TL , agreed
I would build it and prove the doubter wrong. If you have access to an O-scope you could verify transient regulation, also ripple and noise reduction with some chokes I mentioned.
I would build it and prove the doubter wrong. If you have access to an O-scope you could verify transient regulation, also ripple and noise reduction with some chokes I mentioned.
...but I hear it is making a comeback with downloadable hi-rez files and software features like ReplayGain that give control of the volume knob right back to the user. Some of these files even sound like actual music. There is hope, still. I would certainly not design a system around the pathetic 3-5 dB dynamic range that some "modern" recordings have.
I'm a bit surprised that nobody has picked up on the speaker sensitivities here. Just as a back-of-the-envelope calculation, suppose that you want a midrange peak SPL of 105 dB in your listening position. You would need something like 108 dB @ 1 m from one loudspeaker, making various assumptions about both listening distance and room gain, but you have two speakers, and each will need to deliver 105 dB @ 1 m. That is 14 dB above the 1 W rating of your midranges, so you need 25 W at each loudspeaker. Given the 4 ohm rating, that would be 2.5 A and 10 V RMS. Suppose this sonic boom hits close to the speaker resonance frequency, and that the reactive phase angle causes twice the nominal current to be drawn, 5 A. For a sine wave, peak is 1.4 times RMS, so we are at 7 A and 14 V peak delivered from the amp. At power levels like this, the LM3886 dissipates another 30 W as heat, or about 1.25 A at the 24 V supply voltage. Then you would need 8.25 A from the supply, or 16.5 A for both amps and speakers. That is 113 % of the nominal current capacity for the supply, which may or may not be within its instantaneous capacity when supported by a couple of reservoir caps. Redoing the calculation, a 102 dB peak should be well within capacity, drawing approx 10 A from the supply for both amps. In fact, it looks like you could be playing today's hypercompressed pop dreck at 100+ dB average level and still have headroom for whatever peaks there might be in that sort of thing.
My advice: Give it a try. You are more likely to destroy your speakers or ears before running out of current.
Last edited:
And to put a bit more perspective on the last post, I grabbed the old Radio Shack SLP meter, blew off the dust just to get a quick idea of the SPL level from my listening position (approx 1 meter). BTW.....there are DIY nearfield monitors that I use for enjoyment listening as well as the occasional music recording. With that said, what I got for SPL was from 89db to 95db, A weighted. THIS is why I want to open up some room and not use the Kenwood Basic M2A (350wpc) and the NAD2100 (60wpc) Slight overkill 😱
1 m listening distance means that you can add 3 dB to the SPL calculation above. Assuming that your average listening SPL is around 92 dB from two speakers, then you will have some 15 dB headroom for dynamics. Very few recordings have that much dynamic range. 8-12 dB is more common even in quality recordings, so you should be able to reproduce pretty much every CD ever produced, and still have approx 3 dB headroom to spare before running out of current. The exceptions are specialist items like the Sheffield Labs Drum & Track Record with a dynamic range of 17 dB.
Loudspeakers with 90+ dB sensitivity are very different animals from the typical 83-85 dB boxes. Let's see, hypercompressed recordings and "lifestyle" speakers that try to give bass extension from tiny boxes - no wonder that most people do not listen to music any more.
Loudspeakers with 90+ dB sensitivity are very different animals from the typical 83-85 dB boxes. Let's see, hypercompressed recordings and "lifestyle" speakers that try to give bass extension from tiny boxes - no wonder that most people do not listen to music any more.
Last edited:
I do believe I'll give it a go. What would be the worst thing that would happen? The SMPSs go into "Protect Mode" if over driven?
No Chance
No chance of that happening. Two S-350 Meanwell SMPS are well capable of melting a pair of 3886. I doubt you will ever even get their fans to come on. And, I have found with my 2050 amps that they sound much better than my best linear supply.
No chance of that happening. Two S-350 Meanwell SMPS are well capable of melting a pair of 3886. I doubt you will ever even get their fans to come on. And, I have found with my 2050 amps that they sound much better than my best linear supply.
4 melted
Right. I meant to say stereo pair. The S-350 SMPSs will boss them around under full control with by far the best sound for the money. I would look for the 36v Meanwells on ebay, which can adjust down to 31v if your chips get too hot, in order to get more power from your chips than you will have with +-24v supplies. Add some storage of up to 1-3000uf at each chip and you are set.
Sorry Sendler , Not the best advice.
Your advice would be good for an 8 ohm driver, but don't forget TL is running these on 4 ohm mid-bass (Re~ 3.4 ohms) which if you study the LM3886 data sheet shows that running at higher rails leaves less output power for this lower Rl configuration. Also the Meanwells current limit on the higher voltage units are set lower. AFAIK the 24 V outputs are adjustable up to 27 Vdc with the same higher current.
Now that the power supply thing is sorted. On to the PCBs for the LM3886s. Who's is the best? I was thinking getting the boards from Chipamp.com and populating them with parts from Mouser.
I'd suggest the best one is the one with the best layout and grounding. With two SMPSUs you'll want one that allows for two independent supplies - that means having two ground connections for the power inputs which meet at the star point on the PCB.
Turned up
Reports from the 2X100 thread seem to indicate that the Meanwells sound less good turned up than they do at the nominal voltage or turned down. Not sure why. I haven't tried it. My supplies are 36v units running at 32v.
With the 4ohm nominal mid/woofer the highest I can run is 24V-25V per rail. So, is Chipamp.com boards populated with parts specified from Chipamp.com and ordered from Mouser a good choice? All of BrianGT's kits the come with components also come with PSU parts and boards. That's why I would be ordering just the boards from him.
I think the Chipamp.com PCBs are designed to use their power supply boards esp for multiple amp setups. AFAIK they leave out some key parts and/or call them "optional". I'd agree with abrax's post above, also adding that the speaker returns go to their respective star ground on each amps PCB. I think its best to work out the wiring diagram with ground scheme and proposed amp schematics /parts list before placing the order. It's important to provide an image of the PCB routing to make an informed decision for the star ground and various capacitor sizes and lead spacing.
Like I started out saying in my first post, I know only slightly more than nothing ( might as well assume nothing) about electronics. Here is a pic of the Chipamp.com PCB for just the amp. You tell me. If anyone has a better option,please pass it on.
An externally hosted image should be here but it was not working when we last tested it.
Yes those should work. Double sided copper so the ground star is there on top side at the junction of the two supply caps. Use 4 twisted pairs of wires for off board connections. You probably want to consider the non-isolated tabbed version LM3886 due to higher currents on the 4 ohm side. Also don't forget to add an input signal DC blocking cap ( off board if it's a film type ) it's really good insurance even if you don't have DC on the pre/crossover outs. This input cap really should NOT be optional esp if the amps inputs are user accessible. Some folks with out them have blown up chip amps and speakers.
I'll just get the same components to populate the board as what Chipamp.com lists as what is all in their kit.
Case design I'm still playing with. Separate power supply case or all in one? Input pots for each channel? Things like that.
Case design I'm still playing with. Separate power supply case or all in one? Input pots for each channel? Things like that.
Hi,
get a PSU working then add an amp & get that working.
Do this all "open air" on the work bench.
Add second channel and get that working.
Now decide if the second channel needs different components to suit you. You have channel one to compare to modified channel two.
Listen for a while in this "exposed" arrangement.
When you are happy with the final component arrangement, only then think about how you can place the assemblies next to/above each other and work on "packaging" the amplifier allowing for ventilation.
Then buy a case/s to suit.
get a PSU working then add an amp & get that working.
Do this all "open air" on the work bench.
Add second channel and get that working.
Now decide if the second channel needs different components to suit you. You have channel one to compare to modified channel two.
Listen for a while in this "exposed" arrangement.
When you are happy with the final component arrangement, only then think about how you can place the assemblies next to/above each other and work on "packaging" the amplifier allowing for ventilation.
Then buy a case/s to suit.
Very good advice... I wish I'd thought about that before starting my amp project... I've made so many changes along the way that my enclosure is full of drill holes, I've also wasted lots of shrink tubing and tech-flex making it look pretty, just to rip it out to make changes
- Status
- Not open for further replies.