I built a pair of LM3886 monoblocks back in 2011, each with a 120VA 2x25V transformer & 2x 8200uF filter capacitors. They power a pair of 5" 2-way bookshelf speakers from ~80Hz up while I have a separate TDA8920BTH amp (bridged) powering an 8" sub under 80Hz.
I would like to downsize this setup - the LM3886 amps currently sit in a huge 3U 19" enclosure, with another 1U 19" for the active crossover & then a separate project box on top of the 6U 19" rack for the TDA8920 amp. It's kinda ridiculous.
My plan is to build a third LM3886 amp (specifically two LM3886 in parallel) to power the sub & then to house all three LM3886 amps along with a crossover PCB into a single enclosure.
My question is whether I could realistically use the two 120VA transformers I already have to power all three LM3886 amps - using the first to power both of the single LM3886 amps I already have & the second to power the new parallel LM3886 that I plan to make (so each 120VA transformer would be powering two LM3886).
Realistically speaking, I will never want more than ~20W of clean power from each of the single amps & ~50W from the parallel amp - & even that would be on rare occasions. These are all 8Ω speakers. My understanding of the VA to W relationship isn't the best, so can anybody weigh in on whether this is realistic?
If this idea isn't even remotely realistic, I can dismantle the TDA8920 amp for its 225VA 2x18V transformer & use that for the parallel LM3886, but if it's plausible to run the whole setup from the two 120VA transformers I would prefer to do that & keep the size/weight/complexity of the project to a minimum.
I would like to downsize this setup - the LM3886 amps currently sit in a huge 3U 19" enclosure, with another 1U 19" for the active crossover & then a separate project box on top of the 6U 19" rack for the TDA8920 amp. It's kinda ridiculous.
My plan is to build a third LM3886 amp (specifically two LM3886 in parallel) to power the sub & then to house all three LM3886 amps along with a crossover PCB into a single enclosure.
My question is whether I could realistically use the two 120VA transformers I already have to power all three LM3886 amps - using the first to power both of the single LM3886 amps I already have & the second to power the new parallel LM3886 that I plan to make (so each 120VA transformer would be powering two LM3886).
Realistically speaking, I will never want more than ~20W of clean power from each of the single amps & ~50W from the parallel amp - & even that would be on rare occasions. These are all 8Ω speakers. My understanding of the VA to W relationship isn't the best, so can anybody weigh in on whether this is realistic?
If this idea isn't even remotely realistic, I can dismantle the TDA8920 amp for its 225VA 2x18V transformer & use that for the parallel LM3886, but if it's plausible to run the whole setup from the two 120VA transformers I would prefer to do that & keep the size/weight/complexity of the project to a minimum.
I did a quick search for a calculator, but couldn't find one I could understand :3 My intuition said that 120VA should be fine for 50W of output, especially toward the upper end of the IC's voltage handling & thus efficiency, but it's good to hear that your intuition is on the same track!
The datasheet states 84V as the absolute maximum operating supply voltage (94V with no signal) & IIRC I measured something like 76V when I built them back in 2011. I remember this being higher than I was expecting, but I probably didn't realise at the time that the transformers ratings assume 230V across the primary & the mains where I live is usually more like 245V. They've been running fine like this since 2011 so I'm not concerned about voltage, just whether the VA rating of the transformers will be sufficient.
Just measured the mains voltage in my current apartment & it's steady at 247V, which means just a hair under +/-38V on paper before any voltage sag. As I said previously, I remember metering around 76V from the output of the rectifiers when I built them back in 2011 at a different apartment. So they've been running for 7 years like this with zero problems & have always sounded lovely 
So if nobody thinks it is implausible to power two LM3886 from each 120VA transformer for ~50W max load, I think I will go ahead & start planning a new enclosure?
So if nobody thinks it is implausible to power two LM3886 from each 120VA transformer for ~50W max load, I think I will go ahead & start planning a new enclosure?
You've got more than enough transformer on hand. The calculator sheet is at the bottom of that page: LM3886 chip power amplifier power supply design. The section on crest factor is well worth reading.
Thanks, I'll give that a look. I'm familiar with the impact of crest factor from working a lot in the past with pro audio/PA gear, where in the last decade or so there has been a big shift from transformer based amps with power outputs specified using band-limited pink noise, to SMPS based amps with comparatively exaggerated power outputs specified assuming program material with a 6dB crest factor... even though the compression of popular music mastering these days means that it has substantially less crest factor than it used to!
Funny thing is that squashed (low crest fact.) audio will be easier on classic PSU (trafo based) when the playback volume is low, but much heavier when playback is loud (rails will sag a lot more compared to higher crest fac. audio).
I built myself small amps to "model" that situation when I was doing multimedia audio (which naturally had very low crest ... and some years before this became a norm in mainstream audio).
I built myself small amps to "model" that situation when I was doing multimedia audio (which naturally had very low crest ... and some years before this became a norm in mainstream audio).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.