Building a high current step down transformer, Size rect. diodes?

[Beamin]

I'm looking to build a power supply that can run car stereo amps of 120VAC.

What size rectifier diodes should be used? I have the typical gray/black cylinder ones with the gray band (not sure part # don't have box) that are about 4 times bigger then standard: The outer dia is about 1/2". This amp has dual 60 amp fuses so I'm going to size the transformer off that. The leads to these diodes are quite thick. Not much info sorry.

[Zero999]

I'm looking to build a power supply that can run car stereo amps of 120VAC.

What size rectifier diodes should be used? I have the typical gray/black cylinder ones with the gray band (not sure part # don't have box) that are about 4 times bigger then standard: The outer dia is about 1/2". This amp has dual 60 amp fuses so I'm going to size the transformer off that. The leads to these diodes are quite thick. Not much info sorry.

The fuses are there just to protect the cable and probably have a much higher rating than necessary: double isn't unusual.

How big is the stereo? What's the true output power? Note that this is often very different to what the marketing wank says, which is often grossly exaggerated.

I'd probably get a switched mode power supply for this. A computer PSU would probably be ideal.

[Beamin]

I know a computer PSU will be way to small. These amps were made in America (about 12X18 in and weigh about 20lbs each) and each came with a print out of the power output. I'll just be running one class D two channel as they take too much power to run both of them. I think it was 400 or 500 watts RMS, can't remember, but it was a real number; I bought a couple of these amps and each one had slightly different print outs. This was when you could get quality stuff or had the option of buying Chinese and paying less. Now you can only get Chinese and still pay top dollar.  The diodes I found out are 6 amps, so if they are in a rectifier 4 diode config. the most I can put through is 12 amps because two will be switched on at each time correct?

[bson]

Very unlikely 400 or 500W rms.  More likely 500W peak, which for music with a nominal crest factor of 12dB is about 32W average.  This is what the supply needs to provide; peaks are averaged out by caps and inductors and shouldn't be seen by the supply (unless you consider a cap bank part of the supply).

[Beamin]

Very unlikely 400 or 500W rms.  More likely 500W peak, which for music with a nominal crest factor of 12dB is about 32W average.  This is what the supply needs to provide; peaks are averaged out by caps and inductors and shouldn't be seen by the supply (unless you consider a cap bank part of the supply).

These are some crazy amps I wish I had the original info but they are not at my house at the moment. When I bought them I had to upgrade the alternator in my car AND put in two 1F caps. the base would stall out the engine when I tried to put one in a Saturn I had (stereo was worth way more then the car, so I upgraded the car).

[Zero999]

I know a computer PSU will be way to small. These amps were made in America (about 12X18 in and weigh about 20lbs each) and each came with a print out of the power output. I'll just be running one class D two channel as they take too much power to run both of them. I think it was 400 or 500 watts RMS, can't remember, but it was a real number; I bought a couple of these amps and each one had slightly different print outs. This was when you could get quality stuff or had the option of buying Chinese and paying less. Now you can only get Chinese and still pay top dollar.  The diodes I found out are 6 amps, so if they are in a rectifier 4 diode config. the most I can put through is 12 amps because two will be switched on at each time correct?

If the amplifier is really are that powerful, then it will have boost converter built-in. Ohm's law says it's not possible to build an audio amplifier that can put over 18W (36W peak) sine wave RMS into a 4 Ohm speaker at 12V. To drive a 4R load with 500W RMS, you need a peak voltage of 63.25V, so the amplifier will need to run off a higher voltage than this internally, hence the boost converter on the input.

If you're prepared to do some reverse engineering and modify the amplifier, then you could bypass the internal boost converter and replace it with a suitable mains powered switched mode power supply with the correct output voltage. It will work out more efficient, than powering the internal boost converter from a 12V supply.

As far as the diodes are concerned: yes, you can run them at a higher current than the continuous maximum rating, if the duty cycle is 50%, but it won't be double because the voltage drop will be higher.

A big bulking transformer + rectifier won't be ideal. It will need a soft start, otherwise the breaker will blow when the power is first applied and the power factor will be poor, so the actual current drawn from the mains will be higher than you'd expect.

