You say "no load". Is that with absolutely nothing connected to the supply? If yes, expect some droop once you hook up the Parallel-86.
As Andrew said ±42 V is ABS MAX (with input signal). You want to ensure that the supply voltage remains below this value even if the mains voltage varies to the max.
Tom
Here is a peek at my (finally) completed build. It was configured to drive my LXminis. Into an eBay BZ4309 enclosure I've packaged:
-an Antek AS-3222 transformer
-2x Power 86 (one for each speaker channel)
-2x Modulus 86 (one for each full range driver)
-2x Parallel 86 (one for each low frequency driver)
As you can see it's tight but workable. So today after installing a 6.3A slo-blow fuse and passing the quick check on Q4 supply voltages I populated the IC sockets, connected it to the LXminis and fed it some music from my Surface 3 Pro via jRiver Media Center 21 through an Emotiva Stealth DC1 preamp/DAC.
The sound was superb, and this is in my ~700 square foot shop with concrete walls, concrete floor and very little in the way of absorption. There was ample drive (I used the 26dB gain resistors) and after an hour of listening at the high end of comfortable the heat sinks were just perceptibly above ambient temperature at maybe mid-80s F. It was dead quiet between passages, even with an ear at the FR driver. So next I will pack it all back into the family room where the LXminis do duty both as TV speakers and for musical program. Tom...I am very impressed with both the final product, and with the no-issues build thanks to your superb documentation!
-an Antek AS-3222 transformer
-2x Power 86 (one for each speaker channel)
-2x Modulus 86 (one for each full range driver)
-2x Parallel 86 (one for each low frequency driver)
As you can see it's tight but workable. So today after installing a 6.3A slo-blow fuse and passing the quick check on Q4 supply voltages I populated the IC sockets, connected it to the LXminis and fed it some music from my Surface 3 Pro via jRiver Media Center 21 through an Emotiva Stealth DC1 preamp/DAC.
The sound was superb, and this is in my ~700 square foot shop with concrete walls, concrete floor and very little in the way of absorption. There was ample drive (I used the 26dB gain resistors) and after an hour of listening at the high end of comfortable the heat sinks were just perceptibly above ambient temperature at maybe mid-80s F. It was dead quiet between passages, even with an ear at the FR driver. So next I will pack it all back into the family room where the LXminis do duty both as TV speakers and for musical program. Tom...I am very impressed with both the final product, and with the no-issues build thanks to your superb documentation!
Attachments
Dang! That's a very nice and compact build. Nicely done, Sir!
Thank you. I'm glad everything worked out well for you and you like the amp.
Tom
I'd double check that it's below ±42 V and hook it to the Parallel-86. See how much droop you get. Measure the supply voltage with the PAR86 connected. Measure the AC line voltage. Do the math to see how much supply voltage you can expect at high mains.
In the ideal world, you'd have ±35 to ±36 V with the amp idle at nominal mains.
You need to stay below ±42 V at all times.
Tom
In the ideal world, you'd have ±35 to ±36 V with the amp idle at nominal mains.
You need to stay below ±42 V at all times.
Tom
I've noticed so many mention LXmini speakers and I assumed they were small monitor sized speakers like LS35As. After spending a little time on the SL site I found they are floor standers and need to be away from walls. I can't imagine why they are called minis but, SL says they are magical and are reference quality, so I am interested.
While I am interested, they will take up the same floor space as my tower speakers. My only concern is lack of the low end. I have a Velodyne powered sub, I don't use, and would appreciate any thoughts on using it with the Lxminis.
Thanks,
henrylrjr
While I am interested, they will take up the same floor space as my tower speakers. My only concern is lack of the low end. I have a Velodyne powered sub, I don't use, and would appreciate any thoughts on using it with the Lxminis.
Thanks,
henrylrjr
Hi henryljr,
AFAIK these came out first:
LX521 Description
Dr. Linkwitz came up with the LX-Mini as an entry level solution for folks on a smaller budget and smaller listening space... probably hence the "Mini".
AFAIK these came out first:
LX521 Description
Dr. Linkwitz came up with the LX-Mini as an entry level solution for folks on a smaller budget and smaller listening space... probably hence the "Mini".
That sounds like a great question to ask in one of the speaker forums. The multi-way forum would probably be the best place.
While the Modulus-86 and Parallel-86 work well with both the LXmini (that's what I use), Orion, and LX521 the discussion about how to address the various shortcomings of the LXmini, how to address them, and whether to use a passive XO would be better served elsewhere. You may also find some insight in the OPLUG forum.
Tom
Perfect!
I suggest measuring the mains voltage and the DC supply voltage. Then do the following math:
Measured mains voltage: Vmains_meas
Max. mains voltage: Vmains_max
Measured DC supply voltage: Vsupply_meas
Max. DC supply voltage: Vsupply_max = (Vmains_max / Vmains_meas) * Vsupply_meas
Verify that Vsupply_max is below 42 V.
Tom
Last edited:
To be a bit more specific, the 253V Vmains_max suggested carries increased risk of amplifier failure. In the real world (unlike the Andrewverse) overvoltage events happen. Tom's ±35 to ±36 V target accommodates overvoltages of 17-20%. Reasonable but arguably a bit to the low side as 120% mains events do rarely occur in most first world locations (there'll be additional or more frequent variation in less solid grids). It's a personal choice as to how conservative to be but a 276V Vmains_max (120%) is worth considering when doing the maths.
A simpler way of dealing with this (and pick up a number of other benefits) is to use an SMPS-86 type approach, though that's no help in reusing an existing linear trafo.
Assuming the mains voltage doesn't sit constantly at 276 VAC but only goes there during short-duration power surges, one could add a MOV (aka surge suppressor).
The LM3886 power supply spec is 84 V (±42 V) ABS MAX with signal applied. 94 V (±47 V) ABS MAX without signal applied. I generally recommends staying at or below ±35 V for a linear, unregulated supply. ±36 V if using a well-regulated supply. The few pB of extra power a higher voltage would provide aren't worth the reliability hassles. I'd rather err on the side of caution and have a reliable amp.
Tom
The LM3886 power supply spec is 84 V (±42 V) ABS MAX with signal applied. 94 V (±47 V) ABS MAX without signal applied. I generally recommends staying at or below ±35 V for a linear, unregulated supply. ±36 V if using a well-regulated supply. The few pB of extra power a higher voltage would provide aren't worth the reliability hassles. I'd rather err on the side of caution and have a reliable amp.
Tom
Or just plug one's kit into a surge suppressor. The terminology is loose but amongst big E folks (grid engineers, power plant operators, and so on) most events referred to as overvoltages or undervoltages have durations of seconds to a few minutes. These will generally defeat the MOV+reservoir cap type lowpass filtering used in suppressing shorter duration events like voltage spikes and surges. Between the source impedance of the mains, transformer inductance, and large reservoir caps most linear power amp supplies don't do too badly at filtering spikes and surges on their own. A one or two second overvoltage is usually below the cutoff, though.
Fortunately overvoltages beyond 110% are infrequent and rare beyond 115%. On the 24V end of things one's substantially more likely to encounter degraded PSRR due to the control device regulators entering dropout during brownouts.
Fortunately overvoltages beyond 110% are infrequent and rare beyond 115%. On the 24V end of things one's substantially more likely to encounter degraded PSRR due to the control device regulators entering dropout during brownouts.