Hi team.
I'm looking to build my first chip amp, either using the LM3886 or LM3875.
I've inherited a 300VA 50V toroidal transformer from an old colleague. This thing is pretty hefty.
Is there any way that it could be useful in a project, or are the outputs just too high voltage?
Any ideas on what I could use it for?
Regards
Ethan
I'm looking to build my first chip amp, either using the LM3886 or LM3875.
I've inherited a 300VA 50V toroidal transformer from an old colleague. This thing is pretty hefty.
Is there any way that it could be useful in a project, or are the outputs just too high voltage?
Any ideas on what I could use it for?
Regards
Ethan
50 volt single winding or 50-0-50 or two each of 50-0 and 50-0. There is a big difference.
If the first then a single rail AC coupled design (LM3886 included) could be a good option. If the middle option then using just one half is still viable. To make a dual rail design from 50-0-50 is pointless with a 300va toroid unless you accepted that you would never reach full voltage swing under load. Not an ideal option all things considered.
If the later option then you can parallel the windings and make a single rail amp using the full 300va available, or two mono amps essentially with isolated supplies.
So lots of options.
If the first then a single rail AC coupled design (LM3886 included) could be a good option. If the middle option then using just one half is still viable. To make a dual rail design from 50-0-50 is pointless with a 300va toroid unless you accepted that you would never reach full voltage swing under load. Not an ideal option all things considered.
If the later option then you can parallel the windings and make a single rail amp using the full 300va available, or two mono amps essentially with isolated supplies.
So lots of options.
Do build yourself a Mains Bulb Tester.
It will prevent damage to your transformer when you first power ON.
With all those options there is much leeway to getting the wiring a bit wrong and irretrievably damage the transformer beyond repair.
It will prevent damage to your transformer when you first power ON.
With all those options there is much leeway to getting the wiring a bit wrong and irretrievably damage the transformer beyond repair.
Cheers for the quick reply.
Fairly sure it's the last option, two each of 0-50V. I don't have any documentation for the transformer other than what is written on it. Last line is O/P: 0---50V 0---50V (with the corresponding 4 colours listed underneath).
I'll have to do some more reading on single rail amps, but would a dual mono use two separate rectifier boards to each power a single channel?
That does sound like two separate windings. So you have the option of making two independent amplifiers, each a single rail AC coupled design (actually a great option) that are effectively each fed from a 150va winding. Nice 🙂
Second option is paralleling the windings and making a single power supply feeding a pair of single rail amps.
Third option I wouldn't recommend is a high voltage discrete amp (because you would be running on -/+70 volts, so 140 in total). The potential performance of the amp would be unachievable with a 300va transformer and you have the added problems of higher dissipation under normal use at normal volume levels.
And definitely do as Andrew suggests and use a bulb tester.
Second option is paralleling the windings and making a single power supply feeding a pair of single rail amps.
Third option I wouldn't recommend is a high voltage discrete amp (because you would be running on -/+70 volts, so 140 in total). The potential performance of the amp would be unachievable with a 300va transformer and you have the added problems of higher dissipation under normal use at normal volume levels.
And definitely do as Andrew suggests and use a bulb tester.
The lm3886 datasheet has information on building a single supply rail version. Read it carefully and come back with questions on anything/everything you are unsure about.
The single rail topology can be applied at any dual polarity amplifier by using the same AC coupling of the three in/out/feedback connections.
Thanks to both of you.
I'll most likely go the recommended single rail coupled design. Will have a good look over the datasheet.
Anything important to consider having a higher voltage secondary?
I'll most likely go the recommended single rail coupled design. Will have a good look over the datasheet.
Anything important to consider having a higher voltage secondary?
The LM3886 is rated for 94 volts as a maximum and so your 50 vac winding (just over 70 vdc) is well within spec. The speaker coupling cap would need to be rated for that voltage which is really the only real downside to the amp.
AC coupled amps with a properly rated coupling cap are speaker safe as the amplifier can never apply DC across the speaker.
AC coupled amps with a properly rated coupling cap are speaker safe as the amplifier can never apply DC across the speaker.