If you don't want to hack, just get a 12V 50A power supply of a decent brand, see ebay link below.
http://www.ebay.com/itm/OpenBox-MEAN-WELL-SE-600-12-AC-to-DC-Power-Supply-Single-Output-12V-50-Amp-6-/172735116210?epid=1073103913&hash=item2837d0b3b2:g:8e0AAOSwDrNZRGGF

[Beamin]

I know a computer PSU will be way to small. These amps were made in America (about 12X18 in and weigh about 20lbs each) and each came with a print out of the power output. I'll just be running one class D two channel as they take too much power to run both of them. I think it was 400 or 500 watts RMS, can't remember, but it was a real number; I bought a couple of these amps and each one had slightly different print outs. This was when you could get quality stuff or had the option of buying Chinese and paying less. Now you can only get Chinese and still pay top dollar.  The diodes I found out are 6 amps, so if they are in a rectifier 4 diode config. the most I can put through is 12 amps because two will be switched on at each time correct?

If the amplifier is really are that powerful, then it will have boost converter built-in. Ohm's law says it's not possible to build an audio amplifier that can put over 18W (36W peak) sine wave RMS into a 4 Ohm speaker at 12V. To drive a 4R load with 500W RMS, you need a peak voltage of 63.25V, so the amplifier will need to run off a higher voltage than this internally, hence the boost converter on the input.

If you're prepared to do some reverse engineering and modify the amplifier, then you could bypass the internal boost converter and replace it with a suitable mains powered switched mode power supply with the correct output voltage. It will work out more efficient, than powering the internal boost converter from a 12V supply.

As far as the diodes are concerned: yes, you can run them at a higher current than the continuous maximum rating, if the duty cycle is 50%, but it won't be double because the voltage drop will be higher.

A big bulking transformer + rectifier won't be ideal. It will need a soft start, otherwise the breaker will blow when the power is first applied and the power factor will be poor, so the actual current drawn from the mains will be higher than you'd expect.

If you don't want to hack, just get a 12V 50A power supply of a decent brand, see ebay link below.
http://www.ebay.com/itm/OpenBox-MEAN-WELL-SE-600-12-AC-to-DC-Power-Supply-Single-Output-12V-50-Amp-6-/172735116210?epid=1073103913&hash=item2837d0b3b2:g:8e0AAOSwDrNZRGGF

Nothing instills more confidence in a buyer then statements like this run on sentence:

All our Opened Box items are in great condition each of them is tested and verified by professional logistic teams, in most cases you'll hardly notice the difference between our opened box items to equivalent new items, sometimes the product box do not meets brand new items standard, but besides that you will own a great product and in a competitive price.

Basically they are saying that the product can be defective but you will be ok because you are "owning a great product at a competitive price". It doesn't work but I saved a ton of money and also it has greatness that was approved as great by a professional logistic team; a 10 year old Chinese kid in a rural hut plugging it into a generator to see if the light turns on. Only thing better would be "110% service grate guaranteeding for valued!"

[Mr.B]

...Basically they are saying that the product can be defective but you will be ok because...

It does not say that at all.
Loose translation: The box may not be in pristine condition, but there is nothing wrong with what is inside.

[Beamin]

...Basically they are saying that the product can be defective but you will be ok because...

It does not say that at all.
Loose translation: The box may not be in pristine condition, but there is nothing wrong with what is inside.

I thought open box meant that the stuff didn't come with a box. When circuit city was around thats what they meant; some one bought something used it for a month then brought it back..

[Zero999]

If you don't feel comfortable with that seller then you can go else where. Have you tried Googling the manufacturer and part number?

I did and found the same product from Jameco, for less and a 2 year warranty.
http://www.jameco.com/z/SE-600-12-Mean-Well-AC-to-DC-Power-Supply-Single-Output-12-Volt-50-Amp-600-Watt_374011.html

Not happy with that? Fine, there are plenty of other suppliers.

Don't like Mean Well? Again, there are plenty of other manufacturers of 600W 12V power supplies.

[Beamin]

If you don't feel comfortable with that seller then you can go else where. Have you tried Googling the manufacturer and part number?

I did and found the same product from Jameco, for less and a 2 year warranty.
http://www.jameco.com/z/SE-600-12-Mean-Well-AC-to-DC-Power-Supply-Single-Output-12-Volt-50-Amp-600-Watt_374011.html

Not happy with that? Fine, there are plenty of other suppliers.

