That is a center-tap transformer, not a dual secondary.
Remove one of the rectifier bridges and tie the secondary center tap to the ground bus, like figure 2 at the bottom of this page:
60-80W Power Amplifier
Other than that lots of power is wasted in the regulators....better to get the correct transformer and not use any regulation?
Edit: If you must go with regulators, do not forget the protection diodes around the regulators.
Remove one of the rectifier bridges and tie the secondary center tap to the ground bus, like figure 2 at the bottom of this page:
60-80W Power Amplifier
Other than that lots of power is wasted in the regulators....better to get the correct transformer and not use any regulation?
Edit: If you must go with regulators, do not forget the protection diodes around the regulators.
Last edited:
I agree, substantial power will be dissipated in the regulators, especiallty the first set with the +/-35V output that presumably powers the chipamp. Looking at the transformer voltage, I wonder what the max input-output differential voltage on the LM317 - it looks like a lot, and keep in mind that when first turned on the +35V output will actually be at zero volts.
I'd rather eliminate these and use a transformer that gives the desired output directly. I presume the +/-15V is to drive a few opamps, and those regulators sholdn't need nearly as much heat sinking as the others would.
The schematic shows a center-tapped transformer with TWO bridge rectifiers. Tracing through it shows the transformer output gets shorted through the ground connection. It either needs two separate secondary windings, one for each bridge, or a single bridge between the end winding connections and the + and - rails with the transformer center tap connected to ground.
Puffin, presuming you're asking about the protection diode, a regulator cvan be damaged if the input voltage drops below the output voltage (which is especially likely with the second set of regulators), and the backwards diode between input and output prevents that by conducting when the input drops lower than the output.
I'd rather eliminate these and use a transformer that gives the desired output directly. I presume the +/-15V is to drive a few opamps, and those regulators sholdn't need nearly as much heat sinking as the others would.
The schematic shows a center-tapped transformer with TWO bridge rectifiers. Tracing through it shows the transformer output gets shorted through the ground connection. It either needs two separate secondary windings, one for each bridge, or a single bridge between the end winding connections and the + and - rails with the transformer center tap connected to ground.
Puffin, presuming you're asking about the protection diode, a regulator cvan be damaged if the input voltage drops below the output voltage (which is especially likely with the second set of regulators), and the backwards diode between input and output prevents that by conducting when the input drops lower than the output.
A center-tapped transformer only needs one bridge. I'm not entirely sure it will do anything but release the magic blue smoke if you connect it as drawn.
Why on earth are you making it so intensely complicated? And again, as we have all said before, this (Chipamp PSU) is not the right application of a 317/337 regulator.
Use a 2nd transformer for the tone control circuit.
Why on earth are you making it so intensely complicated? And again, as we have all said before, this (Chipamp PSU) is not the right application of a 317/337 regulator.
Use a 2nd transformer for the tone control circuit.
Complicated? Maybe, but...
The operative word is "cheap", i.e., utilizing things I already have or can get free or cheaply. For example, a Hammond 56v -1.8a transformer is roughly half the price of a 50v - 2a transformer. So I have to go with the 56v unit. But that means I have to knock the DC voltage down from 39.5v per side to 35v. That means regulators. And I have or can get those free. I have lots of resistors, caps, diodes, etc. And I have heatsinks for the TO-220 chip pieces. See my thread in this section on the 317 regulator, part 48, for a longer explanation. I don't mind using lots of parts, if I already have them. Yes, some people say use a transformer from a wall wart for the 15v op amps, but others say knock it down with a regulator from the 35v rails.
I used to have a sign in my office that said "Sooner or later the time comes to shoot the engineers and build the *%&!#* thing." So if I finally get a design the will work, I'll build it. I don't care if it's complicated or not esthetically pleasing, or the latest technology. Just so it works.
I don't mean to put down or belittle anyone, and I certainly thank all of you for your advice, and I hope you will continue to respond to my obviously amateur attempts.
Thanx
The Happy Hippy
The operative word is "cheap", i.e., utilizing things I already have or can get free or cheaply. For example, a Hammond 56v -1.8a transformer is roughly half the price of a 50v - 2a transformer. So I have to go with the 56v unit. But that means I have to knock the DC voltage down from 39.5v per side to 35v. That means regulators. And I have or can get those free. I have lots of resistors, caps, diodes, etc. And I have heatsinks for the TO-220 chip pieces. See my thread in this section on the 317 regulator, part 48, for a longer explanation. I don't mind using lots of parts, if I already have them. Yes, some people say use a transformer from a wall wart for the 15v op amps, but others say knock it down with a regulator from the 35v rails.
I used to have a sign in my office that said "Sooner or later the time comes to shoot the engineers and build the *%&!#* thing." So if I finally get a design the will work, I'll build it. I don't care if it's complicated or not esthetically pleasing, or the latest technology. Just so it works.
I don't mean to put down or belittle anyone, and I certainly thank all of you for your advice, and I hope you will continue to respond to my obviously amateur attempts.
Thanx
The Happy Hippy
Nothing wrong with cheap.
But do realize that you will be power limited by the 1.5 A regulators and 1.8 A transformer.
Lets say the transformer and regulators can (optimistically) push 2 A peak for short periods, that is 16 W into an 8 ohm speaker, assuming a separate power supply for each channel.
Also realize that for 16 W into 8 ohm, 20 V rails are quite sufficient and the 35 V rails will never be fully utilized.
But do realize that you will be power limited by the 1.5 A regulators and 1.8 A transformer.
Lets say the transformer and regulators can (optimistically) push 2 A peak for short periods, that is 16 W into an 8 ohm speaker, assuming a separate power supply for each channel.
Also realize that for 16 W into 8 ohm, 20 V rails are quite sufficient and the 35 V rails will never be fully utilized.
As somebody else mentioned, here is the proper connection of a center tapped transformer as a bipolar supply --
(From Elliot Sound Products)
An externally hosted image should be here but it was not working when we last tested it.
(From Elliot Sound Products)
Last edited:
Where does the 16 watts come from. 35 volts a side = 70 volts, x 1.5 amps = 105 watts. And what is "each channel"? There is only one channel. I'm beginning to think we are talking apples and oranges here. Are we talking about the same thing?
Thanx
The Happy Hippy
Watts are calculated using the RMS values of the voltage and current.
So that would be more like 70/2/sqrt(2) x 1.5/sqrt(2) = ~25 V x ~1 A = ~25 W, ignoring amp and rectifier dropout.
BUT, that would only apply if your load (speaker) was 25/1 = 25 ohms.
If the load is 8 ohms, the power is limited by the current (in this case) = 8x1^2 = 8 W.
I was generous and used 2A peak for the power supply so 8x2^2/2=16 W....
(Note that an 8 ohm speaker will actually dip down to ~5 ohms at certain frequencies, making the situation somewhat worse.)
Without the regulators you are in a slightly better situation, even with the 1.8 A transformer, since the transformer will not 'hard-clip' at 1.5 or 2 A like the regulator.
It will quite happily supply current peaks well in excess of its rating, as long as the average stays low, like with typical music.
Many cheaper commercial amps rely on this.
You will of course still have to drop the voltage a little (few 3A diodes?)
Having the wrong voltage vs. current vs. load transformer is like going on the hi-way with a tractor, or plowing the field with a sedan.
It can be done, but there are compromises to be made.
Last edited:
"Read" does not equate to "fully understand"
Contrary to popular belief, I DO read the data sheets and other information available. However I do not YET understand all the technical terms. For example, on the data sheet for my rectifiers, Fairchild GBU4J, is says "Maximum Forward Voltage Drop, per element at 4.0 A" is "1.0" volts. What is the definition of "element"? Is that the whole chip, or is it each diode in the chip?
If you object to people trying to learn, then you ARE wasting your time. If so, then I suggest that YOU stop reading my posts. If not, then please try to help me understand the terminology.
I am posting the latest schematic of my proposed power supply, two versions. One is 48v and one is 56v. Notice that I have dropped the voltage regulators for the mains, but I have kept the voltage regulators powering the op amps in the preamp, as I prefer to use a dual power supply rather than a single power supply for that purpose, and I just can't persuade myself to use the guts from TWO wall warts, and I can't find a 30v - 35v wall wart.
Thanx
The Happy Hippy
Contrary to popular belief, I DO read the data sheets and other information available. However I do not YET understand all the technical terms. For example, on the data sheet for my rectifiers, Fairchild GBU4J, is says "Maximum Forward Voltage Drop, per element at 4.0 A" is "1.0" volts. What is the definition of "element"? Is that the whole chip, or is it each diode in the chip?
