Yaqin MC-5881A amplifier improvements

[DanDini]

Thanks for all the advice.

I've done some measurements, but please bear with me if they're not what you were after.

The heater voltage is between 7.35 and 7.42Vac.

There's 35.4 Vdc on one side of the 500ohm 6 watt wirewound and 8mv on the other side, with respect to ground.

Thoughts?

Dan

[Salas]

Bingo! Both suspicions confirmed. The heater voltage is devouring your amp and the bias is too high. To bring back the heaters to normal, you are going to get 2X 0.33 Ohm 10W resistors and wire them in parallel.
Leave a little space between them. Then insert the combo resistor in the start of the heater wire feed as it comes from the mains transformer, in series with one phase (one lifted wire, resistor in series). Take care so the resistor combo is off the PCB and has all possible space around it without touching cabling. This measure is going to relax the heaters to nominal 6.3V AC or about. Measure to confirm. Then see again if the 35.4V DC drop on the 500Ohm combos on PCB has dropped due to normal heating, and how much is it. As it is now, it indicates 70mA per tube. So to recap, drop the heaters to normal, feel if it has cooled acceptably, check the DC voltage drop, and post findings. We will ***** again.

Regards

[sorenj07]

It actually looks like there might be room for a choke or two if you get creative. They're between $10 and 20, and Hammond lists the dimensions. I'd pick one or two up and see if you can fit them in - it'd really really be worth it. Even the smaller ones should do something. check out the 156R at this site:

http://www.hammondmfg.com/153.htm

That's very small. Surely it could work? $10 a pop from Angela.

[DanDini]

I've got some 0.33 ohm *5w* resistors in 'stock' can I test with those?

I've also got some 0.47ohm 5w...

Dan

[DanDini]

OK, I tested with those and although they do get pretty hot, they're doing ok for testing.

results are: 6.28-6.42Vac on the heaters

35.35Vdc to 6.7mV drop on the 500ohm resistors.

The hum is still there

Dan

[DanDini]

I went for 3 x 0.47ohm to give me as much disapation as possible. Have ordered some others for a permanent fix. Do you think the chinese output tubes could be making the noise?

Dan

[Salas]

Hello. Very good result on the heaters. Calculation proved spot on. They get so hot (your 5W resistors) because they get to dissipate almost 7W as a whole. That is why I specified 2X10W. In general, for no long term problems, a resistor must have a rating 3X the actual dissipation and over. So 20W total will be OK.

Your output tubes still consume 70mA each by the reading you gave. How is the heat now? Did it back off a tad? Will let you know how to bias lower if you tell me that the heat is still unacceptable.

No, the valves alone cannot hum. DHT valves can be a PITA to run humless, not 5881s. And there are dead quite SE DHT amps out there too. There are ways to tackle everything. Noise is related to PSU, layout, and grounding issues mainly. Your amp has 40dB gain that pushes such issues also. Tell me, is there a cable from the PCB attached to chassis? Is there a cable from the mains earth 3rd prong attached to chassis? What is the sensitivity of your loudspeakers?

[DanDini]

Hello. Very good result on the heaters. Calculation proved spot on. They get so hot (your 5W resistors) because they get to dissipate almost 7W as a whole. That is why I specified 2X10W. In general, for no long term problems, a resistor must have a rating 3X the actual dissipation and over. So 20W total will be OK.

Yep - sorted - thanks!

Your output tubes still consume 70mA each by the reading you gave. How is the heat now? Did it back off a tad? Will let you know how to bias lower if you tell me that the heat is still unacceptable.

They are cooler but I think that is still too high. Do you have any advice on lowering it overall (other than running off a variac!)?

Tell me, is there a cable from the PCB attached to chassis? Is there a cable from the mains earth 3rd prong attached to chassis? What is the sensitivity of your loudspeakers?

The earth is wired directly to chassis from the IEC plug. It is also wired in with a load of other earths (eg gnd out for speakers).

I don't know the sensitivity of the current speakers (goodmans double maxim) but I have tried it on various ones including the not sensitive (insensitive?) Kef Cresta 1 and they all hum.

Thanks,

Dan

[Salas]

Lift one 500 Ohm from each pair. This will get the bias down from 70ma to 40 mA. Right now each tube dissipates 26W (overspec). You will measure around 40VDC on each 500Ohm left in circuit, and feel the heat backing off seriously enough. Do that and let me know how many cables go from PCB to chassis and from where off the circuit gnd points.

[DanDini]

Lift one 500 Ohm from each pair. This will get the bias down from 70ma to 40 mA. Right now each tube dissipates 26W (overspec). You will measure around 40VDC on each 500Ohm left in circuit, and feel the heat backing off seriously enough.

OK, I've done that and get 41.7Vdc down to 3.9mV on one side and 40.8Vdc down to 4.1mV on the other.

Any idea where this voltage inconsistency comes from as the transformers appear to deliver an even supply. Is it enough to be an issue?

let me know how many cables go from PCB to chassis and from where off the circuit gnd points.

Ummm... Well none go from PCB to chassis directly. They are linked at the point the mains comes in. They are all twisted together in a big lump on the PCB. There appears to be another unused take off on the opposite (volume side) of the PCB.

If you look at the picture here you can see the twisted gnd points behind the two electrolytic caps in the middle and the unused gnd point just in front of the plastic socket/plug arrangent at the top middle.

Thanks,

Dan