Statement II help needed, big dip in frequency response.

[RookieBuilder]

Looking for some help or opinions here. What could be causing the big valley in the midrange?

Somewhat recently I was playing with REW and my statement II's and I have a major dip in the response, centered at 1300 hz.


I disconnected the mids and took a measurement, blue is all drivers and purple is the just the woofers and tweeter.


Then measured just the mids. The mids are much lower in amplitude than the woofers and tweeter. Red is just the mids and purple is the woofers and tweeter.


Checked and rechecked the crossover, its wired per spec. Attached that just in case, picture doesnt clearly show the source and ground wire runs, but I didnt document the build and its the best I have. Also was some sloppy soldering...


Mucked about checking phase on individual drivers, that all seems to check out. Tried removing the foam from the tunnels, blocking the tunnels, and pulled the speaker out to about 4' into the room. No significant change in the measurement.
All measurements are taken in room, after the initial seating position measurement that highlighted the issue further measurements were on one speaker, at tweeter level at approximately 1 meter. Both speakers measured about the same doing this. The dip was not replicated with a different speaker I tried.

[Jim Holtz]

This is a Curt question. You'll have to send him an email. He's very busy and doesn't frequent the forms very much.

Sorry I can't be of any help.

Jim

[Juhis]

I would check that the mid's have right ohm rate. There are 4 and 8 ohm versions around.
Also use ohm-meter to check that your resistors have correct values.

[RookieBuilder]

This is a Curt question. You'll have to send him an email. He's very busy and doesn't frequent the forms very much.

Sorry I can't be of any help.

Jim

Thanks for the input Jim, I will likely contact Curt, however as per below it looks like I have more troubleshooting to do on my end.

I would check that the mid's have right ohm rate. There are 4 and 8 ohm versions around.
Also use ohm-meter to check that your resistors have correct values.

Ordered the 8 ohm variant, just checked one driver 6.7 ohm. I suspect this checks out. I did check each resister when I was assembling the crossover, they should be in spec.

Did a full range sweep with one mid connected directly to the amp, and got this, red being a single mid and blue a full range sweep of the statement.


Looking at that graph, I believe it highlights a crossover issue... Now I'm wishing I took detailed photos when I was building these, best I can see in the picture I have everything is hooked up correctly.

[roadrune]

Tweeter wired out of phase maybe?

[5th element]

The null is too high in frequency for this to be the tweeter out of phase. I mean it's still a possibility considering there's something wrong, so anything could be, but there's definitely something up with the mids.

Which version of the xover are you using? Because there's a simpler updated version on the website.

http://speakerdesignworks.com/Statement_II_4.html

There's definitely something wrong as the mids are playing 15-20dB quieter than they should, neither are they rolling off to the tweeter correctly either.

[RookieBuilder]

I am using the latest version that you linked. Since both speakers have the same issue I will need to pull a crossover and take a detailed look over. Hopefully I can spot an issue otherwise I will post photos, hopefully some kind person can spot an issue at that point.

I will also reverse the tweeters, they wired up correctly per markings, but if fountek has an error...

[RookieBuilder]

A sketch of what I built, with values, all of the positive lines are tied to the red terminal and the negative lines tied to the black terminal. Personally I am not seeing my mistake, any input on the drawing below. Resister values on the tweeter are written wrong they are 2 x 12 ohm for 6 ohm total, and 2 x 16 for 8 ohm total.

[Zvu]

.....

[5th element]

Why the...?

I just looked over your midrange xover from the photo above and it does seem appropriately connected. I'm assuming you've checked that the resistors are the correct values and that you haven't ordered, or been sent, something that's out by a factor of 10? I've had that happen to me, been sent 0.47ohm resistors when I ordered 4.7ohm or something.

Edit, I see you checked that on assembly. Can you measure the speakers impedance?

[csmielke]

Check your midrange crossover as it does not match the schematic from Curt's speaker design works website for the Statement II. You show two resistors in series which does not match.
Chris

[cochinada]

Check your midrange crossover as it does not match the schematic from Curt's speaker design works website for the Statement II. You show two resistors in series which does not match.
Chris

2.7+1.8=4.5

Are you sure both mids are properly soldered in parallel?

