posted
I thought some of you guys possibly would be interested in this.
A little background on me first. I'm a Test Development Engineer at Analog Devices, (we design a lot of Operational Amplifiers and DSP's for audio components among many other things.) so.... I hope that shows I know a bit of what I'm talking about.
One channel of my Mac, (100x2 MC427), has been acting up lately, so since it was out of warranty, I decided to order the schematics directly from McIntosh instead of sending the amp back and $150. It cost me about $20 for the schematics and shipping. It's about 12 pages of material.
Some notes on building construction of the amp:
1. Generic +/-15V voltage regulators that are supplied from the +/-36V rails. These +/-15V rails supply the preamp section of the amplifier.
2. The +/-15V lines do not have any capacitance to ground. Without capacitance, these supplies can droop and affect audio quality. A couple of 0.1uF caps can cost about a quarter and possibly improve sound.
3. From what I remember, (I'm at home and the schmatic is at work), the +/-36V rails have 1000uF on each side of the supply. This is the main supply of the amp and it is vital to have as much capacitance here as possible. It is even more important to have large cap here than those, "stiffening", caps on the +14.4V supply of your car battery. If this +/-36V supply suffers, your output power will suffer with it.
4. The preamp audio amps specs are of average quality. The preamp op amps are made by a company called Rohm. It's a generic op amp with nothing special about it. It is made for audio applications though. These op amps are of the "dual" variety, (2 op amps in an 8 pin package). McIntosh chose to use each amp in each package for separate channels. Since multiple amps are used in the schematics in series, my opinion is that they should use the same 2 amps for the left channel and the other 2 for the right channel. But, they could have reasons for not doing this.
5. This amp has two pass throughs for the audio signal to connect preamp level outputs to the input of a subwoofer that is crossed over inside the McIntosh. These subwoofer crossovers are simply an RC circuit that provides 6dB per octave filtering. There is nothing special here. Since it has two crossover points, (80 and 120Hz), two different RC combinations are used here.
After opening this thing up, I think 90% of the cost of this amp was dedicated to a few chips, the power supply and the heatsink and exterior. This amp retails for $500, (well, they don't make them anymore), and I bet it cost them $50 to make it.
Now I know why a Krell can cost $20,000. But, I've yet to open up one of those.
Oh, and I'm still working on the amp. It seems to be a preamp stage op amp. I've yet to replace it, because work has been a bit busy.
Comments? Questions? I'll update this thread when I get some more analysis done.
In my opinion, you are paying for the McIntosh name and not the product.
Shawn Ironsides@goplay.com
------------------
Posts: 68 | From: San Jose, CA | Registered: Oct 1999
| IP: Logged |
posted
Couple of simple thoughts, although you probably know quite a bit more about this than me.. The point of McIntosh equipment is simply to provide the least distortion... Therefore, it can probably overlook a lot of other things Also, could the capacitance on the ground, although its absence affects audio quality, cause noise were it present?
I like the fact that you are trying to disband a myth, here... Perhaps some blindfold ear-tests are in order now?
ShadowStar
------------------
Computers are a lot like air conditioners.. Open Windows and they become useless..
The day Microsoft makes something that doesn't suck is the day they make vacuums.
Posts: 2578 | From: Somewhere In the Northeast | Registered: May 1999
| IP: Logged |
* The capacitance to ground is not in the signal path. What I mean is placing a capacitor between the +15V rail and GND, and the -15V rail and ground, (2 total caps). This will beef up the supplies of the op-amp and keep it from ever drooping. If it does droop, (due to excessive power demand), the op-amp will not work and the signal will clip.
* Blindfolded ear tests, huh? Well, my problem is that my amp is broke and I'm fixing it. I also have a pair of PPI A600.2's and they haven't broke yet. But, I won't dig into them until they break. (If it ain't broke, don't fix it, right.) If I did have the time and the amplifiers though, I might just take up that idea.
