Tweet
Axis Sound NH-100 - baffling problem
HELP!!! I'm on the verge of losing my marbles here.
First, the background:
For the last couple of weeks I've been working on a friend's 1969 Axis Sound NH100 amplifier: giving it a complete overhaul, as it hasn't been used for over 10 years (and the last time he did use it, it toasted the screen grid resistor on one of the output valves, causing a small fire!
).
There's almost no technical info to be found about these amps. The only thing I've seen is a comment from a Plexi-Palace forum member that they were "Simms-Watts clones". In the case of this particular example, though, even that description doesn't hold true, as it appears to have been modified a fair bit over the years. For one thing, the Simms-Watts amps used an ultra-linear OPT, whereas this one has a standard type.
Here's a link to a schematic I've drawn which shows the actual amp that's lying on my bench now:
NH_100 Schematic
To download the file, right-click the link and choose 'Save target as...' (Windows) or Ctrl-click it and select 'Download linked file' (Mac OSX).
I've been through the beast from top to bottom and got it working. I've also installed a few little refinements (such as adjustable bias for each half of the push-pull). At the point of clipping it's producing a pretty healthy 140W.
Now for The Problem:
As you can see from the schematic, the EL34s are run with 534V on the anodes. According to all the usual sources (such as Weber's bias calculator etc), in order to set it to a nice 70% dissipation for typical Class AB operation, the quiescent current per valve should be around 33mA.
However, when this amp is running 'normally', the QC figure that shows up on my Fluke meter is scarily high.
I mean, how does 75mA per valve grab you?
"Back off the bias!" I hear you cry - but if I do that until the quiescent current is down to the 33mA level, I end up with the output stage in extremely cold Class B, with little output power and lots of dog-leg crossover distortion.
The oddest part of all is that if I adjust the output stage to run at its 'normal' 75mA per valve, it behaves as if it's perfectly happy! Even when it's being driven to the point of clipping, there's no sign of stress to the output valves: no redplating...nothing.
At one point, I wondered if my Fluke 25 meter was giving false readings, but I've checked it against another meter and they both tell the same story. Those valves really are passing that much current - a situation which, according to the textbooks, shouldn't be possible. I mean, just do the maths: these babies are running at 40W plate dissipation even before you apply any signal!
FWIW, I checked each of the output valves on my AVO CT160 tester before starting work on the amp, and they're all healthy.
So that's it, folks. I'm stumped. Have a good hard look at that schematic, will you? See anything suspicious? Any little detail/blindingly obvious fault that I've missed? For heaven's sake, someone: throw me some ideas before I go nuts..!
First, the background:
For the last couple of weeks I've been working on a friend's 1969 Axis Sound NH100 amplifier: giving it a complete overhaul, as it hasn't been used for over 10 years (and the last time he did use it, it toasted the screen grid resistor on one of the output valves, causing a small fire!
).There's almost no technical info to be found about these amps. The only thing I've seen is a comment from a Plexi-Palace forum member that they were "Simms-Watts clones". In the case of this particular example, though, even that description doesn't hold true, as it appears to have been modified a fair bit over the years. For one thing, the Simms-Watts amps used an ultra-linear OPT, whereas this one has a standard type.
Here's a link to a schematic I've drawn which shows the actual amp that's lying on my bench now:
NH_100 Schematic
To download the file, right-click the link and choose 'Save target as...' (Windows) or Ctrl-click it and select 'Download linked file' (Mac OSX).
I've been through the beast from top to bottom and got it working. I've also installed a few little refinements (such as adjustable bias for each half of the push-pull). At the point of clipping it's producing a pretty healthy 140W.
Now for The Problem:
As you can see from the schematic, the EL34s are run with 534V on the anodes. According to all the usual sources (such as Weber's bias calculator etc), in order to set it to a nice 70% dissipation for typical Class AB operation, the quiescent current per valve should be around 33mA.
However, when this amp is running 'normally', the QC figure that shows up on my Fluke meter is scarily high.
I mean, how does 75mA per valve grab you?
"Back off the bias!" I hear you cry - but if I do that until the quiescent current is down to the 33mA level, I end up with the output stage in extremely cold Class B, with little output power and lots of dog-leg crossover distortion.
The oddest part of all is that if I adjust the output stage to run at its 'normal' 75mA per valve, it behaves as if it's perfectly happy! Even when it's being driven to the point of clipping, there's no sign of stress to the output valves: no redplating...nothing.
At one point, I wondered if my Fluke 25 meter was giving false readings, but I've checked it against another meter and they both tell the same story. Those valves really are passing that much current - a situation which, according to the textbooks, shouldn't be possible. I mean, just do the maths: these babies are running at 40W plate dissipation even before you apply any signal!
FWIW, I checked each of the output valves on my AVO CT160 tester before starting work on the amp, and they're all healthy.
So that's it, folks. I'm stumped. Have a good hard look at that schematic, will you? See anything suspicious? Any little detail/blindingly obvious fault that I've missed? For heaven's sake, someone: throw me some ideas before I go nuts..!
Comment