# Thread: Schematic question for LM138 voltage Regulator

1. Transistor
Join Date
Jun 2008
Posts
36
Rep Power
25

## Schematic question for LM138 voltage Regulator

Hey everyone. In order to use just one power source for my robot, I'm building several voltage regulators to distribute power as needed. At the moment I have two LM138s (http://www.ti.com/product/lm138), which I'm going to use to run 7.5v to the four motors driving my bot (one LM138 to two motors). I've successfully soldered on regulator circuit, according to this diagram here:

Now my question is, when I solder the second circuit onto the board (sharing the same source voltage and ground obviously), am I going to replicate that schematic again exactly, or will I have to make changes to it? I don't know much electrical engineering, but that 0.1 uf capacitor is essentially going to be shared by both circuits, since it's just bridging the source voltage to ground. Replicating that diagram again will put another 0.1 uf capacitor in parallel with the first one, and I'm not sure how it will affect my circuits, if at all.

2. ## Re: Schematic question for LM138 voltage Regulator

It will somewhat improve the ability of the voltage regulator to reduce any voltage fluctuations/ripple on the input voltage source making it through to the output. You could probably do without it, but it would not hurt to include it.

Out of curiosity, what is the main power source that feeds the regulators and how much current do the motors require? Linear regulators create a voltage drop between input and output by essentially being a well controlled vyou will almost definitely need some sort of heatsinking to keep the LM138 from failing.ariable power resistor (best description I can give without going into the actual transistor circuitry design) and get rid of the excess voltage as heat (P = I*V). The more current you need and the larger the voltage drop, the more heat you have to dissipate from the device (usually assisted with a heatsink and possibly forced air cooling or even water cooling in really high power situations). The more the device heats up, the less current it can provide to the output. So the better the heat transfer out of the device, the longer the life of the device and the more power it can dissipate without damage (thus more current at regulated voltage to the load).

The TO-220 package LM317 that I wired up for providing +9V to the wireless router on darsha is powered by either a 11.1V LiPo or a 12V SMPS and is usually loaded to less than 0.5A. The power required to be dissipated from the regulator is ~1.5W depending on conditions and gets pretty warm even with a large-ish heatsink.

3. Transistor
Join Date
Jun 2008
Posts
36
Rep Power
25

## Re: Schematic question for LM138 voltage Regulator

Actually I'm glad you brought that up. I'm going to have a small fan to suck the hot air out of my robot but it looks like I might need a lot more. My plan was to have two of those regulator circuits in parallel providing 7.5 volts to a single motor controller, which would then be driving four motors. Their stall torque is 1.6 amps each, for, in the worst case scenario, 6.4 amps total, or 3.2 amps per regulator (Their max amperage is 5A). The battery is going to be a 12v NiCad (Though in reality it varies from 12-15 volts).

But my next question is related: If I have the Vo of both regulators hooked up to a single device, will the totally amperage that device draws be split approximately evenly between the two regulators?

Also, would Radioshack conductive glue suffice to attach a heatsink the tops of the regulators? I'm wondering if it might act like a thermal paste as well as binding heatsink to the regulator casing
Last edited by wireframewolf; 07-26-2012 at 04:26 PM.

4. ## Re: Schematic question for LM138 voltage Regulator

Originally Posted by wireframewolf
Actually I'm glad you brought that up. I'm going to have a small fan to suck the hot air out of my robot but it looks like I might need a lot more. My plan was to have two of those regulator circuits in parallel providing 7.5 volts to a single motor controller, which would then be driving four motors. Their stall torque is 1.6 amps each, for, in the worst case scenario, 6.4 amps total, or 3.2 amps per regulator (Their max amperage is 5A). The battery is going to be a 12v NiCad (Though in reality it varies from 12-15 volts).
Is there any particular reason you are not using a battery pack with a voltage more appropriate for the motors and motor controller that you are using? Using a linear regulator to dump nearly half the output voltage (and thus nearly half the overall power) of a battery as heat is not very efficient and maybe even a potential burn and/or fire hazard. Seriously, the TO-220 package can go as high as 125[C] and the TO-3 package can go as high as 150[C] before shutdown, both of which could easily cause a nice burn or melt certain plastics. (This warning brought to you by an accident prone guy with a now two day old burn from liquefied solder being splashed onto his finger when tinning a wire as well as multiple scars from previous burns and cuts. Fortunately, on both occasions when I stepped on a nail, in a single ~10 minute time period no less, I managed to keep from putting all my weight on the at risk foot so no tetanus shot for me - still pushed through the shoe and sock to bare skin).

