• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Harman Kardon F-3/FA-3000X/XAM-3B

All,

I did some tests on this receiver and noticed that the voltages are lower than according to the schematic.

Somewhere in the past, it seems that a dropping resistor R106 was changed from 220 ohms to 1 kohm. The voltage drops from 340v to 278v instead of 340v to 325v (per the schematic).

Esteemed diyers... Is there a reason that someone might intentionally reduce the voltage?

1. Compensate for higher AC line voltage?
2. Increase the life of the tubes?
3. Support 6V6GT tubes vs. the original 7408 tubes?
4. Support a "fixed bias" modification for the power tubes (see my first post) vs. the original Cheap Charlie method?

Thank you in advance for your insights.

ikonw8
 
Upon further checking, the dc voltage at the grid is behaving differently on this tube. Once the amp is warm, the other three tubes measure ~10mv DC voltage. At this tube I measure ~100 - ~200mv of DC and it keeps rising. Strange. I have ordered a replacement set of 6v6GT tubes, and will probably put the 1kohm resistor back in place to drop the voltage for the time being.
 
Referencing the schematic, look at C49, C50, C67, and C68 (0.05uF 400V).
Those are the interstage coupling capacitors from the split load phase inverters (V9) to the grids of the four 6V6s. One of those caps might be leaky. If they're paper caps then they're likely to be leaky. Probably good preventive maintenance to replace those with decent quality poly film caps, 0.047uF 400V or 630V.
 
I know this may be obvious, and if it is I apologize in advance, but...

Have you swapped the 6V6 tubes around into different positions to see if the problem is always with whichever 6V6 is plugged into that tube socket, or does the problem 'follow the tube'?
 
All 6V6 tubes are sharing the same self bias circuit (3 12A_7 tube filaments), the bottom one puts the grids @ ~ +12V, the cathodes @ ~ 36V, for ~ 24V self bias.

The maximum spec for grid resistor, Rg, for self bias is 500k.
Rg is 470k in your circuit.
When all 4 tubes share the single self bias resistance, it is a little closer to fixed bias, because each tube contributes more or less current.
A really strong tube will tend to hog the current, get hot, and . . .

The maximum Rg for fixed bias is 100k.

You really need very good 6V6, preferably a fairly well matched Quad of them.

As rongon suggested, does the red-plate follow one tube, or is that socket the only one with the red plate when tubes are moved.
Are the cathode balance pots centered?
 
Thanks all for the feedback. This amps circuit has been modified to delete the 12A_7 heater biasing... in favour of a cathode bypass on each pair of tubes. It uses a 32uf + 240 ohm 5W resistor in parallel from the wiper of the bias variable to ground.

The red plating has stopped, but I am not sure why. Since it stopped red-plating, I have not tried swapping tubes around.

I have since lowered the voltage by replacing the 220 Ohm dropping resistor with a 1000 ohm resistor. (no replating of course, but lower watts).

I have matched quad set of tubes on the way...
 
Last edited:
Another question...

This amplifier had .033uf Sprague green drop capacitors paralleled across the main power supply capacitors. They are not in the schematic.

Any suggestions why this modification would have been done? Less RF noise? I have removed them and I don't notice any difference (yet).

Thanks all for taking the time to respond...
 
All...

The stereo indicator on this receiver will not light even when on a strong stereo signal. According to the schematic the indicator is a neon lamp ACOM 7337. No results for that part number.

20210610_200436.jpg

I tried a couple different types of new neon lamps, NE-2. One had a firing voltage of 90V DC, the other 50-60 VDC. No luck.

I measure 30VDC arcoss the neon lamp, which is likely why the new lamps are not firing. I also measure .3mA, which is not enough to drive a red LED (about 20ma or better);

I recently aligned the tuner, and it provides an excellent FM stereo signal with plenty of separation.

It is entirely possible that the circuit should drive higher than 30V, but I wouldn't drive anywhere near what I need for 50 or even 90vdc.

I would like to either find a neon lamp that fires at a lower voltage, or modify the circuit to use an LED instead.

Of course, I could just leave it alone since it sounds good anyway.

Has anyone replaced these type of indiciators? Find a source for them?
How about modifying to use an LED?

thanks,

View attachment H-K_FA3000X_sch1.pdf
 
Last edited:
As I said in post #24 it ends up a 19KHz signal, as you can see one end is connected to T1=GLCOM6167.

VERY little current is used in those neons because in this case its not just a striking voltage required to activate it but a superimposed audio signal --don't confuse this circuit with electrical mains derived indicators .

FYI - radio hams made use of neons to "tune up " their HF transmitters with NO physical connections just the RF field.
 
Thanks everyone for the input!

I checked C45, and it has been replaced sometime in the past with a .068uf 600V Sprague. Measures 66.62uf at .1% Vloss.

Via antenna, at 98Mhz, 19khz pilot on, no L or R signal. TRIO SM-301

At T1
206.4 VDC
25 V P-P AC
38khz (sinusoidal wave)

At Pin 3 of 6M11
175.7 VDC
.26 Vp-p
19 Khz (looks like a sawtooth wave)

At Pin 3 with audio signal
L+R .8vpp
R .4
L .7
L-R .45

The resistors check out too
R45 = 66.4 k, R44 = 45.4k
 
Last edited: