Helium - a true micromonitor

[mga2009]

Just leaving a picutre of my Heliums 8ohm Rev. 2.0 WIP. Probably finishing them within this week.

I don't know how it will turn up, they are tuned at 65hz with a slot port, using the XO of the Helium soundbar.

[Dunk_c]

Hi
Is it possible (or even necessary) to make a passive high pass filter that is inline with the speaker cables to reduce low frequency signals (below those that these speakers can faithfully reproduce) to these speakers where they have been teamed up with a separate sub base speaker that has no pass through filter? Wondering if this is easy, effective and worthwhile? The sub has a low pass filter but the Heliums still get full range signals from the amp.

I found these but they are signal level not speaker level
Harrison Labs FMOD Inline Crossover Pair 100 Hz High Pass RCA on PE

[LOUT]

Is it possible (or even necessary) to make a passive high pass filter that is inline with the speaker cables to reduce low frequency signals (below those that these speakers can faithfully reproduce) to these speakers where they have been teamed up with a separate sub base speaker that has no pass through filter? Wondering if this is easy, effective and worthwhile? The sub has a low pass filter but the Heliums still get full range signals from the amp.

I found these but they are signal level not speaker level
Harrison Labs FMOD Inline Crossover Pair 100 Hz High Pass RCA on PE

Yes, I think it should be possible because it should act just like a series high-pass at the start of both crossover parts where it'll have little/no effect on the tweeter section because that'll already have crossed much higher and it'll taper down the woofer's lows about how you'd expect such a high-pass to do. You'll still want to test or model it to tune how the highpass behaves around the woofer's impedance peak/s.
I'm assuming it's easier to build externally rather than inside in this case?

It should be pretty easy and effective for an amp that's passing full-range, but expect the cap/coil to be fairly large and expensive.
It looks like a standard 2nd order low-pass using a ~500uF-750uF capacitor followed by a 4mH inductor to ground (with a resistor added to the inductor to increase total resistance to around 3-4ohms) might work.

[Dunk_c]

How about a simple inline first order filter.
So, for 100hz cutoff. The plot on page 1 shows 15ohm impedance at 100Hz, Using an online calculator, 100 uF cap required. (For 70Hz, 150uF). I guess 60V woukdbe plenty. Is this about right?

[Chris Roemer]

By itself, a 100uF series cap will cut the bottom end by about -4dB from 300-200Hz, but only [email protected] Below that you'll be -6 near 100Hz, -12 from 50-80Hz, only -5dB @ 40Hz, but dropping below that. (The effective of a 1st-order filter (just a series cap) will give an output that mimics the appearance of the vented box's impedance curve: dual peaks w/a "valley" in-between near 70Hz.)

If you also run Dayton's 10mH iron-core in parallel (so, a 2nd order high-pass), you'll be -3 from 300-200, up to [email protected], then dropping below that: [email protected], -18dB near 60Hz, and down.

An 80uF series cap (and 10mH shunt coil) puts you at -4dB from 300-200, -1 near 130Hz, [email protected], -12 near 90, -18 near 70, and -24dB @ 45Hz.

I think if you just plugged the port (and used no additional filtering) your bottom end should be down (compared to vented) [email protected], [email protected], [email protected], [email protected], and -4dB(again)@50Hz.

These ND drivers CAN take quite a bit of abuse, so...
I'd try them "as is" (and see if you have issues).
Then try blocking the port (your sub integration might be OK just like that).
If you still have excursion/distortion trouble (@100Hz and below), I'd try the 2nd order w/the 80uF cap and 10mH (iron) coil. This will be the most expensive option.

[Dunk_c]

OK, not a trivial mod then with that much impact on SPL due to uneven impedance across the freq range.
Can I ask a basic question. Even though the speaker gets effectively worse at producing sound below a threshold, it does still receive the full signal at the lower end, what impact does this have on the sound produced within its freq range? Crossovers (eg 120Hz) set in HT AV receivers are supposed to relieve the small speakers from bass duty allowing them to do better without the bass-burden. Without a DSP that is all I was hoping to achieve with a simple inline cap (I guess those car audio boys aren’t worried about HiFi reproduction so much when they use simple bass blocking caps).

[LOUT]

Here are a couple VituixCAD sims of the ND91 in a 1.2L ported box:

First, unprotected.


Then with a 2nd order HP using a 500uF cap and 4mH coil (with total 3-4ohm resistance on the coil).


Then a simpler 1st order series HP using only a 200uF capacitor with an 8ohm resistor parallel with the cap.


These are kind of aiming to keep the little speakers playing as close to 70-80hz F3 as possible while still protecting them a fair amount...kind of going on the assumption the subwoofer will be placed somewhere it sounds nice but isn't necessarily near the speakers and crossed low because of it.
If the sub will be somewhat between the speakers where it can be crossed pretty high (closer to 200hz) then you could probably just use a 50-80uF capacitor by itself and the sub would help fill in the wide valley around 150hz that the plain cap cuts out from the ND91. The plain capacitor kinda cuts a wide path of destruction all the way from 75hz to 400hz, so the subwoofer would need to fill in on low-mids as well in that situation.

The plain 80uF capacitor models like this:


Adding the resistor parallel with the larger cap instead (shown in the 3rd cluster of images) helps to keep most of that valley around 3-4db louder, though it's still around 3-4db down compared to the 2nd order HP or the unprotected model.

[scottsehlin]

This is some nice work and very helpful to anyone contemplating further low frequency protection to the Heliums.

The original question by Dunk_c did ask about whether it was necessary and that is an interesting question in itself.

I would definitely agree that keeping extremely low frequencies out of the Helium's would result in cleaner midrange and likely better tonal balance particularly when playing loud.

One thing the modeling above does show is that even unfiltered, the Helium's playing at 98 dB, which is quite loud are still well below Dayton's published xlim of 25 mm (which is pretty insane for a 3.5" driver). This correlates pretty well to what I have observed at various events in larger venues where we have tried to challenge the Helium's deep and loud and it has met the challenge surprisingly well every time. Those original Helium's are still alive and kicking today. Being above xmax, but below xlim isn't ideal from a distortion perspective, but is likely survivable.

The ND-91 is a remarkable driver and the T/S parameters drive it to a tuning that does not promote trying to play too low in too large a box, so this contributes greatly to power handling. This might be one of those rare cases where you have to worry about the voice coil giving out before you have to worry about over excursion taking out the woofer.

[Dunk_c]

Thanks for that Guys. Scott, what are your thoughts on HMM integration with a sub bass speaker for fuller range sound reproduction, as opposed to speaker protection implications of a bass blocking filter?

[scottsehlin]

I think bass blocking has the potential to help the midrange/midbass sound cleaner. How much difference it will make depends on how loud they are playing. It will be hard to kill them, even running unfiltered, so you can try it out as-is and decide if you want to add a filter later.

I once had a setup with a pair of bookshelf speakers with an F3 around 50 Hz that I ran wide open, then brought a sub in below that just to add extension. I felt it added something without changing the tonality of the main speakers - so was worth doing. That was a small system in a small to medium sized room mostly for personal listening, so I really wasn't straining the woofers too much.

[Dunk_c]

Interesting topic!
They do sound excellent with or without a small sub (Rel T-Zero is what I am using, I posted an amateurish freq response in post 429, and the sub is only doing duty up to 40Hz).
So as long as the amp is not clipping when sending bass signals to the HMM that it cannot reproduce, all should be OK and bass blocking filters might not have a significant effect on sound quality, especially as I am only listening at <75dB at 4 meters.

[kineticmac]

Hey guys,

This is my first post on this forum and this is also the first sourround system i'm building!
I'm looking for a surround speaker build and this Helium speaker looks promising. It needs to be as small as possible so i can sell this to the wife.
I do have a few questions:

1. I've input the driver parameters into winISD and i can't seem to make it work. If i select a port of 3/4inch i get a length of 5.27inch and a port velocity of over 64!
The box is 0.04cubic feet en it is tuned to 71.48Hz.


































Isn't the port velocity way too high? What am i doing wrong?

2. The distance from listening position to speaker is around 3 feet. Will this speaker be loud enough to watch a movie (not music)?
The front speakers are 160W 91db distance10 feet.
3. I know the surrounds will never play anything like the front speakers but i'm wondering what a good spl level is for a surround speaker and i can't seem
to find it on google. What's a good level?

Any help is welcome!

[djg]

Seen these?

https://www.parts-express.com/passive-aggressive-mini-speaker-kit-pair--300-7148

[kennyrayandersen]

So, I'm trying to go through this myself, fer my own edumacation. I don't need to use the parts express box, and in fact prefer to just build my own because it's just easier to get what I want. In any case, I got to thinking that first I would check the box alignment. Parts express lists a volume fir the ND91-4 of .03 ft cubed (I believe they used to list it as .04 ft cubed), but i a really small box that's a lot of round-off; so, I broke out the Winisd and started looking at various box volumes and tuning frequencies.

First, it starts off by giving you a 3 liter volume and a nice curve, but 3 liters is .106 cu ft -- far larger than the one listed by Parts Express, which claims to use BasBox Pro 6. That volume is also larger than the original box used in this thread. OK, so I says to myself -- self, let see how small we can get the box. Long about 2.36 liters, it really starts losing F3. So then I tried using a passive radiator, in fact, I used 2 different passive radiator configurations. The first is with a single ND105-PR, and the second was with dual ND90-PRs. The box volume seems to work about right for either the passive, or the port, but the F3 of the ported is about 78 Hz and the F3 of the passives even higher (on fact the passives seem to under-perform compared to the port) (see attached screen grab). If I want to drop the F3 to 70 Hz or below the box has to be at least 3 liters, and the woofer appears to like a box that is somewhat bigger than that, but the gains in F3 don't seem to be justified over the 3 liter volume (marginal improvement in F3 with increasingly larger changes in box volume).

So, What gives? Is Winisd not good for small boxes? Not good or rather not particularly accurate compared to BassBox 6 Pro? BassBox Pro 6 is saying I can get an F3 of 73 Hz with a box volume of .03 cu ft (51 in cu or .84 liters) which is even SMALLER than the Parts Express project box. So, hep me!

I'll also attach what the response looks like for configurations with the 3 liter ported box, the other boxes were 2.5 liter for the ND105-PR and 2.8 liters for the dual ND90-PRs. That seemed to be the 'right' volume for each.

I'll try horn response next -- is there something else I should be modeling the box alignment with?

Thanks Guys!

[Chris Roemer]

You know, T/S parms DO tend to change somewhat over time (due to mfring. alterations). What values are you using for Qts, Fs, and Vas (since 2016 I've been using 0.41, 74, and 0.05cf)? The "initial" parms for this driver were different.
In a .1cf (2.8L) vented box, I think a tuning in the upper 60s (1"x4") looks high. I prefer a 5" tube (Fb in the LOW 60s). F3 (using MY parms) is in the mid 50s. ;-)

- "So, What gives? Is Winisd not good for small boxes? Not good or rather not particularly accurate compared to BassBox 6 Pro? BassBox Pro 6 is saying I can get an F3 of 73 Hz with a box volume of .03 cu ft (51 in cu or .84 liters) which is even SMALLER than the Parts Express project box. So, hep me!"

WinISD works FINE for small boxes.
I believe the main reason PE uses BassBox is that it's not free (like WinISD), AND, THEY are a seller - so they make money on it. I've never found it to be "better" at box sims than WinISD.
You DO know that a woofer in a (tuned) box can have an infinite number of rolloff curves - ANY of which can be "perfect" - depending on what the designer intends to do with it?

- "I'll also attach what the response looks like for configurations with the 3 liter ported box, the other boxes were 2.5 liter for the ND105-PR and 2.8 liters for the dual ND90-PRs. That seemed to be the 'right' volume for each.
I'll try horn response next -- is there something else I should be modeling the box alignment with?"

You can't ALWAYS match a ported box FR using a PR(s). Often you can get very close, but sometimes not.
Do you have something against a 4" long 1" port tube for this project (I'd feel better w/a 1.25"id, 1.5 would be even better)?
I'll do quite a bit to avoid (the expense of) using a PR.
I haven't run all your box sims, but it seems like you know what you're doing.

Another (free) option is a slot port.

Once you get up to 0.10 cf, the TCP115 drivers model well (at a significantly lower co$t), and (IMO) they look much better.