Redesign or retrofit! #poll-notice
|
Samuel
Thanks for the feedback. I don't have an easy way to generate polars with accuracy. Since the drivers for the most part are of similar size and/or have similar polar responses within the region of crossover and I don't plan on altering the driver placement or crossover any more than is necessary, knowing the exact polar response is of secondary usefullness. You comment on ferrofluid in the DT94 is interesting. While I have heard for years it negatively impacts sound, can you describe what it does to the sound, or perhaps provide measurement. my main concern removing it is that this may not be very reproducible or something many people want to do. I would be interested in giving it try though. I am also concidering the Tymphany?D19TD05-08 though that would require a 2-3 element impedance compensation network. The OEM mid are the tough ones to replace. If they and tweeters were not so stiff due to drying adhesives and coatings I would just leave them alone. ?One of my tweeters and mids had failed voicemail leads that were repaired. Can you confirm that both the properly functioning mids and tweets have dual impedance peaks, like vented enclosure? |
|
Richard
My piezos work but I can not hear them with the crossover and by themselves. I can hear the difference when the supertweet is disconnected. The reason -the ST is partially out of phase with the tweeter and hence CANCELS some response from the tweeter in the crossover region. See the response below taken at the listening position, some 8ft way, which is at at the DQ10 sweet spot. The sweet spot is VERY small and is the axis at which all the drivers are in phase and sound the best. The orange is without the ST, the red with.You will see a dip of about 3dB at 13.5kHz WITH the ST. You can also see the ST response rises above15khz so it not only partially out of phase it is also too hot. If you unbolt the ST and move it forward this dip goes away at some point BUT since the ST is now sticking out of the baffle it creates diffractions in the response of the other drivers. I believe it best to just eliminate the ST and go with a tweeter that has extended high frequency response.? FWIW running the piezo directly from the amp is not valid test of a DQ10 deployed ST. Without a crossover it will have response down the low kHz range which I hope is audible to all of us. It does prove they are working though.? |
|
Shown seems to sound ok. Bink has many tracks for sweeps and phasing.? They still sound as good or better then my B&w 802d. Never liked ported towers trying to get deep bass from porting or 8” drivers. That said dq10’s without subs and active xo arnt great below 70hz for me. I like to hear bottom organ pedal tones . On Sun, Jan 16, 2022 at 4:14 PM Charlie Conger <ctconger@...> wrote: Richard |
|
Sorry for the long delay in updating the project status.
I have determined that the above drivers work and make a roughly comparable replacement for the original. They are also inexpensive. I do not believe I can say they sound or measure better, unless the original drivers are defective. The Visaton DT94 tweet requires removal of some or all of the ferrofluid as has been mentioned elsewhere to sound it's best. Also these tweeters vary enough that you have to measure and adjust the resistor values on each one to get it to correctly match the dome mid. Because fo these complications I can not wholeheartedly recommend this pair as an improved solution. My current plan is to evaluate other tweeters and midranges as alternatives. So far i have found a good tweeter candidate but no good replacement midranges that are good enough to be better and still reasonably priced and more importantly not on LONG backorder (some more than 6 months). So the next phase is likely to take 6 months or longer. Because my DQ10's ?are so old and had several previous owners I found there was a lot of changes necessary in order to restore them to like new stock condition.? This is a list of the items I found wrong. Dahlquist DQ10 changes ?
That's it for now. |
开云体育Are the two drivers wired out of phase the same as the schematic’s? If I’m correct that is done on some speakers for timing and dispersion? I noticed it on the specs and inspected mine which were wired per the schematic. I have a pair of Siegfried Linkwitz designed LX521.4’s which is a dipole type 6 driver per speaker which I have a precision ASP for the crossover for the drivers. Two woofers one pointing forward and the one pointing towards the rear is out of phase to the other. The speaker then has a lower midrange driver and an upper midrange and two tweeters which one is pointing for and the other towards the rear which is also phase inverted. Every driver has it’s own dedicated crossover design and each have a dedicated 365 watt amplifier channel except for the tweeters which share a channel. I am using my DQ-10’s which I will overhaul soon and my DQ-20i’s which have been restored and a pair of DQ-20’s I’m restoring with a rear mounted crossover for easier access. I have the silver plated Speakon connectors for and the original binding posts have been upgraded for use if needed to use one or the other.?Bill Meyerhoff? On Mar 8, 2022, at 18:18, Charlie Conger <ctconger@...> wrote:
|
|
Charlie you were reading my mind - I was today going to ask if there was an update! Thanks for this. A few follow up questions:
|
|
Bill, yes the drivers are NOW wired out of phase ( which is CORRECT) as shown on the most current schematic. They have to be because each section of the crossover introduces 90 degree phase shift for a total of 180 degrees. This is the same as inverting one of the drivers.
