HiVi DIY 3.1 Revisited

ScottS

It's been a pretty cold winter, which has delayed some planned projects.  This situation has presented an opportunity to dig further into the HiVi DIY 3.1's that I picked up shortly before DIY Iowa in 2017 and brought along for a quick demo.  My pair painstakingly finished in primer for the show,is shown here.This kit is still available on Amazon for $249/pair including knockdown cabinets with free shipping.  Some of the reviews indicate that the overall sound is a bit bright, but can be corrected by increasing mid and tweeter resistor values.  Rich (Turn 2) had done this to a pair that he also brought to the Iowa show, but a wiring error led to some strain on the mids - so we stopped the demo early.  Rich indicated that he had modified the resistor values.

I wanted to start by measuring the stock FR.  This actually looks almost identical to what is posted on HiVi's website (but using a more revealing scale).  I didn't splice in any low frequency nearfield response, so anything below about 200-300 Hz is not accurate.  I guess one could call the response 89 +/- 4 dB without really lying - but most of the 8 dB variation is a steady rise from 85 dB in the 200-300 Hz area to 93 dB around 12 kHz.  I would expect this to sound a little bright and, to my ears, it does.

The next step was to unhook the individual drivers from the crossover and see what we are working with.  The woofer is the L6-4R (4 ohm), which is the same model number sold by Madisound and Meniscus.  The in-box FR is remarkably well behaved with a very smooth rolloff starting between 2 and 3 kHz.


The in box impedance plot shows that the cabinet is tuned to 50 Hz.  There is little else remarkable to note.
The midrange is the DMN-B.  I used the DMN-A in my recent Indium 7 project and really liked it.  Physically the only difference I can readily observe is a 4 screw flange vs. 6 screws for the DMN-A.  6 screws would have limited the overlap between the mid and tweeter flange, so maybe that is the reason.  The frequency response is very similar to what I measured for the DMN-A with minor differences likely due to cabinet geometries.Impedance is also similar to the DMN-A, including the published Fs of 850 Hz.The tweeter is the RT-1.3B.  I have heard the 1.3 before, but have never worked with it.  The faceplate is different from the standard RT1.3, presumably to help with the physical overlap between the mid and tweeter flanges.  The frequency response is pretty well behaved, which isn't a given for a planar tweeter.  There is a slight peak at around 12kHz, which also showed up in the overall speaker response.Impedance is pretty much a flat line at around 6 ohms, as one would expect from this type of tweeter. The next step will be to use the measurements to simulate what it would take to balance out the response. 

Turn2

I got mine to a point where there's no embarrassing on-stage honk from the mid.   The sound is tamed enough to my ear to close down the project. Only the woofer coil and the series caps to the mid and tweeter aren't stock. It will be interesting to see what you conclude using more sophisticated tools.

ScottS

The step of applying my measurements to a sim of the stock crossover is complete.  This is a sanity check so that I can start changing things with some confidence that the sim resembles reality.  Looks good so far except that i had to reverse the polarity of the woofer - I must have reversed it when measuring.

The first thing to try should be either to adjust the resistor values or try just adding resistors in parallel with the mid and tweeter to pad them down.  I will try leaving the stock crossover alone and just adding a resistor across the driver terminals for the mid and tweeter - as this would be a really easy fix.  A reasonable first attempt is a 10 ohm resistor across the DMN-B and a 5.6 ohm resistor across the RT-1.3B, which results in a simulated response that would fall into a +/- 2 dB window around an average of about 86.5 dB at 2.83V, 1m.  The primary trouble spots look to be a little extra emphasis  between 500 Hz and 1.5 kHz and the 12 kHz peak.  Time for a listen.

