I'm going to try and design a small SMPS for powering op-amps in an automotive application where I intend to use an op-amp to convert a balanced audio signal (from headunit) to unbalanced (to amplifiers).
I was thinking of a push-pull design based on the UCC38085 PWM regulator IC. It's a current-mode 8-pin dip with 500ma source and 1000ma sink outputs. The datasheet is here.
For the switchers, a TN0604 N-ch in a TO92 package. The datasheet is here.
Opinions?
I was thinking of a push-pull design based on the UCC38085 PWM regulator IC. It's a current-mode 8-pin dip with 500ma source and 1000ma sink outputs. The datasheet is here.
For the switchers, a TN0604 N-ch in a TO92 package. The datasheet is here.
Opinions?
+/-15V are pretty standard voltages. Why not just use a ready-made DC/DC converter module? Or do you really want to roll your own.
+/-15V, +/- 190mA
+/-15V, +/- 190mA
How many Op Amps will you power?
Or, to be more precise, how many mA do you need?
Your design might work with this:
ICL7662 CMOS, Voltage Converters - Maxim
FWIW I commercially make battery (12.6V 7AH) powered guitar amplifiers and use a similar circuit (in fact a home designed one with a 555 oscillator) to turn available +12V into -12V , to run my TL072 based preamps ... in fact same preamps I use in the "large" (60/100/200W) line powered ones.
Op Amps are happy working with +/- 12V by the way.
Or, to be more precise, how many mA do you need?
Your design might work with this:
ICL7662 CMOS, Voltage Converters - Maxim
An externally hosted image should be here but it was not working when we last tested it.
FWIW I commercially make battery (12.6V 7AH) powered guitar amplifiers and use a similar circuit (in fact a home designed one with a 555 oscillator) to turn available +12V into -12V , to run my TL072 based preamps ... in fact same preamps I use in the "large" (60/100/200W) line powered ones.
Op Amps are happy working with +/- 12V by the way.
I'd recommend a good ol' TL494 push-pull circuit with a Center-Tap secondary winding....
Use a FERRITE filter toroid salvaged from old CRT. Wind Primary with 12+12Turns, Secondary with 15+15T. Use 4 high speed diodes.
Just connect the feedback resistor to the positive rail output, the negative rail should follow. Use a 1K load resistor for each rail or 1K resistor + LED for dummy load.
Don't use an iron powder (yellow) toroid for the transformer, but you can use it for post-filtering after the output capacitors, followed by another pair of capacitors.
Use a FERRITE filter toroid salvaged from old CRT. Wind Primary with 12+12Turns, Secondary with 15+15T. Use 4 high speed diodes.
Just connect the feedback resistor to the positive rail output, the negative rail should follow. Use a 1K load resistor for each rail or 1K resistor + LED for dummy load.
Don't use an iron powder (yellow) toroid for the transformer, but you can use it for post-filtering after the output capacitors, followed by another pair of capacitors.
Last edited:
The yellow one in the photo has a 6+6 CT primary and 8+8 CT secondary. Haven't tried it yet. Load would be a pair of MC33078 here.
EWorkshop - Did you take a look at the UCC28085 data sheet in the first post? Seems like it would function much like one might use a TL494.
Here's the example schematic from it.
EWorkshop - Did you take a look at the UCC28085 data sheet in the first post? Seems like it would function much like one might use a TL494.
Here's the example schematic from it.
Attachments
> balanced audio signal (from headunit) to unbalanced (to amplifiers).
These power-amps surely do not need more than 2Vrms of signal drive? So to slam them you only need 6V p-p, plus 2V each side for opamp loss, 10V total supply voltage.
Which could easily, and cleanly, be dropped from 12V with an R-C filter.
I'd fret more about the quality of your differential amp. Many of the classics are pretty marginal in a real world. And a car is very "real".
I'd try the inverter-inverter diff-amp. The input impedance is symmetrical (not so for the one-opamp "diff amp"). It isn't infinite, but can be much higher than all likely sources of un-balance. At car-amp level the hiss of 50K is well below the self-hiss of any head-amp. The opamp has no common-mode input, and light load, so fancy opamps are not demanded. And a particular feature: the overall gain can be trimmed above and below unity with one resistor.
Resistors may be nearest standard value, though the "*" Rs should be within a percent of each other (resistor-packs are good). 5K+500pFd makes 1dB down at 30KHz and 20dB down at the bottom of the AM band. "Vref" is a half-supply bias, typically two 10K and a 100uFd cap to ground. With the light load of one or two TL072s, the 12V filtering can be 470 Ohms and 470uFd. The chip can stand over 36V short-term, and your costly power amps probably can't, so I see little point in voltage limiting against alternator load-dumps and double-battery jump-starts.