* In the future, I do plan on opening up and fixing stuff as I go on, but I don't think I'll be upgrading my components in my amps, *yet*. My biggest problem now for my car is road noise of my cheap tires and bearing noises, so I have bigger things to deal with first before I go tweaking! Plus, I'd be instantly moved into the "Pro" category at IASCA more than likely if I modified my amps.
Questions? Comments?
Shawn ironsides@goplay.com
------------------
Posts: 68 | From: San Jose, CA | Registered: Oct 1999
| IP: Logged |
posted
I plan to put extra capacitance in my Zues across the main +/- supplies (around 52v), am I going to cause any problems with the power supply since it is a switching power supply that you can think of? I like to make imperfect things better. I think this is one of my biggest problems. Why cant I ever just relax and be satisfied. also I love to open up all my amps to. Thats usually the first thing I do. I have opened up rockford, orions, Hiphonics, and others but non where even close and I mean even close to as nicely designed as Harman Kardon. I have never been able to get any company to give me schematics though. but you can tell by the quality of parts and the board layout which they at HK obviously spend a lot more time on. P.s. Ironsides, Analog devices has a place in southern Cal, right? is that where your at? If so how do you like it there?
------------------
Posts: 1259 | From: Fullerton. CA ,USA | Registered: Oct 1999
| IP: Logged |
I plan to put extra capacitance in my Zues across the main +/- supplies (around 52v), am I going to cause any problems with the power supply since it is a switching power supply that you can think of?
Nope, no problems will come up. The original idea came from my boss. He put the caps on his amp's rails. He hasn't had a problem, but cosmetically it looks like crap. He has a bank of 4 caps lying next to his amp. I'm sure a good installer can solve that problem. The big issue though is that you have to have your +/- 52V supplies exposed, so this isn't a tweak for "Joe Consumer".
You said:
P.s. Ironsides, Analog devices has a place in southern Cal, right? is that where your at? If so how do you like it there?
It's in Santa Clara CA, (Bay area). I like it a lot. I deal with test engineering in Power management, (i.e. for laptops or cell phones).
Shawn ironsides@goplay.com
Posts: 68 | From: San Jose, CA | Registered: Oct 1999
| IP: Logged |
posted
How did those caps work out for your boss? This Zues seems to have like a delayed over dampening on some of the transients. I dont know if its the feedback network or the power supply cant hang. it's wierd if its turned up to slightly clipping for a few seconds and then back down it almost seems to work ok for a little bit? I don't know I plan on just replaceing it with an HK.
------------------
Posts: 1259 | From: Fullerton. CA ,USA | Registered: Oct 1999
| IP: Logged |
posted
sounds like beefing up your power supply with a little extra capacitance certainly wont hurt. I met Helmut (owner of HiFonics) He was hitting on my best friends mom at finals in Tulsa... Damn that was funny...
------------------ STILL waiting to hear a GOOD pair of HLCD's!!
If it dont fit, force it... if it breaks, it needed to be replaced anyway...
posted
The caps worked great for my boss. I haven't heard any complaints. If you think about what the caps do, (store energy), then you can see why they'd do well on the power rails. The rails on amps tend to sag at full power, (just bench test it and watch them sag), so adding capacitance woudn't hurt.
You can't use caps for car audio for the rails I hope you know, (they're rated at low voltages), so you have to drop by an electronics supply store and get some that are rated at over your 52V. ESR, (Equivalent Series Resistance), isn't too big of an issue for rails, but I'd definitely consider checking into it. I think he had 44,000 uF per side.
Shawn ironsides@goplay.com
Posts: 68 | From: San Jose, CA | Registered: Oct 1999
| IP: Logged |
posted
I've noticed a number of companies using something as tiny as .01 uf caps at the 15 volt regulators and then a significant number of 10uf (more or less) on each op amps + and - supply pins. Each pair of caps at each op amp decouples it from the rest of the circuit and the cummulative effect stabilizes the regulated circuit overall. Is this not the case here?