I determine/ confirm correct phase by inverting one of the drivers and then measuring the summed response. If the response drops 4-25dB in the crossover region when one driver is inverted the phase was correct before the change. ?If the repose does not drop but increases that indicates the driver are NOW correctly in phase. BUT, unless it is the woofer or supertweeter we are talking about the phase of the next driver will be wrong if the it was correct to begin with.? For most it is easier, but not trivial, to trace out the wiring and make sure that the drivers are wired according to schematic.? ? I am familiar with the 521's and yes the drivers must move either toward the listener or toward the front wall simultaneously in order to be correct. Since they are on opposite sides of the baffle one must move towards the magnet while the other moves away from it. This technique reduces distortion from suspension and magnet circuit nonlinearities as the non linear motion cancel. Speakon connectors ?are great!!!? |
|
Robert, Nice lab.
I agree?HiVi??is great value for the money. Your analysis of increased dome diameter impacts looks correct. I did a similar rudimentary analysis but I determined what is best by measuring. I did not do a complete power response analysis but rather just a listening window measurement. My DQ10 exhibits a very large amount of driver overlap. For example both the cone and dome range contribute equally over the range 760 hZ to 2.8 kHz. Additional the mid dome rolls off with asymmetrical crossover slopes at close to 18dB /octave slopes while the tweeter slope is closer to 6dB per octave. See the image below. the corrected measured repose of the HiVi dome on the baffle is quite close to the original dome mid. the system response also changed only very slightly. ?I do not critically listen off axis much at all. After careful measurements and listening I finally found the DQ10 sweet spot it is basically on the mid dome axis. This is the only axis where all the drivers will acoustically sum and the axis has a very narrow angle. At my listening position, about 9 ft away the sweet spot is only about 3" high and 3" wide. This equotes to a total angular variation of 1.6 degrees. Think almost head in a vice type of requirement. If anyone is interested I can describe how to locate the sweet pot on your DQ10. My woofer is now wired per the schematic and that is correct. It must be wired out of phase because and inductors, shift the phase 90 degrees. Since L1 is a low pass for the woofer and L2 is a high pass for the cone mid they shift in opposite directions for a total electrical phase shift of 180 degrees which is the same as reversing the leads (phase) o on of the drivers. Thus they must be wired out of phase electrically. You are correct that, when present, this affects the required phase of the driver next frequency band up or down, so it tricky to get them all playing together. ?I have confirmed that if you wire as shown on the latest schematic you will be correct .? See below for a measured ACOUSTIC response of my system the may help clarify : Just a couple of comments. Notice that the mid cone and dome operate at the same time over the range 760-2800Hz. Because fo the lateral displacement of those drivers I believe this and the very smooth summed response in this region is a big contributor to the DQ10 sound. Notice also that the woofer has large peak at 2.8kHz that affects the summed response. A similar issue occurs at 17kHz with the mid dome and the tweeter. I doubt that the later is audible for most of us over 60. The last comment is the RED is measured at the listening position and all the other measurements are taken very near the individual drivers. this is necessary because the DQ10 has a hybrid series /parallel crossover networks so you can not easily measure one drive at time. You must get very close and use absorbers to try to best isolate individual drivers.? |
|
we
1. Yes there were. The system was very dirty and had be sitting in agar for 10 years. I believe many of the errors were caused by a local third party repair service. if you are the original owner or know the OO your may not have most of these problems.? 2. Yes if you are having problems. it is very hard to tell what the drivers originally sounded like because these, except for the woofers, are 45 years old. i have done everything I can do to restore them to as close to original condition as possible.? 3. You can use your ear and a toilet paper tube to isolate the sound of each driver and hear defects like not working, ?distortion, ?very low in level. a schematic and good light / magnifier can help you locate wiring defects and poor soldier joints. A multimeter in ohms mode and knowledge how to use it finds most bad and intermittent connections. To find drivers that have been repaired and not properly restored is hard and requires more advanced tools and skills. in this case replacing the driver(s) may make sense but will take time and money.? 4. Parts Express but I also use Madison and Meniscus. Parts availability is a BIG issue these days and many of the drivers i wanted to try are long or indeterminate backorder.? 5. As I have time yes but see also #4 6. Yes, BYMMV probably depends a lot on how high you can hear. I can not hear the supertweeter at all when playing normal music at normal levels. With test signals at higher levels I can hear it. I can easily measure it. The ST is partially out of phase with tweeter so it never really sums correctly with it. This is matter of how the ST is mounted. But mounting it so it would sum correctly would introduce more problems than it solves. I plan on designing it out. To test the effect for yourself just cut one lead on R6 ( 35 ohm resistor) so that you can solder it back. This will kill the ST so you can hear if you can tell the difference. I will add that although I can not hear the ST on music I can hear it when it is disconnected. This is because it is out of phase with the tweet so it is partially cancelling its response at between 10 and 14kHz. So when you disconnect the ST the response goes UP in that range! |
|
Robert I measure the complex response of speaker which includes both magnitude and phase, not just the magnitude alone. ?I actual rely heavily on Group delay as one of my design analysis parameters. I also use step response as quick indicator of phase linearity. The DQ does credible job on both fronts which come as no surprise since they are both calculated from the impulse response. I believe that the acoustic phase which is represented by the summed pressure response is linear then the speaker is linear phase.?
I have also sent a separate private request for more information.? |
Charlie Conger
Bill Meyerhoff
SpongeSound
