The AM Forum
March 28, 2024, 12:03:41 PM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1] 2   Go Down
  Print  
Author Topic: Voltage Regulator IC question  (Read 13873 times)
0 Members and 1 Guest are viewing this topic.
k4kyv
Contributing Member
Don
Member

Offline Offline

Posts: 10057



« on: May 24, 2010, 12:03:47 PM »

The specs for the National Semiconductor 3-terminal adjustable voltage regulator indicate 1% output voltage tolerance and max. 0.01%/V line regulation.  But they also state that since the regulator is “floating” and sees only the input-to-output differential voltage, supplies of several hundred volts can be regulated as long as the maximum input to output differential is not exceeded, i.e., avoid short-circuiting the output.

But doesn't that imply that the only thing the chip regulates is the input-to-output differential voltage, not the actual voltage from the output terminal to ground as would be the case with a conventional three-terminal fixed regulator with a common connection to ground for input and output?  Wouldn't this merely pass input voltage variations on through to the output, maintaining only a constant difference between the input and output voltages, similar to having a zener diode or a gaseous VR tube in series with a circuit?
Logged

Don, K4KYV                                       AMI#5
Licensed since 1959 and not happy to be back on AM...    Never got off AM in the first place.

- - -
This message was typed using the DVORAK keyboard layout.
http://www.mwbrooks.com/dvorak
WA1GFZ
Member

Online Online

Posts: 11152



« Reply #1 on: May 24, 2010, 12:18:18 PM »

which one. Most have a resistor to ground to monitor output voltage
Logged
WD5JKO
Member

Offline Offline

Posts: 1996


WD5JKO


« Reply #2 on: May 24, 2010, 02:24:59 PM »



Don,

  Adj regulators like the LM317 provide 1.2V output when the adjustment pin is grounded. They usually use two resistors to make a voltage divider when the desired output is higher than 1.2V.

  Floating the LM317 can be done, but as you say, don't short out the output! Also when the output is high, don't short out the input either unless you have a reverse connected diode from input to output.

  An alternative device is the Supertex LR8:

www.supertex.com/pdf/datasheets/LR8.pdf

   Here the device can take 450v (vin - vout), but the current capability is very low (< 20ma). Still, you can use the LR8 to feed a FET source follower for more current.

  So what is your application?

Jim
WD5JKO
Logged
WA1GFZ
Member

Online Online

Posts: 11152



« Reply #3 on: May 24, 2010, 02:40:07 PM »

I don't think the 317 will regulate voltage without a reference to ground. The adjustment pin is 1.2 volts below the output. I usually use a 1.2K resistor between the output and adjustment pin so it comes out 100 ohms per 1/10 volt for the second resistor.
A 317 floating may work as a current reguator if the sense voltage is across a current sense.
Logged
k4kyv
Contributing Member
Don
Member

Offline Offline

Posts: 10057



« Reply #4 on: May 24, 2010, 03:30:06 PM »

The application is to monitor a remote antenna tuner from the operating position. The tuner will be adjusted using a  reversible DC motor. The motor-driven tuning control will be ganged to a 10K or 20K pot.  The pot is a special precision wirewound unit that has no mechanical stop, so the shaft can be rotated an unlimited number of full 360° turns without damage. With approximately 24 volts on the pot I will monitor the voltage at the wiper contact to indicate the position of the tuning adjustment. Using a shunt diode across half the multiplying resistance and the meter, and a meter bias voltage of about 12v, I can make the meter read 0-12v over 180° of rotation. The meter is one of those special 6" diameter industrial jobs with a pointer and scale that indicate over about 270° of rotation, so I can get almost as good resolution as I could with a 0-100 mechanical dial scale mounted directly on a manual tuning control. The linearity of the pot is rated at +/- 0.1%.

I need to regulate the 24 volts so that line voltage variations won't alter the indication, since on some bands the setting of the tuning capacitor is fairly critical.  I want the output of the regulator to be adjustable so that exactly 180° of rotation can be set to take the meter from 0 to exactly full scale. The meter bias voltage will be obtained from a high quality 500Ω wirewound pot across the output of the voltage regulator, with a lockdown on the shaft to hold the setting once it is determined.
Logged

Don, K4KYV                                       AMI#5
Licensed since 1959 and not happy to be back on AM...    Never got off AM in the first place.

