Playing with the Fostex more they are excellent sounding and do some things very well however they will not play as loudly or as dynamic as the ScanSpeak. I'll live with them for now and mean while continue to play with the ScanSpeak as free time permits. Maybe I can still get to the bottom of it.
@System7 I'm positive I have not damaged the voice coils. Tube amp doesn't have high damping factor anyway nor any DC getting past the transformer.
It's most likely as others have pointed out that I have a peak somewhere which I've not been able to detect due to still learning how to use this mic & software properly.
I feel like blaming the drivers but agree it's not likely.
@System7 I'm positive I have not damaged the voice coils. Tube amp doesn't have high damping factor anyway nor any DC getting past the transformer.
It's most likely as others have pointed out that I have a peak somewhere which I've not been able to detect due to still learning how to use this mic & software properly.
I feel like blaming the drivers but agree it's not likely.
I'm not suggesting you have damaged the voicecoils at all. But catastrophic breakup is something you encounter with difficult loads and highly tuned filters IMO.
What might help is wiring the SS mids in series. It can produce an easier load to the amp and avoids some resonance issues with parallel wiring. Damping factor is usually a bass concept, but applies right across the range in reality.
You also ask yourself what impedance is the speaker seeing looking back towards the amp and filter and second driver in an MTM arrangement. For sure there are some loads that some amps just get into trouble with, especially when you wick up the volume.
What might help is wiring the SS mids in series. It can produce an easier load to the amp and avoids some resonance issues with parallel wiring. Damping factor is usually a bass concept, but applies right across the range in reality.
You also ask yourself what impedance is the speaker seeing looking back towards the amp and filter and second driver in an MTM arrangement. For sure there are some loads that some amps just get into trouble with, especially when you wick up the volume.
Sorry I forgot to mention that yes my last experiment was to wire them in series but they sounded exactly the same at 16 ohms. No difference at all. I also tried putting in a series cap for 200hz passive cut off with no difference.
I also used not highly tuned crossovers using low Q 6db to med Q 24 db and everything in between with no difference. Varied upper frequency from 1800hz to 3600hz no change. Varied low filter from 180 to 360 no change. (I'd thought I might be able to 'slip' under it with the 1800hz but the anomaly has to be under 1800 somewhere)
Okay so I'm on my second box already because I needed to solve baffle step problem with my first smaller box. Baffle step problem aside I had exactly this same issue with no change in the other boxes which are probably 1/3rd internal volume of the boxes I'm using now.
My plan is to put them back into the smaller boxes (baffle step won't matter for testing this) and get back to measurements others have requested me to obtain.
Mean while just enjoy the Fostex until I can solve this. They do have surprisingly good midrange but I think would prefer a crossover a little higher than I wish to run 12 inch woofers.
Yeah I owe it to those assisting me. I'll get back to it when I have time. I work 12 hour shifts and my wife is starting to get pissed at me for getting nothing else done.
Hey I wanted to update something just in case anyone else might be curious.
I spent most of yesterday measuring and tweeking the Fostex FE108EZ drivers I'm trying for midrange units. I wanted to get them setup right for proper evaluation and comparison to the ScanSpeaks.
They do a lot of things right and sound damn good actually. However they are missing the dynamics that the ScanSpeak mids have. Even crossed over 24db at 360hz I can't play them as loud.
I probably should have figured that out with the low Xmax. Anyone who doesn't like to crank up the volume now and then would probably be happy with the Fostex. That's not me so I'll be back to working on the ScanSpeaks when time permits.
@System7 I searched on and did some reading about "rocking mode" phenomena you mentioned. It's interesting for sure. I would "think" when I put the 200uF caps in series at four ohms and also tried the units at 16 ohms that would probably rule that out. I also mentioned I tried with my transistor amp and had the same issue. I wonder though if the underhung motor might be more susceptible to this phenomena and perhaps I should persue further.
Anyway thanks to everyone. I'll get around to better measurements. Wife is pissed at me right now because I haven't been getting my other stuff done around the house or taking her out enough.
Mark
I spent most of yesterday measuring and tweeking the Fostex FE108EZ drivers I'm trying for midrange units. I wanted to get them setup right for proper evaluation and comparison to the ScanSpeaks.
