My GF-777Z ** Project 2 OHM STABLE

[Cpl-Chronic]

Hello kids. Watcha doin'? Well, I'm here to tell you I've acheived good bass from the GF-777Z & in that endeavour I've broken the 3 ohm barrier & loaded BOSE subwoofers from a Cadillac that have a DC resistance of 2.1 OHMS. Typically, under normal conditions, this would roughly translate into about 2.5 ohm nominal impedance.

I've talked to many knowledgable people about running the GF at 2 ohms on the super woofer amps & ALL of them cautioned me that the Hitachi IC's will blow if I put too much load on them, like say, 2 OHM woofers.

So I looked up the Hitachi PDF for the HA 1392 & I found some interesting things. As yo can see in the snapshot below, the spec shows 2 ohm operation at the bottom limit which implies it can run at 2 ohms. :

The max voltage input is 20v DC & the max peak current output is max. 4A which translates to a few more watts @ 2 ohms. Also, the spec mentions thermal protections so ultimately, if the experiment goes awry, the thermal trip should help me stay out of trouble. Ok, so taking that info into account & looking at the physical circiuts of the GF-777Z:

You can see the massive size of those sinks to wick away heat, compared to the size of the IC's & that is also a BIG plus for taking on this experiment & that is why I decided to risk the farm on a bet that it would sound good with the 2 OHM BOSE woofers. I also took into account that BOSE usually makes a well behaved transducer that is not hard on the amp's damping factor so the load to the transistors won't be that gruesome, either. The old subs I used were too current happy on the absolute bottom octaves.

Here's a few pics of the woofers in their new homes as 'Super-Woofers' in my GF-Dirty-DISCO-Destroyer:

 

[Cpl-Chronic]

LOUD VIDEO!!! WARNING!! Turn down your speakers before playing this DEMO!!

 

[Fatdog]

Very interesting.

 

[blu_fuz]

Some day I will have a ghetto rigged 777 to mod .





That thing must be SOOOOO heavy now

 

[Superduper]

Don't do it Bobby.

 

[oldskool69]

My question is this...how are the amps set up in the GF-777 again? Someone refresh my memory. I'm thinking combined loads and if I'm right...then the math ain't right...

Got something for me on that one Norm?

 

[Superduper]

My question is this...how are the amps set up in the GF-777 again? Someone refresh my memory. I'm thinking combined loads and if I'm right...then the math ain't right...

Got something for me on that one Norm?

It's got dual HA1392 amp chips, each chip is set up in SE configuration. Each HA1392 contains stereo amps when operated SE. So the first one feeds the full range drivers on the sides. Then an attenuated signal is taken from the output of the first amp, highs shunted to ground via high-pass capacitor, and the filtered signal is fed into the second amp chip to power the center super woofers.

In SE configuration, the recommended output load is 4-ohms impedance. The thing is that impedance varies based on frequency, and actual (4-ohm speaker) impedance under dynamic usage probably will vary between 2.5 to 9 ohms depending upon the frequency imparted on the driver. If you start at 2-ohms, then what's the low impedance the amp will see during usage? One point something? Or maybe point something or other. In a lot of electronic circuits, a 1.X ohm or 0.X ohm is not much different than a short circuit. In fact, install a 5-watt 1 or 2 ohm resistor in series with any electronic circuit.... heck install it into the power lines to the boombox itself. Chances are, you won't even notice that the resistor is in there. I'll bet you install a 2-ohm resistor in series with your cassette deck motor and it won't even slow down enough to perceive.

As for the amps themselves, this particular chip burns out faster than any other chip out there common to boomboxes. Mostly they fry when in the C100's. Second is the BA521/BA532. They burn out quite readily too but almost always in the M9994 due to the M9994 using 2.5 ohm drivers. It is important to note that while the heatsink in the 777 looks beefy, they aren't that efficient as heatsinks go since there are no fins. It's basically a flat aluminum bar bent U-shaped. It does well to absorb the heat but without good surface area (in the form of fins) it does not do that good of a job of radiating and releasing the heat stored. Each heatsink has what is called thermal resistance. Good heatsinks are rated in their capacity to dissipate power (C/W) or the number of degree C rise per watt of power dissipated. The higher that number, the lower the capacity to release that energy. Even if a heatsink had infinite cooling power, a perfect heatsink if you will, there is still the problem of junction thermal resistance. That is the ability of the IC itself to transfer that heat through the small footprint to the substrate. I'm not going to get into the fine details, but I do presume that the boombox as produced by Sharp adheres to the recommended amp-chip manufacturers ratings of 4-ohm loads. Since even 4-ohm drivers will delve into the 3's and perhaps even into the 2's ohms ranges, the ohm/power chart does shows the graph of the relationship between load impedance and voltage to power output.

I'm not going to fall for and get sucked in again re getting into the arguments for/against a particular mod. All I can tell you is that I have replaced a whole lot of HA1392 chips and a whole lot of BA532 chips. I am quite certain that the Rohm chips burn out in the M9994 because of the low 2.5 ohm impedance of the drivers used. In the C100, those chips are configured in BTL mode, which means it requires DOUBLE the impedance (compared to SE usage) to be stable and safe. Does anyone know what impedance drivers are used in the C100? Anything less than 8-ohms would once again, be akin to using drivers that are less than 4-ohms in SE config.

 

[baddboybill]

My question is this...how are the amps set up in the GF-777 again? Someone refresh my memory. I'm thinking combined loads and if I'm right...then the math ain't right...

Got something for me on that one Norm?

It's got dual HA1392 amp chips, each chip is set up in SE configuration. Each HA1392 contains stereo amps when operated SE. So the first one feeds the full range drivers on the sides. Then an attenuated signal is taken from the output of the first amp, highs shunted to ground via high-pass capacitor, and the filtered signal is fed into the second amp chip to power the center super woofers.

In SE configuration, the recommended output load is 4-ohms impedance. The thing is that impedance varies based on frequency, and actual (4-ohm speaker) impedance under dynamic usage probably will vary between 2.5 to 9 ohms depending upon the frequency imparted on the driver. If you start at 2-ohms, then what's the low impedance the amp will see during usage? One point something? Or maybe point something or other. In a lot of electronic circuits, a 1.X ohm or 0.X ohm is not much different than a short circuit. In fact, install a 5-watt 1 or 2 ohm resistor in series with any electronic circuit.... heck install it into the power lines to the boombox itself. Chances are, you won't even notice that the resistor is in there. I'll bet you install a 2-ohm resistor in series with your cassette deck motor and it won't even slow down enough to perceive.

As for the amps themselves, this particular chip burns out faster than any other chip out there common to boomboxes. Mostly they fry when in the C100's. Second is the BA521/BA532. They burn out quite readily too but almost always in the M9994 due to the M9994 using 2.5 ohm drivers. It is important to note that while the heatsink in the 777 looks beefy, they aren't that efficient as heatsinks go since there are no fins. It's basically a flat aluminum bar bent U-shaped. It does well to absorb the heat but without good surface area (in the form of fins) it does not do that good of a job of radiating and releasing the heat stored. Each heatsink has what is called thermal resistance. Good heatsinks are rated in their capacity to dissipate power (C/W) or the number of degree C rise per watt of power dissipated. The higher that number, the lower the capacity to release that energy. Even if a heatsink had infinite cooling power, a perfect heatsink if you will, there is still the problem of junction thermal resistance. That is the ability of the IC itself to transfer that heat through the small footprint to the substrate. I'm not going to get into the fine details, but I do presume that the boombox as produced by Sharp adheres to the recommended amp-chip manufacturers ratings of 4-ohm loads. Since even 4-ohm drivers will delve into the 3's and perhaps even into the 2's ohms ranges, the ohm/power chart does shows the graph of the relationship between load impedance and voltage to power output.

I'm not going to fall for and get sucked in again re getting into the arguments for/against a particular mod. All I can tell you is that I have replaced a whole lot of HA1392 chips and a whole lot of BA532 chips. I am quite certain that the Rohm chips burn out in the M9994 because of the low 2.5 ohm impedance of the drivers used. In the C100, those chips are configured in BTL mode, which means it requires DOUBLE the impedance (compared to SE usage) to be stable and safe. Does anyone know what impedance drivers are used in the C100? Anything less than 8-ohms would once again, be akin to using drivers that are less than 4-ohms in SE config.

Norm I believe the C100 uses 5.4 ohm speaker

 

[Superduper]

Thanks Bill. That explains it. 5.4 ohm drivers would be the same as using 2.7 ohm drivers in a SE amp setup. So, it's just a matter of time....... Other boomboxes have already proven a tendency to blow amps when sub 4 ohm drivers are employed. I'd play it safe myself.

 

[baddboybill]

Thanks Bill. That explains it. 5.4 ohm drivers would be the same as using 2.7 ohm drivers in a SE amp setup. So, it's just a matter of time....... Other boomboxes have already proven a tendency to blow amps when sub 4 ohm drivers are employed. I'd play it safe myself.

