharging small 1.2v lithium cells with a light bulb..."
A lithium cell that has dropped to 1.2volt can't be revived anymore.
Lithium cells might ballance better "top end" (stop cells getting >4.2volt).
NiCad cells were occasionally discharged before charging, to reduce memory effect, but NiMh cells don't need that.
Need more info to answer your question.
Leo..

Hi,

Hi, I am looking for some pointers on how to perform a controlled discharge of a 200-230 VDC battery pack (i know - danger!!). I need to do this to properly balance the pack. I already have a charger, it is just the discharge side I need, as it needs to be very precise to avoid cell damage.

How many cells and will you be monitoring each cell to make sure the balance takes place and you don't irreversibly kill a cell?

What battery type is the pack, how many Ah?

Thanks. Tom. :slight_smile:

Light bulbs are among the cheapest power resistors you can get, in the 200V range. Pick your wattage and let it glow!

(Are you sure you don't need to discharge cells separately?)

This is mostly irrelevant, but it is ~220v 6.5ah NiMH, there are already a few good examples on handling different chemistry. I am already set on using a couple 100w 120v bulbs for load, not interested in balancing the discharge, that will be handled on the charger, and yes it should be fine to discharge the whole pack together.

The big problem is finding the proper relay and monitoring circuit, some sort of variable voltage attenuator to drop down to 5v range readable by Arduino, that can automatically range between 240v-15v. With each discharge taking several hours, I need to set different programmable stop levels all the way down to around 16 volts.

Hi,

This is mostly irrelevant, but it is ~220v 6.5ah NiMH,

This is very relevant as it helps to decide on the discharge current and the discharge level.
How are the cells configured and do you want to measure each cell assembly to check voltage balance?

Do you have a idea of how you want to lay your project out?
If so can you please post a basic block diagram showing how?
A picture of a hand drawn circuit in jpg, png will do the trick.

Can you please tell us your electronics, programming, Arduino, hardware experience?

Thanks.. Tom.. :slight_smile:

I dont need to monitor cells, just need to drop the pack down to X voltage and disconnect the load relay. This pack is accustomed to 50a discharges during normal use, I plan to do a 1.5-3a slow discharge in 3 steps to prevent polarity reversal, it will be more than safe for it, I just need to disconnect or step at the right volts.

I already know what volt levels I need to step down at, and have already tried it manually.

People have been doing this without Arduino using 3 methods, I feel it can easily be automated on an Ardiuno and a high volt dc relay, using the same loads that people have already proven.

Many people do 4-6 cells at a time using multi-channel RC balancers like hitec x4, while dummy proof and automated, this is extremely time consuming and messy with the need to unwire the whole pack, and still takes several days to perform because these are only able to handle individual cells, not 25-30 in series.

The "manual" method many use: a set of 120v bulbs, starting with a pair of 200w bulbs in series, once the pack dips to 140v a 75w down to 85v, then a 25w bulb down to a 20v final cutoff.

I am no expert, but I can pick up a lot with a little guidance. I have had a few arduinos/RPi for 4-5 years, one is running a remote controlled car, it was a kit, but I had to correct a lot of broken code shipped from china to make it work right, and I am teaching my daughter how to code sketches. I have a couple other little projects, far from expert, but my feet are wet. I work for a company that does commercial/residential automation and IT services, I help manage ~100 servers both windows/linux, I code a number of scripting languages. I have built and installed on and off-grid solar systems (not professionally). I have a couple of soldering irons and 2 shaky hands. When my hands were younger I soldered a ZIF socket in my Honda ECU and burned my own EPROMs for custom tuning, I know how to use Digikey/Mouser ;). I am not great on circuit design, but I have built my own LED lighting from scratch, and nothing burned. So I feel like I have basic electro/mechanical skills and coding skills necessary to do something basic, and with a little guidance, maybe something more complex.

I have a few options for design:

  1. Set it to disconnect/wait at each voltage level and do a manual swap of the load attached to the relay (200w, 75w, 25w).
  2. Use 3 relays (pricey)with 3 loads, and I switch on each relay automatically at the desired voltage.
  3. Build a complex circuit that steps down the load at each configured voltage point, this is ideal, but may be taxing my abilities unless someone really spells out the exact circuit for me. I could probably figure it out on my own after a LOT of trial/errors, time spent would be a deterrent.

I found this 240vdc relay for $65, any cheaper suggestions?... hoping to do the whole project for ~$75.
http://www.electricmotorsport.com/solid-state-relay.html

Here is an example that someone else designed, it does exactly what I need, just at a lower voltage, I have removed a couple items, and the shunt is even optional, I dont necessarily need to track the AH, but I think I might already have a 250v shunt somewhere. The sketch for this circuit already does 99% of what I need, just needs a couple of items trimmed and add my voltage cuts.

totalimpact:
I am looking for some pointers on how to perform a controlled discharge of a 200-230 VDC battery pack...
I need to do this to properly balance the pack.

totalimpact:
I dont need to monitor cells, just need to drop the pack down to X voltage and disconnect the load relay.