- - -
This message was typed using the DVORAK keyboard layout.
http://www.mwbrooks.com/dvorak
WA1GFZ
Member

Online Online

Posts: 11152



« Reply #5 on: May 24, 2010, 03:40:43 PM »

Don,
Why not use 12 volts on the pot and eliminate sources of error. Use good RF bypass around the regulator or RF will cause offsets. 24 volts will be a lot of dissipation on a 500 ohm pot.
Logged
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2648


Making AM GREAT Again!


« Reply #6 on: May 24, 2010, 04:15:13 PM »

Don,

I have something similar in one of my linears...  A Harris RF-103..  Schematics are available online.

It has a rotary inductor that is driven by a DC motor..  A 10 turn hi precision is ganged to it, and drives a meter on the front of the amplifier to show 'tuning'.

It also has provisions to either remote control the amp or operate in preset mode (up to 64) by using multiple 10 turn precision pots and a comparator circuit (this is where mine is having issues).

Might be worth taking a look at how Harris accomplished it...  Schematics are available online.

--Shane
Logged
k4kyv
Contributing Member
Don
Member

Offline Offline

Posts: 10057



« Reply #7 on: May 24, 2010, 04:30:04 PM »

Don,
Why not use 12 volts on the pot and eliminate sources of error. Use good RF bypass around the regulator or RF will cause offsets. 24 volts will be a lot of dissipation on a 500 ohm pot.

The pot rotates nearly a complete 360° turn to cover its full range, but the bread slicers in my ATU go from minimum to maximum capacitance in only 180°.  With 12v across the pot the meter would drop only to half scale at minimum capacitance, and then to zero on the other half-turn as the capacitor is rotated the other 180°, which only duplicates the capacitance range of the previous 180° of rotation.  I can double the indicator resolution to use the full meter scale by having the meter go from 0 to full scale with 180° of rotation.  The shunt diode keeps the negative voltage on the meter at  less than 0.5v throughout the redundant 180° of rotation.

Even with 12 volts I would still need a regulator to eliminate the effects of line voltage variations.  Our mains voltage is very flaky and jumps round all over the place, especially in winter when a lot of people are running electric heat and in the peak of summer when air conditioners are dragging the mains voltage down. I have complained to the power co.  They have checked it and admit that the voltage does vary a lot, but they say it is still within specs and they are not going to do anything to improve it.

I am sure I'll have to shield and by-pass the shunt diode and voltage regulator to keep rf from affecting things. The electronics will all be in the shack.  The only thing at the ATU will be the motor and pot, and I picked  up about a 250' spool of antenna rotator cable at Dayton to connect them to the unit in the shack.

The wirewound 500Ω pot will draw about about 50 milliamps at 24 volts, and dissipate 1.2 watts.  It is rated at 4 watts. The indicator pot will draw only about 2.5 mills if I use the 10K section, or 1.25 mills if I use the 20K section.  In any case the whole thing will draw less than 100 mills from the regulator, and even the smallest version of the National Semiconductor regulator will handle several times more current than what I need.

The precision ww pot is actually two separate pots in the same case, one 10K and the other, 20K.  That gives me a spare in case one opens up.

I might use a similar circuit to the Harris with comparator for running the switching function, but the tuning indicator is very simple.  Two wirewound pots, a motor, cable, a diode, two 30K precision multiplier resistors, 0-200 microammeter, a voltage regulator and a bare-bones basic DC power supply, plus whatever by-pass capacitors and rf chokes I'll need to keep the thing immune to rf.
Logged

Don, K4KYV                                       AMI#5
Licensed since 1959 and not happy to be back on AM...    Never got off AM in the first place.

- - -
This message was typed using the DVORAK keyboard layout.
http://www.mwbrooks.com/dvorak
WA1GFZ
Member

Online Online

Posts: 11152



« Reply #8 on: May 24, 2010, 04:53:05 PM »

Don,
All you need is a pot in series with the meter to calibrate full scale. Just set up the pot so the travel is in the bottom of the range so it tracks from 0 to 12 or so volts over 1/2 turn. Diode junction voltages drift a lot over temperature.
Then just reference the meter to the low side of the pot, not ground. You are better off floating the whole circuit so ground currents don't upset the reading. In aircraft we always float position pots. This way the power supply can be in the shack.   
I would keep the dissipation in the pot low so the wire doesn't heat up and drift. .5 watts is more than enough.
Logged
k4kyv
Contributing Member
Don
Member

Offline Offline

Posts: 10057



« Reply #9 on: May 24, 2010, 06:03:43 PM »

How would I keep the meter from being bombarded with double the full scale voltage on the opposite half turn?  Diode junction voltage whould be a  factor only if the diode is in series with the meter. I plan to run the diode in shunt with the meter and part of its multiplying resistance, so that the multiplying resistor does double duty as a current limiting resistor when the voltage goes negative.  The diode would not conduct until the wiper voltage goes negative respective to the meter bias voltage. Diode junction voltage would only determine how negative the voltage to the meter would go, but I don't think it would allow it to go negative enough to bang the meter against the pin hard enough to damage the movement, regardless.  OTOH, twice the full scale positive voltage might damage the meter if I carelessly let the capacitors rotate over the wrong half turn.