Don't like Mean Well? Again, there are plenty of other manufacturers of 600W 12V power supplies.

The point is more to build a project rather then just buying something and plugging it in. These amps have really good sound quality and were very expensive back in their day so I would rather not start messing around with the internal circuitry, that would probably be a bad idea, unless I was a very accomplished EE.

[Zero999]

If you don't feel comfortable with that seller then you can go else where. Have you tried Googling the manufacturer and part number?

I did and found the same product from Jameco, for less and a 2 year warranty.
http://www.jameco.com/z/SE-600-12-Mean-Well-AC-to-DC-Power-Supply-Single-Output-12-Volt-50-Amp-600-Watt_374011.html

Not happy with that? Fine, there are plenty of other suppliers.

Don't like Mean Well? Again, there are plenty of other manufacturers of 600W 12V power supplies.

The point is more to build a project rather then just buying something and plugging it in.

But buying a power supply is the most economical solution. Even an accomplished EE would do that, unless there was a very specific reason not to.

These amps have really good sound quality and were very expensive back in their day so I would rather not start messing around with the internal circuitry, that would probably be a bad idea, unless I was a very accomplished EE.

Just buying a power supply doesn't involve messing around with its internal circuitry. There's more of a risk of messing it up by hacking together your own power supply.

Just for educational purposes, suppose you want to make your own power supply. Presumably it doesn't need any voltage regulation, just a transformer + rectifier and smoothing capacitor. Lets look at the cost of the components:

1) Transformer. It will need to be over-rated because the poor power factor of a bridge rectifier generates greater copper losses. Wow $190.59, over twice the price of the pre-built PSU! I suppose you could rewind a microwave oven transformer, to save cost.
https://www.digikey.com/product-detail/en/signal-transformer/MPI-900-20/MPI-900-20-ND/1118256

2) Capacitor. I've selected 270 000µF 25V low impedance, high ripple current. You'll need two for a low ripple voltage. They're $96.3415 and you need two so that's $192.683
https://www.digikey.com/product-detail/en/cornell-dubilier-electronics-cde/CGS274U025X5L/CGS274U025X5L-ND/1596026

3) Rectifier. Schottky is best for low voltage drop. It was a pain to find one which wasn't very expensive or a huge minimum order. You'll need 4 at a total of $16
https://www.digikey.com/product-detail/en/ixys/DSS60-0045B/DSS60-0045B-ND/1651434

4) Now you'll need something to limit the inrush current, so it doesn't trip the breaker. A timer relay, which bypasses a resistor in series with the transformer's primary, will do. You can make it or buy one.

5) Power switch, cable, fuse.

5) Enclosure.

I got bored looking through Digiky at #4. Please buy one and save your time for more interesting projects.

[Beamin]

If you don't feel comfortable with that seller then you can go else where. Have you tried Googling the manufacturer and part number?

I did and found the same product from Jameco, for less and a 2 year warranty.
http://www.jameco.com/z/SE-600-12-Mean-Well-AC-to-DC-Power-Supply-Single-Output-12-Volt-50-Amp-600-Watt_374011.html

Not happy with that? Fine, there are plenty of other suppliers.

Don't like Mean Well? Again, there are plenty of other manufacturers of 600W 12V power supplies.

The point is more to build a project rather then just buying something and plugging it in.

But buying a power supply is the most economical solution. Even an accomplished EE would do that, unless there was a very specific reason not to.

These amps have really good sound quality and were very expensive back in their day so I would rather not start messing around with the internal circuitry, that would probably be a bad idea, unless I was a very accomplished EE.

Just buying a power supply doesn't involve messing around with its internal circuitry. There's more of a risk of messing it up by hacking together your own power supply.

