Vintage JBL Pro Cinema Build Rev 2.0

[miscrms00]

The adventure continues....
http://www.audiokarma.org/forums/in...ankenstein-monster-system-on-a-budget.677257/

Brief recap, took the plunge and bought up all the old speakers from a local concert hall when the opportunity came along.

Ended up with a functional and enjoyable setup, but knew it would ultimately be temporary while tracking down the gear required to take it to the next level.

Coming up on two years now of trading, selling, scrounging, and repairing gear, I'm finally starting to assemble the next generation. This will likely involve some experimentation and testing, but excited to be getting started!

Equipment list (will probably not all get used in this setup):
Amps:
2x Altec 1608A (replaced marginal electrolytics, replaced burned output stage / driver resistors on one, debugged flaky driver card instability, most likely staying on the test bench)
2x Altec 9440A (replaced marginal electrolytics, upgraded coupling cap to film, ceramics to film / silver mica)
3x JBL 6260 (repaired flaky relay on one)
1x JBL 6290
1x Yamaha P-2200 (needs repair on intermittent meter, maybe electrolytics)
1x McIntosh MA-230 (replaced failed/marginal electrolytics, diodes, thermistor, modified for pre-out and post-in, still needs some cleanup, likely going to upstairs vintage system)
1x HK AVR7000 (failed input card, modified for 5ch power amp, needs PS cap replacement, moving out to backup status)

Speakers:
2x JBL 4648A w/dual 2226H 15" drivers
4x JBL 2446H w/ 2365A CC biradial horns & 2x 2506B brackets
2x JBL 2404H tweeters
2x JBL 8330 surrounds
1x JBL 4660 w/2346 "everest" defined coverage horn
1x JBL 4646A (previous center LF, likely moving to 4660)
1x JBL 2446H w/2380A FF horn (previous center HF, likely moving to 4660)
1x Intersonics ServoDrive Contrabass (made belt set, replaced bad motor, reglued separated cone/surround)

Sources / Control / Other:
1x Sherbourn PT-7030 - Pre/pro, hdmi switching, bass management, parametric EQ
2x JBL 5235 active crossover, cards 3x 500 CCBREQ, 1x 800 FFBREQ, 1x 80Hz, 1x blank
4x UREI 539 Room EQ
1x UREI 535 Stereo EQ
HDMI sources: Mac Mini (Plex/Itunes), PS3, PS4, WiiU
Media Server: Dell T3500 (CentOS 7), PERC 6/E SAS to MD1000 (16TB RAID 50)
Other sources: Denon DCD 3520, ELAC 50H (needs work), ELAC 750 (now working) - these are now in use in upstairs vintage system w/JBL 4411s
Panasonic P54G10 54" plasma (probably reusing for different room)
Panasonic P65VT50 65" plasma (probably replacing 54" plasma and projector)
Sony VW40 projector, tab tensioned 4:3 draper accessV screen (probably going away)

Test Equipment / Misc:
HP 4192A Impedance Analyzer
HP 35665A Dynamic Signal Analyzer
Tektronix 2465 Analog Oscilloscope
AudioTools iOS w/UMIK-1
GTC CM100 mA scale AC/DC clamp current meter
Fluke 87IV DVM
Misc. Cheap DVMs
Dale 5.8Ohm 200W resistor load
Collection of various vintage/nos caps (rel, wonder, sprague, Siemens, motor start/runs, silver mica, etc), coils, resistors, magnet wire, diodes, etc
temp controlled solder station (vintage Hexacon)
Various simulators (Cadence Spectre, ADS, Spice, etc)

 

[miscrms00]

Started by making some measurements to answer some questions that I've had.

The HP 35665A Dynamic Signal Analyzer is a pretty capable piece of audio frequency test gear, so I was excited to get a chance to try it out. I was very lucky to find one locally at a reasonable price a while back, as they still cost at least a few $k online. Primarily its an early 1990s FFT based spectrum analyzer that runs from ~0Hz to 112kHz. But it also has an integrated signal generator and two channel input, so it can be run as a network analyzer by dividing the signal output between the DUT and channel 1, and then comparing the channel 2 DUT output to the channel 1 DUT input, and it has an instrument mode to automate swept sine bode plot generation. In theory it can also be set up with a reference mic for room EQ. It has a vector based input detector, so it can measure both gain/amplitude and phase. In this mode it can measure the transfer curves of amplifier chains, filters, etc and because it has a high input impedance (rather than a 50Ohm input termination) it can be used to measure devices in situ without affecting the signal path or having to break connections. Inputs can be configured as DC or AC coupled (5Hz low frequency cutoff) and ground referenced or floating. It can measure speaker level outputs, but is limited to 4V peak to peak so it can't run at full volume without a resistor divider.

1) The Sherbourn PT_7030 is a consumer piece of gear, and not incredibly well documented from a technical standpoint, so I was curious how it stacked up in terms of gain and output levels when trying to drive into pro-audio gear. It turns out that it does quite well. I measured as much as 4Vrms output in my initial testing with the Fluke DVM, which is enough to drive pretty much anything I'm aware of. The input sensitivity on the Altec 9440As is 0.6Vrms, the JBLs are 1.1V (assuming peak to peak?).

Test setup: HP35665A Source Out -> PT-7030 Analog RCA input (Right Front) -> PT-7030 XLR Output (Right Front) -> JBL 5235 Ch 1 Input XLR / HP35665A Ch1 Input

Observations:
- 100Hz test tone at 1V rms fed back from source output to Ch 1 input reads 0dBV as expected.
- With PT-7030 Volume at 78, output also reads 0dBV indicating this is 0dB gain (gain=1) setting for differential output
- With volume at 85, output reads +6dBV (2Vrms) differentially, or 0dBV measured one side to ground (or RCA output). So PT-7030 XLR is a true differential output.
- Increasing volume for greater than +6dBV output is possible, but harmonics start to increase indicating clipping/distortion is probably beginning. So 2Vrms seems to be about the max linear output.
- Reduced input amplitude to 0.1Vrms, reads -20dBV fed back to input as expected.
- With volume set to 98 (100 is max) output is back to 0dBV. So volume notches appear to be ~1dB even though they are not labeled that way, and max gain from analog inputs appears to be 22 dB to differential output or 16dB single ended.

2) Measure 5235 500Hz CCBREQ (2360 series horn EQ) measure frequency response and compare to simulation.

Test Setup: Same as above, but source output split and returned to Ch1, JBL 5235 input (PT-7030 output), hi and low outputs connected to channel 2 with power amps connected in parallel for transfer curve measurements. PT-7030 volume set at 85, -20dBV (100mVrms) input. Measured 15.625 Hz to 32kHz, log sweep with 24 points per octave.

Simulated result based on schematics found on Lansing heritage forums
http://www.audioheritage.org/vbulle...5235-Active-Crossover-Filter-Cards&highlight=

Seems like good agreement between reality and simulation, which is always nice 5235 hi output is set at 10, which appears to give it 6dB more gain relative to sim, and the LF boost function is also configured as recommended to extend the low end range of the 4648A LF boxes. Phase difference is ~90 degrees through the crossover region as expected. One of the things I would like to look at down the road is building a custom set of more modern 4th order Linkwitz-Riley filter cards for the 5235 that retain the CC horn EQ but improve the phase alignment through the crossover region and double the slope for more separation between the LF and HF sections.

 

[Nightwisher]

 

[Drugolf]

What kind of crazy man would buy up all the gear from an old theater?........
oh wait...
AT least you are putting yours to use. I just like looking at and listening to the big efficient speakers.

 

[mark warren]

awesome!

 

[Rome]

Way beyond my pay grade!
Looks like an outstanding build!

Rome

 

[whoaru99]

Nice.

On the Yamaha P-2200 reflow the connections on the meter driver module (or whatever it's called) and surrounding parts. That'll probably fix the flakey meter reading.

 

[miscrms00]

Nice.

On the Yamaha P-2200 reflow the connections on the meter driver module (or whatever it's called) and surrounding parts. That'll probably fix the flakey meter reading.

Thanks! Good to know.

Rob

 

[miscrms00]

After trying a number of combinations, here is what I ended up with for a first cut:

The center channel performance was better with the 4660 than previous, but still did not match the front R/L very well. In particular it was quite lacking in low end by comparison, so I added the UREI room EQ to try and balance it out. When I tested it previously I had been boosting the low end with the analog tone controls on the 1608A. Still working on the profile, but it does seem to help quite a bit.

 

[miscrms00]

After going through and verifying polarities, doing a rough balancing of section and channel levels, and playing with the PT-7030 parametric EQ a bit, decided to give it a little "stress test" . Watched Metallica "Through the Never" with the pre/pro volume set at 74. Fired up AudioTools with the Umik-1 to make sure my ears weren't in immediate danger (RTA peak hold, SPL real time). I can honestly say this is the most concert like experience I've ever had in a home setting

Still got more work to do, but so far the upgrades are making a big difference in the overall experience! I'll have to check on the iPad and see if I captured an RTA response with pink noise, as I recall it is flatter than the above source material based result.

 

[miscrms00]

Amplifier channel location was done by what sounded best, or seemed to produce the appropriate levels at each channel most effectively. After the fact I went back a traced through gain / signal levels to see if that made sense based on the amplifier specs. I chose to do the numbers based on the max output from the PT-7030 as measured above. In a number of cases this would drive amps into clipping in reality. In that case max linear continuous amplifier output power is assumed. This is also enough power to exceed limits on some speakers. What I list as SPL "max" assumes the speaker can handle this power level to see how balanced the amplifier chains are in reality where input signal level will be considerably less. The SPL "cont" is what the speakers can actually put out at their spec limit if less than what the amp can provide to give an idea of max continuous SPL each speaker is capable of in the system.

Note this is everything at max, in reality volume / gain levels are used to balance channels further. But I thought it was interesting how well many things were balanced to begin with. The front horns are pretty hot, as expected, and the surrounds a bit weak. But the center and subwoofer are closer than I would have though. Note that the mid range to the horns is down 12dB from the max value shown, so they are actually a lot closer to the super tweeters and LF sections than it looks. I guess that's part of the magic of the crossover cards

 

[miscrms00]

Also discovered I need a power sequencer

I had noticed the inrush current when the power amps are switched on via the PDU was pegging the meter >15A and creating about a 10V voltage drop before settling to an idle current of ~1A and normal voltage. What I hadn't noticed until last night in the dark is that the main switch on the PDU is arcing significantly as it closes!

Ordered a 3 relatively inexpensive 3 stage sequencer from monoprice. Hopefully that will spread out the inrush from the various amps to a more reasonable level.

 

[jim steil]

a little slice of heaven you have there, nice

 

[jim steil]

The Golden Ticket