50VAC secondaries is going to give 68-70VDC when rectified and smoothed. This is way beyond most chip amp's safe operating voltage. If you really want to build an amplifier with this transformer, you're generally looking at a discrete amp with at least 4 pairs of output transistors, and a big heat sink.
You COULD build a chip amp based on a single, non symmetrical supply I guess.. using just one secondary per amplifier (assuming youre building stereo), but it seems like a lot of extra work to me. Additionally, you would have to use speaker coupling capacitors which is detrimental to sound quality.
You COULD build a chip amp based on a single, non symmetrical supply I guess.. using just one secondary per amplifier (assuming youre building stereo), but it seems like a lot of extra work to me. Additionally, you would have to use speaker coupling capacitors which is detrimental to sound quality.
Just because you can do something doesn't mean that you should.
I think in the end, if you make your goal be "build optimal amp to drive X Speakers", then source parts accordingly to meet that goal, you will be happier than having your goal be "use this power supply" and then finding all of the work-arounds to make it possible.
"I just happened to have some hot-dog relish around, and I wanted to be able to use it, so I put it on my hand-cut, organic, grass fed, New York Strip Steak"
I think in the end, if you make your goal be "build optimal amp to drive X Speakers", then source parts accordingly to meet that goal, you will be happier than having your goal be "use this power supply" and then finding all of the work-arounds to make it possible.
"I just happened to have some hot-dog relish around, and I wanted to be able to use it, so I put it on my hand-cut, organic, grass fed, New York Strip Steak"
well, bit fidly but you could make a regulated supply out of it for a paralell tda 7294 amplifier. it should be okay. but verry verry wastefull.
you could unwind a few turns from the secundary coils, that would make it mutch more usefull i guess.
not sure what you could use it for in its current form tho..
i don't now of any amplifer chip that could utilise say.. 150 volt rail to rail...
maybe a tube amplifier could use the high voltage, however for many designs actually 150 VDC is not sufficient.
maybe you can trade this traffo with someone for a more suited one.
you could unwind a few turns from the secundary coils, that would make it mutch more usefull i guess.
not sure what you could use it for in its current form tho..
i don't now of any amplifer chip that could utilise say.. 150 volt rail to rail...
maybe a tube amplifier could use the high voltage, however for many designs actually 150 VDC is not sufficient.
maybe you can trade this traffo with someone for a more suited one.
Poor choice of words.
94V is ABSOLUTE MAXIMUM rating with NO LOAD , meaning going a little above it means destruction and in any case is unusable.
not "well within", but the top practical voltage suggested.
In fact it´s the suggested voltage to get 50W RMS into 8 ohms (forget 4 ohm loads) with excellent performance and not stressing it beyond what´s expected.
A fine choice for what you have.
True, but spending a few bucks on an extra capacitor (one per channel) compared to the savings on a way more expensive transformer is still a good deal 🙂
True and a nice justification fo the small extra caps cost 🙂
In any case cheaper and safer than any "speaker protector" implementation.
What kind of music do you listen to?
What impedance speakers do you have?
LM3886 can't put out more than 50 W maybe 100 peak, with 50 v supplies. It has an internal current limiter.
70 v supply is perfect for a AX6 discrete amp with ONE pair of output transistors, with 8 ohm speakers. For classical music
If you keep the average power below 25W, you have some reserve to put out a 200 W peak occasionally on peaks of classical music. Like the cannon shots in 1812 overture.
Pop, rock, techno, house music, have average level close to maximum level. The soa of the single pair output transistors won't let put out more than the rated 50W. You'd need the 3 pairs of output transistors mentioned above to put out 70-100 W continuously. Plus a massive heat sink, or fans.
Build thread is here: http://www.diyaudio.com/forums/solid-state/236256-retro-amp-50w-single-supply-20.html
post 1 or so has the board layout, my point-to-point version with **** heatsinks and fans is towards the end.
I'd like to point out to all above, the "extra" cap cost of a speaker cap in a single supply design is much cheaper and more newby proof for not destroying speakers, than any relay speaker protection circuit . Another thread current now has the protector board cost down to $10 a pair with a **** off brand relay, plus of course $8 freight of an e-bay purchase for 2 items . If you buy a speaker protection cap, they are about $6 for two and the freight is free from a major distributor in the box with all your other parts.
What impedance speakers do you have?
LM3886 can't put out more than 50 W maybe 100 peak, with 50 v supplies. It has an internal current limiter.
70 v supply is perfect for a AX6 discrete amp with ONE pair of output transistors, with 8 ohm speakers. For classical music
If you keep the average power below 25W, you have some reserve to put out a 200 W peak occasionally on peaks of classical music. Like the cannon shots in 1812 overture.
Pop, rock, techno, house music, have average level close to maximum level. The soa of the single pair output transistors won't let put out more than the rated 50W. You'd need the 3 pairs of output transistors mentioned above to put out 70-100 W continuously. Plus a massive heat sink, or fans.
Build thread is here: http://www.diyaudio.com/forums/solid-state/236256-retro-amp-50w-single-supply-20.html
post 1 or so has the board layout, my point-to-point version with **** heatsinks and fans is towards the end.
I'd like to point out to all above, the "extra" cap cost of a speaker cap in a single supply design is much cheaper and more newby proof for not destroying speakers, than any relay speaker protection circuit . Another thread current now has the protector board cost down to $10 a pair with a **** off brand relay, plus of course $8 freight of an e-bay purchase for 2 items . If you buy a speaker protection cap, they are about $6 for two and the freight is free from a major distributor in the box with all your other parts.
Last edited:
Pretty sure I didn't say I absolutely have to use this transformers, hence why I started this thread in the first place. I wanted to know whether or not I could use it, not necessarily finding a place to force it in.
Rock, electro, wife listens to pop, pretty much anything.
Speakers are 80 W 6 ohm Kenwood S-3Ms.
I'm open to building any type of amp (I want to regardless of being able to use this transformer or not), chip amps were just what I stumbled across a while ago as being relatively friendly for a newby and easish to build.
Thanks everyone for the input.
The LM3886, AC coupled seems the perfect choice for simplicity to me.
Your biggest expense will be two reservoir caps (assuming you go for dual mono construction) and two speaker coupling caps. You needn't go silly with the values either, anywhere from 4700uf to an absolute maximum of say 8200uf would be OK and would keep costs down.
You could make a very nice power amplifier with what you have.
Your biggest expense will be two reservoir caps (assuming you go for dual mono construction) and two speaker coupling caps. You needn't go silly with the values either, anywhere from 4700uf to an absolute maximum of say 8200uf would be OK and would keep costs down.
You could make a very nice power amplifier with what you have.
6ohms and 50Vac for the lm3886 is a heavy load.
If that 6ohms is really a 4ohms midbass and 8ohms treble, then it will be even worse.
What do you have on the primary side?
If it is dual primary, can you put the windings in series and end up with 25V?
If it is dual primary, can you put the windings in series and end up with 25V?
> The speaker coupling cap would need to be rated for that voltage
Strictly, the output cap gets *half* the voltage of the main DC supply, and "can" be half voltage. Both should be rounded-up generously. So 100V for main cap, 50V for output cap.
But more (to a point) is always better.
Output caps in HEAVY bench-testing can be highly stressed. A physically large cap is wanted. This may include a high voltage rating.
Output caps in DIY experimental work CAN be stressed to full supply voltage. (Mis-bias the input, the output goes high.) So a full-Volt cap may be wise for self-protection.
Strictly, the output cap gets *half* the voltage of the main DC supply, and "can" be half voltage. Both should be rounded-up generously. So 100V for main cap, 50V for output cap.
But more (to a point) is always better.
Output caps in HEAVY bench-testing can be highly stressed. A physically large cap is wanted. This may include a high voltage rating.
Output caps in DIY experimental work CAN be stressed to full supply voltage. (Mis-bias the input, the output goes high.) So a full-Volt cap may be wise for self-protection.
Last edited:
- Status
- Not open for further replies.