Even if the bias pot heats up and drifts value, it shouldn't change anything since it is working as a voltage divider, and I would assume the temperature coefficient of every turn of the entire wirewound element would be approximately the same. This is why I would use a much lower resistance in the bias pot than in the indicator pot; current drawn by the indicator pot would have negligible effect on the voltage output of the divider resistor pot.

I'll draw a diagram and post it as soon I get a chance. Another thought is to use a digital voltmeter and let the voltage run through full range and not worry about it.  I have seen 3" panel meters with a LCD or LED digital scale, that mount similarly to old fashioned Weston 301's.  I assume they require some kind of external power source to make the electronics work.

I have the analogue meter but would have to buy a digital one.
Logged

Don, K4KYV                                       AMI#5
Licensed since 1959 and not happy to be back on AM...    Never got off AM in the first place.

- - -
This message was typed using the DVORAK keyboard layout.
http://www.mwbrooks.com/dvorak
WA1GFZ
Member

Online Online

Posts: 11152



« Reply #10 on: May 24, 2010, 08:34:14 PM »

OK I see the problem your motor only turns in one direction. Well why not just put a 12 volt zener in series with the meter and set the actuator pot to max out at 24 volts.
I would put transorbs on the pot leads back at the shack to protect against induced lightning. Float the pot at the tuner.
Logged
WA1GFZ
Member

Online Online

Posts: 11152



« Reply #11 on: May 24, 2010, 08:41:00 PM »

OTOH you could run the actuator pot on 12 volts and map both dips either side of the rotation
Logged
k4kyv
Contributing Member
Don
Member

Offline Offline

Posts: 10057



« Reply #12 on: May 24, 2010, 10:18:56 PM »

No, the motor is reversible.  Switch polarity and the motor turns the other way. I could just be careful to avoid running the capacitors beyond the correct half-turn, but there is always the possibility that, due to human error, or some malfunction, it would keep turning into the wrong 180° range.  The 0-12v meter is subject either to 0-24v or -12 to +12v over the full 360°, but all I need to run the variable capacitors from minimum to maximum is 180° of rotation, while the indicator pot requires a little less than 360° to run the wiper arm from zero to maximum. So the meter could be hit with double its full scale voltage, or its full scale voltage at the wrong polarity. In the former case the meter would be pinned in the positive direction; in the latter it would be pinned in the negative direction.  I don't like to pin analogue meters.

The actual voltage is arbitrary; I could use a  6 or 12 volt supply or a 50 volt supply; it would just be a matter of adjusting the multiplying resistors for the meter, and making sure both wirewound pots stay within their rated wattage ratings. I want this thing to be foolproof and not subject to damage when some visitor unbeknownst to me pushes the tune button just to see what happens, or I get sidetracked myself and run it beyond normal range.

The indicator pot has a tap at 10% from one end.  Too bad that tap isn't at 50%.

I had devised a method of preventing rotation beyond approximate 180°, using a couple of microswitches, but that would require fabricating yet another mechanical assembly and running an additional 4 conductors of cable out to the tower.

I don't see anything wrong with my design, but I just want to be sure the 24 volts is well regulated. Here is the National Semiconductors circuit.  I have to assume that the voltage divider formed by R1 and R2 would make the device act as a series regulator to maintain Vout at a constant value despite variations to Vin, within the limitations of the range of the regulator, and that the  ratio of R1 to R2 would be set to determine the actual  value of Vout.  The smallest regulator, the LM317MDT, is rated at 500 mills, more than enough to run the indicator.

The reversible motor runs at 12 vdc, and I have measured its current at between 350 and 450 mills, depending on physical load.  No need to regulate its voltage.  I plan to use separate DC supplies for the motor and indicator circuit. The rotor cable has enough conductors to leave both DC supplies floating without having to ground either one.


* voltage regulator circuit.jpg (14.84 KB, 300x200 - viewed 386 times.)
Logged

Don, K4KYV                                       AMI#5
Licensed since 1959 and not happy to be back on AM...    Never got off AM in the first place.

- - -
This message was typed using the DVORAK keyboard layout.
http://www.mwbrooks.com/dvorak
k4kyv
Contributing Member
Don
Member

Offline Offline

Posts: 10057



« Reply #13 on: May 25, 2010, 01:14:27 AM »

Here is the tentative circuit diagram of the indicator unit.  Voltages shown are approximate.

* TNR002.pdf (1422.6 KB - downloaded 265 times.)
Logged

Don, K4KYV                                       AMI#5
Licensed since 1959 and not happy to be back on AM...    Never got off AM in the first place.

- - -
This message was typed using the DVORAK keyboard layout.
http://www.mwbrooks.com/dvorak
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 8308



WWW
« Reply #14 on: May 26, 2010, 01:52:44 AM »

That is an interesting way to do it. http://www.kk5dr.com/Thor.htm has some really decent info on mods/hack for that amp too but the pic are all crapped up with some noise.
Logged

Radio Candelstein - Flagship Station of the NRK Radio Network.
k4kyv
Contributing Member
Don
Member

Offline Offline

Posts: 10057



« Reply #15 on: May 26, 2010, 12:32:06 PM »

I'm working on a prototype right now.  Last night completed the power supply.  I don't have the regulator IC yet, so just threw together a manual voltage adjuster using a rheostat and a couple of resistors.  I'll probably finish the rest of it to-day and see how well the indicator works, then hook it up to the motor.  

I have two reversible motors each with its own gear box, one that turns about 12 RPM and the other that turns about 60 RPM. They have enough output torque that I can't stop either one by hand.  The output of the gear box will run the worm drive that turns the capacitors.  With the slow one, it takes about 5 minutes to turn the capacitor a full 360°, but that allows a precise setting; that's the one I used years ago with my remote 160m tuner. The faster one should make 180° in about 30 seconds.  I'll try the faster one and use it if I can, but if it's too difficult to precisely set the capacitors, I'll go with the slow one.  The two motor/gearbox units are drop-in interchangeable so trying both is no problem.  I have only one slow one, but several spare fast ones.

When finished, the unit will simultaneously turn four or possibly five ganged-together bread-slicers, one for each single-band tuner. I have already tested the ganged capacitors, complete with worm drive and two 90° 1:1 ratio gear drives linking everything together, and everything rotates without difficulty.  The only problem so far has been keeping all the set screws tight enough that they don't slip.  I think I have solved that problem, but if not, Plan B is to drill holes and insert pins to fix the shaft couplers to position.

So far, I have been able to find every necessary part except for the regulator chip in my junk collection; just proves the point that it is "better to have it and not need it than to need it and not have it".
Logged

Don, K4KYV                                       AMI#5
Licensed since 1959 and not happy to be back on AM...    Never got off AM in the first place.

- - -
This message was typed using the DVORAK keyboard layout.
http://www.mwbrooks.com/dvorak
WA1GFZ
Member

Online Online

Posts: 11152



« Reply #16 on: May 26, 2010, 12:45:21 PM »

Don you could always put a 2:1 pulley or gear ratio between the cap and pot.
LM317s are good to have around.
Logged
ka3zlr
Guest
« Reply #17 on: May 26, 2010, 01:05:33 PM »

 Smiley Yea they are and they're Adjustable I just fixed a 1.5 amp bench DC supply I picked up
at Barnos for 5 bux all the problem was bad heat sinking on the regulator now I have
a bench supply with direct 5 volt, 1.6 to 16.9 positive dc and 1.6 to 16.9 negative dc
both adjustable.. Smiley

73

Jack.
Logged
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2648


Making AM GREAT Again!


« Reply #18 on: May 26, 2010, 01:43:48 PM »

I know it's more complexity, Don, but what about putting a PWM on the motor?  Would that allow you to slow down the fast ones?

Seems a shame they run too fast, to have bunches laying around Smiley


--Shane
Logged
k4kyv
Contributing Member
Don
Member

Offline Offline

Posts: 10057



« Reply #19 on: May 26, 2010, 01:49:21 PM »

PWM= pulse width modulator?
Logged

Don, K4KYV                                       AMI#5
Licensed since 1959 and not happy to be back on AM...    Never got off AM in the first place.

- - -
This message was typed using the DVORAK keyboard layout.
http://www.mwbrooks.com/dvorak
w3jn
Johnny Novice
Administrator
Member

Offline Offline

Posts: 4620



« Reply #20 on: May 26, 2010, 02:12:49 PM »

Don, here's how TMC did it with the ATS remote ant tuner.  The pot in the tuner is a multiturn that tracks the coil, the resistors on the switch tracks the configuration switch (6 steps on the meter).

