120hz hum in new build

[beasleybodyshop]

Hey everybody!

I just recently fired up a new build - A circuit clone of the Matchless DC30 - and am struggling with finding a solution to a loud 120hz hum that the amp produces.

Backstory:

1. Circuit is note for note clone of the old Matchless DC30 schematic that's been floating around : https://www.thetubestore.com/lib/thetub ... ematic.pdf
2. The only changes that were made was omitting the FX loops, switchable Hi/lo power for a fixed single resistor/cap combination, and adding a Post Phase Inverter Master Volume between the PI and output tubes. I also upped the grid stoppers on the power tubes themselves to 5.6K.
3. Power supply nodes are within 10% of spec listed in the schematic as well.

So when you flip off the standby, there is a noticeable 120hz hum (measured with an RTA app on my phone) that, if you do nothing else other than flip OFF standby, hums loudly. Even with the master volume turned all the way down. BUT. After about 40-50 seconds the hum slowly starts to get softer and softer until its pretty well manageable. FWIW, measured from the RTA app the noise goes from around -30dB to -58dB within a span of 30 seconds to a minute after flipping off standby.

So when you turn up the master volume, the hum gets louder.

With this in hand, these are the steps i've performed to attempt to resolve this:

1.Pull PI tube, hum goes away. Removed both preamp tubes for the top boost channel. Hum mostly goes away. Pull just the cathode follower tube of the top boost channel. Hum mostly goes away. Pulled Input tube for Top Boost channel and leave CF tube in, hum is there.
2.Moved power tube cathode connection to the negative tab of the cap can that filters both the plate and screens for the power tube section (using an F&T 32/32 can cap here for this)
3.Ground input to CF tube with .1uf cap - no change. Ground output of actual Cathode Follower circuit (it drives a tone stack) - 120hz hum mostly goes away.
4. Lifted heater CT and replaced with pair of 100R resistors from heaters to ground - no change.
5. Floated Heater reference via pair of 100R resistors from heaters to top of power tube cathode resistor (floats AC about 14 volts DC) - no change.
6. Replaced both CF and Top Boost channel preamp tubes with 2-3 other known good tubes - no change.
7. Clipped ground for CF filter cap from ground buss direct to chassis - no change. Clipped all ground connections for CF stage together and off the buss, onto chassis. No change.
8. Audited circuit for component values and connections. No seemingly missed connections or incorrect parts values.

Damn, at this point I'm stumped. I've tried all the tricks that in the past, seemed to mitigate this kind of issue. Does anyone have any thoughts or suggestions? Thanks.

Matchless-DC30-Old-Schematic.pdf

[R.G.]

Your description makes me think that the bias supply is charging slowly, giving it high ripple for a short time during standby.
Edit: I should have looked at the schematic first. Cathode biased, so no bias supply. The description still sounds like an issue charging a capacitor up after turning off standby. I would 'scope the filter caps for what they're doing when standby is flipped off.

[beasleybodyshop]

RG,

Thanks for chiming in. I forgot to mention that this amplifier is cathode biased.

[rangdipkin]

Can you share some gut shots of the new build? the more detail the better, a few fresh sets of eyes on it might help identify something

[beasleybodyshop]

Will share some pics here in a bit, thanks.

I did clip in a 100uF cap in parallel with the node that feeds the Cathode Follower stage - this did reduce the hum from a measured -28dbFS to -38dBFS.

Clipping in the same cap onto the mains reservoir for the output tubes also reduced the hum by about that same amount.

Seems like perhaps a bad filter cap? I am using a 32/32uF 500v can for the plates/screens. also using a 32uF cap for the CF node. I know AC30s never had much in the way of filtering but seems like this would have been enough?

power supply ripple at the mains reservoir is roughly 200mv. measured with the additional 100uF clipped in parallel. Not sure if that seems high?

[beasleybodyshop]

With stock filtering applied, it measures around 340mv of ripple at the mains node.

Measured ripple at the CF node with stock filtering is perhaps 20mv.

[Stevem]

Maybe you have a bad tube socket.

Are you seeing any heater voltage bleed into any of the grids on that channel that's acting up?

[beasleybodyshop]

Interior shots.

Don't mind the detritus. I will clean up once she's sorted

[beasleybodyshop]

I replaced the 32/32 cap can with a 50/50uF 500v one i had laying around, replaced the CF node with a 40uF one i had as well.

Filter ripple at Main B+ is now 215mV. CF node is now maaaybe 10mV

Hum is still pretty loud when i turn up the master.

[rangdipkin]

nothing jumping out at me but lets get more eyes on this gal! Around the phase inverter, looks like one of the output coupling caps to your PPIMV is sitting below the rest of the terminal strip, is that right against a hummy HT supply node? unlikely to be the source of that much hum but ya never know...

EDIT: have you already measured current flowing through the power tubes? Wondering if you have a bad power tube throwing the power amp out of balance. Might even be worth pulling a pair and troubleshooting with only 2 tubes operational. Sift through em to rule out a bad single.

[beasleybodyshop]

nothing jumping out at me but lets get more eyes on this gal! Around the phase inverter, looks like one of the output coupling caps to your PPIMV is sitting below the rest of the terminal strip, is that right against a hummy HT supply node? unlikely to be the source of that much hum but ya never know...

EDIT: have you already measured current flowing through the power tubes? Wondering if you have a bad power tube throwing the power amp out of balance. Might even be worth pulling a pair and troubleshooting with only 2 tubes operational. Sift through em to rule out a bad single.

The hum goes away when I turn down the master volume - I would think that the power tubes are OK if they don't make any noise with the inputs effectively grounded. I suppose I can try pulling a pair and seeing

[bepone]

very sensitive wires going from MV to grids of out tubes passing right over the noisest point in the amp - filter capacitor?
lets move this first away

[bepone]

where is filter choke? if there is filter choke there will be no ripple to the preamp stages , so i assume that choke was ommited in this build?

if so, nodes in the preamp must have additional RC filter stage..

[bepone]

120Hz humm is for sure from the rectification or B+ line so all concentration must be there

[beasleybodyshop]

where is filter choke? if there is filter choke there will be no ripple to the preamp stages , so i assume that choke was ommited in this build?

if so, nodes in the preamp must have additional RC filter stage..

filter choke is indeed in place between the plate and screen nodes.