I've got two toroidal transformers, each having 2x40vac windings, left over from a past project. I'm going to use them in a new power amp project for which there isn't really a need for separate transformers (one bigger one would likely have been more efficient), so I need to decide how best to assign them. Their job will be powering a pair of 3-way speakers with active crossovers (so 6 amps total), but with a twist - the midwoofer+tweeter cabinets will be separate (from the bass cabinets) and sometimes used alone as 2-way speakers.
Tentatively in descending order of my preference, but it seems a bit arbitrary.. so I'd be interested to hear what other people think about the choices or if anyone has better options?
Cheers, Kev.
So, one logical answer would be to use one trafo to supply two stereo amps for the mids+tweeters and the other to supply one stereo amp for the bass drivers. With two possible benefits: firstly any big, depleting bass sounds wouldn't affect the mid+high drivers and secondly if/when only the 2-ways are being used then the bass trafo and associated supply can remain switched off (or even not present at all, if built into a separate enclosure).
Another reasonable option would be to use one trafo for the left and the other for the right channels. In theory this could improve channel separation, though I'm not sure that would be very noticeable if the amps were fed by their own dedicated capacitor banks anyway (which they would be).
Or, the two trafos could be wired together to act as one, either entirely in parallel or else the two windings of each parallelled then the two trafos in series. But that would seem to glean no benefit from having two, whilst being probably less good than one big one.
Tentatively in descending order of my preference, but it seems a bit arbitrary.. so I'd be interested to hear what other people think about the choices or if anyone has better options?
Cheers, Kev.
In my opinion, (1), then (2) but not (3), as there could be local circulating currents between the two transformers. However, you may parallel them after the rectifier.
The first idea sounds most logical, but you have another problem unless these are 500VA+ transformers. 2x40VAC means a 113V rail-rail voltage and ~200W/channel into 8 Ohms, 400W/channel into 4 Ohms. 200W x 4 channels /60% efficiency is 1,333 Watts, 666 Watts for two 8 Ohm channels. You do not want huge amps for mids and tweeters. 2x20VAC would give you almost 50W/channel @ 8 Ohms, plenty for mids and highs.
Thanks for the thoughts, very useful! Yes the potential for currents between the two transformers were my main reasons for disliking option3, but I'd not thought about doing it after the rectifiers. It doesn't seem the strongest option still, but done like that it would certainly be viable if wished so that is good.
Cheers,
Kev
Cheers,
Kev
Yes Steve, they are 500VA transformers and would make +/-56Vdc rails. Although the amps are to be mosfet types (ESP prpject101) so the output power is a little bit less; supposedly they'll deliver 150W peak into 8ohms - although comments on here suggest less in practice (I don't have any 4ohm drivers). Also, the aim is to have headroom for nice big clean peaks (rather than deafening continuous SPLs); so continuous power would be considerably less. If the peaks were 20dB above average then the bass and mid-woofer would normally only draw a couple of watts each. So all in all the trafo sizes should be okay, in theory anyway.
Though it does still raise an interesting point. I think the plan may be okay for the mid-woofers: these are rated to 60Wrms (100h noise test), and actually are the least sensitive of all three drivers, so need the most power for the same SPLs. Plus (somewhat unusually) when the setup is used as small 2-ways, they will be responsible for all the bass. But the tweeters are another matter, as they are 7dB more sensitive than the mid-woofers and so certainly don't need such high voltage rails; it was really just for economy that they were to share the same trafo.
I'm not sure if it would actually be a problem to do so though? I reasoned that the tweeters would be crossed over at some khz, and so never draw much continuous power even if available (the amps would only need one pair of output devices, whereas the mid and bass amps would have two pairs). So perhaps it is more a question of the higher voltage being a problem in some way?
Cheers,
Kev
To answer the question, (1).
steveu is right though in that these are probably a bit high-voltage unless you are building an extremely powerful system, e.g. with large PA drivers. I built an amp a bit like that and used a 35+35V for the woofers (2x100W/8R) and a 25+25V for the mids+tweeters *4x50W/8R).
steveu is right though in that these are probably a bit high-voltage unless you are building an extremely powerful system, e.g. with large PA drivers. I built an amp a bit like that and used a 35+35V for the woofers (2x100W/8R) and a 25+25V for the mids+tweeters *4x50W/8R).
Thanks, so it is looking like a general consensus then, which for once coincides with what I was thinking; good news 🙂
Hmm, though it does seem to be slightly over-powered from what people are thinking. Not sure that I want to replace the existing transformers as the intention was to mostly use what i already have; maybe I could instead consider how to limit or protect things if necessary.
However, it wouldn't cost a lot to buy an 'additional' smaller transformer just for the tweeters, especially as I'd already intended to have separate capacitor banks for them anyway (IIRC i have some modestly sized but fairly well specified ones for tweeter amps). That might well be worth it; if so, that would mean a trafo for the bass amps, one for the mid-bass amps and one for the tweeter amps; quite a lot of transformers, but there would be a certain logic to it.
