rlbrenn@vlsi.waterloo.edu (Robert Brennan) (11/15/89)
Here is a circuit for a split 12 volt power supply that I have used
(and I'm sure most people have used) successfully in typical
low current applications (at end of article).
That is until now...
I have recently built a pre-amp for my audio system as a front
end to a home-made 600W amp (Now, if I just had decent speakers to
handle the power ... but that's a different issue).
The system works great except for an extremely loud start-up
pop out of the speakers. The fast voltage rise on the power
supply rails seems to be the culprit. It appears that the pre-amp
circuit is coupling power supply transients to the output. Without
redesigning the pre-amp circuit (really, I just want to use it ... it was
a canned design), how can I go about 'softening' the abrupt voltage
rise on power-up? I can't think of any off-hand that wouldn't
also sacrifice regulation.
By the way, I tried a slow turn on circuit using LM317 and LM337
regulators and a circuit from the National Linear Guidebook. It
didn't solve the problem because the voltage didn't start from zero.
It ramped up starting at about 2 volts. Even that much of an offset
created a massive power-up click. I don't think the best answer
lies with a complicated circuit (that's why I'm going back to
the simple circuit below), I just need some advice from some of
you who might have had similar problems. I'd like to have the
start-up sound like a nice expensive system (very muted).
Depending on interest, I'll repost replies.
___ ___________
~ BB _____|____ _______
BB |bridge | Pos. | |
AC BB-+ +-|rectifier|---+------|7812 |-----+----------- +12 V
BB | | |_________| | |_____| |
~___BB_|_ | _____| __|__ | __|__
| | _____ 1000uF| _____
| | | | |10uF
+------------------+---------+--------+-----------+ Ground
| __|__ | __|__
| _____ 1000uF| _____
| | ___|__ |10uF
| | | | |
+---------------+------|7912 |-----+------------ -12 V
Neg. |_____|
Thanks for your help
--
Robert Brennan (U of Waterloo VLSI Group, Waterloo Ont.)
rlbrenn@vlsi.waterloo.{cdn,edu,bitnet}
rlbrenn@vlsi.UWaterloo.caankleand@mit-caf.MIT.EDU (Andrew Karanicolas) (11/16/89)
In article <184@vlsi.waterloo.edu> rlbrenn@vlsi.waterloo.edu (Robert Brennan) writes: >I have recently built a pre-amp for my audio system as a front >end to a home-made 600W amp (Now, if I just had decent speakers to >handle the power ... but that's a different issue). >The system works great except for an extremely loud start-up >pop out of the speakers. The fast voltage rise on the power >supply rails seems to be the culprit. It appears that the pre-amp >circuit is coupling power supply transients to the output. Without I take it that you have exonerated the power-amp by turning on the power-amp without the pre-amp connected to it to see if there is a glitch. One of the simpler ways to solve your problem may be to use a circuit that delays switching on AC power to the power amp relative to the pre-amp. This would allow the pre-amp output to settle before it can cause trouble. Of course, you can also switch on the pre-amp and the power-amp manually to avoid loud glitches. A friend of mine had trouble with power up glitches on his Denon receiver, the solution there was to use the speaker switches to connect the speakers after the receiver was powered up and to disconnct them before it was powered down. Many receivers have a delay built in to the speaker protector circuit at the output of the power amp (the 'click' you hear with many receivers and amplifiers shortly after AC power is applied is the relay in this protection circuit). If your power amp has such a speaker protector, then you could modify it to include this delay. If it does not include one, the problem you are experiencing is one of the many good reasons to include one. There are audio purists whose beliefs are incongruent with such circuits though and rely on fuses. In such a case, delaying AC power application to power-amp would be an easy way out. I guess you could also delay connecting the pre-amp output to the power-amp input but I've never seen the problem dealt with in that manner. This was a little longer than needed perhaps, good luck and have fun! Andrew N. Karanicolas Microsystems Technology Laboratory ankleand@caf.mit.edu
jgd@rsiatl.UUCP (John G. De Armond) (11/16/89)
In article <184@vlsi.waterloo.edu> rlbrenn@vlsi.waterloo.edu (Robert Brennan) writes: >Here is a circuit for a split 12 volt power supply that I have used >(and I'm sure most people have used) successfully in typical >low current applications (at end of article). > >That is until now... >I have recently built a pre-amp for my audio system as a front >end to a home-made 600W amp (Now, if I just had decent speakers to >handle the power ... but that's a different issue). >The system works great except for an extremely loud start-up >pop out of the speakers. The fast voltage rise on the power >supply rails seems to be the culprit. How 'bout a real KISS solution. Actually, two. One is less KISS than the other. The KISS solution: Simply bridge to ground the OUTPUT of the preamp with a DPDT mercury-wetted reed relay. Connect the NC terminals to the output of the preamp such that when the relay is deenergized, the output is shunted to ground. Connnect the NO contacts to the output jack such that when the relay is energized, the signals are passed straight through. Energize the relay from the rail through a time delay network that consists of a capacitor across the coil and a resistor in series. Choose appropriate values to give a couple second delay. I'd also suggest that you bridge the resistor with a reverse biased diode to provide a quick discharge path for fast recovery during short power transients. Be sure and put a back-biased diode across the three term regulator to protect it from the discharge current. You probably know that :-). The slightly less KISS solution is basically the same as above except that you fire the relay with a comparator set up to trigger when there is no DC component on the output of the preamp. This scheme ensures that only enough time to stabilize the preamp is used during startup. Handy when you just GOTTA have that quick music fix :-) This solution has many benefits. The MW reed relay will last forever, will not introduce any noise, thermal emf or any of the other effects imagined by the golden ears, is very fast and totally silent. It will not be destroyed by accidental transients fed into the output. PS: If you need some relays, let me know... John -- John De Armond, WD4OQC | Manual? ... What manual ?!? Radiation Systems, Inc. Atlanta, GA | This is Unix, My son, You emory!rsiatl!jgd **I am the NRA** | just GOTTA Know!!!
ISW@cup.portal.com (Isaac S Wingfield) (11/17/89)
If your preamp "pop" problem is like mine, it also occurs at power off; after lots of head scratching, my preamp controller 1) turns on the preamp 2) delays 5 seconds, and turns on the power amp 3) at turnoff, turns off the power amp 4) delays 30 seconds, and turns off the preamp BTW, the controller and preamp power supply are in a separate box located about 4' from the low level electronics. The power switch is on the preamp; when it's off, all power is off (including the controller). When it's on, there is *no* 60 Hz or other power line component in the wires to the power switch. Took a little work, but hum is not a problem. Isaac isw@cup.portal.com
derek@hppad.HP.COM (Derek Schuurman) (11/18/89)
I had this problem once and all I needed to do was place a little inductance in my supply lines - this actually also improved my regulation. How much inductance you need and how practical it is to implement is another story - you're in the VLSI group - you should be able to figure it out! :^) I'm still struggling as an undergrad EE at waterloo! :^) derek.
tomb@hplsla.HP.COM (Tom Bruhns) (11/21/89)
Yet another way to avoid the problem of turn-on transients causing output that might drive your speaker cones through the wall (600W??? Per channel???) is to simply leave the preamp on all the time. This doesn't help during power outages (unless you have battery backup :-), but if you supply is fairly reliable, it's a good way to go. In starting the system, you can manually sequence it on as another poster suggested doing automatically.
whit@blake.acs.washington.edu (John Whitmore) (11/22/89)
In article <184@vlsi.waterloo.edu> rlbrenn@vlsi.waterloo.edu (Robert Brennan) writes: >Here is a circuit for a split 12 volt power supply that I have used >[in] low current applications (at end of article). > >The system works great except for an extremely loud start-up >pop out of the speakers. The fast voltage rise on the power >supply rails seems to be the culprit. It appears that the pre-amp >circuit is coupling power supply transients to the output. > > ___ ___________ ~ BB _____|____ _______ BB |bridge | Pos. | | AC BB-+ +-|rectifier|---+----+-- ----|7812 |-----+------ +12 V BB | | |_________| | R1 \___/ |_____| | ~___BB_|_ | _____| __|__ |____|_ | __|__ | | _____ | -+- C1 | _____ | | | R2 --- | |10uF +------------------+----+----+--------+--------+-------+ Ground | __|__ | __|__ | _____ 1000uF| _____ | | ___|__ |10uF | | | | | +---------------+------|7912 |-----+------------ -12 V Neg. |_____| If your only trouble is the abrupt turnon of the power supplies, ramping the input to the regulators will help; I show an example above, in a modification of the positive regulator. The R1/C1 pair form a turnon delay; about 1 second for R1= 1k ohm, C1= 1000 uF. R2 should be much larger than R1 (30 k ohm) and just discharges C1 when the preamp is powered down. The symbol above C1 is intended to be a NPN transistor (or Darlington, in which case R1= 10k ohm, C1= 100 uF) which is connected as a "capacitor multiplier" to C1. The emitter is connected to the 7812; for the negative side, use PNP transistor. It is important that the input to the 7812 be above 15Vdc, so the transistor (which "steals" two volts) will only be biased correctly for a bridge rectifier output of circa 17Vdc. I once made a version of this circuit with a power MOSFET for the transistor (N-type for positive), but that requires the pulled-up voltage on the gate to be at least 5V higher than the required output; unless there are other power supplies available, that will not work. A more elegant scheme would be to use a single regulated power supply to control both + and - voltages (i.e. a tracking regulator), with a slow-start connection to the voltage reference that drives the regulators. I have tested the ramp-the-input scheme shown above, and the output does not ramp as smoothly (the input goes from six to eight volts while the output goes from zero to eight volts) as the input. If the same scheme is used with a LM317 regulator, it should be more effective than with a 7812 (because the LM317 operates at lower voltages, it will turn on before six volts is applied). ... I am known for my brilliance, John Whitmore by those who do not know me well.
eegauthe@cybaswan.UUCP (o) (11/24/89)
Hi, The idea I'm suggesting here is not mine: I've pinched it from a French magazine. Anyway the result is impressive: I hardly see my boomers moving at power down, not at all at power up, and that's the most noisy!. The secret lies in the power supply: it is L146 based (LM723 clone, with a higher maximum voltage rating and a slightly higher reference), using the frequency compensation to slow down the rise at power up; big capacitors supply the energy for a slow fall (I actually use the power supply of the amplifier for the preamp as well, so 20mF on both the +/-32V!). It is designed to supply up to 6 Amp continuously, which might be luxury for only a preamp (:->). If you don't need much current, 2x1mF might do. Only the slow rise bit is presented down here: ___________________________________________ | | | _________________________ | | | | / | | V. out|-------|< | | L146 | v =10mF | f.comp | |________________+32V | |_______________________| | | | | | | | \ = 10uF | | >|___/\/\/\___| = 10mF | v 100k | | | | GND GND GND Good luck! Olivier -- Olivier GAUTHEROT,// Postgraduate // Electrical Engineering Dept. University of Wales // Swansea, SA2 8PP, U.K. (+44 792 205678 ext 4564) UUCP : ...!ukc!pyr.swan.ac.uk!eegauthe JANET : eegauthe@pyr.swan.ac.uk "A lovely girl in a hammock is a hanging garden." A. Allais