These power-amps surely do not need more than 2Vrms of signal drive? So to slam them you only need 6V p-p, plus 2V each side for opamp loss, 10V total supply voltage.
Which could easily, and cleanly, be dropped from 12V with an R-C filter.
I'd fret more about the quality of your differential amp. Many of the classics are pretty marginal in a real world. And a car is very "real".
I'd try the inverter-inverter diff-amp. The input impedance is symmetrical (not so for the one-opamp "diff amp"). It isn't infinite, but can be much higher than all likely sources of un-balance. At car-amp level the hiss of 50K is well below the self-hiss of any head-amp. The opamp has no common-mode input, and light load, so fancy opamps are not demanded. And a particular feature: the overall gain can be trimmed above and below unity with one resistor.
Resistors may be nearest standard value, though the "*" Rs should be within a percent of each other (resistor-packs are good). 5K+500pFd makes 1dB down at 30KHz and 20dB down at the bottom of the AM band. "Vref" is a half-supply bias, typically two 10K and a 100uFd cap to ground. With the light load of one or two TL072s, the 12V filtering can be 470 Ohms and 470uFd. The chip can stand over 36V short-term, and your costly power amps probably can't, so I see little point in voltage limiting against alternator load-dumps and double-battery jump-starts.
Attachments
PRR - Tried your schematic in the emulator and it works, until I connect the input ground. Then the output drops in half it seems.
How did you feed it?
PRR circuit runs off single 12V ,not sure about what "input ground" you are referring to.
PRR circuit runs off single 12V ,not sure about what "input ground" you are referring to.
Some Saab owners has used a trafo in their car's, balun.
Amp's like TPA3116 has balanced inputs. Use a car laptop
power supply and drive the amp att 19 Volt.
Amp's like TPA3116 has balanced inputs. Use a car laptop
power supply and drive the amp att 19 Volt.
My balanced signal in ran through a twisted pair that's shielded. Shield is grounded.
This is what I don't understand, why grounding a shield in an emulator "drops signal in half".
Shields have no influence on signal, except blocking external intereference, something simulators have no clue about.
In the emulator, it's the 0v/common terminal of the signal generator.
EDIT - Never mind. I accidentally had an incorrect value in the lower half of my reference voltage divider.
EDIT - Never mind. I accidentally had an incorrect value in the lower half of my reference voltage divider.
Last edited:
Ok, the image is somewhat small and I can't clearly see the generator.
1) which waveform is selected?
The squarewave on the right looks somewhat darker, don't know if that means something.
2) I see 3 terminals:
left: "+"
center: unmarked
right: "-"
3) are + and - signals symmetrical and, of course, out of phase?
Do they have same value relative to Gnd?
Please apply your Audio Millivoltmeter straight to the generator and measure:
these 3 with center terminal ungrounded
a) + voltage relative to center terminal
b) + voltage relative to center terminal
c) + voltage relative to -terminal
then ground center terminal and repeat as above
d)
e)
f)
What I want to know is whether you are losing signal/gain in the differential (balanced) input preamp or the generator loses half the signal.
1) which waveform is selected?
The squarewave on the right looks somewhat darker, don't know if that means something.
2) I see 3 terminals:
left: "+"
center: unmarked
right: "-"
3) are + and - signals symmetrical and, of course, out of phase?
Do they have same value relative to Gnd?
Please apply your Audio Millivoltmeter straight to the generator and measure:
these 3 with center terminal ungrounded
a) + voltage relative to center terminal
b) + voltage relative to center terminal
c) + voltage relative to -terminal
then ground center terminal and repeat as above
d)
e)
f)
What I want to know is whether you are losing signal/gain in the differential (balanced) input preamp or the generator loses half the signal.
1) Sine
2) Center is common
3) Yes they are symmetrical. No offset.
The program has an AC volt meter.
With the center (common) terminal disconnected...
A) + to center .7v
B) - to center .7v
C) + to - 1.4v
With center (common) terminal grounded...
D) + to center .7v
E) - to center .7v
F) + to - 1.4v
The measurement tools shown in the schematic are minimized. The first pic is what they look like with the sig-gen common terminal grounded. The second pic is ungrounded.
2) Center is common
3) Yes they are symmetrical. No offset.
The program has an AC volt meter.
With the center (common) terminal disconnected...
A) + to center .7v
B) - to center .7v
C) + to - 1.4v
With center (common) terminal grounded...
D) + to center .7v
E) - to center .7v
F) + to - 1.4v
The measurement tools shown in the schematic are minimized. The first pic is what they look like with the sig-gen common terminal grounded. The second pic is ungrounded.
Attachments
Last edited:
- Status
- Not open for further replies.