Thanks again,
Kev
Hmm, though it does seem to be slightly over-powered from what people are thinking. Not sure that I want to replace the existing transformers as the intention was to mostly use what i already have; maybe I could instead consider how to limit or protect things if necessary.
However, it wouldn't cost a lot to buy an 'additional' smaller transformer just for the tweeters, especially as I'd already intended to have separate capacitor banks for them anyway (IIRC i have some modestly sized but fairly well specified ones for tweeter amps). That might well be worth it; if so, that would mean a trafo for the bass amps, one for the mid-bass amps and one for the tweeter amps; quite a lot of transformers, but there would be a certain logic to it.
Thanks again,
Kev
Note that there is nothing fundamentally wrong with using an amp at lower power than it is capable of - consider most listening is at about 1-5W while we've all got boxes capable of 200W+. The amp isn't going to sound worse, it will just run warmer than a more-modest amp would.
So on second thoughts and since you have the transformers already, option (2) is good. Three power amps per channel, all rated for that 120V rail, and with one beefy enough to put 200W into the woofer, and the other two a bit less so, say 75W each. Same design but fewer final output devices on 2/3 of the channels.
It will run warmer than strictly necessary but you'll get an honest 300W RMS from each channel across the three bands, at least until the transformers overheat.
With option (1), the woofer transformer would overload first and go into regulation, you will be limited to about 150W/ch for the woofers while the mids+highs will be running very lightly. Option (2) lets you borrow some of the power from the upper bands for the woofer.
So on second thoughts and since you have the transformers already, option (2) is good. Three power amps per channel, all rated for that 120V rail, and with one beefy enough to put 200W into the woofer, and the other two a bit less so, say 75W each. Same design but fewer final output devices on 2/3 of the channels.
It will run warmer than strictly necessary but you'll get an honest 300W RMS from each channel across the three bands, at least until the transformers overheat.
With option (1), the woofer transformer would overload first and go into regulation, you will be limited to about 150W/ch for the woofers while the mids+highs will be running very lightly. Option (2) lets you borrow some of the power from the upper bands for the woofer.
More useful thoughts there, thanks again!
I do like the idea of any driver/frequency having access to the majority of a transformer's power, if other frequencies are using it lightly at the time. I'd not seen it quite like that until now, but it makes a lot of sense. Both of course happen, but I'd guess it is much more common to have big sounds at some frequency-range(s) than to have big sounds entirely on the left or right channels.
Or looked at differently, In my case the midwoofer's lower sensitivity might make them the most power-needy in some ways, but that will still be set against the bigger demands of lower frequencies that tend to occur. So if i avoid dividing trafos according to frequency range it makes things more flexible and less necessary to predetermine.
I may have just revised my preferences 🙂
Cheers,
Kev
I do like the idea of any driver/frequency having access to the majority of a transformer's power, if other frequencies are using it lightly at the time. I'd not seen it quite like that until now, but it makes a lot of sense. Both of course happen, but I'd guess it is much more common to have big sounds at some frequency-range(s) than to have big sounds entirely on the left or right channels.
Or looked at differently, In my case the midwoofer's lower sensitivity might make them the most power-needy in some ways, but that will still be set against the bigger demands of lower frequencies that tend to occur. So if i avoid dividing trafos according to frequency range it makes things more flexible and less necessary to predetermine.
I may have just revised my preferences 🙂
Cheers,
Kev
Okay, so one 500va transformer for left and the other for right channels is the final decision; not what I'd thought earlier up the thread, so thanks again for the thoughts Laplace. The final reasoning is similar to above, just with a bit more confirmation:
So, that is decided. Thanks for the help!
Cheers,
Kev
Whilst there might arguably be more sonic power in the bass section, there is still a lot in the mid-range and the reduced sensitivity of the mid-woofers mean that they will actually need as much or more electrical power to keep up. The peaks are another aspect, that are more restricted by rail voltage than power handling; these peaks might arguably be slightly lower in the mid-range for a lot of music, but not always and again the mid-woofer's lower sensitivity means that they need as much or more rail voltage than the bass driver to reach them. So all in all the mid-woofers would not want a particularly smaller power or lower voltage supply (than the bass driver), plus of course in this case the mid-woofer will sometimes effectively become the bass driver (when used just as a 2-way).
So, my original idea of having a trafo dedicated to the big bass driver would make power availability less flexible across the frequencies, and at times even make one trafo redundant (in the 2-way configuration), so I've abandoned that idea now. One trafo for left and one for right mean they're both always available for any frequencies that want them and in either of the 2-way/3-way configurations.
I 'could' also get a smaller trafo for the tweeters, as their sensitivity mean they don't need as high voltage rails (and there is relatively little power needed too). But as there is no actual need to reduce the rail voltage (it is simply unused headroom) it would seem to make more sense just to go with the trafo supply that is already available. Maybe just reducing the number of output devices, given the lower power needed, though again more won't matter (and may help with bias current). They'll have their own rectifiers and capacitor banks so will be semi-isolated to at least some degree from things going on at lower frequencies.
So, that is decided. Thanks for the help!
Cheers,
Kev