Amplifier Hum due to Vacuum Tube failure?? - Electronic Circuits Projects Diagrams Free

Rustydawg · 26th February 2006, 06:24 PM

Greetings Electrotechies,

I'm brand new to the site and I've got a question / scenario to throw out there. I've had my hands into elecronics for a long time but I'm not terribly familiar with vacuum tubes and their various failure modes.

I have been restoring the power supply & power amplifier chassis from an old console-style Hi-Fi system (circa 1964). The system consists of a tuner/preamplifier up top and the power amplifier/supply in the bottom. Due to it's simplicity, I started with the power chassis.

I have replaced the majority of the small value capacitors except for those that tested fine. I have also added a bunch of filter capacitors on the B+ to compliment the lowly 10uF and 40uF in the original design.

The power amplifier uses a 12AX7 (double triode) amplifying both L&R channels, feeding two separate single-tube stages in class A.

There is a great deal of hum at the speaker output (over a volt p-p) on both channels. If I pull the 12AX7 tube, the noise clears up.

My question is this: Can a failing tube actually introduce hum? If so, where can it be getting the 60 Hz signal from - the filament?

All theories are welcome... Tube guys get on board!

**Nigel Goodwin** · 26th February 2006, 07:34 PM

Quote:

Originally Posted by Rustydawg

Greetings Electrotechies,

I'm brand new to the site and I've got a question / scenario to throw out there. I've had my hands into elecronics for a long time but I'm not terribly familiar with vacuum tubes and their various failure modes.

I have been restoring the power supply & power amplifier chassis from an old console-style Hi-Fi system (circa 1964). The system consists of a tuner/preamplifier up top and the power amplifier/supply in the bottom. Due to it's simplicity, I started with the power chassis.

I have replaced the majority of the small value capacitors except for those that tested fine. I have also added a bunch of filter capacitors on the B+ to compliment the lowly 10uF and 40uF in the original design.

The power amplifier uses a 12AX7 (double triode) amplifying both L&R channels, feeding two separate single-tube stages in class A.

There is a great deal of hum at the speaker output (over a volt p-p) on both channels. If I pull the 12AX7 tube, the noise clears up.

My question is this: Can a failing tube actually introduce hum? If so, where can it be getting the 60 Hz signal from - the filament?

Yes it could do?, but what have you got connected to the input? - if the inputs are floating then it's going to pick up mains hum. By the look of just the chassis sat there, this could well be the reason!.

Rustydawg · 26th February 2006, 07:44 PM

The hum is there with the inputs floating, tied straight to ground, or through 1K Ohm to ground. That's what las me thinking the tube is inducing it. I wish I had another 12AX7 to try...

**Nigel Goodwin** · 26th February 2006, 07:55 PM

Quote:

Originally Posted by Rustydawg

The hum is there with the inputs floating, tied straight to ground, or through 1K ohm to ground. That's what las me thinking the tube is inducing it. I wish I had another 12AX7 to try...

As it's on both channels it's got to be something common to both, the valve is one possibility - so if the HT decoupling to the preamp stage.

Most double triodes are the same connections, so if you have another type you could try that? - UK numbers are ECC81, ECC82 and ECC83 - with the 83 being the low noise audio one, the 81 and 82 are often used as drivers. But any of them will work for you!.

JimB · 26th February 2006, 10:04 PM

Consider the possibility of a heater-cathode short circuit in the 12AX7. Or at least low resistance.

JimB

Rustydawg · 27th February 2006, 05:38 PM

Hmmm... The tube checks OK for low impedance between any pins (other than across filament leads) so I'm leaning towards the decoupling of the pwr supply to this preamp stage as Nigel suggests.

Perhaps the addition of the modern computer caps to the power supply have made things worse, what with their lower ESR.

I should also mention that I added 1N4007 diodes between the HV transformer and the rectifier tube just in case the tube wasn't 100%.

I have checked for isolation between the filament windings of the transformer and other windings, it tests OK.

Russlk · 27th February 2006, 08:50 PM

Heater to cathode leakage in the 12AX7 is the most likley cause of the hum, but it is possible the the filter caps that you added may have their ground returned to a point near the audio circuit. If the chassis is ground, that should not be a problem, but if there is not a power transformer, and the chassis is floating, look for a ground problem.

OK, I reviewed the picture, there is a power transformer, so replace the tube. ( I have some, let me know if you need one).

Rustydawg · 1st March 2006, 11:19 PM

Indeed the tube is the culprit. Even though there was no low-Z path between the filament and the cathode (at least within the DMM's 20 Mohm ability) it is inducing hum from the filament's AC. Even a capacitance test showed only minute capacitance; about the same few pF as the loose leads of the capacitance meter by themselves.

I dug into the schematic for the tuner/pre-amp and found two more 12AX7 tubes to try (that I hadn't thought of before!) and they were much quieter than the original.

At this rate I suspect there will be more capacitors and tubes in need of replacement in the pre-amp chassis too... Thanks for the input, guys.

JimB · 3rd March 2006, 06:18 PM

Quote:

Originally Posted by Rustydawg

Indeed the tube is the culprit. Even though there was no low-Z path between the filament and the cathode (at least within the DMM's 20 Mohm ability) it is inducing hum from the filament's AC.

Heater-cathode leakage tends to show up when the heater is hot, but shows OK when cold.

JimB

ljcox · 3rd March 2006, 09:08 PM

Is one side of the filament supply connected to ground? If so, try removing the gnd. You could also try transferring the gnd to the other side.

Another possible solution would be to install a 6 or 12 Volt DC supply for the filaments. This would avoid you having to throw away the 12AX7.