In a worst case scenario, you are talking about (15[V]-7.5[V])*6.4[A] => 48[W] wasted as only heat in a presumably confined space. That is more power than the adapter of my netbook can produce without failing due to an over-current fault (does not require even 40[W] when powering the netbook under heavy load while simultaneously charging the battery). The netbook itself is pretty frickin' large compared to a single TO-220 voltage regulator and it still gets hot enough to be painful (without anywhere even close to all of the input power being wasted as heat).

Originally Posted by wireframewolf
But my next question is related: If I have the Vo of both regulators hooked up to a single device, will the totally amperage that device draws be split approximately evenly between the two regulators?
On the top of page 16 of the datasheet you linked it actually has a schematic for creating a 10A regulator with two LM338 and an Op-Amp to ensure the output voltage of both regulators is identical and the current mostly equally distributed between them (with a minimum load of 100mA for accurate voltage regulation). There is a similar Op-Amp controlled 15A regulator on the top of page 15 (you should be able to add or remove extra regulators with out major modifications although I am not in electronics analysis mode right now, so I cannot say for certain). There is also a "quick and dirty" way to connect multiple regulators in parallel on the bottom of page 13 (10mA minimum load and +/-100mV matching between all regulators). The "quick and dirty" method is cheap with needing only two resistors to control all the regulators (each regulator with a single diode between the adjust pin and the output pin), but the +/-100[mV] range of voltage matching between all the regulators means that there will be some excess power loss due to "fighting" between the mismatched regulator output voltages.

Originally Posted by wireframewolf
Also, would Radioshack conductive glue suffice to attach a heatsink the tops of the regulators? I'm wondering if it might act like a thermal paste as well as binding heatsink to the regulator casing
I doubt the graphite filled conductive glue from radioshack is going to be a very good choice. Thermal adhesives are not terribly expensive (\$15 for 7 grams of Arctic Silver Thermal Adhesive at newegg with 4-7 day free shipping) and are intended explicitly for bonding (permanently) heatsinks to electronic devices while minimizing the thermal resistance between the two and maximizing the electrical resistance (may be good or bad depending on device and heatsink designs and operating environment - good if the heatsink might accidentally come into contact with metal object that would short output to ground - bad if the heatsink is intended to be used as the electrical output terminal of the regulator). The LM317 that I used for the router is connected to its heatsink (also from radioshack) with leftover Arctic Silver 5 and a #4-40 screw, but I kind of wish I had gone with adhesive since the tab of the LM317 from radioshack is so thin it bends and does not maintain great contact with the heatsink except at the screw.
Last edited by tician; 07-26-2012 at 08:49 PM.

5. ## Re: Schematic question for LM138 voltage Regulator

You would replicate that circuit directly.

However, I don't think driving motors with a linear regulator is the right thing to do. You will reduce efficiency of the entire system, and you will only be able to drive small motors. Note that the regulator has to be able to continually supply the full stall current of your motor, which may be quite high and require significant cooling for the regulators. It sounds like you've thought some about this already, and perhaps it's the right solution for your particular situation. If you want to try something else, I have some suggestions for what might work better:

1) A high-power switching regulator. This can get very efficient -- even the cheap versions will get over 90% when designed right. You can buy switching regulators that support several amps at pololu; for higher currents you'll need "DC DC converters" which can be bought ready-made, or built from parts.

2) Using PWM for driving the motors, and driving directly with whatever voltage you have available. As long as it's "in the ballpark," you can drive the voltage straight into the motor, and simply regulate the heat loss in the motor (which is what you really want to accomplish with regulating the voltage) using the PWM duty cycle. For bonus points, use a current sensor, and regulate current, not voltage, like a "chopping" motor driver.

3) Use a "chopping" motor driver directly, and don't regulate voltage at all. That's probably what I would do :-)

6. Transistor
Join Date
Jun 2008
Posts
36
Rep Power
25

## Re: Schematic question for LM138 voltage Regulator

I took all of your suggestions into consideration, and played around with a few things. I found that I can run my motors on 6v. Not at full power of course, but there are four of them, and it's a small robot, so I think it will be okay. What I'll probably do is invest in a few 6v batteries and run the power directly to the servos and motor controller. Switching regulators look nice but considering the potential amount of amps I could be using between four motors and seven servos, it would be quite expensive.

7. ## Re: Schematic question for LM138 voltage Regulator

Note that a "6V battery" is really more like a "7.5V to 4.5V battery." This is why the RC community uses "UBECs."
For regular servos, a 5A UBEC (at about \$12) per servo might be an option.
And an UBEC is really just a switching regulator, anyway.

There are currently 1 users browsing this thread. (0 members and 1 guests)

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts
•