They do a lot of things right and sound damn good actually. However they are missing the dynamics that the ScanSpeak mids have. Even crossed over 24db at 360hz I can't play them as loud.
I probably should have figured that out with the low Xmax. Anyone who doesn't like to crank up the volume now and then would probably be happy with the Fostex. That's not me so I'll be back to working on the ScanSpeaks when time permits.
@System7 I searched on and did some reading about "rocking mode" phenomena you mentioned. It's interesting for sure. I would "think" when I put the 200uF caps in series at four ohms and also tried the units at 16 ohms that would probably rule that out. I also mentioned I tried with my transistor amp and had the same issue. I wonder though if the underhung motor might be more susceptible to this phenomena and perhaps I should persue further.
Anyway thanks to everyone. I'll get around to better measurements. Wife is pissed at me right now because I haven't been getting my other stuff done around the house or taking her out enough.
Mark
Last edited:
I haven't done much yet. I did do a sine wave frequency sweep using an audio frequency generator (real one not PC based) and no significant peaks are showing up. One thing I didn't like is the baffle step dip which showed up on my edge analysis program (between approx. 650 to 900hz) looked a lot deeper than the program projected. I don't think that dip would have anything to do with my problem though. Note this graph is midrange drivers only and my 290 - 360hz lower crossover point is just top of the baffle step thus that isn't an issue.
I'll get to the additional measurements. Free time has not been good to me.
I'll get to the additional measurements. Free time has not been good to me.
Attachments
Not finished but I have an update to this. Sorry I don't have all the measurements yet but I've spend countless hours using my ears which are I think very valid. I'm a novice when it comes to speaker building but I'm 60 and been an avid audiophile since very early 20's. I've owned a lot of good speakers over the years from Acoustat 2+2's to B&W 801's etc.
I need to apologize to those who were telling me it's in the crossover. It is although I think I have two additional issues which made more difficult my analysis of this.
Some of my experiments were made in hast due to time constraints and I encountered what I think is another problem which masked the first.
I hzve two of the Pass Labs xvr-1 crossover units. I had an old one and also bought a new one last summer. Very expensive so I refused to believe it could be from the crossover. I bought two Bryston (10b std and 10B-Sub) to compare. No change they had the same issue.
I'm scratching my head about how or why but I have to be hearing crossover ringing but only through the midrange drivers. I can't use even a 6db med Q setting without hearing it just a little. The more poles I add the worse the anomaly becomes. However I can get rid of it when I use the 6db settings with Low Q. It's gone on that setting.
However then when I turn it up I'm hearing a different sounding anomaly which I "think" is the peak up around 4.5 to 5.5 K. Thus I need a steeper setting. Also 6db is not steep enough to keep lows from over driving the units at high levels either. I think I'm starting to exceed xmax when I play it loud.
My thoughts are most likely going to be to put a simple 6db band pass filter on the midrange units at approx. 300hz and 2.5K and use the 6db active settings at the same. 290hz and 2.5k which will give me my desired 2nd order.
I'm really not understanding why I'm having this trouble. My prior midrange was the focal 5k4411 which I had no problems with. None and it was a good sounding midrange. Wishing I had just kept it.
Anyway I've made some mistakes in this design. Live and learn I guess. If I can't resolve it with the ScanSpeaks soon I'm going to try the Eton 4 inchers Madisound has on sale right now. It's almost an exact fit and looks like it has twice the Xmax.
I'm wondering if some of this trouble is the ScanSpeaks underhung coil with limited Xmax issues amplified by the fact I chose to use two of them and like to play fairly loudly at times.
I need to apologize to those who were telling me it's in the crossover. It is although I think I have two additional issues which made more difficult my analysis of this.
Some of my experiments were made in hast due to time constraints and I encountered what I think is another problem which masked the first.
I hzve two of the Pass Labs xvr-1 crossover units. I had an old one and also bought a new one last summer. Very expensive so I refused to believe it could be from the crossover. I bought two Bryston (10b std and 10B-Sub) to compare. No change they had the same issue.