Rustydawg · 4th March 2006, 12:51 AM

Len, I'm with you on the DC supply. What with all the small 5V switching supplies that I've acquired, I could likely tweak one up to 6V to run both the pre-amp/tuner and power amp chassis. The 12AX7 is configured with two 6V filaments tied in common, so it runs on 6V. There's a good chance that there are more tubes with the same problem on the tuner chassis, so this may be an economical option - although rather new-age. The goal was to keep the chassis looking as original as possible, that's why I hid the filter caps under the aluminum chassis. Perhaps I can hide a switcher under there too.

Oh yeah - the heaters are all run off a single, isolated 6V winding of the power transformer.

Jim B, I was wondering if the leakage would only show up with the filament running! I'll try testing it further with the filament hot and see if it can be measured... Perhaps one of the parameters checked on the old 'tube testers'?

p.s. The larger filter caps and the replacement tube mean almost no audible hum on the output, most impressive. With 5W/ch on tap too!

JimB · 4th March 2006, 01:00 AM

Quote:

Originally Posted by Rustydawg

Perhaps one of the parameters checked on the old 'tube testers'?

Yes the better ones did test for heater-cathode leakage.

JimB

ljcox · 4th March 2006, 08:03 AM

Quote:

Originally Posted by Rustydawg

Len, I'm with you on the DC supply. What with all the small 5V switching supplies that I've acquired, I could likely tweak one up to 6V to run both the pre-amp/tuner and power amp chassis. The 12AX7 is configured with two 6V filaments tied in common, so it runs on 6V.

Jim B, I was wondering if the leakage would only show up with the filament running! I'll try testing it further with the filament hot and see if it can be measured... Perhaps one of the parameters checked on the old 'tube testers'?

Remember that the filaments will need a lot of current.

You could test for the F - K leakage by measuring from the filament line to gnd with a DC voltmeter (with the power on of course). If there is a leakage, you sould see some DC.

eblc1388 · 4th March 2006, 08:52 AM

In the old days, the 12AX7 filament connections is not earthed but parallel with a 1K pot and the wiper earthed.

Info in those construction articles mentioned turning the POT to position of least hum.

**Nigel Goodwin** · 4th March 2006, 09:51 AM

Quote:

Originally Posted by eblc1388

In the old days, the 12AX7 filament connections is not earthed but parallel with a 1K pot and the wiper earthed.

Actually, that wasn't in the 'old days', that was a reasonably 'modern' technique 8) usually only applied to tape recorders where you have very high gains and small signal levels.

There's no reason, or need, to do that in a power amp - if the valve is faulty?, throw it away and replace it!.

Valve gear is very crude and simple, repairing it is childs play!, this thread is really way over the top! :lol:

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

26th February 2006, 07:44 PM	(permalink)
Rustydawg	floating inputs The hum is there with the inputs floating, tied straight to ground, or through 1K Ohm to ground. That's what las me thinking the tube is inducing it. I wish I had another 12AX7 to try...

26th February 2006, 10:04 PM	(permalink)
JimB	Consider the possibility of a heater-cathode short circuit in the 12AX7. Or at least low resistance. JimB __________________ Experience is directly proportional to the value of the equipment ruined.

27th February 2006, 08:50 PM	(permalink)
Russlk	Heater to cathode leakage in the 12AX7 is the most likley cause of the hum, but it is possible the the filter caps that you added may have their ground returned to a point near the audio circuit. If the chassis is ground, that should not be a problem, but if there is not a power transformer, and the chassis is floating, look for a ground problem. OK, I reviewed the picture, there is a power transformer, so replace the tube. ( I have some, let me know if you need one). __________________ see my website: www.geocities.com/russlk

1st March 2006, 11:19 PM	(permalink)
Rustydawg	12AX7 is to Blame Indeed the tube is the culprit. Even though there was no low-Z path between the filament and the cathode (at least within the DMM's 20 Mohm ability) it is inducing hum from the filament's AC. Even a capacitance test showed only minute capacitance; about the same few pF as the loose leads of the capacitance meter by themselves. I dug into the schematic for the tuner/pre-amp and found two more 12AX7 tubes to try (that I hadn't thought of before!) and they were much quieter than the original. At this rate I suspect there will be more capacitors and tubes in need of replacement in the pre-amp chassis too... Thanks for the input, guys.

3rd March 2006, 09:08 PM	(permalink)
ljcox	Is one side of the filament supply connected to ground? If so, try removing the gnd. You could also try transferring the gnd to the other side. Another possible solution would be to install a 6 or 12 Volt DC supply for the filaments. This would avoid you having to throw away the 12AX7. __________________ Len

4th March 2006, 12:51 AM	(permalink)
Rustydawg	Len, I'm with you on the DC supply. What with all the small 5V switching supplies that I've acquired, I could likely tweak one up to 6V to run both the pre-amp/tuner and power amp chassis. The 12AX7 is configured with two 6V filaments tied in common, so it runs on 6V. There's a good chance that there are more tubes with the same problem on the tuner chassis, so this may be an economical option - although rather new-age. The goal was to keep the chassis looking as original as possible, that's why I hid the filter caps under the aluminum chassis. Perhaps I can hide a switcher under there too. Oh yeah - the heaters are all run off a single, isolated 6V winding of the power transformer. Jim B, I was wondering if the leakage would only show up with the filament running! I'll try testing it further with the filament hot and see if it can be measured... Perhaps one of the parameters checked on the old 'tube testers'? p.s. The larger filter caps and the replacement tube mean almost no audible hum on the output, most impressive. With 5W/ch on tap too!

4th March 2006, 08:52 AM	(permalink)
eblc1388	In the old days, the 12AX7 filament connections is not earthed but parallel with a 1K pot and the wiper earthed. Info in those construction articles mentioned turning the POT to position of least hum. __________________ L.Chung