I don't see how these two statements can go together.

totalimpact:
I plan to do a 1.5-3a slow discharge in 3 steps to prevent polarity reversal...

Polarity reversal occurs at any current.
When the battery is flat, the other ones in the string start charging it with reverse polarity.

totalimpact:
The "manual" method many use: a set of 120v bulbs, starting with a pair of 200w bulbs in series, once the pack dips to 140v a 75w down to 85v, then a 25w bulb down to a 20v final cutoff.

Discharging a 220volt pack to 20volt?
I think deep discharging was sometimes done for NiCad, but never for NiMh.
AFAIK service life of NiMh degrades quickly below 0.9volt/cell.
Leo..

Hi,
That SSR is;
Crydom Solid State Relay 25A 240V DC IN.
It is an AC switching SSR, which is not suitable for your application.
You are switching DC.

Tom... :slight_smile:

Wawa:
I don't see how these two statements can go together.
Polarity reversal occurs at any current.
When the battery is flat, the other ones in the string start charging it with reverse polarity.
Discharging a 220volt pack to 20volt?
I think deep discharging was sometimes done for NiCad, but never for NiMh.
AFAIK service life of NiMh degrades quickly below 0.9volt/cell.
Leo..

This is not AFAIK info, it is fact, thousands of people are dropping voltage BELOW 20v, in fact multiple commercial products have been engineered and are sold that do exactly this. A company in Australia ships world wide, and Prolong chargers/dischargers ship world wide out of the USA. I don't claim to be a battery genius, but I don't have to be an expert to follow someone else's proof of concept. This has been proven forwards and backwards, I don't need to know anything further about the numbers, they work.

Forget about .9v, dramatic capacity improvements are found by using proper stepped amperage, all the way down to .1v per cell. The cell capacity and overall lifespan are both improved. This is not done on the first cycle, it is done on 3 charge discharge cycles. First a shallow discharge and recharge, then a deeper, and recharge, then all the way down to .1 volts/cell on the dot.

The bank is cooled while cycling, and following proper precise step levels there is little risk of cell damage. People are literally doing this in mass the stone age way with a volt meter and a few light bulbs.

If someone can point me to a proper DC relay for this application I might be able to figure out the rest. I have seen several DC relay in the $400 area, but I can buy an already built product for a lot less, so someone must make the relay for a lot less.

I would rather use a high voltage mosfet than a relay to switch 200volt (0-3A resistive) DC.
No logic level mosfets for that voltage, so you also need a mosfet driver.
No experience there.
Leo..

That sounds like an excellent idea, thanks so much, I am looking in to mosfets/drivers.

I found TSM70NB1R4CP at Mouser that has 2-4v gate, 700v/3amp drain, 1.1ohm Drain-Source Resistance. Would this be safe to move 240vdc/3amp?

Finding something with a 5v gate is extremely difficult, there is nothing available in the 200-300v range.

HI,
I think you had better explain what you are building.
It is a "Battery Reconditioner", this is not designed for regular discharge cycles of NiMh battery packs used in some Hybrid Cars

https://hybridautomotive.com/pages/recon

When attempting to recycle NiMh the terminal discharge level is monitored and stepped.

https://hybridautomotive.com/pages/sd#termination

I would not be doing this to a good pack of NiMh battaries especially a 240cell battery pack (240 * 1.25 = 300Vdc )down to 24V,

A single cell down to 0.1V.

NOTE:

  • You should fully charge & balance the battery pack before any discharge session. This will ensure the cells will be balanced when beginning the process.

They are saying that the battery pack should be balanced before reconditioning, this reco-process does not appear to fix reverse charged cells in a series pack, just rejuvenate a poorly performing pack.

Tom... :slight_smile:
PS. Why didn't the OP put a link to this to properly explain what he/she wanted to do?

totalimpact:
Finding something with a 5v gate is extremely difficult, there is nothing available in the 200-300v range.

I smell the common misunderstanding of Vsg(th).
This is the point where the mosfet turns off.
Drive (turn on) voltage needs to be several times that turn off voltage.
Just look at max drain voltage and DS resistance, not at Vgs(th).
You need a driver running on 12volt for any of these mosfets, for a gate drive voltage of 10-12volt.
Leo..

off topic: I can explain just 1 thing for sure, I am not a circuit engineer, this is why I ask questions, but you have posted a working example and say it cannot work?? These are being made in the US and Australia by 2 different companies, and same procedure also being performed manually by hundreds of people with long term success... sounds like it works pretty well.

back to the topic: correct, I quickly confused VGS... although nowhere on the mouser spec sheet does it mention 10v, but of course its right there on Digikey.

So I am thinking maybe something like a 4N25 opto-isolator to drive the gate would be best??