They use 105V regulated from a VR tube.  There's also a humidity sensor that's part of the bridge circuit shown in the control unit schematic.

http://jptronics.org/radios/TMC/manuals/ATS2/ats2.schem.pdf
Logged

FCC:  "The record is devoid of a demonstrated nexus between Morse code proficiency and on-the-air conduct."
k4kyv
Contributing Member
Don
Member

Offline Offline

Posts: 10057



« Reply #21 on: May 26, 2010, 03:12:02 PM »

Don, if you have any problems getting regulated 24V out to the motors through 900' of cable to the dawg house then try the LT3080 in the TO-220 pkg. It's a floating device and has no ground pin Wink

From some of your other topics I assume you are going through a very long length of rotor cable or something? The conventional LM317 might fall on its face.

I'm not running regulated DC on the motor.  I'll just force feed the cable carrying the motor current with enough voltage so that +/- 12 volts shows up at the motor terminals.

The regulated DC operates the indicator circuit, which it totally isolated from the motor circuit.  The only purpose for the regulator is to maintain a constant meter reading unaffected by sags and spikes in the line voltage.  The indicator pot at the tower (I believe the correct technical term is "sector pot") pulls only a couple of mills, with either the 10K or the 20K resistance element bridged across 24 volts regulated. Since everything following the regulator is done with voltage dividers, temperature should have minimal effect on the readings.  The special pot I am using appears to be hermetically sealed, so hopefully, humidity won't be an issue.
Logged

Don, K4KYV                                       AMI#5
Licensed since 1959 and not happy to be back on AM...    Never got off AM in the first place.

- - -
This message was typed using the DVORAK keyboard layout.
http://www.mwbrooks.com/dvorak
WA1GFZ
Member

Online Online

Posts: 11152



« Reply #22 on: May 26, 2010, 09:42:11 PM »

Just put at least 10 uf on the input and make sure you have the input at least 3 volts above the output, not rocket science
Logged
k4kyv
Contributing Member
Don
Member

Offline Offline

Posts: 10057



« Reply #23 on: May 27, 2010, 02:06:39 AM »

I got the prototype test setup running this evening.  It pretty much works as planned.  There is a slight (but anticipated) non-linearity because the 60K load from the meter circuit causes the voltage at the wiper arm of the 10K pot to sag slightly, maybe 1-2%, when the meter reads at mid-range on the scale, but I don't think that is serious enough to warrant purchasing a higher ohms-per-volt analogue or digital voltmeter. The important thing is that the reading be stable and consistent.
Logged

Don, K4KYV                                       AMI#5
Licensed since 1959 and not happy to be back on AM...    Never got off AM in the first place.

- - -
This message was typed using the DVORAK keyboard layout.
http://www.mwbrooks.com/dvorak
k4kyv
Contributing Member
Don
Member

Offline Offline

Posts: 10057



« Reply #24 on: June 02, 2010, 02:38:10 PM »

My prototype has been built and successfully tested, using manual control on the voltage source.

Here's what I have come up with, using the available meter and precision multiplying resistors on hand:

With exactly 120 volts a.c. input to the DC power supply,

Input voltage to regulator = 38.4 v

Output voltage = 25.7 volts.

The line voltage here sometimes rises to 125 volts.

It looks like either the National LM-317 or the LT3080 would do the job.  I looked at the 3080  specs and it looks they make both a SMT and a 5-lead TO-220 package.

One concern I have is that it looks like my input voltage may equal or slightly exceed the absolute maximum recommended according to the manufacturers specs.  What I don't understand is that since these are essentially series regulators with  control voltages derived from the current through a network of resistors to ground, why there should be a strict limit to input voltage, as long as the maximum differential between input and output voltages is not exceeded.  It looks to me that the value of the resistor to ground could be slightly increased to keep the current the same, and that way, the input voltage could be increased  to a higher value without affecting the circuit internal to the chip, but I have not carefully studied every word on the application sheet to the 3080.

Can anyone recommend an easy source for purchasing either one of these regulators?  I would prefer to avoid minimum order or "sell-to-commercial business-only" hassles, if possible, although I have long used a phantom "consulting service" as my business title when necessary.
Logged

Don, K4KYV                                       AMI#5
Licensed since 1959 and not happy to be back on AM...    Never got off AM in the first place.

- - -
This message was typed using the DVORAK keyboard layout.
http://www.mwbrooks.com/dvorak
Pages: [1] 2   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.066 seconds with 19 queries.