Printable View
I gave in and picked up a beat up 604E. Sent it off to Bill to have him give it the GPA treatment.
Woof T/S:
https://live.staticflickr.com/65535/...001e3376_b.jpg
Vas:
https://live.staticflickr.com/65535/...6d0b71ac_b.jpg
Horn T/S:
https://live.staticflickr.com/65535/...3aeab064_b.jpg
Cool. Let us know how it is when it gets back from Oklahoma.
Nice! Please keep us posted. I'll bet that speaker sounds amazing after you get it back from Bill/GPA. BTW, I'm inventing a triplex speaker. Altec 15" Biflex woofer with h.f. driver/horn (not to be confused with a triflex?).... ;)
Congrats
Welcome to the Duplex family
Are your magnets alnico or ceramic
I noticed the woofer R is about 7 Ohms. I expected to see about 12. What V.C. Are you getting on the new GPA?
Don
Ummmm,
I got it back from GPA yesterday - those measurements reflect the "GPA treatment"...
It's a huge mother - 604E so it's Alnico.
And yes, Don, I expected the same thing from the woofer VC - Bill told me he was putting in 16 ohm VCs for both HF and LF...
I only had time last night to very briefly hook it up through the Behringer dcx2496, as I don't have a XO for it yet. No cabinet either. It sounded...meh. The mids were very clear, and I think I got a glimpse of what everyone talks about that "point-source" sound - some extremely revealing details from the recordings, but as you'd expect I guess from an unmounted speaker it was utterly lacking any low end. I need to throw together a passive XO (plan to use Markwarts design as a starting point). But it is curious that the Woofer R is reading like an 8ohm VC is in there. Some of the other T/S params match very well with what is posted on Todd's list, while others are off by quite a bit. Does the MM number look right? I thought it was curious that DATS came back with a number that essentially equaled the amount that I applied to the cone to calculate the Vas. The math for that doesn't work out... :confused:
Ah, those are after - sorry! I'd say Mms looks low, that's a light cone. But I'm not sure. I'll see if I can dig up files on some GPA reconed 604s (ferrite) I had here.
From the plot, it looks like an ohm woofer and horn to me.
Quote:
Originally Posted by AlienShore
From your measurements, it appears the "GPA treatment" is to turn a 16 ohm 604 into a 10 ohm 605 ( if that DAT measurement is to be believed ) .
Time will tell if your measurements hold water ( fyi, others using WT3, which was the G1 version of DATS , had many problems with it, maybe DATS is better ??? ) .
The current measurements tune up quite well ( if accurate ) using a 9.5 cu' enclosure tuned to 25hz .
:)
Yep - that's what it looks like. I sent GPA an email - we'll see if they reply.Quote:
Originally Posted by Earl K
In the meantime, do you recommend another (more reliable) method for testing T/S?
WT2 is more reliable than WT3/DATS ( different maker with a different way of measuring ) .
REW has a TS parameter measurement area ( you need to wire-up a small box ) that uses the same method as WT3 .
Personally, I wouldn't bother GPA with this type of thing until you've gotten similar measurements from at least one other measuring program.
:)
- Every once in a while it is worth the diy hobbyists time to check ( recalibrate ) the validity of ones measurements by using another method ( really , it's simply part of the scientific method ).
- For TS parameters ( when checking these "all-in-one" devices ) one can use the old standby method outlined here !
- I believe that verifying Re, Fs & Qts ( Qms, Qes ) is enough to know where ones device stands .
:)
Thanks Earl. I wanted to talk to Bill again simply because when I spoke with him just prior to GPA shipping the speaker back to me, he indicated that they used 16 ohm parts for the re-cone and new 'phram. I have been looking at some of these time-aligned crossover designs for the 604, so I'll just need to adjust for 8 ohm voice coils.Quote:
Originally Posted by Earl K
I love the sound.westhost.com site. So much great information. [edit: westhost was taken down in 2016. I linked the internet archive copy here]Quote:
Originally Posted by Earl K
Totally agree, and while I have no reason to doubt the accuracy of DATs, I will perform a validation/calibration on one of my drivers to be sure.
Been out of town the last week or so, but will try to get back on track this week.
While the electrical investigations continue, I'm trying to move the cabinet research forward as well.
GM often relays the opinion of the MLTL being the best design for the 604 for home listening.
There's the
GM MLTL
Bill's Corner cabinets, and
6moons
upon a quick search...
Hey Mike - I built these cabs that GM sim'd for Jay Fisher - which is a variant of the 6moons build and am well pleased with them - http://www.hostboard.com/forums/f700...-604-mltl.html
I ended up pulling all the stuffing for a raw 1" BB ply cab, putting a piece of ply painted to match the carpet under them (essential, since they are bottom-firing) and with 3R on the woofer xo, and am very well pleased with them. They sound very nice with 3 watts!
In retrospect, the only thing I would have done different is make them wider and less deep on the cross-section, as they are a bit too 3:2 and could use a bit more baffle and would sit a bit more nicely in the room.
Thanks 'nipper.
Did you do plans for your build? In retrospect, what would you change? Would a forward-firing port be better (e.g., to avoid the need for the reflective plywood on your floor)? Do you have a schematic for the crossover you used?
GM did the original sims, and I got plans from Jay, which I slightly modified for size. If you want, send me a PM with your email and I can send you an info pack I have sent a few others on the build.
A couple good related threads:
Ping GM!
GPA 604 and Hiraga
Like I said, I would make them slightly wider across the front and less deep (keeping the same x-sectional area). I would also offset the drivers L/R in the baffle, since I neglected to do that. I would definitely not do the build like the 6Moons one - where they used the narrow edge of the box.
If I did it again, I might make some of the changes discussed in this thread with GM, Scooter, and RC Daniel:
GPA 604-8H-III Build - Sneak Peak
I dont have a schematic for the xo, but it is described by the builder, Rick Craig, as thus:
"I didn't really pay any attention to the stock crossover or specifications when creating the new design. The drivers were tested for frequency response and impedance and then everything was imported into our software optimization program.
This combination presents a difficult challenge in that there's very little overlap of the driver response curves due to the fairly small horn and large diameter woofer. The woofer has a significant breakup in the 1.8K-3K area The tweeter also has resonance issues in the 1-2K area. The shaping of the filters around the crossover point is critical and choosing the place to cross isn't easy.
I started first with a lower parts count crossover and experimented with different topologies. The objectives were to maintain good phase response in the crossover region, minimize the resonance and frequency response issues, maximize horizontal dispersion coverage, and provide plenty of headroom to keep the horn distortion low.
The crossover's acoustic slope target around the crossover point was 24db/octave. There are different ways to achieve this and a third order electrical filter was implemented for the tweeter. The woofer filter is a first order combined with a parallel resonance trap. Both are far from textbook values and designed to work in tandem with the driver responses to achieve the desired acoustical transfer function.
I also added a conjugate filter which creates a smooth system impedance. This is important for tube amps, in particular SET designs which typically have high output impedances. The less reactive load will provide better results and insures compatibility with virtually any tube amp."
The full post is here: GPA/Altec 604 Series 3 Crossover
As I have said elsewhere, this xo is far superior to the GPA one (which I also have) and Rick's build is competitively priced to the GPA xo.
But you have a 604E, so this may not be so relevant. Have you looked at this? http://home.earthlink.net/~jmarkwart/id22.html
Mike
Calibration of your test equipment is a great idea.
The impedance curves can be verified by testing a resistor in place of the speaker; try an 8 or 10 ohm .
This should produce a flat line at that value.
Don
OK, performed an impedance curve using a calibrated Vac test rig, and it tests out very close to the DATs test...
https://live.staticflickr.com/65535/...bb83f436_c.jpg
I didn't test 1 Hz intervals so the peak at Fs isn't as high with this rig, but this confirms that DATS is working well!
Mike
What is the freq of the peak on the horn T/S curve? The blue line
Thank you
Don
Here's the T/S measurements and free air impedance sweep on the HF horn:
https://www.hostboard.com/forums/hbm.../2012/09/1.jpg
If you can't make it out from the pic:
Re = 6.5471 Ohms
Fs = 1832 Hz
Qts = 1.412
Qes = 2.217
Qms = 3.886
Le = 0.122 mH
At Fs, impedance is right at 18 Ohms.
At 1500 Hz, impedance is 13.21 Ohms.
Hows the project coming?
Coming along. Been doing a lot of impedance testing, math, modeling, etc.Quote:
Originally Posted by cradeldorf
Don P. is helping me immensely - he took it on himself to construct a filter for us to test. When that arrives, I'll be putting it through it's paces, for proper impedance and FR, and once we have that tweaked, it will be onto adjusting the time delay stuff to synch up the HF and LF.
Here's a very detailed look at an impedance sweep (400-5000 Hz) on the woofer with the zobel added. It's hard to see here because this is such a up-close view, but the zobel really flattened out the impedance of this thing!
https://live.staticflickr.com/65535/...875a6564_c.jpg
I built this filter with 3 caps (8.2 + 8.2 + 2.0) and 1 resistor (10.0). The real-world measurement of them was 18.33uF and 10.14 ohms. I need to buy some modeling software.... doing this stuff all by hand with excel spreadsheets is tedious :)
I love that PCD from jeff bagby. so easy to do and free too!!! I used my .ZMA files from DATS. :) I just redid my crossovers this morning and I'm so happy with them it's crazy. I really like the way they bring out every little nuance of the movie soundtracks. 1.5mh erse super q on the LF and a zobel consisting of a 10.3 cap and 20 ohms of resistance, The HF is a 5.1 cap hooked to a 3 ohm resistor and then a 1.1 mh coil in parallel. I also wired in my attenuators and have them set at 2.5.
Thanks man. I'll download it tonight.Quote:
Originally Posted by cradeldorf
PanomaniacQuote:
Originally Posted by Panomaniac
What ohms ? on the woofer and horn
Thank you
Don
If you need help let me know, I'd be glad to talk you through it. :)Quote:
Originally Posted by Alien_Shore
Here's an FR graph of the 604E. These FRs are "barefoot" as Don likes to say, with no XO (just a 4uf HP on the horn).
I used gated impulse, and aligned the gate time for both measurements LF and HF:
https://live.staticflickr.com/65535/...99963d76_c.jpg
There are some crazy phase curves going on with these, but I left them out of this chart since it clutters it up so badly. That, and I have no clue what I'm looking at or how to interpret phase graphs yet. :p
Here are the FR graphs with Don's prototype TD XO on the LF (HF has a simple HP and L-Pad). I used the same settings in HolmImpluse as with the previous measurements:
http://imageshack.us/a/img856/2664/604efrwxo.png
I had issues with mine... After a couple months the software is the problem IMO. It kept crashing on my computer..So I sold it.. And worked fine for the gentleman I sold it too.. So if you have something on your computer (Other Audio Software) you could have some conflicts... I opted for the WT2 from Smith & Larson.. 10x better IMHO... And it worth the extra $60. A must for the DIY hobbyist IMO&E.Quote:
Originally Posted by Earl K
The mad scientist experimentation lab has been getting a workout:
https://live.staticflickr.com/65535/...6ebdf304_c.jpg
And today, I wound my first coil... Pain in the ass, let me tell you. But really, it's amazing how accurate the Shavano DIY coil calculator is...I made my own form with pvc pipe and rolled that sucka LoL. ...
Anyway, here's the ugly thing - I affectionately call it the 2.287 mH Christmas coil!
https://live.staticflickr.com/65535/...601a2b78_c.jpg
DIY coil winding is an interesting hobby, but I think to really get into it, you'd need something that can count turns or feet automatically for you, and some kind of oscillating line alignment gadget on there like a spinning reel does. I don't know, maybe I'll try it again and see, it is nice to be able to create precisely the inductance that your circuit needs, but it seems awfully time-consuming for what you get.
Congrats on your first coil! I've seen worse - I own worse. ;)
Wow, that's a lot of caps back there - what going on?
Tell me that speaker is not just balancing there like that!!! :eek2:
Thanks Pano!Quote:
Originally Posted by Panomaniac
That's the prototype LF filter. It's a 10th order Bessell, so yeah, it's got quite a lot of components.
My cat likes to play with it, he thinks it's an oversized top! I'll get it spinning real good and he tries to hang on ;)Quote:
Originally Posted by cradeldorf
I am really hoping to get a cabinet built for this speaker over the holiday vacation, so I can start some real listening tests. The preliminary tests with the free-standing driver are pretty amazing. I just lay it on it's side to get some good floor loading and put on my Nat King Cole CD. I'm anxious to get this HF filter built and then hook up the LF and the HF together...it's kinda scary how revealing this speaker is. I can hear the engineer tweaking individual mics on the recording... unreal.
I know what your saying, I have heard things in recording that were clearly not supposed to be there. :) I can't wait till you hear it in a cab. They are impressive.
Alien Shore
Mike:
Hope you had a good holiday.
How is the new speaker cab. coming?
You will be surprised when you hear it all tuned up
Don
Thanks Don! Yep - Holidays were great. Our son came down from NYC to visit for a few days and we hooked up with some friends and family.
I am working on an Altec 620 clone cabinet to test with. Should be done within a week. I have built the HF filter portion to mate up to your time-delay Besell filter for the woofer...so I think we're all set to test this thing out once the test cabinet is finished.
OK, 2 months later that I thought, but I finally have my 604E mounted in a test box:
https://live.staticflickr.com/65535/...07a375ab_c.jpg
Right now, its a sealed box. I will probably put a PVC vent in it with threads so I can adapt the size of the vent easily, or cap it off if I want it sealed. So far so good, oh, and before you guys start ragging on my crappy construction let me just say it's my first cabinet ever so go easy. :p
that was one fast week. :D
box looks ok . nothing a beltsander can't fix
That box looks better than mine. and trust me your gonna be making a lot of boxes. :D
@Alien_Shore
Mike, nice to see you kicking around these old haunts again. :)
If ( when ) you see this, can you answer a question or 2 please?
A good decade back you displayed a pic of the response of your 604E using the N1500a as the crossover.
"Was that with your GPA refurbished 604E's after Bill ( GPA ) had used the lower than expected impedance voice-coils ( more or less 8 ohms versus 16 ohms )" ??
If so, I'm really intrigued at how well your non-604E ( 8-ohm ) elements actually modelled for LC values originally chosen ( by Altec ) for the N1500a and it's higher impedance drivers.
Also, do you remember what distance you used for the off-set of the 2 drivers Acoustic Centers ( within PCD )??
Thanks! EarlK
PS: ( link included below for a much larger version of the pic >> if a member at HFH )
Log in | HiFi Haven
I'm looking through some of my old posts, computer files, notes, etc. to see if I can answer your questions. Yeah, it's been a minute hasn't it!?
Hi Earl and thanks - sorry for the lag in replying here - the holiday rush has had me strapped for time.Quote:
Originally Posted by Earl K
The graph I posted was not from a FR test, it was the sim output from a model that I had done with PCD. In that model, I had imported the FR curves and impedance data, so I presume that data was from the drivers with the refurb'd VCs. Glad to publish those data if you want to see it! Looking at the sim in excel, it appears that a parallel two-way config was used. Driver offsets were zero. The woofer design in the sim is 2nd order L/R crossing at 1.5 kHz 70 dB. There is a series RC (Zobel) with C=13 uF and R=10 Ohm, followed by the lowpass 2nd order section with L2=1.5 mH and C2=8 uF. For the HF section, it was 2nd order parallel with C9=6.0 uF and L9=1.0 mH, with a 0.3 Ohm DCR. There is a 3.0 Ohm series resistor after the HP section AND there is an 8.0 Ohm resistor in parallel leg 1. It will take me a bit to draw this out in a schematic, as I'm sure this description is confusing :pQuote:
Originally Posted by Earl K