If you object to people trying to learn, then you ARE wasting your time. If so, then I suggest that YOU stop reading my posts. If not, then please try to help me understand the terminology.
I am posting the latest schematic of my proposed power supply, two versions. One is 48v and one is 56v. Notice that I have dropped the voltage regulators for the mains, but I have kept the voltage regulators powering the op amps in the preamp, as I prefer to use a dual power supply rather than a single power supply for that purpose, and I just can't persuade myself to use the guts from TWO wall warts, and I can't find a 30v - 35v wall wart.
Thanx
The Happy Hippy
I would assume each element is an individual diode within the case.
The voltage of either transformer is to high. They are saying that you do not have enough current for the LM3876, even before you regulate it.
LM337 is not the proper device, unless you find a way to control a higher amperage transistor with it. Even then it gets complicated.
I started to go down this path, then I just decided to buy two of these 25v dual secondaries from this eBayer. He has 3 left, no affiliation.
120/240VAC PRIMARY 25 V 3.2A DOUBLE SEC POWER VOLTAGE TRANSFORMER LOW PROFILE | eBay
He is open to "offers"
I got 2 of them delivered for a very good price. Less then 2" thick.
They are 160va each, I am going to use one per channel for a total of 320VA for the system.
The voltage of either transformer is to high. They are saying that you do not have enough current for the LM3876, even before you regulate it.
LM337 is not the proper device, unless you find a way to control a higher amperage transistor with it. Even then it gets complicated.
I started to go down this path, then I just decided to buy two of these 25v dual secondaries from this eBayer. He has 3 left, no affiliation.
120/240VAC PRIMARY 25 V 3.2A DOUBLE SEC POWER VOLTAGE TRANSFORMER LOW PROFILE | eBay
He is open to "offers"
I got 2 of them delivered for a very good price. Less then 2" thick.They are 160va each, I am going to use one per channel for a total of 320VA for the system.
More definitions, explainations, etc.
On my original thread for the amplifier, L6L pointed out to me that according to the LM3876 data sheet, I need a minimum of 33-34v per side to get a full 50+ watts out of the chip, for an 8 ohm speaker.
Accxording to TI, the LM337 is the low-cost version of the LM137, which is a negative version of the LM 117. In other words, matching voltage regulators. There is nothing needed for the LM337 that isn't also needed for the LM317. No transistors.
I am new here, and I cannot figure out how to refrerence my other threads onto this thread. So that may be part of the problem. Basically, I want to make an amplifier,around 50 watts, for my bass guitar. I have the speaker/cabinet constructed, and it drownds out everything else in my church when I turn it up full. But I'm currently orrowing a friend's modified PA amp, so I need one of my own. OK?
Thanx
The Happy Hippy
On my original thread for the amplifier, L6L pointed out to me that according to the LM3876 data sheet, I need a minimum of 33-34v per side to get a full 50+ watts out of the chip, for an 8 ohm speaker.
Accxording to TI, the LM337 is the low-cost version of the LM137, which is a negative version of the LM 117. In other words, matching voltage regulators. There is nothing needed for the LM337 that isn't also needed for the LM317. No transistors.
I am new here, and I cannot figure out how to refrerence my other threads onto this thread. So that may be part of the problem. Basically, I want to make an amplifier,around 50 watts, for my bass guitar. I have the speaker/cabinet constructed, and it drownds out everything else in my church when I turn it up full. But I'm currently orrowing a friend's modified PA amp, so I need one of my own. OK?
Thanx
The Happy Hippy
No need to get snippy.
35 +/- is the maximum you can use for 8ohm speakers, not the minimum.
Lower voltages will allow for lower ohm speakers like 4ohm or 6ohm. (22v or 18v transformer)
Having one or two channels is does not matter as they would be connected in parallel, needing the same voltage from the transformer. Current would double for two channels.
A LM337 is a 1.5amp device, even if you use the complimentary device I doubt you will have anywhere near the current handling needed.
The transformer I link you above would be suitable for a single channel. Even it is 3.5Amp PER SECONDARY. So even a frail 160VA 25v transformer would pass 7amps total. My math is rough and loose, just trying to give you an idea.
If you look at the regulator made by this guy- Roman Black's gainclone amp you can see he uses 2 2SA1265. These are 10amp devices controlled by a LM317(s).
35 +/- is the maximum you can use for 8ohm speakers, not the minimum.
Lower voltages will allow for lower ohm speakers like 4ohm or 6ohm. (22v or 18v transformer)
Having one or two channels is does not matter as they would be connected in parallel, needing the same voltage from the transformer. Current would double for two channels.
A LM337 is a 1.5amp device, even if you use the complimentary device I doubt you will have anywhere near the current handling needed.
The transformer I link you above would be suitable for a single channel. Even it is 3.5Amp PER SECONDARY. So even a frail 160VA 25v transformer would pass 7amps total. My math is rough and loose, just trying to give you an idea.
If you look at the regulator made by this guy- Roman Black's gainclone amp you can see he uses 2 2SA1265. These are 10amp devices controlled by a LM317(s).
Mr Hippy. I have built a regulated GC using LM338's (5A capability) (see link for application)
You can use these in a split rail supply. I am using 45v per rail, just on the limits of the LM3875. The 338's are heatsinked, but they break no sweat at all and remain cool. I don't think that you will find the same with the 317/337 you intend to use as you will be dropping 15 - 18v per rail (depending on which version you build)
Building a Gainclone chip amp with a regulated power supply (PSU).
You can use these in a split rail supply. I am using 45v per rail, just on the limits of the LM3875. The 338's are heatsinked, but they break no sweat at all and remain cool. I don't think that you will find the same with the 317/337 you intend to use as you will be dropping 15 - 18v per rail (depending on which version you build)
Building a Gainclone chip amp with a regulated power supply (PSU).
Questions
Quoting Globug
"35 +/- is the maximum you can use for 8ohm speakers, not the minimum."
Where does that information come from? I have seen no voltage max on my speaker datasheet. And I am already using the 50 watt, 8 ohm speaker: with the proper cabinet design it puts out more sound energy than some 100W speakers.
"A LM337 is a 1.5amp device, even if you use the complimentary device I doubt you will have anywhere near the current handling needed."
At 60+ volts, I only need 1amp.
Quoting Globug and Puffin
"gainclone amp"
Most gainclone amps that I have seen are for PA systems. This is for a bass amp. I'm not sure it qualifies as a gainclone anymore.
Quoting Puffin
"You can use these in a split rail supply. I am using 45v per rail, just on the limits of the LM3875. The 338's are heatsinked, but they break no sweat at all and remain cool. I don't think that you will find the same with the 317/337 you intend to use as you will be dropping 15 - 18v per rail (depending on which version you build)"
I'm using an LM3876, not an LM3875. I don't know the specs on the LM3875, but the LM3876 will handle up to +/-35v (70v total) easily, with minimal distortion, according to the data sheet. And the TL072 op amp uses a minimum amout of current. That means less than 1 watt, so heat should not be a problem. However, I fully intend on running this up on a breadboard first. So if necessary I will go to larger capacity regulators.
I thank all of you for your input, and I am learning something from everyone. Please continue your lessons to me.
Thanx
The Happy Hippy
Quoting Globug
"35 +/- is the maximum you can use for 8ohm speakers, not the minimum."
Where does that information come from? I have seen no voltage max on my speaker datasheet. And I am already using the 50 watt, 8 ohm speaker: with the proper cabinet design it puts out more sound energy than some 100W speakers.
"A LM337 is a 1.5amp device, even if you use the complimentary device I doubt you will have anywhere near the current handling needed."
At 60+ volts, I only need 1amp.
Quoting Globug and Puffin
"gainclone amp"
Most gainclone amps that I have seen are for PA systems. This is for a bass amp. I'm not sure it qualifies as a gainclone anymore.
Quoting Puffin
"You can use these in a split rail supply. I am using 45v per rail, just on the limits of the LM3875. The 338's are heatsinked, but they break no sweat at all and remain cool. I don't think that you will find the same with the 317/337 you intend to use as you will be dropping 15 - 18v per rail (depending on which version you build)"
I'm using an LM3876, not an LM3875. I don't know the specs on the LM3875, but the LM3876 will handle up to +/-35v (70v total) easily, with minimal distortion, according to the data sheet. And the TL072 op amp uses a minimum amout of current. That means less than 1 watt, so heat should not be a problem. However, I fully intend on running this up on a breadboard first. So if necessary I will go to larger capacity regulators.
I thank all of you for your input, and I am learning something from everyone. Please continue your lessons to me.
Thanx
The Happy Hippy
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.