GM-70 All done

[astouffer]

time.

 

[gaborbela]

Congratulation!
You have (built) a very nice amp!
Greetings G.

 

[djn]

Is it normal to have a resistor in line with a coupling cap going to the output tube? I am planning on using a PIO 1.5uf with no resistor. Comments?

 

[kevinkr]

Is it normal to have a resistor in line with a coupling cap going to the output tube? I am planning on using a PIO 1.5uf with no resistor. Comments?

That's a grid stopper resistor, and yes it is good practice with the GM70 - should be mounted right at the socket. (I used 1K grid stoppers in my IT coupled GM70 amps)

 

[MelB]

Alrighty, back with some Visual Analyzer results. I bought a sound card interface from Pete Millett. This is into a 6.8 ohm 50 watt resistor. Output transformer 5000 ohm primary using 4 ohm tap.
Output THD
5 watts 0.97
10 watts 1.38
15 watts 1.75
20 watts 2.07
25 watts 2.64
30 watts 3.2%
And attached 1000Hz at 15 watts image. No third harmonic???

 

[cotdt]

No 3rd harmonic... there could be cancellation effects with the driver tube. Try a different driver tube?

 

[MelB]

Nope simpler than that. Visual analyzer is not doing the spectrum capture properly. Doh! What I see on the screen and what it captures are not even close. Best look into it.

 

[MelB]

OK...actual screen capture. Visual Analyzer not terribly good at it.

 

[cotdt]

That looks perfect. Exactly what one would want!

 

[kevinkr]

Yes, nice plot..(Each harmonic significantly lower than the preceding one.)

 

[popilin]

Wonderful !

Very nice looking, high quality building, clever circuits, outstanding measurements.

All that one wants for an amplifier.

BTW. I am pleased to see you used DC on all heaters.

Well done !

 

[the_manta]

Like the finish. Especially the hammer finish on the caps.
Spectrum is typical triodish - nice.

 

[MelB]

Thanks guys! The culmination of years of thinking and parts gathering.

 

[deafbykhorns]

I know this is an old post but wondering if you could share a copy of your pic file.
I'm doing some experimenting and 1st time pic user

 

[MelB]

20K grid choke vs. 90K resistor. I swapped between using a 20k grid choke or a 90K resistor to bias the GM70 just to see what difference it made. This is a 10khz square wave.

The blue trace is probe on the GM70 grid. (Ooops was a 100x probe and scope set to 10x so voltage/division should be 10 not 1)

Yellow trace is the 4 ohm output of output transformer with 6.8 ohm load.

 

[wrenchone]

My only beef with the layout is the big hunk of magnetics (output XFMR?) that looks as if it's terribly close to the GM70. It certainly won't freeze to death...

 

[MelB]

Yeah the picture is actually very deceiving. Here is another angle..it's a long ways away actually.

 

[wrenchone]

Yes, much better from that view point, though it still won't freeze to death.

 

[celestar32]

Looks great. Good job and nice built there.

 

[wrenchone]

My inclination would be to to house the GM70 in its own pyrex/polycarbonate tubular shroud and cool it with pressurized airflow from a small centrifugal blower located below decks. This is easier to implement for the GM70, as there's no plate cap and associated lead to get in the way of aesthetic design. The heat gets blown up, up, and away, and doesn't trouble adjacent components.
Understand - this is just my senile rumblings - if the present configuration works, more power to it. Having said that, though, this is how I'm going to work with an 813-based monoblock I have in mind. In addition to the cooling benefits, the cooling shroud will keep inquisitive fingers away from the plate cap. For those folks contemplating amps with tubes that require a heat dissipating (but hideously dangerous) aluminum plate cap (100TH, 100TL, etc.), the cooling shroud offers a approach that makes engineering sense, but salves the conscience as well...