Can we make this amp sound better?

GordonW

GordonW

Speakerfixer
Subscriber
Well... I finally have stable operation on my Heathkit AA-100. 465v on the plates... a bit higher than stock, but I'll live with it (Scott 208s live with 477 on the plates with a very similar circuit)... the output tubes are averaging around 30ma of plate current each (I modded the board and put in 10 ohm cathode "bias sense" resistors on each output tube), too... a bit low, but in the 'target range'.

Now... I am noticing that this amp seems to start clipping/distorting a bit earlier than the Scott LK72 or the Fisher 500 that I worked on recently. So... it's time to find out why. As such, I'm coming to the collective experience here, for suggestions... for once, you guys tell ME what to do... :D

My thoughts are probably pre-amp related... I don't see anything amiss with the power amp section, at a first glance. But, I have my doubts about the preamp. Apparently, others have come to the same supposition... especially THIS guy:

http://home.netcarrier.com/~rstevens/aa-100.html

However... if you read through his whole page, I'm wondering if a LOT of the problems he had, were related to the UTTERLY STUPID bias supply resistors (and the bias being all over the map as a result)... I have updated mine with 220K resistors instead of the insane 470Ks it came with, supplying negative bias voltage to each output tube grid... bias is stable and correct.

That said... I do wonder about some of the choices. For example... 100K anode resistors on 12AX7s in the preamp. Sure, that's OK in a Fender guitar amp (they do it all the time in those), but isn't that a bit low for a hi-fi amp? I'm wondering if the 12AX7s are just loaded too hard? I do have, apparently, the ability to bump up the B+ to the preamp tubes a bit... so, if larger anode resistors (and larger cathode resistors to restore the gain a bit) might help, I can do that...

BTW: Here's a link to the schematic for this amp... it's quite big, so I won't just post it here:

http://web.archive.org/web/20040913112841/www.alligator.freeserve.co.uk/diagrams/aa-100_sm.gif

Anyone care to make suggestions? I would prefer to NOT change out tubes. The 7247s used in the mod from the first link are rather uncommon... and the amp has Mullard 12AX7s in it now! I'm sure we can figure out SOME way to keep 'em in there?

EDIT: BTW: I am not totally opposed to the idea of yanking that balance/separation pot out, and replacing it with a NORMAL balance pot... having 100K in series with the grid of a 12AX7? Maybe not the greatest idea?

Regards,
Gordon.
 
Register to hide this ad
Do you have any trouble with the B+ climbing over 500 untill the tubes start conducting? I asume you're using 500v filter caps.
 
jaymanaa said:
Do you have any trouble with the B+ climbing over 500 untill the tubes start conducting? I asume you're using 500v filter caps.
Click to expand...

Only very briefly, and only a small amount. Maybe 510 volts peak, and only for a few seconds. Having a 5AR4 (slow start) helps...

Regards,
Gordon.
 
OK. It's probably not the end-all for this amp, but this is officially a "DUH" moment.

Note that the volume control is AFTER the preamp gain stages. And that there are gain trimmer controls on every input.

Well... I turned down the gain trimmers, from all-the-way-up, down to about 2/3s of the way up. WAY less distortion.

That said, this amp needs a volume control, akin to that in a Marantz 7... DUAL-gang per channel, one section BEFORE the gain stage, one section AFTER the gain stage. Cuts down the input AND output levels. Still minimizes hiss and noise (due to the one at the output) while protecting the preamp from clipping (the one at the input)...

But, as long as you trim it down to where the source doesn't clip it... it sounds quite a bit better...

That does make me wonder if biasing the front end tubes with bigger cathode resistors (less current, more bias voltage) might help...

Regards,
Gordon.


Regards,
Gordon.
 
I will give you a quick and dirty response. I have several Scott amplifiers, mostly the LK-72 variety. I agree, the Heathkit will not produce nearly as nice a sound as the Scott. I have a Sherwood. It is about the same. Can't say that I find any of the 7591 amps that perform as well as the Scott. Don't know the problem but I like my Scott...
 
Anybody know where I can get a quad-gang, 1M pot? :scratch2:

Or even better- quad-gang, two sections 1M, the other two sections 250K or 200K? Wow, I'll be shocked stunned and amazed if THAT exists! :D

Regards,
Gordon.
 
UPDATE: May not NEED to change the pot.

I thought "hey, with that low of an anode resistor... why not try a 12AT7?"

Did it... DOUBLED the bias voltage on the input stages. 6 dB more headroom. Same gain. Definitely more "punchy" and "ballsy" sounding...

Again, maybe not the end-all... but BOY, did it seem to work...

EDIT: Went and modeled the 12AT7 vs. 12AX7 in TubeCad, given the anode and cathode resistor values and B+ voltage. The gain only goes down about 1/2 dB with the 12AT7, in theory, according to the program... from about 26dB for the first stage, down to 25.5 dB. Not even enough to notice...

Regards,
Gordon.
 
After listening, I think this thing may have actually surpassed the Scott LK72/299D a bit. Just more "luscious" sounding.

Not up to the level of the Sherwood 5000-II yet. I'd put it at about par with the Fisher 500C now, though...

Of course, that's just my opinion... and I have some pretty specific things I like...

Regards,
Gordon.
 
Sounds like you've cracked the code on that one Gordon. :thmbsp:
 
Glad that you guys are having fun! I guess that I am biased (pun intended!). I still like my LK-72/299C,D amps the best. Of course, the Sansui 1000A ranks way, way up there. My Sherwood, to me, doesn't quite sound as good.

Guess that is what makes the world go around...
 
u might tinker with a dissimilar like the 12dw7 there used in the eico stereo 40's .
A bit hard to find nos but half 12ax7 and half 12au7 has possibilities.
 
If you look in the archives of the Vintage Asylum forum, the AA-100 is often defined a "dog". So it seems not to be the best sounding amp around.
Biggest problem seems to be a very poor printed circut board that doesn't age well. The OPT's seems to be very good instead, and many people buy the AA-100 only to use the irons for a better DIY project.
Maybe this link can help for improving the amp (maybe you already know it): http://home.netcarrier.com/~rstevens/aa-100.html
 
This might have been the same sort of fight engineering had with marketing on the eico st40. The engineers didn't want the loudness switch
marketing pushed for it and in the end the techs built it not to function properly.
 
You must log in or register to reply here.
Back
Top Bottom