Just for educational purposes, suppose you want to make your own power supply. Presumably it doesn't need any voltage regulation, just a transformer + rectifier and smoothing capacitor. Lets look at the cost of the components:

1) Transformer. It will need to be over-rated because the poor power factor of a bridge rectifier generates greater copper losses. Wow $190.59, over twice the price of the pre-built PSU! I suppose you could rewind a microwave oven transformer, to save cost.
https://www.digikey.com/product-detail/en/signal-transformer/MPI-900-20/MPI-900-20-ND/1118256

2) Capacitor. I've selected 270 000µF 25V low impedance, high ripple current. You'll need two for a low ripple voltage. They're $96.3415 and you need two so that's $192.683
https://www.digikey.com/product-detail/en/cornell-dubilier-electronics-cde/CGS274U025X5L/CGS274U025X5L-ND/1596026

3) Rectifier. Schottky is best for low voltage drop. It was a pain to find one which wasn't very expensive or a huge minimum order. You'll need 4 at a total of $16
https://www.digikey.com/product-detail/en/ixys/DSS60-0045B/DSS60-0045B-ND/1651434

4) Now you'll need something to limit the inrush current, so it doesn't trip the breaker. A timer relay, which bypasses a resistor in series with the transformer's primary, will do. You can make it or buy one.

5) Power switch, cable, fuse.

5) Enclosure.

I got bored looking through Digiky at #4. Please buy one and save your time for more interesting projects.

To me that is interesting. I like winding transformers and inductors its therapeutic. So maybe I should buy some used PSU's or other things with power supplies off ebay and take them apart and build my own. Replacing the shitty parts with better ones. I tend not to buy stuff like that off digikey way too expensive. I was looking for some RF connector adapters and they were 25.00 each at quantity! I got them off ebay with same bandwidth for 2.00 each. Checked SWR and they weren't causing any noticeable loss vs coax. Unless you have a really good idea of what is a "more interesting" project? There is something satisfying about building stuff yourself even if its from a kit. Most things I build are from other recycled electronic some are 100% recycled parts that cost me almost no money and were fun to build. 

[Zero999]

If you don't feel comfortable with that seller then you can go else where. Have you tried Googling the manufacturer and part number?

I did and found the same product from Jameco, for less and a 2 year warranty.
http://www.jameco.com/z/SE-600-12-Mean-Well-AC-to-DC-Power-Supply-Single-Output-12-Volt-50-Amp-600-Watt_374011.html

Not happy with that? Fine, there are plenty of other suppliers.

Don't like Mean Well? Again, there are plenty of other manufacturers of 600W 12V power supplies.

The point is more to build a project rather then just buying something and plugging it in.

But buying a power supply is the most economical solution. Even an accomplished EE would do that, unless there was a very specific reason not to.

These amps have really good sound quality and were very expensive back in their day so I would rather not start messing around with the internal circuitry, that would probably be a bad idea, unless I was a very accomplished EE.

Just buying a power supply doesn't involve messing around with its internal circuitry. There's more of a risk of messing it up by hacking together your own power supply.

Just for educational purposes, suppose you want to make your own power supply. Presumably it doesn't need any voltage regulation, just a transformer + rectifier and smoothing capacitor. Lets look at the cost of the components:

1) Transformer. It will need to be over-rated because the poor power factor of a bridge rectifier generates greater copper losses. Wow $190.59, over twice the price of the pre-built PSU! I suppose you could rewind a microwave oven transformer, to save cost.
https://www.digikey.com/product-detail/en/signal-transformer/MPI-900-20/MPI-900-20-ND/1118256

2) Capacitor. I've selected 270 000µF 25V low impedance, high ripple current. You'll need two for a low ripple voltage. They're $96.3415 and you need two so that's $192.683
https://www.digikey.com/product-detail/en/cornell-dubilier-electronics-cde/CGS274U025X5L/CGS274U025X5L-ND/1596026

3) Rectifier. Schottky is best for low voltage drop. It was a pain to find one which wasn't very expensive or a huge minimum order. You'll need 4 at a total of $16
https://www.digikey.com/product-detail/en/ixys/DSS60-0045B/DSS60-0045B-ND/1651434

4) Now you'll need something to limit the inrush current, so it doesn't trip the breaker. A timer relay, which bypasses a resistor in series with the transformer's primary, will do. You can make it or buy one.

5) Power switch, cable, fuse.

5) Enclosure.

I got bored looking through Digiky at #4. Please buy one and save your time for more interesting projects.

To me that is interesting. I like winding transformers and inductors its therapeutic. So maybe I should buy some used PSU's or other things with power supplies off ebay and take them apart and build my own. Replacing the shitty parts with better ones. I tend not to buy stuff like that off digikey way too expensive. I was looking for some RF connector adapters and they were 25.00 each at quantity! I got them off ebay with same bandwidth for 2.00 each. Checked SWR and they weren't causing any noticeable loss vs coax. Unless you have a really good idea of what is a "more interesting" project? There is something satisfying about building stuff yourself even if its from a kit. Most things I build are from other recycled electronic some are 100% recycled parts that cost me almost no money and were fun to build.

Yes Digikey are expensive and is there for professionals, who's time is so costly, it works out cheaper than faffing around searching for it elsewhere and waiting for it to turn up.

There are some parts which are inherently costly. After the transformer, there are the capacitors, which need to be good. You can try your luck on Amazon and ebay but you can never be certain they're any good. The key features you need are low ESR and high ripple current rating.

If you want a challenging project, then you could build a 12V 50A switched mode power supply, but it's not a beginner friendly project because it involves capacitors charged to mains voltages and having to design transformers and a PCB with adequate insulation from the mains. You'll probably also need to get yourself a massive isolation transformer for test purposes.

[dr.diesel]

What duration do you need to run your off vehicle testing?

A deep cycle battery and say cheap 10amp charger work for you?

[Beamin]

Yes Digikey are expensive and is there for professionals, who's time is so costly, it works out cheaper than faffing around searching for it elsewhere and waiting for it to turn up.

There are some parts which are inherently costly. After the transformer, there are the capacitors, which need to be good. You can try your luck on Amazon and ebay but you can never be certain they're any good. The key features you need are low ESR and high ripple current rating.

If you want a challenging project, then you could build a 12V 50A switched mode power supply, but it's not a beginner friendly project because it involves capacitors charged to mains voltages and having to design transformers and a PCB with adequate insulation from the mains. You'll probably also need to get yourself a massive isolation transformer for test purposes.

Yeah I'm not going to do that. People seem to have a disconnect between what is ideal and what is practical. They often tell you to buy the best test equipment for some thing you are going to do once or recommend things that a beginner would try but then get frustrated and give up once they realize its way too expensive and learning how to do it would put you half way to a BS in EE. Its like running a business, there is a middle ground of practicality and isn't found in any text book but comes from common sense. Often great engineers are shitty business people because they can't step back and say "Yes I would do that, but when I was just starting or an average mechanically inclined person would it make sense to do that if it was even possible?" Sometime the quick and dirty solution is the best solution when all that matters is the outcome and not how you got there. But conversely when the outcome isn't as important (its not to make a financial gain or just get something working) what is of value is learning from the process. No one blinks an LED on off because their house needs strobe lighting or they own a blinking led company its to learn how it works, even if it doesn't work hopefully they learned what not to do.

[Zero999]

Yes Digikey are expensive and is there for professionals, who's time is so costly, it works out cheaper than faffing around searching for it elsewhere and waiting for it to turn up.

There are some parts which are inherently costly. After the transformer, there are the capacitors, which need to be good. You can try your luck on Amazon and ebay but you can never be certain they're any good. The key features you need are low ESR and high ripple current rating.

If you want a challenging project, then you could build a 12V 50A switched mode power supply, but it's not a beginner friendly project because it involves capacitors charged to mains voltages and having to design transformers and a PCB with adequate insulation from the mains. You'll probably also need to get yourself a massive isolation transformer for test purposes.

Yeah I'm not going to do that. People seem to have a disconnect between what is ideal and what is practical. They often tell you to buy the best test equipment for some thing you are going to do once or recommend things that a beginner would try but then get frustrated and give up once they realize its way too expensive and learning how to do it would put you half way to a BS in EE. Its like running a business, there is a middle ground of practicality and isn't found in any text book but comes from common sense. Often great engineers are shitty business people because they can't step back and say "Yes I would do that, but when I was just starting or an average mechanically inclined person would it make sense to do that if it was even possible?" Sometime the quick and dirty solution is the best solution when all that matters is the outcome and not how you got there. But conversely when the outcome isn't as important (its not to make a financial gain or just get something working) what is of value is learning from the process. No one blinks an LED on off because their house needs strobe lighting or they own a blinking led company its to learn how it works, even if it doesn't work hopefully they learned what not to do.

Sorry, I don't see the point, you're trying to make?