ScottS

The parallel resistor fix actually sounds pretty good.  Tried with some better quality material and cheesy 80's music.  The midrange presentation is slightly forward, but it is over a broad range, so it isn't really peaky or irritating.  The 12 kHz peak is down enough to line up with the midrange.  The recessed area in between correlates with a BBC dip and is also pretty broad and shallow, so it doesn't really feel like a lot is missing.  Bass/midbass is present at the right level, so BSC appears set about right.  For a cheap, easy fix, this looks like a win.  My initial comparisons were done against the Matrix Revolutions.  I will want to listen to these head-to-head vs. the Indium 7 to see if there are any obvious tonal balance differences.  I recall the Indium 7 having at least a similar midrange presentation (which might not be all that surprising as they use essentially the same mid).

If one is starting out assembling these, it might be preferable to change the values of the existing series resistors rather than add parallel resistors.  In that case, it looks like a similar amount of padding can be acheived using a 6 ohm resistor on the tweeter and a 3 ohm resistor on the mid.Unfortunately, is approach gives a slightly less linear response than the parallel resistor approach.  I suspect this is why Rich decided to increase the woofer inductor as an additional modification.  If I had to guess, i would suspect he used a 2 mH inductor for L4.  Here is how that looks:
With this crossover topology, there is another tool that can be used to help with the midrange peak - R3.  

Changing R3 from 1 ohm to 2 ohms helps bring the midrange peak down.  The value can be increased further.  By the time the value gets to 3 ohms, the midrange peak is lower than the 12 kHz peak - which experience tells me won't be desirable.  Further padding the 12 kHz peak at that point will shift the overall balance slightly more bass-heavy.  I would probably settle on something in the range 2 to 2.5 ohms and call it good.

The real limitation of the stock crossover approach winds up being the shape of the high frequency response.  If that could be flattened out, then one could make the tweaks needed to really flatten the rest of the response without the 12 kHz area sticking out as an overall peak and sounding "icy" for lack of a better term.

Here is a modified crossover that would allow reuse of the 3 largest inductors, 2 largest caps, and the PC board itself.  The woofer filter is the same as above.  The 6.8 uF cap in the midrange filter is changed to 5.6 uF, which addresses the area between 1 and 2 kHz.  The tweeter crossover has all new components, but they are small ones...

The result is a much smoother overall response.  I've still got some comparing to do, but it appears there are several viable options

1.  Keep the stock crossover as is.  Add a 5.6 ohm resistor in parallel with the tweeter and a 10 ohm resistor in parallel with the mid.  

2.  Replace all the resistors.  R1=6 ohms, R2=3 ohms, R3=2 to 2.5 ohms.

3.  Replace mid and tweeter resistors and the woofer inductor.  R1 = 6 ohms, R2 = 2 Ohms, L4 = 2 mH.

4.  The perfectionist option.  Replace R2 and R3 (both 2 ohms), replace C3 with 5.6 uF.  Redo the tweeter circuit (R1=5.6 ohms, C1=1.5 uF, C2=2 uF, L1=1.5 mH). 

I will report back with impressions as I work through the options. #1 is quite respectable.

MartyH

Great reporting, Scott!  Being a fan of the infamous BBC dip, I'd be curious to see what would happen if you brought C3 down to 5.1uf and R2 to 3.5 ohms.  Good fix on the tweeter too!  BTW how do the reverse nulls look?  Naturally I defer to your being there and your impeccably designed work that I've heard in the past.  You have a damn fine ear!

ScottS

Thanks for the kind words Marty.  Decreasing C3 further starts to push the response up above 3 kHz - probably not the direction you would want to take.  For the perfectionist mod, R2 starts at 2 ohms.  Increasing that brings in a nice little BBC dip.  Season to taste...  I think 2-3 ohms looks like the right range to be in. .  

Turn2

Very good work. For the record I changed:

L4 from 1.5mH (DCR 0.4 ohms) to 1.5mH (DCR 0.22 ohms)
R2 from 1 ohm to 2 ohms
R1 from 1.5 ohms to 3.0 ohms

Judging from your work mine should still sound a little hot. I know that I've made some improvement and I do have 60-year-old ears,

Tom_S

Scott - I'll admit I really liked what I heard even from the stock crossover at Grinnell. No major warts and the treble actually seemed quite smooth. I was prepared for a screeching upper mid after seeing some preliminary reports on the forums. Now I wasn't anywhere close to being on-axis - I was walking up the left aisle during the demo. But I was still kicking myself for not jumping on these when the price was so low. I hope I never hear your revised crossover #4. The last thing I need is another speaker in this house...and that's about the only thing keeping me from hitting the Add to Cart button right now. 