So why would the manufacturer not match speakers to amp

 

[oldskool69]

Thanks Bill. That explains it. 5.4 ohm drivers would be the same as using 2.7 ohm drivers in a SE amp setup. So, it's just a matter of time....... Other boomboxes have already proven a tendency to blow amps when sub 4 ohm drivers are employed. I'd play it safe myself.

So my hunch was right. And I agree Norm, anyone's mod is anyone's mod and their business. But like you, I, when doing the math knew something was amiss, and thanks for confirming the setup of the 777.

 

[Superduper]

So why would the manufacturer not match speakers to amp

In an era and world of More-Power wars, this was a way to get a step ahead of the pack but does so at the expense of reliability. I've repaired several myself. Also a few members have posted examples where my gut feel is that the amps are likewise toast. What was the warranty.... 90 days? Or maybe 1-year. After that, you would be on your own. Could take years before the amps fail.

One thing I should mention.... component junction failure is probably far more common than catastrophic melt-down type of failure. By component junction failure, what I mean is the internal connection between the active component material and the leads. When a component is used, it begins to heat up. When turned off, it cools down. This heat cycle causes dissimilar materials to expand and contract at different rates. Over time, the internal connections begin to fracture and you'll get intermittent operation, sometimes heat related. Eventually, it fails completely. The HA1392 amp has a junction thermal resistance of 4 degrees C/Watt. This is separate than the heat dissipation of the heatsink itself and a limit that can not be overcome regardless of the type/size heatsink used since we are talking about the internal ability of the component to shed heat to the tab where heat is then passed off to the heatsink. If you were making the amps push 10wpc, 20watts total, and multiply that by 4, you would get an 80degree C rise. That converts to 176 degrees F above ambient. If the ambient temperature was 80 degrees, then the temperature at the junction would be 256 degrees F. How many 256 degree F heating/cooling cycles will it take for the amps to finally fail? I don't know that answer. What I do know, however, is that the greater the differential between max/min temperatures and the greater the number of heat cycles, the greater the failure rate.

And I do know that there is definitely a correlation between low speaker driver impedance, and amplifier failure rates. Don't mean that they will all fail when 3 ohm (or under) speakers are used. But the opportunity and chances are greater for that to happen.

 

[kraftmatic]

Looking forward to Cpl Chronic's performance report!

 

[Cpl-Chronic]

Well, I understand what you are saying SuperDuper & yes more cautious members should stay away from this type of mod unless you are willing to risk a 777Z's HA1392 to push the envelope. As for the math, I don't quite understand the 'doesn't add up' part. The BOSE speakers are 2.1 OHM DC resistance & yes they are thumpin' just fine & probably only do 2 OHMS at the bottom end of the range, if at all. I really doubt they go below 2 ohms at any frequency or they would be unusable as factory car spealers & probably present a pretty well behaved load to the amps so I'm betting it will be OK for the long haul as long as I don't take it out in 110 degreee weather & crank it to the ceiling. Oh yeah, they sound friggin' awesome too.

If the IC's blow, I'll man up & admit it like I have when other points have been proven, & even apologize, as I have in the past but so far I've been crankin' them pretty hard, every day for about 2 to 4 hours at a time & so far no burnt smell, no heat marks on the sinks & the caps look good too.

 

[tshorba]

Nominal impedance 2ohm (AC). All drivers vary impedance (AC) with frequency, this is fact.
DCR 2.1ohm, reading DC resistance on an AC device good to test if the are open circuit, that's about all

On the use of these in a car, a number of amps for that application are 1-2ohm stable now days.

You are putting a lot of faith in a company who will not publish a single specification on any piece of gear they make

 

[Superduper]

On the use of these in a car, a number of amps for that application are 1-2ohm stable now days.

Excellent point. In fact, all modern IC amp modules used nowadays are different and more robust -- the HA1392 now obsolete. Also should point out that many of Bose speakers are intended to be installed as an array and might be used in series as part of multi-speaker array. So 2x 2-ohm drivers = 4-ohms in series. Bose is known for having drivers with all kinds of unusual impedance. I have a set of Bose 901 series VI speakers for sale.... if anyone wants, LMK. BTW, each speaker has 9 drivers I think. Guess what, they are 1-ohm speakers. My point is that just because low impedance drivers are available doesn't mean they should be installed as a replacement for another driver in another application. Install these as a boombox replacement speaker and see how long the amp lasts....

http://www.parts-express.com/pe/showdetl.cfm?partnumber=290-922

Also, it's not enough to say I did it, nothing burnt, so all is well. As I said, most failures are stress related compounded over time, and not acute or catastrophic burnt smell, melted, or other such damage. Anyone can do any mod they wish, just like Freddie said. I myself have done many mods that other's cringe on. However, my point is more to let folks know that there's another side of the story so if they wish to follow along these lines, they should know and understand that there is risk involved. Sometimes, the exact same mod on 2-different boxes will have different results. One blows up, the other doesn't. How to explain that?

 

[Cpl-Chronic]

Well, again you are right, SD, to caution people about the risks involved & yes I am taking a big risk by doing this but I'm confident that I won't regret this decision to use the BOSE woofers. You may also be right that after a short period of time, they will blow the IC's but I'm pretty sure It'll work out Ok & I really do believe they over-engineered the amp stages to handle the 2+ ohm loads, including the massive sinks. After all, we are talking about the flagship of the SHARP line & everything about the triple7 screams durability & servicability compared to others like Lasonic, etc.

That's why I said that I would be the brave 1 & push it to see if it would work because others don't want to risk their tri[ple7 for the pursuit of better bass. That's whst makes it exciting for me. I want to see if I can go beyond the conventional thought & try something different.

 

[Cpl-Chronic]

Looking forward to Cpl Chronic's performance report!

My Report: Sounds nice & punchy with lots of deep bass & much louder overall. 808 tracks like Hashim, Planet Rock, etc absolutely sound AWESOME with these woofers as well as modern hip-hop sounds deep & cavernous with a sealed back. Also, the bass of the BOSE speakers fits quite nicely witht he 3.2 ohm outer woofers & the muddy 'boominess' is not an issue like other woofers so, 'thumbs up all around'.

Only time will tell if SuperDuper is right, which he may well be. I trust his expertise in this issue but still want to see if I may be on to something here that many may find a viable route for upgrading the puny bass of the GF.

I would say six months would be a good testing period to settle the point for good.

 

[tshorba]

I want to see if I can go beyond the conventional thought & try something different.

Why not use a couple of T-amp's to power the speakers. You could tap off the existing pre-amp stage of the 777 and remove the odd woofer signal input. The box could then be modded with SLA/Gel cell battery to give longer playing time.

If you do end up cooking your chips there are not many replacements available (what is available could be fakes) which could render the box useless, except for mods/parts. If using the T-amp approach you could always revert back to stock form.

 

[Cpl-Chronic]

I want to see if I can go beyond the conventional thought & try something different.

Why not use a couple of T-amp's to power the speakers. You could tap off the existing pre-amp stage of the 777 and remove the odd woofer signal input. The box could then be modded with SLA/Gel cell battery to give longer playing time.

If you do end up cooking your chips there are not many replacements available (what is available could be fakes) which could render the box useless, except for mods/parts. If using the T-amp approach you could always revert back to stock form.

You bring up some good points for NOT doing what I did. I am not disagreeing in any way with you, SD or anyone who wants to say that I'm a mixed up kid who shouldn't play with electricity. I want to make that clear. I do understand that I'm running my IC's at their bottom limit of the impedance range & risk catastrophic failure. Normally, I wouldn't chance it but looking at the PDF, it shows 2 ohm operation, it has thermal protection, the open design of the sinks, their MASSIVE size, the rugged build quality, etc., I figured it would be worth a try to see what can be done with the mighty triple7-ZEEE. I really love this tank & think it is under-rated & a bit under-powered. Do alot of 777's have IC failure compared to the others?

To answer your question:

Well, I did think of using a car amp with speaker level inputs. That way, I could run the original super-woofer wires to the amp's speaker level inputs & run the woofers from the amp giving me a good 20w per side or more, depending on how I set the input gain. Using speaker level inputs would avoid messy splicing into the pre-stage, present a higher impedance load to the original woofer amps & lower the load to the HA's. The down side to that plan is space. It's really tight in there already & any added amps would have to be attached externally to the back-panel. I have no idea what a T-amp is tho.

In the end, I scrapped the amp idea as too messy & I have faith that 2.5 OHMS will be doable & give me some modest gains in the BASS area. So far, they sound great & add more dynamic range to music & I doubt that the BOSE woofers dip down below their DC resistance point with an alternating current. Most websites tell you to take the DC resistance & multiply it by a factor of 1.15 to 1.3 depending on what article you read, for AC impedance. I peg it at 1.2 for my calc. & 2.15 OHMS x 1.20 = 2.58 OHMS nominal or 'averaged'. I bet BOSE labels them 2 ohms even tho 'nominal' is closer to 2.6 to err on the safe side.

It'll be interesting to see what happens over time so stay tuned to the Chronic Channel to find out.