In this case the most idea/practical solution is to buy a suitable switched mode power supply, for your amplifier. I agree that designing one is not feasible for a beginner.

There isn't much you will learn by building a PSU, using a huge transformer, rectifier and capacitor. It would just be the same small power supply just scaled up. It's also inferior to using a switched mode power supply, in just about every single way. Quite often people avoid switched mode power supplied for audio applications, but in this case, it's a car amplifier, which will have plenty of filtering and a switched mode power supply built-in, to increase the voltage to a level, to deliver the required power into an 8 or 4 Ohm speaker.

A good pen and paper exercise would be reverse engineering the existing amplifier and performing some tests on it, to check if it really meets its advertised specification. Quite often manufacturers grossly exaggerate the specifications for their audio amplifiers and it would be interesting to see whether this is the case here.

[Beamin]

Yes Digikey are expensive and is there for professionals, who's time is so costly, it works out cheaper than faffing around searching for it elsewhere and waiting for it to turn up.

There are some parts which are inherently costly. After the transformer, there are the capacitors, which need to be good. You can try your luck on Amazon and ebay but you can never be certain they're any good. The key features you need are low ESR and high ripple current rating.

If you want a challenging project, then you could build a 12V 50A switched mode power supply, but it's not a beginner friendly project because it involves capacitors charged to mains voltages and having to design transformers and a PCB with adequate insulation from the mains. You'll probably also need to get yourself a massive isolation transformer for test purposes.

Yeah I'm not going to do that. People seem to have a disconnect between what is ideal and what is practical. They often tell you to buy the best test equipment for some thing you are going to do once or recommend things that a beginner would try but then get frustrated and give up once they realize its way too expensive and learning how to do it would put you half way to a BS in EE. Its like running a business, there is a middle ground of practicality and isn't found in any text book but comes from common sense. Often great engineers are shitty business people because they can't step back and say "Yes I would do that, but when I was just starting or an average mechanically inclined person would it make sense to do that if it was even possible?" Sometime the quick and dirty solution is the best solution when all that matters is the outcome and not how you got there. But conversely when the outcome isn't as important (its not to make a financial gain or just get something working) what is of value is learning from the process. No one blinks an LED on off because their house needs strobe lighting or they own a blinking led company its to learn how it works, even if it doesn't work hopefully they learned what not to do.

Sorry, I don't see the point, you're trying to make?

In this case the most idea/practical solution is to buy a suitable switched mode power supply, for your amplifier. I agree that designing one is not feasible for a beginner.

There isn't much you will learn by building a PSU, using a huge transformer, rectifier and capacitor. It would just be the same small power supply just scaled up. It's also inferior to using a switched mode power supply, in just about every single way. Quite often people avoid switched mode power supplied for audio applications, but in this case, it's a car amplifier, which will have plenty of filtering and a switched mode power supply built-in, to increase the voltage to a level, to deliver the required power into an 8 or 4 Ohm speaker.

A good pen and paper exercise would be reverse engineering the existing amplifier and performing some tests on it, to check if it really meets its advertised specification. Quite often manufacturers grossly exaggerate the specifications for their audio amplifiers and it would be interesting to see whether this is the case here.

These amps were made by A/D/S/ and Kicker when they used to build the amps out of Stillwater OK. They even send a print out of each amps performance. Yes modern day junk out of china, especially if it's directly from them, will have a made up number and an amp that can't deliver the current to keep up. Sony back in the day was the worst offender, those amps produced no current so their watts rating was all on paper delivering high voltage which you don't want for driving a linear motor. Their subs sounded like shit too was all about being loud and sloppy. I had dual 15" that sounded like a single 10". Precision and volume was my goal.

[Zero999]

The cheapest option is to get a 12V computer PSU with 600W+ rating. A decent computer PSU will have almost all its output to be 12V. It's common to see a 600W ATX PSU can provide 550W+ on 12V rail itself.
However, to meet EU safety regulation, a single 12V rail can't go beyond a certain current, I think 30A or 40A, so many PSUs provide dual or triple 12V without parallel operation, each at 20A or so. You CAN'T USE THEM.
Find a single 12V, 600W+ PSU, and that's all you need.

That's interesting. I've not heard of sue an EU regulation. Presumably it just applies to computer power supplies?