ScottS

I've been running option 4 head to head against the Indium 7's lately.  Tonal balance is very similar.  The Indium 7 bass digs deeper due to the shallow sealed roll off.  I think there might be some subtle differences in the lower and upper midrange, but they are similar enough that it is not always easy to tell which is playing.  

Maybe the modified DIY 3.1 will become the Indium 5.5 (the Infinity Kappa 6 had a 10" woofer and the Kappa 5 was a two way with the EMIT tweeter)...

That would be a good outcome for me as the modified 3.1's could serve as surrounds for the Indium 7 mains in the man cave.

ScottS

ScottS said:

The parallel resistor fix actually sounds pretty good.  Tried with some better quality material and cheesy 80's music.  The midrange presentation is slightly forward, but it is over a broad range, so it isn't really peaky or irritating.  The 12 kHz peak is down enough to line up with the midrange.  The recessed area in between correlates with a BBC dip and is also pretty broad and shallow, so it doesn't really feel like a lot is missing.  Bass/midbass is present at the right level, so BSC appears set about right.  For a cheap, easy fix, this looks like a win.  My initial comparisons were done against the Matrix Revolutions.  I will want to listen to these head-to-head vs. the Indium 7 to see if there are any obvious tonal balance differences.  I recall the Indium 7 having at least a similar midrange presentation (which might not be all that surprising as they use essentially the same mid).

If one is starting out assembling these, it might be preferable to change the values of the existing series resistors rather than add parallel resistors.  In that case, it looks like a similar amount of padding can be acheived using a 6 ohm resistor on the tweeter and a 3 ohm resistor on the mid.Unfortunately, is approach gives a slightly less linear response than the parallel resistor approach.  I suspect this is why Rich decided to increase the woofer inductor as an additional modification.  If I had to guess, i would suspect he used a 2 mH inductor for L4.  Here is how that looks:
With this crossover topology, there is another tool that can be used to help with the midrange peak - R3.  

Changing R3 from 1 ohm to 2 ohms helps bring the midrange peak down.  The value can be increased further.  By the time the value gets to 3 ohms, the midrange peak is lower than the 12 kHz peak - which experience tells me won't be desirable.  Further padding the 12 kHz peak at that point will shift the overall balance slightly more bass-heavy.  I would probably settle on something in the range 2 to 2.5 ohms and call it good.

The real limitation of the stock crossover approach winds up being the shape of the high frequency response.  If that could be flattened out, then one could make the tweaks needed to really flatten the rest of the response without the 12 kHz area sticking out as an overall peak and sounding "icy" for lack of a better term.

Here is a modified crossover that would allow reuse of the 3 largest inductors, 2 largest caps, and the PC board itself.  The woofer filter is the same as above.  The 6.8 uF cap in the midrange filter is changed to 5.6 uF, which addresses the area between 1 and 2 kHz.  The tweeter crossover has all new components, but they are small ones...

The result is a much smoother overall response.  I've still got some comparing to do, but it appears there are several viable options

1.  Keep the stock crossover as is.  Add a 5.6 ohm resistor in parallel with the tweeter and a 10 ohm resistor in parallel with the mid.  

2.  Replace all the resistors.  R1=6 ohms, R2=3 ohms, R3=2 to 2.5 ohms.

3.  Replace mid and tweeter resistors and the woofer inductor.  R1 = 6 ohms, R2 = 2 Ohms, L4 = 2 mH.

4.  The perfectionist option.  Replace R2 and R3 (both 2 ohms), replace C3 with 5.6 uF.  Redo the tweeter circuit (R1=5.6 ohms, C1=1.5 uF, C2=2 uF, L1=1.5 mH). 

I will report back with impressions as I work through the options. #1 is quite respectable.

