Need some help. I soldered up one of these this evening. Using LM317T regulators off Ebay. Problem is the regulators are not pulling any current. The inline watt meter only shows ~2 watts being pulled from the 4s lipo pack. That is from the fan on the heat sink.

I put a volt meter between the positive leg on a regulator and the heat sink. Read 15+ volts. So it seems like the connections are correct, right? Any ideas what I did wrong?

See attached pics. Thanks.

Kcobra,
Everything looks OK. Can you check to make sure you have continuity(short circuit, 0 ohms) between the regulators adj and Vout pins (soldered together) and the negative lead where it is soldered to the metal mounting strip? If you don't have a ohmmeter, you can try putting a jumper from the negative lead to one of the regulators at the point where the 2 terminals are soldered together (adj and Vout) to see if you pull a couple of amps. Who is the manufacturer of the LM317T's you used?

Spog: Thanks for the advise. I used the continuity function on my multi-meter to confirm there is a continuity between the two soldered pins of each regulator and the negative lead soldered to the mounting strip.

Not sure on the manufacturer. At this point I'm thinking I got a bad batch of regulators. I got a bag of them from a cheap Hong Kong seller. I tested out each un-soldered regulator by jumping the negative lead of a battery to the adj and Vout pins while touching the positive lead to the input pin. Almost all of them read 0 watts. Only a couple pulled some watts but those two immediately started to smoke. So all of that got me thinking that maybe the input pin was on the opposite side of these particular regulators. Doing my test with the reverse connection did actually pull some watts but it was low and varied from regulator to regulator. Some showed only a couple watts while others showed 6 to 8 watts. Never higher than that though.

Anyways, sounds like I'm a victim of the old saying, "You get what you pay for". Anyone recommend the cheapest place in the US to get some reliable regulators? Thanks.

I got mine from Digi-Key. Digi-Key ships first class mail which is great for small items such as this.

http://www.digikey.com/product-searc...ds=LM317TFS-ND

Did you add the cap too? I recall one example of the schematic I saw indicated a cap is required.

http://www.digikey.com/product-searc...ds=445-8532-ND

Good point about the capacitor Ohmic. It should always be there to prevent regulator instability. Did you get your discharge load completed?

I need to order a TAP. Will order one this week so I can complete mine. Been meaning to do this for a while now.

Originally Posted by jglenn

Look at a chopper with fets, I am using OmniFets which have current limit and
temp protection. That way you can control the current. And only have one
or two TO-220 devices. For 50A.

Jglenn,
Do you have a schematic of your TO-220 setup?
Thanks...
Gus...

Gus
There is no need for a chopper (PWM switching) to complicate things. A FET can be operated in linear mode and dissipate most of the power, while still maintaining tight current control. There is a thread here somewhere in the DIY electronics with a schematic if you're interested. It's a more expensive setup than the LM317 design, but it provides current regulation which is required for battery testing. For a simple discharge to storage device, or 3S and below battery testing, the LM317 design is tough to beat for cost/power. Both designs provide as much power as you want, limited only by your heat dissipation.

Totally agree. The energy has to go somewhere. It either gets dissipated in the FET operating in linear mode or in an additional external load if you operate the FET as a switch ( which adds a lot of circuit complication).

The thing to be really careful off is that MosFETs are not really designed for linear operation and most are poor for that. Don't just take current and voltage ratings and assume they will cope. You MUST use a FET that quotes a genuine DC line in the SOA curve and stay within it AND ensure you have a heat sink that keeps the die temperature below the quoted rating. It's not all that easy - I'm speaking from practical experience.

The lovely thing about the LM317 solution is you can ignore all that. The magician inside the chip looks after it.

Originally Posted by spog

Gus
There is no need for a chopper (PWM switching) to complicate things. A FET can be operated in linear mode and dissipate most of the power, while still maintaining tight current control. There is a thread here somewhere in the DIY electronics with a schematic if you're interested. It's a more expensive setup than the LM317 design, but it provides current regulation which is required for battery testing. For a simple discharge to storage device, or 3S and below battery testing, the LM317 design is tough to beat. Both designs provide as much power as you want, limited only by your heat dissipation.

A wise warning from jj604. If you operate a FET out of it's SOA, bad things will happen, like explosion and fire. Been there, done that.

JJ604 & Spog,
Thanks for the quick answers.
Gus...

Decided to build my own discharge boxes after having 2 iCharger 4010's fail after storage discharging. While the failures now seem to have been caused by a firmware issue I figured it was a good idea to storage discharge into something cheaper rather than dump all that heat into the 4010s.

I initially planned on building the Cell Log circuit using an 1157 lamp array for the load but the LM317 solution is smaller, simpler, and can cope with a wide range of voltage without any changes. Since I use packs between 2S and 6S the LM317's made more sense than different arrays of 1157s to cope with the input voltage difference.

I used the latching relay diagram and a Cell Log described in the other thread and the load is a CPU heatsink/fan and 10 LM317s. The LM317's are attached to the heatsink with 6-32 socket head cap screws and there is a thin smear of Arctic Silver thermal compound between the LM317 heatsink and the CPU heatsink.

The CPU fan is powered off the DC-DC buck/boost converter which is set for 14V output.

Below the banana jacks is a AGC fuse holder for short circuit protection in the event something goes amiss inside the discharge box, the positive lead from the packs goes immediately into the fuse holder before powering anything else inside the discharge box.

The 5 switches on the front are wired as follows:

1. With all switches off a single LM317 (~25W) is connected. This "primary" LM317 also has a 0.1uf cap connected between positive & negative.

2. The lower 3 switches each connect to individual LM317s, so each lower switch controls an additional ~25W.

3. The 2 upper switches are each connected to 3 LM317s for ~75W.

With this switch & wiring combination I can select between 25, 50, 75, 100, 125, 150, 175, 200, 225, and 250W discharge load with different combinations of the 5 switches. I'm concerned the heatsink & fan combo will not be adequate to keep the LM317s below their thermal limits for the 200-250W load cases but I haven't tested a sustained 200W+ discharge yet. Most of the time it will be used at the 25W, 50W, and 150W settings. I've tested it up to 150W discharge so far and at 150W the heatsink & fan combination was more than adequate.

It works great although I'm still tweaking with the low cell voltage alarm settings on the Cell Log for my various packs such that after shutoff the pack voltage "rebounds" to ~3.85V/cell.

Originally Posted by Kiba

Decided to build my own discharge boxes after having 2 iCharger 4010's fail after storage discharging. While the failures now seem to have been caused by a firmware issue I figured it was a good idea to storage discharge into something cheaper rather than dump all that heat into the 4010s.

I initially planned on building the Cell Log circuit using an 1157 lamp array for the load but the LM317 solution is smaller, simpler, and can cope with a wide range of voltage without any changes. Since I use packs between 2S and 6S the LM317's made more sense than different arrays of 1157s to cope with the input voltage difference.

I used the latching relay diagram and a Cell Log described in the other thread and the load is a CPU heatsink/fan and 10 LM317s. The LM317's are attached to the heatsink with 6-32 socket head cap screws and there is a thin smear of Arctic Silver thermal compound between the LM317 heatsink and the CPU heatsink.

The CPU fan is powered off the DC-DC buck/boost converter which is set for 14V output.

Below the banana jacks is a AGC fuse holder for short circuit protection in the event something goes amiss inside the discharge box, the positive lead from the packs goes immediately into the fuse holder before powering anything else inside the discharge box.

The 5 switches on the front are wired as follows:

1. With all switches off a single LM317 (~25W) is connected.

2. The lower 3 switches each connect to individual LM317s, so each lower switch controls an additional ~25W.

3. The 2 upper switches are each connected to 3 LM317s for ~75W.

With this switch & wiring combination I can select between 25, 50, 75, 100, 125, 150, 175, 200, 225, and 250W discharge load with different combinations of the 5 switches. I'm concerned the heatsink & fan combo will not be adequate to keep the LM317s below their thermal limits for the 200-250W load cases but I haven't tested a sustained 200W+ discharge yet. Most of the time it will be used at the 25W, 50W, and 150W settings.

It works great although I'm still tweaking with the low cell voltage alarm settings on the Cell Log for my various packs so it shuts off at ~3.85V/cell.

Kiba,

That is some SERIOUSLY impressive stuff right there. VERY nicely done. *thumbs up*


-Shane

Another very nice discharger, awesome work Kiba.
I think that heatsink will handle all 10 regulators. Please let us know how it works out for you.
Nice to see a fuse.
Where did you get that heatsink?

Originally Posted by Kiba

Decided to build my own discharge boxes after having 2 iCharger 4010's fail after storage discharging. While the failures now seem to have been caused by a firmware issue I figured it was a good idea to storage discharge into something cheaper rather than dump all that heat into the 4010s.

Awesome piece of work! Congratulations!