Home built amp shutting down with loud hum after 10 minutes

[Clinkous]

Hi everyone,

I built an ax84 high octane amp that performs well with almost no noise for about 10-15 minutes and then shuts down with a full volume hum. It does this whether I'm playing with it maxed out or with no input and all controls zeroed.

2 12ax7's and either a 6v6 or an el84 power tube. Pulling the preamp tubes has no effect. Pulling the power tube reduces the volume of the hum considerably but it is still present with no power tube. I have checked all ground points to the input chassis ground and all less than .1 ohm. Have resoldered the OT and PT leads but I don't SEE any bad solder joints. Checked all connections starting at the AC plug, pilot light, standby switch and connections to the board.

I am using a variable external bias pot and have the 6v6 at 46ma and the el34 at 40 ms with a plate voltage of 255V.

I know I'm missing something and it might be obvious to someone else. Any help would be appreciated!

Thanks,
Chris

[Stevem]

And does that plate voltage of 255 change when the amp starts to hum?

[gui_tarzan]

6V6 OR EL84? The schematic only shows an EL84.

[Clinkous]

NO. The plate voltages dropped to around 197 and checking with the meter changed the hum. Had higher voltages at the filter caps and checking them affected the hum also.

I added the optional octal socket to use either 6v6 or el84 singly.

[Clinkous]

I checked again and had 257 on the OT lead on the el84 socket (the tube I'm using). The hum got quieter the higher the voltage reading went.

[strelok]

What kind of bias pot are you using? If its too low wattage it may be getting hot and shifting the bias.

[Clinkous]

I just disconnected the OT primary leads and am getting 129 ohms across the primary, shouldn't this be the stated 4500-10000 ohm value? After resoldering several joints it got worse and worse and now it simply hums on startup. Could this be the OT slowly dying or is that reading normal.

The bias pot came with the kit it's a 2w 1k. The tubes never red plated or anything. It never slowly faded to a hum, it just would play like a champ and it would go to a full volume hum like a switch was flipped.

[romberg]

Pulling the power tube reduces the volume of the hum considerably but it is still present with no power tube.

I would then guess that you are getting 120hz buzz in straight from the rectifier to to PT. If It were mine, I'd suspect the first two filter caps C1 and or C2.

Mike

[Leo_Gnardo]

I just disconnected the OT primary leads and am getting 129 ohms across the primary, shouldn't this be the stated 4500-10000 ohm value?

You're measuring the DC resistance of the primary with your ohmmeter. The value you read isn't unusual. Those k-ohm specs include the induction factor, remember those transformer windings are made of thousands of turns of wire and the inductance is further increased by the steel core. Nothing to worry about here.

[Stevem]

With the tubes in and the amp on set your meter for low ac and measure the ripple voltage at the first filter nide, does it rise up when the issue kicks in?
If so you have a bad filter!
If that the case unsolder your second filter and jump that to the first one as long as it can take the voltage, does the issue stop, if so that proves out once again that the filter is bad.

[Clinkous]

Thanks guys,

I removed the first filter cap to check it and just re-clean and solder everything again. Cap checked out and amp was dead quiet again(except for guitar!) and I left it alone with no input. Lasted about five minutes and swithched to loud hum for one second and then went away and is just generally AC noisy like bad wiring...go figure.

I realized I've been doing all this with recently bought lead free solder (bought unknowingly... thought it was 60/40 but it's Sn99.3/Cu.7 and I wonder if this just won't work for the tube amp parameters. Any thoughts, because I've worked on my older amps and built pedals and done general guitar repairs and have never had a problem with bad solder joints.

Going to go ahead and re-solder with 60/40 and see. I'll report back and I REALLY appreciate the help!

Thanks again,
Chris

[John_P_WI]

On my phone so this will be brief, look closely at the bias supply after the pot. Sounds like the pot is opening and you are losing bias.

[dorrisant]

Can you post some pictures of this build? Closeups of the filters and bias supply would be helpful.

[R.G.]

It would make a great deal of sense to suspect the rectifier of overheating and then semi-shorting.

The time delay is telling you one piece of information. Time delays in electronics stuff tend to be limited to seconds or a small fraction of a minute if you work hard at it, because of the range of components - we usually don't get to work with farad- or tens-of-farads caps, nor giga-ohm resistors much. So things that happen over minutes tend to be heating related. Something is overheating and going bad.

The advice here is good: overheating caps, overheating bias pot, etc. Removing the tubes and getting a similar hum puts this outside the normal range of bad tubes, although you did the right thing in checking tubes first. Always suspect tubes first! There's a reason they're in sockets.

Overheating caps could do it to you, with the cap going open after a while. So could a bias pot, and so could a bias power supply. However, these would tend to vanish if you pulled the power tubes.

Hum while no tubes are present tends to say a power supply issue is still getting some AC voltage to the OT. That could be BIG ripple, as with an open cap, or it could be a bum rectifier not blocking well.

There is a sneaky troubleshooting trick I use sometimes when I suspect overheating. I get out my $30 Harbor Freight infrared pistol thermometer. It reads emitted infrared directly. It's the same test I used to do on low voltage stuff by running my hands over the circuit, but I hate to use hands-on in a live tube circuit - you get false-positives when you touch high voltages.

Another step that is really insighful is to scope the high voltage when the hum happens. Of course, you have to have an oscilloscope to do this.

[Clinkous]

Thanks everyone,

I haven't checked for ripple yet, just resoldered joints with known good stuff. Filter caps check out disconnected with a meter for what it's worth.

Amp did this before I installed extra parallel octal socket and bias pot so it did it with fixed bias on the el84. I just figured a bad solder joint and cleaned them all up when I added the other socket. I have resoldered the bias supply and pot anyway.

The amp played quieter than it has a right to for about ten minutes at low volume and then BAM the hum turned on. Full volume with standby on and 1/4 volume with standby off.

I want to check for the ripple Stevem mentioned and the rectifier diodes per R.G.

The filters are Nichicon and all parts are new. Amp has less than 5 hours on it.

I'm posting a pic (if I can figure it out) and then my family has informed me that we are NOT working on amps all day!

I'm learning a lot so thank you all for putting up with me!