I'm scratching my head about how or why but I have to be hearing crossover ringing but only through the midrange drivers. I can't use even a 6db med Q setting without hearing it just a little. The more poles I add the worse the anomaly becomes. However I can get rid of it when I use the 6db settings with Low Q. It's gone on that setting.
However then when I turn it up I'm hearing a different sounding anomaly which I "think" is the peak up around 4.5 to 5.5 K. Thus I need a steeper setting. Also 6db is not steep enough to keep lows from over driving the units at high levels either. I think I'm starting to exceed xmax when I play it loud.
My thoughts are most likely going to be to put a simple 6db band pass filter on the midrange units at approx. 300hz and 2.5K and use the 6db active settings at the same. 290hz and 2.5k which will give me my desired 2nd order.
I'm really not understanding why I'm having this trouble. My prior midrange was the focal 5k4411 which I had no problems with. None and it was a good sounding midrange. Wishing I had just kept it.
Anyway I've made some mistakes in this design. Live and learn I guess. If I can't resolve it with the ScanSpeaks soon I'm going to try the Eton 4 inchers Madisound has on sale right now. It's almost an exact fit and looks like it has twice the Xmax.
I'm wondering if some of this trouble is the ScanSpeaks underhung coil with limited Xmax issues amplified by the fact I chose to use two of them and like to play fairly loudly at times.
While you are experimenting, I would not bother trying a hybrid approach. Do a 12 db/octave on the active unit and measure the results. My guess is that you'll find significant ringing left over. Compensate for that with a resonant filter. If you go hybrid, use a Zobel filter to help smooth out your response, and then you will have an easier time with cutting out the ringing in the high frequency. Once you have figured out how to get a nice smooth upper end decide on whether to use an active or hybrid approach.
Not trying to be snarky, but this is what is taking you so long to dial them in.
<<Updated>>
I know some will be disappointed but I gave up on using the two 12MU's. I don't know what the hell. I played for many many hours with the omnimic. Distortion and other measurements I got were all very similar to what I found for that driver on the internet. Nothing bad showed up.
I have one theory and again it's only a guess. I play my music loud at times. The anomaly I was trying to solve only showed up at fairly loud volume. To me not to loud but kind of loud. Turn it down a little an the anomaly disappeared. The 12MU only has 3.5mm Xmax. Possibly I was pushing Xmax at higher volume?? I don't know.
I switched to a single 18WU/4741T and everything is solved. It sound VERY good. Seriously I'm happy with them and can live with them.
Pictures below but please no making fun of my cabinets. I'm not much of a wood worker but they are nice and solid with bracing and what not.
If anyone lives near the north Denver area I'll give you a FREE pair of 12MU's as thank you for the help people tried to provide for me here. The other pair I'm going to use for a smaller system I'm gonna build for the tv down in the family room. The drivers are perfectly fine I didn't damage them in any way or anything. Again free if anyone is close enough to come and get. I can't ship them because I didn't save the boxes.
Mark
P.S. - The equalizers are only on the woofer sections.
I know some will be disappointed but I gave up on using the two 12MU's. I don't know what the hell. I played for many many hours with the omnimic. Distortion and other measurements I got were all very similar to what I found for that driver on the internet. Nothing bad showed up.
I have one theory and again it's only a guess. I play my music loud at times. The anomaly I was trying to solve only showed up at fairly loud volume. To me not to loud but kind of loud. Turn it down a little an the anomaly disappeared. The 12MU only has 3.5mm Xmax. Possibly I was pushing Xmax at higher volume?? I don't know.
I switched to a single 18WU/4741T and everything is solved. It sound VERY good. Seriously I'm happy with them and can live with them.
Pictures below but please no making fun of my cabinets. I'm not much of a wood worker but they are nice and solid with bracing and what not.
If anyone lives near the north Denver area I'll give you a FREE pair of 12MU's as thank you for the help people tried to provide for me here. The other pair I'm going to use for a smaller system I'm gonna build for the tv down in the family room. The drivers are perfectly fine I didn't damage them in any way or anything. Again free if anyone is close enough to come and get. I can't ship them because I didn't save the boxes.
Mark
P.S. - The equalizers are only on the woofer sections.