[Juhis]

I tried to figure out and it seems that the it's built correctly. (MID part of crossover at least)
I would take ohm measurements from the resistors that I painted in picture below.

Also confirm the source and negative leads.
Cant say from that picture where those really end

[RookieBuilder]

Thanks everyone, to answer the questions.

Speaker measure, for the whole system is 4.0 ohm and 6.7 for a single mid range.

100% sure the mids are connected in parallel.

Juhis, I think your right for the next steps, I will verify the resistance and source leads into the mid range crossover. Stupid question time, can the resisters be properly measured while wired up, or will i need to cut the leads and measure individually?

[ergo]

Some of the resistors can be measured in circuit and some not.

You have to analyze it by keeping in mind
a) if there is a capacitor in loop it will block DC voltage flow, so in that case you can measure the resistors as there is no DC loop forming parallel circuits (I marked samples of such resistors with ok)
b) if there is a coil, or a speaker or another resistor forming a parallel loop, you can not measure the resistor in circuit as the value you measure would show the result with that parallel path included (I marked those two resistors with no ok as there is a loop through the 0.39u coil that will form a parallel DC path)

Hope this helps.... all in all if you disconnect the drivers and maybe release one or two soldering joints you should be able to get all of them tested following the above guidelines.

[Juhis]

Thanks everyone, to answer the questions.

Speaker measure, for the whole system is 4.0 ohm and 6.7 for a single mid range.

100% sure the mids are connected in parallel.

Juhis, I think your right for the next steps, I will verify the resistance and source leads into the mid range crossover. Stupid question time, can the resisters be properly measured while wired up, or will i need to cut the leads and measure individually?

You can measure the resistance without cutting leads.
Both wires going to resistors should be disconnected while measuring and also the Mid-speakers and amplifier should be disconnected from crossover.
No re-soldering needed which is quite nice in this case.

X = should be disconnected during measurement
Dots = Measurements points

[RookieBuilder]

Measured the resisters and found the problem. the 2.7 ohm resister was burnt in both speakers, measuring 27.8 and 39.9 ohm. Jumper'd them out until I can get new ones. Everything else in the mid network measured close to what it should. I will check the rest when i remove them to install the new resister.

My question right now is if these were from a faulty batch, mis labeled from the factory (say they were actually 27 ohm would I burn them up?), or could I have other issues at play??? I checked through my photos and they were at least labelled 2.7 ohm, all markings are burnt off now. I believe I measured everything when new, but that was more than 2 years ago so I may not be remembering.

New measurement missing the one resister, much, much better. Ill post measurements again once I have a new unit installed.

[Jim Holtz]

Excellent news! :T

Jim

[ergo]

If markings have burned and the resistors are brown etc it would indicate they have gotten very hot at some point. One thing you might want to do is to up the wattage by 2x or so (the BOM of Statements does not seem to say what powwr spec are the current ones)

[cochinada]

Or you can always use more resistors combined to spread the wattage among them if that was the case.
BTW, I have had bad experiences with hot glue, namely with resistors getting too hot and since then changed to epoxy.

[Jim Holtz]

I'd suggest you use Zisters from Meniscus. They're rated at 25 watts, are very high quality and of course they sound good. I've also doubled standard 10 watt resistor values to end up with the correct resistance and high wattage capability. The mid circuits are the hardest on wattage requirements IF the speaker is driven hard. Normal use only requires the standard 10 - 12 watt resistors.

HTH

Jim

[RookieBuilder]

I'd suggest you use Zisters from Meniscus. They're rated at 25 watts, are very high quality and of course they sound good. I've also doubled standard 10 watt resistor values to end up with the correct resistance and high wattage capability. The mid circuits are the hardest on wattage requirements IF the speaker is driven hard. Normal use only requires the standard 10 - 12 watt resistors.

HTH

Jim

That solves that, I am using a 10 watt resisters and on occasion like it loud. Would you recommend a single 4.5 ohm Zister, or pair up a 1.8/2.7 as per Curt's website? I will be replacing the whole mid network, as I assume the problem will move on to the next weakest link.

Any thoughts on the AchrOhmiC MIL spec resisters at Solen? https://solen.ca/product-category/re...pec-resistors/ They are 16 watt, I assume best to double them up if t hats the route I go.

[Jim Holtz]

That solves that, I am using a 10 watt resisters and on occasion like it loud. Would you recommend a single 4.5 ohm Zister, or pair up a 1.8/2.7 as per Curt's website? I will be replacing the whole mid network, as I assume the problem will move on to the next weakest link.

Any thoughts on the AchrOhmiC MIL spec resisters at Solen? https://solen.ca/product-category/re...pec-resistors/ They are 16 watt, I assume best to double them up if t hats the route I go.

I would use Zisters wherever possible. They are a simple and affordable elegant solution. Mills are also top notch but expensive when paralleled or in series and I've not found them to be superior in any regard other than construction and measure spot on. I wouldn't be afraid to use Lynk resistors paralleled or series if Zisters aren't available. Honestly, the mids are the place I'd suggest you focus on for higher wattage resistors. They take the abuse. The ribbon will be fine with top quality standard value resistors. However, I usually parallel values there as well just because my OCD kicks in.

I can't comment on the ArchrOhmiC resistors since I've not used or tested them personally. I have read they are very good.

BTW, I'd strongly suggest you measure all of the crossover components if possible. Caps and resistors are always measured during my builds.

HTH

Jim

[RookieBuilder]

I would use Zisters wherever possible. They are a simple and affordable elegant solution. Mills are also top notch but expensive when paralleled or in series and I've not found them to be superior in any regard other than construction and measure spot on. I wouldn't be afraid to use Lynk resistors paralleled or series if Zisters aren't available. Honestly, the mids are the place I'd suggest you focus on for higher wattage resistors. They take the abuse. The ribbon will be fine with top quality standard value resistors. However, I usually parallel values there as well just because my OCD kicks in.

I can't comment on the ArchrOhmiC resistors since I've not used or tested them personally. I have read they are very good.

BTW, I'd strongly suggest you measure all of the crossover components if possible. Caps and resistors are always measured during my builds.

HTH

Jim

Perfect, thanks, zister is in stock. While I'm in there i was thinking about upgrading the 3.9uf and possibly the 100uf cap. Any suggestions here?

Meniscus has clarity CSA or mundorf supreme on the tweeter and clarity PX on the mids. Would any of these or others be worth my while?

[Jim Holtz]

Perfect, thanks, zister is in stock. While I'm in there i was thinking about upgrading the 3.9uf and possibly the 100uf cap. Any suggestions here?

Meniscus has clarity CSA or mundorf supreme on the tweeter and clarity PX on the mids. Would any of these or others be worth my while?

Curt and I are big fans of Clarity caps. PX on the mids and CSA on the ribbon would be perfect.

Excellent choice!

Jim

[sdl2112]

Just a word of caution. Although the Zister resistors show 25W, they don’t show the assumptions. Resistors are all about surface area and temperature. Aluminum resistors of this type get their rating from a suitable heatsink. A comparable Ohmite or TE resistor of similar size may have a rating of 50% or less without the heatsink.

e.g.

https://www.ohmite.com/assets/docs/acl_hs10.pdf?r=false

https://www.te.com/commerce/DocumentDelivery/DDEController?Action=srchrtrv&DocNm=1773035&DocType=DS&DocLang=English

[Curt C]

This seems like a good opportunity to discuss the finite thermal properties of resistors and how to optimize their power dissipation abilities in the confines of a speaker enclosure.

One important consideration using the Vishay Dale, Alcor, Zister or other aluminum housed, chassis mount resistors:
In order for them to meet their published power specifications, they must be mounted on a suitable heat sink. For example, the Vishay Dale data sheet indicates the 25 watt resistors must be mounted on an aluminum heat sink in order to meet their 25 watt specification. For speaker crossover purposes, heat sinks are generally not practical, so we must derate the power capability appropriately. For a 25 watt resistor, typical derating will be 50% for no heat sink. A 25 watt would derate to 12.5 watts, while a 50 watt would derate to just 20 watts with no heat sink. – In spite of this they still ‘look’ cool, so it may be a suitable application anyway. For those looking more for value and function over form, the prosaic 25 watt sandcast wirewounds are inexpensive, plentiful, and require no heat sink to meet their power dissipation specifications.

Which brings us to the next concern: The power rating assumes ‘free air’. While solidly affixing or hot gluing the resistors directly to the crossover board certainly solves some vibratory issues, it also limits the power handling capabilities of the resistor. It is better to mount the resistors so they are ‘flying’. That is spaced above the board by their leads for better airflow around them. An inch or so is good, but just about any spacing will be better than none. Try to avoid covering the crossover boards with foam or stuffing. The resistors cool by convection. They heat the surrounding air, which rises, and is replaced by cooler air from below. If there is no airflow, their cooling can be severely limited. Additional cooling can be accomplished in some cases by placement of the crossover board near the port.

While it is obvious, I would be remiss if I didn’t point it out here. If the resistors are dissipating this much power, they should not be placed near or on combustible materials. This means you should not use that handy piece of cardboard to build your crossover on. Likewise, paper backed fiberglass bats are not an acceptable surface to rest the resistors against. Foam is combustible, as well as the enclosure itself. Now, before you all get worried and start second guessing yourselves, I’ve schlepped some pretty poorly done crossovers in my day, and never had one catch on fire. However, I’m not in the habit of playing my speakers at rock concert levels either. A modicum of common sense here will keep you and your speakers safe.

Generally large power dissipation in the resistors is confined to the midrange crossover resistors. In most home tweeter circuits there is far less acoustic energy required, and consequently far less power dissipated. Due to their low DCR / ECR values, typical crossover inductors and capacitors do not suffer heat issues, so from a practical standpoint, it is safe to ‘bury’ most woofer crossovers under the foam/stuffing, and with some reservations, the tweeter crossover as well. Putting the midrange crossover on its own separate board makes it easier to find an acceptable place for it.

Considering the Statement II circuit in this thread:

Presuming there was not a premature failure of the resistors, we can make the following assumptions: Since the 2.7 ohm resistors opened up, but the 1.8 ohm resistors did not, we presume that the 1.8 ohm resistor dissipated no more than 10 watts. A little off the cuff calculation indicates an average of 2.36 amps was flowing through the resistors. By extension, the 2.7 ohm resistor was dissipating 15 watts, for a total of 25 watts. - More than enough to keep your coffee warm. Common practice is to rate components at twice the parameters they are expected to experience, or 50 watts in the case of a single 4.5 ohm resistor.

One possible solution would be to parallel a 10 ohm 25W resistor with an 8.2 ohm 25 watt. This results in a combined resistance of 4.5 ohms. At the average 2.36 amp current the power dissipated through the 10 ohm resistor would be approximately 11.24 watts and 13.67 watts through the 8.2 ohm resistor. Both would be comfortably dissipating near half of their rated power, but would still need unimpeded air around them in order to dissipate the heat generated. The series solution is easier to conceptualize: Three 1.5 ohm resistors in series will each dissipate 1/3 of the 25 watts in our example, or 8.33 watts each. Note that in either case, the power dissipation is shared equally only when the resistances are equal. A circuit with a 4 ohm and a 0.5 in series would provide the same response, but the 4 ohm resistor would dissipate 89% of the power. –Not a very equitable balance of power, but acceptable as long as the 4 ohm resistor was sized appropriately. - in this case 50 watts.

C

[RookieBuilder]

I promised an update with measurement of the repaired crossover. I had order the Zisters before Curt posted and had less understanding about these things than I thought, I ordered single units to spec instead of multiples to create spec and share the load. So instead of reordering, and waiting, and paying import fees, I installed them on a heatsink. Still not the overall ideal but better than it was and also good reason to keep the volume just a little bit more sane.

Fist, file the zisters base flat, they are terrible out of the box


Then onto the heatsink with thermal compound. It would actually be fairly easy to sandwich the zisters with a pair of heatsinks, but at that point I think it would be wiser to have multiple resisters.


Crossover with Clarity Caps on the mids and tweeter


And the in room response


I still need to do the second speaker, but I will be getting new terminal strips first. The ones I have are a major pain with the cable I used to wire the drivers, I think I spent almost as much time fighting the terminal strips as I did with the rest of the repair...