Helium - a true micromonitor

[scottsehlin]

If you already have the parts to do 14 and 40 ohms, I would stick with those. The 15 and 47 ohm values were chosen to avoid the need to go with multiple parts or parts larger than they need to be. PE used to carry a 5W 39 ohm resistor when the design was first published.

[mmu7]

Just got the crossover parts. Looking at the two schematics the 2 ohm resistor and 4 uF capacitor are switched, does the order of them matter or is there a typo in one of the diagrams?

[scottsehlin]

The order of those two parts doesn't matter. The convention usually is to place the capacitor before the resistor - so that's what I did. In practice, I sometimes switch them if it helps improve the layout of the crossover board.

[Dunk_c]

So are these speakers rated at 30W RMS power handling when finished? If so, do we need about 60W of clean power per channel from the amp to drive them fully?

[scottsehlin]

30W is the thermal power handling. It means that running a sine wave through these at 30W or more continuous RMS power will risk melting the voice coil. Musical content would be much less stressful on the voice coil (amplifier clipping can be an exception).

Usually, small speakers are limited more by excursion than thermal considerations. The ND91 woofer used in the Helium is really quite special. Linear excursion is rated at 4.6 mm, but the woofer reportedly can travel 25 mm one-way (one inch!) before reaching it's mechanical limits. At 30W input, the ND91-4 will output approximately 96 dB at 1 meter. The Helium's enclosure and tuning helps control the woofer's excursion. With that 30W input, excursion hits a local peak of right around 4.5 mm between 90 and 100 Hz, then it drops to a minimum around the port tuning frequency, back to 4.5 mm at 55 Hz, then increasing up to 15 mm at 30 Hz. At 30 Hz, the box and tuning results in the response 25 dB below that 96 dB level at higher frequencies, so the distortion due to the woofer being exercised that much won't be very prominent - and the extremely large amount of mechanical travel before bottoming out means that this speaker will handle the full 30W pretty gracefully - even without a high pass filter. With a high pass filter, even at 80 Hz, this speaker will handle considerably higher peaks. At 75W input, the ND91 in the Helium's enclosure will be outputting about 100 dB and will still be just below 25mm of excursion at 30 Hz, so 75W peaks are about all the Helium would want to see without a high pass filter of some sort for music with deep bass content.

What that all means is that your amp is less than about 75W, they could be played louder if you had the juice. Whether you would want to play them louder depends on the size of your room, how far away your listening position is, background noise, etc... In practice, I find that if I design something that can handle being played at 95 dB, I don't really ever have trouble with over-excursion while listening to music at non-painful levels, even in venues such as DIY events. The Heliums have passed that test on a few occasions.

[Dunk_c]

Thanks for the tech details. Excuse my ignorance, but where does the power handling capability of the tiny 10W tweeter come into the story?

[scottsehlin]

One of the great things about the Xsim program is that it can directly answer this question. I set a simulated amp to output 30W into 4 ohms into a simulation of the Helium design. This is the resulting power dissipation for each of the drivers as a function of frequency (S1 is the ND91, S2 is the ND16 tweeter). The tweeter sees very little power in this design.

[Dunk_c]

So delivering 75W of power into each Helium will give 95dB of audio output and the drivers willl be comfortable. I think this means that the speakers can deliver 77dB at 1.2W which is close enough to the standard reference point, as far as I know. What is the rule that determines how sound level changes with number of speaker units? How many dB will be delivered when 2 speakers are being driven?

Can you roughly speculate on amplifier power requirements in order to get 75W of clean power into Heliums without clipping? I belive that knowing this number is more about protecting the tweeter than preventing audible distortion (could be wrong, just learning). I am guessing 100-150W nominal.

I don’t expect I will ever run them at high outputs - just curious about the specs. In fact, while they sounded good to my ears when played in a large room using a cheap 15W D-class amp, they were obviously no match for the floor standers being used as “speaker stands” so I am now enjoying them as personal speakers in my workshop. This is not a criticism at all, I think I could use them through the house and may make another two pairs for my teenagers when they want their own hifi system (using cheap D-class amplifier boards, because I enjoy the low cost DIY).

[Wolf]

I really doubt the Heliums can handle a full 75W before some kind of mechanical/thermal issues occur. That's a lot of power for a 3" driver to take.
Later,
Wolf

[scottsehlin]

That's what the math says if the 25 mm xlim is real. It couldn't take a sine wave at that power level, but theoretically could take occasional peaks. It is an inefficient 4 ohm speaker, so that doesn't translate into extreme SPL's. They also roll off high enough to provide some protection for the woofer, as opposed to designs that try to extend down into the 40-50 Hz range with a 3-4 inch driver. In reality, I doubt that the suspension can support an inch of travel - so one would hear distress at some point and eventually damage the surround and/or spider.

I haven't done any conclusive testing where I was monitoring the voltage input level, but they have survived some rough treatment at DIY events...

I do think the biggest enemy to the voice coil is clipping, so having clean power is the key to trying to play them closer to their limits. Clean watts can be a challenge to figure out in this era of overstated specs. If you have an amp that can do 50 wpc with less than 0.1% THD from 20 Hz to 20 kHz and is 4 ohm stable, you should be fine. If the power rating is 100 wpc at 10% THD at 1 kHz (which you will see far too often), you really don't have any idea what you have - but it probably isn't very good.

[Dunk_c]

Thanks.
Taking another look at my requirements for a variety of listening setups, using online calculators and info (CrownAudio), here are some calculated amp requirements (assuming that the Heliums are 75dB at 1M and 1W; with 3dB headroom to reduce chance clipping, and probably many other assumptions):

SPL , Distance , Amp Watts/ch into 4ohms with low THD
75dB 1M 2W (desk)
75dB 2M 8W (my small workshop)
75dB 3M 18W (medium sized room, perhaps)
80dB 1M 6W
80dB 2M 25W
80dB 3M 57W
85dB 1M 20W
85dB 2M 80W
85dB 3M 180W
90dB 1M 63W (added just for interest)

Another web calculator determined that for 75dB speakers, with max power handling of 30W, that the maximum output of the speaker would be 90dB. Not sure how this fits in with the data above?

From this I could conclude that when used as nearfiled speakers, hardly need any power at all. I am mucking around with chinese D-class amps and I get plenty of volume (80-85dB, measured with Radioshack SPLM C weighted) from 10-20W (as listed) at 1-2M listening distance in a very small room. This seems to make sense.

If I found an amp with 50-60W clean power that would cover quite a few options.

How about a DTA-120 ? But it only has

[bsl100]

Newbie questions with regard to post #320 and #322 and the overall design in general.

1. What is the load that the amp will see if the drivers in the Helium are 4 ohms and 6 ohms and they are connected in parallel.
2. Are the resistors across the tweeter changing the load on the amp or the work as an L-PAD.
3. What are the F3 and port tuning frequencies.

Thanks in advance.

[Chris Roemer]

Basically, the amp needs to be OK w/a nominal 4ohm speaker. The woofer section will be 4ohm, and the tweeter band near 6ohms. Scott's Z-graph shows a Z-min of about 5ohms near 250Hz and again around 10kHz. Since a "nominal" 4ohm speaker would most likely have Z-mins closer to 3ohms at these freqs, EYE would call this a nominal 6ohm design. (I'll wager that many "8 ohm" amps will be OK with this load.)

L-pad.

A 3/4"id x 7" long port will tune the box into the low 60s (you can SEE this on the woofer's section of the Z-plot, at the bottom of the "valley"), yielding an F3 in the low 90s (using Dayton's latest (2016?) specs in WinISD).

[scottsehlin]

My sim gets an f3 of around 70 Hz using the parameters posted on the PE website. Chris is correct about the impedance. The second post on page 1 includes a plot of impedance as a function of frequency.

For Dunk_c: The DTA-120 should be okay. I had it's predecessor, and it could do about 30 wpc into 8 ohms and closer to 50 into 4 - but there were thermal issues running into a 4 ohm load. Mine eventually died (a friend had it powering a pair of Lithiums). If I recall correctly, the DTA-120 was supposed to address that issue. I used the SMSL Q5 Pro for the Helium soundbar and it has several nice features that the DTA120 doesn't, like remote control, USB DAC, and subwoofer out, but it costs a little more.

[bsl100]

Thanks Chris, Scott.... Am getting the picture that as long as each driver in a 2-way is rated at a nominal / minimum 4 ohm and the amp can handle 4 ohm loads, then it should be fine even if the drivers are connected in parallel.

Scott... I'm going to try your design using the ND91-4 along with an AMT-8 tweeter. The 8 ohm tweeters are about 89db 1W/M and have a frequency response from 3500-40000Hz. They will be powered by a TPA3116D2 amp.

Since I had a couple of 0.9mH inductors with me, was wondering if I could use these for the woofer section. Will it affect the other components for the woofer and change the sound signature? Also, would I need to redesign the entire tweeter crossover for the AMT-8?