Matching Drivers

[Robh3606]

Just wondering if any of you have tried to match drivers into matched pairs?? Looking at Mr. Widgets F/R curves in the Ring Radiator Comparison thread there seems to be some variation within the groups of drivers tested. I have experimented with EQ in my active system and there seems to be a corelation between matching the curves closely say between 200Hz-8K or so and image stability. This to me seems to support the idea that a matched driver set could have advantages beyond uniform F/R with more stable imaging as a bonus. The low end and high end seem to have less affect its the midrange that seems to have the most effect. I was wondering what experiences you have had?? If you have tried to match drivers how did you do it.

Thanks Rob:)

[Lancer]

One reason people buy JBL components is because there is greater unit to unit consitency. If that isn't good enough and you want matched pairs send them back to JBL like Zilch did. See what happens... Another option is to buy a group and sift through them like people do with capacitors. Perhaps one can check with Westlake and see how they interact with JBL, TAD, etc.

The low end and high end seem to have less affect its the midrange that seems to have the most effect.

The high end definitely has an effect due to the plethora of overtones. If you can't hear the difference using a mismatched pair of, say, 066's then you're probably better off and won't need to spend a whole lot of money on hi-fi. :)

[Robh3606]

"The high end definitely has an effect due to the plethora of overtones."

Hello Lancer

By the midrange I meant the 200-8k range. The effect was strange especially if you unbalanced things on purpose like changing amplitude in the lower voice fundimental regions and leaving the top balanced. A center imaging vocal could be made to have the voice biased toward the right of left with the sibilance centered. Kind of disjointed things a bit. If you had speakers that had a bit of variation that were on either side of the spec I can see the imaging being pulled toward the louder source and if the loudness between the pair varies with frequency the imaging could also be hit or miss depending on the frequency content and the actual diferences between the speaker pair. Obviously there is more going on because of room interaction then just driver variability

As far as I know JBL doesn't match pairs like the KEF did in the 70's. As long as the driver variation is not significant then the issue would be moot except for what the room will do.

Rob:)

[Lancer]

By the midrange I meant the 200-8k range.

Right, which a high frequency driver like an 066 would cover whereas an 077 was not intended to do so. There was a wider variation in the older domes. The 035Ti has been real popular due to unit to unit consistency.

As long as the driver variation is not significant then the issue would be moot except for what the room will do.

It is probably best that everyone view it that way. That's one reason why people buy JBL transducers.

Looking at Mr. Widgets F/R curves in the Ring Radiator Comparison thread there seems to be some variation within the groups of drivers tested.

Yeah, it's fun for us to play around with measurement gear and get an idea of what's going. We're hobbyists. Obviously for any comparisons intended for professional publication purposes I'd recommend the measurement facilities at Harman Intl.

[DavidF]

Matched components (speakers, coils, caps etc.) do matter especially as you move from high-quality sound over to high-end sound. Even in mid-to-high quality sound (in the home), matched components (caps, coils, etc. as well as drivers) may go further in providing the good sound as does spending more money on the very best caps, very best coils, and the like.

It would take some amount of measurement accuracy to try this at home, though absolute calibration is not needed since you are looking at only deviation among the units. At one time, many of the D.Y.I. suppliers would sell matched drivers for an extra price. And, as pointed out, you may need several units to compare. I think you may still find that service, but more limited to high-end drivers. This seems contradictory to me in some ways, since, as Lancer says, it starts with very high production consistencies.

I had problems with a first-round 4-way crossover design I am tweaking. In addition to issues with balancing vertically (matching response sensitivities) I had a persistent and annoying sense that the image was tilted off center. Very noticible with vocals since these are mostly centered in the mix. The image tended to float a bit as the image does when drivers are connected out of phase, and the image wander is effected by frequency. So, I thought this was a phase issue (which it may have been to an extent) and spent a lot of time retracing wire, checking polarity, thinking driver response variations, and alignment issues. Should have swapped the crossovers to begin with, as one simple way to determine the source of the problem by elimination. The problem moved with the swap. Rechecking the components I found that I installed a 22mF value cap instead of a 2.2mF on one side. This was a series component to the lower mid-range so the effect was to increase the low end side of this band-pass response curve. You could equate the effect of this component imbalance to a mismatch in frequency response between drivers. Though probably more significant in terms of typical production tolerance, it does point out how a mismatch can effect the response.