It was brought to my attention that I have a typo in my description of the "perfectionist option" near the end of the quoted post.  L1 should be 0.15 uF, as shown in the schematic.

PWRRYD

PhD...... pfffff!


j/k Scott  

ScottS

It was only off by an order of magnitude...

Davemo

I'm building a set of the 3.1s and I want to build the perfectionist option crossovers - but I don't know what quality of components to buy - any recommendations? 

jr@mac

The Dayton Audio coils and Audyn red caps will be more than adequate for this build, Dave. 

Davemo

Thanks for the advice jr... I've built the perfectionist crossover - but substituted a 2.2uF for c2, pe doesn't sell a 2.0 uF audyn red cap. How will this affect the performance of the system? thanks for the help...

ScottS

2.2 uF is fine for c2.  I plugged it into my sim and the impact is very minor, probably nothing audible.

Davemo

Thanks for checking that Scott! Ok, rookie question, everything else being equal, is it ok to mix caps of different "grades" in a crossover? I could have used an audyn silver cap in 2.0 uF for C2, as an example. And, I do indeed realize that 2.2 uF and 2.0 uF are not equal, I just mean in general... 

MartyH

Davemo said:

Thanks for checking that Scott! Ok, rookie question, everything else being equal, is it ok to mix caps of different "grades" in a crossover? I could have used an audyn silver cap in 2.0 uF for C2, as an example. And, I do indeed realize that 2.2 uF and 2.0 uF are not equal, I just mean in general... 

2.2 uF is 2.2 uF. In other words treat each cap as labeled.  However mixing brands of caps is perfectly allowable.  Some prefer the sound of one brand over others and there's nothing wrong with using your ear and budget as a guide.  In fact you'll find that many designers use NPE caps as the 'tank' cap on woofers as they can climb quite high in value, and a 45uF NPE is one-tenth the cost of the least expensive 45uF polypropylene cap. 

Davemo

i've got the 3.1s built and plugged in. They started out a bit icy, but seem to be warming up a bit as they break in. I haven't done any in room measurements yet, just trying to listen and enjoy. So far I'm impressed, for $250/pair they sound pretty darn good!
I've run through kind of blue, some gogo penguin, andrew bird's weather systems, they all sound terrific. The Blade Runner 2049 soundtrack is obnoxious, but is a real showcase for highs and lows, it's dark and dramatic and beautiful, and it sounds great on these speakers. 
They seem to blend better with the Dayton HF 15" sub than the smaller Echos, but I feel like the soundstage is less broad? Regardless, thanks Marty and Scott for the sage advice! Without the crossover tweaks I'm afraid I would have been disappointed. 

Davemo

So, this kit started on Amazon at $174, then settled at $249 since Nov 2017. A popular reviewer on YouTube did 4 videos building the kit and a final review... the price for the kit went up to $289 and promptly sold out. This is why we can't have nice things...

stevenswall

Scott S, your option 4 "perfectionist crossover" is a dream!

I've since built a shopping list from Parts Express for those who aren't familiar with the proper voltage and wattage ratings for the various components, and you can see a finished setup here:



https://imgur.com/JJasE8K

Part numbers and prices in the Reddit thread here:

https://www.reddit.com/r/diyaudio/comments/96brng/help_swan_hivi_31_crossover_circuit_part/

If anyone seeing this doesn't have an account, feel free to ask for help on Reddit.

Davemo: Thank you for posting the original price. I'd buy a few at that price! I'm curious what a front bevel around the tweeter and midrange would do to help the baffle diffraction, and I've also entertained the idea of sticking felt to the front.

jr@mac

Can you paste a link to this thread on that Reddit discussion? It will help Scott get the credit he deserves for this mod.

Nicholas_23

+1

ScottS

stephenswall,

I'm glad the mods worked out for you.  The shopping list and image of the completed crossover should be helpful for other builders.  Thank you for sharing those!

stevenswall

jr@mac said:

Can you paste a link to this thread on that Reddit discussion? It will help Scott get the credit he deserves for this mod.

It is. Also linked it in the description on the Imgur album in case anyone stumbles across it there. Hopefully the SEO machines kick in and this becomes more popular!

Right now I'm trying to make friends at the BYU Acoustic Engineering group... Anechoic chamber measurements would be useful for reference and further mods.

ed_diymaker

Hi! I just built the perfectionist option. 

Thanks Scott for this amazing design!

Thanks Steven for the parts list,
crossover photo etc. These speakers sound amazing! 

Trebbles are faint and could be a bit brighter, just my preference. I only heard bells etc so far from the trebble drivers and they were quite subdued. But I really wouldn’t sacrifice a flat frequency reaponse for higher/too bright trebbles. 

Some Questions:

Q1. I used wood glue for mdf boards, which seems to them together hold fine but would you recommend screws etc to firm up the box? I have this nagging feeling it may fall apart, probably just me worrying.

Q2. Can anyone suggest a good amp for these. 5.2 or 7.2 should be enough for my needs, I am really planning a 2.1 or 2.2 configuration just for music mainly.  My tsr 5830 from Yamaha seemed to be overheating a bit, and it’s rated 6-8 ohms impedence.

Q3. What is the frequency response of the 6.5inch woofers below 100hz? I am trying to determine if subwoofers would add much for music listening (I am not too interested in movie audio effects like thunder etc.)? 

If it would truly help for music, I am thinking of getting a pair of sb2000 subs, which seem to have even frequency response up-to 150hz. Any cheaper (or diy) options that would be nice for music? 


Some contributions back to the community, my two cents for future ameteurs
like me —

- order of build: the kit comes with screw-plugs and screws on which you mount the crossover board to the mdf cabinet floor. I suggest hammering the screw-plugs  to the floor board before gluing the mdf cabinet together. My floor board (one) came off and I had to re-glue it. 

- Hacks probably not recommended, let me know if should redo these steps: I hand-joined the wires from the crossover to the input port and drivers, instead of welding. I also used scotch tape to insulate open wires I hand spliced together. I can easily replace with electric tape if needed. 

- Cut the wires that go from the crossover to inputs and drivers about 4 inches longer than you think you will need. This makes it possible/much easier to connect terminals. 

- I plan to finish with a wood polish. I’ll share photos afterwards.

- I lined both sides (as recommended in instructions) and parts of the back with cotton insulation. Sound is expansive without seeming muffled, but I am no hifi expert on how it should sound.
- my soldering is not the best, but certainly seems to have gotten the job done 🙂

Bryan@MAC

From the Iowa DIY Audio show in Ankeny this past Oct 18-19, here is the man, the legend....the actually very modest guy who is a master designer, but will never admit it......  happy to say he's my bud.....Mr Scott S - 

https://youtu.be/xBmB6gYZ7Uc

Scoobis

Is the included foam ok as dampening material or should something else be used?  

ScottS

I voiced using the included foam.  There are better damping materials to be sure, but one does run the risk of changing the port tuning or tonal balance.

Scoobis

ScottS said:

I voiced using the included foam.  There are better damping materials to be sure, but one does run the risk of changing the port tuning or tonal balance.

Ah ... Thanks Scott. 
I will use the foam as well then.

Frank

Hi guys,
Well I finally finished my HiVi 3.1 perfectionist series, well not exactly, they're still in white primer LOL But they sound much better than I had expected. I used the Foam and I don't think I'm going to mess with them except to finish them off. I mean why get all crazy second guessing everything when its all there out of the box ya know, but do opt for the crossover upgrade, you'll be glad that you did. I've only been running them for about 3 hours but they make me want to listen to music now more so than before. I have a pair of Dayton 10" Sub's that I hooked up to see how that combo would work, and man I have to say that the HiVi's while having a nice low freq punch, the Sub's just filled the gaps with all of that bottom end and I'm really happy with these speakers even more. 

I learned a lot for this discussion, and of course I have to thank Scott for his effort's and sharing his findings with us all. Now its time to move on to another project  

Enjoy the music fella's