Ga-5 Build Schedule 40

[shaun]

Hi All,
built this amp recently and housed it in parts from an old Zenith Console. Also got the spkr (5 ohm reading, so somewhere around an 7-8 ohm spkr), chassis, and PT from the Zenith. The chassis and PT were from the reverb unit in the console - I was surprised to find a reverb unit in a stereo, but there it was.

I'm new to this sort of fun, but here's what I think resulted from the experiment; I welcome feedback.

1: The current draw of the three tubes is almost half an amp higher than was intended for the PT, which only had to contend with a 1.6 amp draw in its previous life. Now, 2 amps is being pulled through it. The PT gets hotter than I'd like it - 165 f. Surprise, surprise.
2: The PT and EZ81 rectifier combination put out a little too much B+ at node one (382v), resulting in high PVs all around.
4: The high B+ resulted in early break-up of signal, so I replaced the 250 ohm cathode resistor on the EL84 with a 470 ohm, which helped a great deal in terms of tone, but bumped up the EL84 pv to 377v. (I believe this happens when increasing the cathode resistor - voltage goes up, current goes down).
5: I matched a mojo771 OT to the EL84 and speaker. SE EL84 output imp = 5000. The mojo771 handles 7k ohms into either a 4 or 8 ohm speaker (I'm using the 8 ohm tap), so that seems to be in the ballpark. Or not. My knowledge is limited, but I'm guessing the high voltage on the output tube increases dissipation and results in an output impedance above 5000 ohms?
7: Plate dissipation is also where my knowledge falters. I used Jim Jones' online calculator chart and came up with 39mA plate voltage and 13w dissipation. I felt lucky to get that close. I got that result by changing the value of the cathode resistor of the EL84. I couldn't get the plate dissipation lower by increasing the cath. res. any further because other things seemed to go more out of balance. As I say, my knowledge falters.
8: I have considered options for a different PT to lower the B+ and to handle the 2A draw, but the chassis is kinda tiny and would require a complete rebuild, so I may hold off until the poor wee thing dies on me.

The upshot is that the amp sounds friggin great - I'm thrilled with it. And I can fry eggs at the same time I'm playing:).

One cool lesson I learned is that there is a beautiful balance gong on in amplifiers, and the builder's job is to find the perfect match for a perfect interplay btwn the various elements of the build. Hopefully, I'll be able to do that one day, once I learn how to obtain perfect plate dissipation. And nirvana.

[shooter]

once I learn how to obtain perfect plate dissipation

You already have, it's a series circuit DC (no signal) so ohm says current goes up, voltage comes down, R goes up, current comes down, voltage goes up.

the intuitive part, adding signal, then everything becomes dynamic and you gotta be real anal to try and figure it out In real time 
better to play and wait for smoke, so you have something to look forward to!

[PRR]

Voltages are very high. I bet it has a buzz also. Add an R-C dropping filter and then re-bias the EL84 to something near 300V 40mA.

[shaun]

Wow, very nice info, PRR. Thank you. I'll give it a try.
The amp does have buzz, but surprisingly not bad.

[ALBATROS1234]

I don't know if this is possible, perhaps prr or another pro can chime in. Can you make a voltage divider and shunt part of the plate feed to ground? I don't know if this will work in an amp situation but theoretically you should be able to do this. Probably need a 10 watt resistor because my guess is this would burn up a small watt resistor. I was curious about the same thing a while back. I ended up just using a 6L6 in order to handle the voltage/ dissipation.

[jjasilli]

once I learn how to obtain perfect plate dissipation

You already have, it's a series circuit DC (no signal) so ohm says current goes up, voltage comes down, R goes up, current comes down, voltage goes up.

the intuitive part, adding signal, then everything becomes dynamic and you gotta be real anal to try and figure it out In real time 
better to play and wait for smoke, so you have something to look forward to!

I don't agree; and suggest that you measure your current.  A 1 Ohm bias sense resistor is a good method. 


Shooter's comment is applicable to a purely resistive (passive) circuit.  But tubes are reactive devices; not merely passive.  As stated, the larger cathode resistor pushes up plate voltage.  The higher attraction of the increased +charge on the plate tends to draw more current through the larger cathode resistor, than might be expected in a passive circuit.  As plate supply voltage increases, at some point cathode resistance becomes ineffective to properly bias the tube; so fixed bias is used.


EDIT:  N.B.:  If voltage plate voltage rises, and current drops commensurately, plate dissipation in Watts remains the same. 


Also, so long as your plate diss is OK for your tube type, I don't think you have a problem.  I.e., 165 degrees F is not too hot for metal.  Bargain basement vintage amps often ran their PT's real hot.

[PRR]

...Can you make a voltage divider and shunt part of the plate feed to ground?...

The power tube is its own shunt element. You need a series element. Yet, it may be high-Watt. I drew this up already.