Attachments
Pass Labs XA-30.8
I thought I'd give a quick plug for these new amps from Pass Labs. Conservatively I'd say I've owned more than 20 high end power amps over the last two years. Some of them 100 lb plus monsters. Prior to now, my favorite amp excluding Bass was an Audio Research Reference 110. It had a sweet midrange that was hard to beat.
The XA-30.8 is heads above every other power amp I've ever owned. I'm running it above 300hz on the ScanSpeak 18MU midrange and D3004/6640 Beryllium tweeters crossed over at 2500 hz.
Last night I cranked the hell out of I-Robot by Alan Parson's project. They hit hard yet the needle didn't even inch up.
This Power Amp is a keeper.
I thought I'd give a quick plug for these new amps from Pass Labs. Conservatively I'd say I've owned more than 20 high end power amps over the last two years. Some of them 100 lb plus monsters. Prior to now, my favorite amp excluding Bass was an Audio Research Reference 110. It had a sweet midrange that was hard to beat.
The XA-30.8 is heads above every other power amp I've ever owned. I'm running it above 300hz on the ScanSpeak 18MU midrange and D3004/6640 Beryllium tweeters crossed over at 2500 hz.
Last night I cranked the hell out of I-Robot by Alan Parson's project. They hit hard yet the needle didn't even inch up.
This Power Amp is a keeper.
They were sealed but of rather large volume. I didn't think it mattered much as my crossover point was around 300hz give or take a little experimenting. (loosely stuffed with wool/fiberglass combination)
Last edited:
Mercury...
I would take your Omni Mic and do some CSD measurements on yor offending drivers AND the ones that sound ok just as a comparison. Run the drivers directly AND with the crossovers to compare.
I have seen drivers that have a similar issue to what you describe that showed no big issue in freq. response or distortion ( at moderate volume anyway) BUT that showed some obvious problems on CSD. Some problems are much easyer to see on CSD .
I would set up with mic about 1 inch away from driver diaphragm (at low volume) and in your case set the widow at about 5 ms. Take some measurements with mic directly at cone center, close to cone edge, and close to mid way between those two areas. Sometimes a cone can misbehave at a specific area and you can miss it if MIc is moved a bit. You are looking for a "tail" in the plot. This will be a section of the response that decays at a much longer time period than everything else and that section of the plot will look like a "tail".
Recently I was looking at a ribbon that had a similar issue that you discribe. Distortion was low and freq response was smooth but had a small wiggle at 4 khz that many would think is not bad. However looking at it with CSD we see a huge "tail" at 4k. Everywhere else on the plot was clean (decay to about nothing within about 2 ms) BUT at 4k it would "ring" out for about 6 ms.
When you said you could hear a sound similar when you hook up a battery that was a giveaway to me. This ribbon was the same. You could tap lightly on it and hear a metallic sound. It was a pourly damped resonance centered at 4 khz that rang forever BUT was only a small wiggle in freq response.
I would take your Omni Mic and do some CSD measurements on yor offending drivers AND the ones that sound ok just as a comparison. Run the drivers directly AND with the crossovers to compare.
I have seen drivers that have a similar issue to what you describe that showed no big issue in freq. response or distortion ( at moderate volume anyway) BUT that showed some obvious problems on CSD. Some problems are much easyer to see on CSD .
I would set up with mic about 1 inch away from driver diaphragm (at low volume) and in your case set the widow at about 5 ms. Take some measurements with mic directly at cone center, close to cone edge, and close to mid way between those two areas. Sometimes a cone can misbehave at a specific area and you can miss it if MIc is moved a bit. You are looking for a "tail" in the plot. This will be a section of the response that decays at a much longer time period than everything else and that section of the plot will look like a "tail".
Recently I was looking at a ribbon that had a similar issue that you discribe. Distortion was low and freq response was smooth but had a small wiggle at 4 khz that many would think is not bad. However looking at it with CSD we see a huge "tail" at 4k. Everywhere else on the plot was clean (decay to about nothing within about 2 ms) BUT at 4k it would "ring" out for about 6 ms.
When you said you could hear a sound similar when you hook up a battery that was a giveaway to me. This ribbon was the same. You could tap lightly on it and hear a metallic sound. It was a pourly damped resonance centered at 4 khz that rang forever BUT was only a small wiggle in freq response.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.