DavidF

[Mr. Widget]

Looking at Mr. Widgets F/R curves in the Ring Radiator Comparison thread there seems to be some variation within the groups of drivers tested.

Yeah, it's fun for us to play around with measurement gear and get an idea of what's going. We're hobbyists. Obviously for any comparisons intended for professional publication purposes I'd recommend the measurement facilities at Harman Intl.

I agree that we do not have the facilities that are available to Harman International... really no one but Harman does. I also agree with criticisms made by several members that the information in those threads was not absolute or complete. That said, what was presented was quite accurate in as far as it goes.

In defense of our Comparison Threads, we did see some clear anomalies in a pair of Zilch's drivers. After sending them back to the factory for a bit of rehab, they came back and measured beautifully and are now quite well matched.

To get accurate measurements at home isn't inexpensive or easy, but I do think it is extremely eye opening and very useful, especially when dealing with older drivers that may no longer be what they once were.


Widget

[Lancer]

In defense of our Comparison Threads, we did see some clear anomalies in a pair of Zilch's drivers. After sending them back to the factory for a bit of rehab, they came back and measured beautifully and are now quite well matched.

There are those who watch this site.

I also agree with criticisms made by several members that the information in those threads was not absolute or complete. That said, what was presented was quite accurate in as far as it goes.

I thought the whole thing was cool! Yeah, I would have liked all the data files and impedance curves, etc. but it was fine. There will be other opportunities. I wish more members would spend the cash for at least the WT2, if nothing else, so they could post such information as their own impedance curves, TS parameters, etc. It's fun!





I think I'm pretty sure what precipitated this thread and I'm not real sure what I think about it yet.

[Zilch]

In defense of our Comparison Threads, we did see some clear anomalies in a pair of Zilch's drivers. After sending them back to the factory for a bit of rehab, they came back and measured beautifully and are now quite well matched.

They sound mightily good, too! :D

Mr. Widget's measurements revealed several instances where response is not well controlled. There are always design/manufacturing/material variables; some designs are inherently more consistent and stable than others.

Mr. Widget and I observed that often in our testing. See, for example, chaos above 10 kHz in 243x drivers. JBL is apparently making improvements in that area however; each new pair that I test seems to be better matched.

Another company that used JBL OEM is Meyer Sound, local here. They rebuilt/enhanced drivers to their own specifications, allegedly, until they were able to build them completely in-house.

The used vintage drivers we all love can often be mismatched, of course. Some level of testing should be employed when incorporating them into systems. Sophisticated equipment is not necessary for that. Something as simple as an SPL meter and tone disk will work; the results don't have to be absolute, as DavidF indicates above.

New diaphragms in everything isn't always the optimum approach, either, unless they are themselves subsequently verified with measurements.

[Lancer]

New diaphragms in everything isn't always the optimum approach, either, unless they are themselves subsequently verified with measurements.

That is correct, hence the JBL Authorized Service Center.
Faulty recone and diaphragm kits are returned to JBL.

[Ian Mackenzie]

I wonder what Unwound, I mean Edgwound has to say about this?

What is the faulty or failure rate on change outs?

Perhaps it comes down to the skill of the reconer.

Speaking of change outs I still piss myself laughing :rotfl: :rotfl: :rotfl: about Boputnam trying to fit an oval port in a round hole of an Alpine car sub. I hope they never offer him an Installer license....

Ian

[edgewound]

Hey Ian...how did you know how my day was going? are you psychic?

The skill of the technician has alot to do with how the finished product sounds because we are trained to look for certain anomalies that might occur in the part...and sometimes we can fix it and sometimes we can't....like Lancer said defective cone kits/diaphragms go back to JBL.

A properly repaired driver...any driver...is every bit as good as a brand new unit...because basically, it is returned to new....sometimes improved specs.

If a driver isn't prepped properly for repair,and done meticulously it's destined to repeatedly fail...sometimes the repaired driver is better than when it came off the assembly line because it should be receiving more attentive care.

Then there's the issue of pilot error..DJ's are notorious for running this stuff into the ground...that'll piss some people off...I know Scotty knows better;) .

Edgewound

[Ian Mackenzie]

Thanks Edgewound,

Have a great weekend. :cheers:

Ian

[edgewound]

:cheers: You're most welcome, Ian....time to tip a few because it's been hotter'an hell here in socal...but...seems like it is most everywhere...our earth is angry:( .


Edgewound

[Robh3606]

Hey Edgewound

I am sweltering here in NY! Here's one to yah!

Rob:cheers:

[louped garouv]

:cheers: You're most welcome, Ian....time to tip a few because it's been hotter'an hell here in socal...but...seems like it is most everywhere...our earth is angry:( .


Edgewound

been hotter than ever was recorded (think the first were like 1890 or thereabout) on the Western Slope of Colorado (Grand Junction) this past week.... forget what the temp/ was though :D

Denver hotter than since I've lived here, but that,s not too long.....



so beyond a basic SPL meter, what other budget measurements would be easily/cheaply implemented by an utter idiot DJ?


all of you guys are deep, sometimes I am just blown away by the audiotalk....

then theres the timess that I chime in..... :slink:

[Robh3606]

Hey louped garouv

Hot by you too?? A couple of days ago you could see the damn air it was so humid with 97F on top of that!

"so beyond a basic SPL meter, what other budget measurements would be easily/cheaply implemented by an utter idiot DJ?'

With me I couldn't live without my RTA. You can set crossover points, verify phase and driver balance, system balance and speaker to speaker balance. Great tool for not a lot of money these days. Can't see doing active without one. On the driver front I think what Lancer recommended the Woofer Tester 2 is a great low cost option. Got to get me one of those.

Rob:cheers:

[louped garouv]

cool -- i'll have to invest in some measurement tools -- I suspect that I may go active for real soon....

I am sorting though my picture as we sit here, (scotch here too) -- been a long hot week hitting over 100 in denver, I think Grand Junction was 106 (but don't quote me), but almost no humidity -- Colorado is very nice in that way, don't know that I could move back to it -- but then again, we ain't got an ocean within driving distance though.... :( and I could see a fire (hopefully a 'controlled' burn) on one of the table mountains in Golden (where the Coors Beer comes from) from my office window...

[louped garouv]

cool -- i'll have to invest in some measurement tools -- I suspect that I may go active for real soon....

I am sorting though my picture as we sit here, (scotch here too) -- been a long hot week hitting over 100 in denver, I think Grand Junction was 106 (but don't quote me), but almost no humidity -- Colorado is very nice in that way, don't know that I could move back to it -- but then again, we ain't got an ocean within driving distance though.... :( and I could see a fire (hopefully a 'controlled' burn) on one of the table mountains in Golden (where the Coors Beer comes from) from my office window...


Record Heat Wave Continues To Grip Area

No Relief Expected Until Monday



POSTED: 1:10 pm MDT July 21, 2005
UPDATED: 5:25 pm MDT July 21, 2005


DENVER -- Denver's weather went into the record books Thursday as the temperature hit 104 degrees, three degrees hotter than the old record for this date in 1981.




Thursday was the third consecutive day of temperatures above 100 degrees for the metro area, approaching the second-longest stretch of 100-degree heat ever. The longest stretch was five days in July 1989. The hot weather was forecast to stick around for a few more days.

Denver wasn't the only place that was hot, either. The northeastern plains town of Sedgwick recorded 114 degrees on Wednesday, matching the hottest temperature ever recorded in Colorado. It was the second time in 51 years that Sedgwick got that hot, according to the Colorado Climate Center of Colorado State University.

[Mr. Widget]

Yeah... that Global Warming thing like Evolution is just theory...
SF today hit 70° F it is warming up tomorrow... 74° F:p


Widget

[yggdrasil]

There just is no justice in the world. 45F here tonight, shiny weather, bright midnight sun and everything.


Have been to a booooring wedding today/-night. I can confirm that it is possible to drink quite a lot without having any fun at all. :cheers: