Anniversary Series amp runs hot

[cclark65]

I have 4 9446 amps, and one of them seems to run quite hot. Even at idle with nothing running it will eventually heat up enough to feel like a small furnace! However, it has always performed fine and never shut down.

I believe this may be a problem with the bias, and I was very happy to see the service info for the 9442, 44, and 46 amps in the literature section, (I have a couple of each model) but they don't really explain where to check the bias voltage, and what to set it to.

Can anybody point me in the right direction for this info, or tell me if I am wrong with my thinking of why it's running hot?

Just for additional info, I am running them into 8ohm loads. They get hot even with the level control set to minimum with no input.

Thanks!

[bfish]

Howdy Charles!

Unfortunately, the library manuals don't detail the procedures for the 9444 or 9446, but it IS described in the service section of the 9442. Although the test points and trimpot locations will differ, the procedure should be essentially the same. There are really 4 separate adjustments to be made, DC offset and bias, on both channels. The voltmeter method is the simplest, only requiring a DMM with a millivolt scale, and is accurate if component values haven't drifted too far, so I'd do it first on a 'cold' amp, then recheck after warmup.

DC offset adjustment is simply setting the proper trimpots for 0.0VDC across each channels' output terminals. Although offset alone shouldn't cause your overheat, offset DOES affect bias (not to mention keeping DC off those unprotected diaphrams!)

Bias adjustment actually controls output stage idle CURRENT, but, (as described in the 9442 lit.) is performed via voltage measurement across a resistor. I prefer to measure and adjust the actual current flow, eliminating the effects of component drift. This requires 'opening' the bias circuit at the test point by unsoldering and lifting one end of the resistor, and clipping your meter (set to DCMA) probes in series with the lifted connection. The older 9440s were done this way, as can others, with some simple ohm's law calculation.

Biasing has a drastic effect on an amps performance (and longevity). Class A amps can operate with Class AB or even B characteristics when improperly biased (for awhile), though proper design doesn't allow such drastic adjustment. It's not unusual to find an amp some tweaker has set for higher-than-design output, at the expense of greater distortion and shortened component life, hence my previous recommendation to perform these adjustments on a newly-acquired, used amp. Also, until these adjustments are made, many of the test-point voltages will be way out of line, and could cause one to start replacing good components.


If your amp gets hot enough for the 'protection' indicators to light (around 140F, I think) you may have other problems. Your fan should also switch from lo to hi at some point of the warmup curve. Also keep "on" time to a minimum when you have the cover off, as the air path changes. Since the OPT collecters are in direct contact with the heat sink(s), (no mica insulators), heat transfer is so rapid, finding a single "hot" OPT is nearly impossible unless you've got a non-contact, IR thermometer.

Somewhere in one of my many "piles", I know I've got a hardcopy of the procedure for the 9444, but I doubt it for the 9446, since I've never had one of them. I've committed the immediate future to sorting those piles, as I can't find the new PS caps I bought for the 9440 on the bench, and they're almost the size of beer cans! I'll contact you if I can find anything that might help.

[cclark65]

Thanks for the info, I will check it out hopefully this next week or so. I do have 3 of the 9440 amps, so I know what you mean by the huge power caps! I'll check the values of them, they throw large caps away here at work sometimes for junk, and I usually grab them, you never know when one might come in handy!

[cclark65]

I do have access to a non-contact thermometer...we have one at work I can borrow. That would make it a bit easier to find the hot spots, either that or take it to work and use the infrared video stuff to look at the entire amp and see the hot spots.

I'll play around with the biasing first though as you described, hopefully that will take care of the problem.