LOUD hum, with master volume turned up

[Gainzilla]

Hi All,


Still at it with my JCA20HV. Haven't done much beyond replacing signal caps with Mallory's, and swapping rectifier diodes with FREDs. I've been noticing this for a while but have been busy on other projects. So here's the deal:

There's a lot of hiss when I crank the gain, which is to be expected. However, when I crank the master I get a LOUD 60 cycle hum and some radio station interference. Even with gain on 0, guitar cable unplugged, cables unplugged from the FX loop. I mean, it's loud. Too loud to even remotely consider ignoring. What could be causing that? Here are some things I've considered:

• The FREDs are not 100% functioning. I had to REALLY get some heat on those solder joints trying to get the new ones in. Could they be damaged?
• Could the power tubes be unbalanced or biased too hot? When I biased the new tubes, I noticed a difference in voltage between the two supposedly matched tubes. I split the difference on the meter with the bias control.
• Bad tube?
• Cold solder joint somewhere?
• Cheap crap Chinese built amp? lol Seriously, though, should I go through the trouble of rewiring the heaters and get them away from the PCB?

That's all I've got. Any ideas? I really want to like this amp, but I've kicked much more expensive amps to the curb for FAR less. I just thought, it's such a simple circuit... How could they screw that up (haha)? 

I do have access to a scope, but I'm not sure if it's up to snuff for tube amp work. Thanks in advance!

[eleventeen]

"I do have access to a scope, but I'm not sure if it's up to snuff for tube amp work. Thanks in advance!"

There is no functioning scope on the planet unsuitable for audio. A surplus WW2 Dumont with 100KHz bandwidth would work just fine as long as it worked, was stable, and you had anything approaching a suitable probe. There are plenty of scopes, even brand new modern ones, that are not rated for > 250 or 300 volts on the input, so you can't just slam the probe onto your B+ without great potential for damage. This is solved by placing a series capacitor of adequate (prefer 600 WVDC, typically .01 or .05 ufd but almost anything non-electrolytic will work) voltage rating in series with the probe. And.....discharging it (by shorting it) after you measure HV!

The hum and the radio pickup are likely to be two different things.

*Did the amp hum with the vanilla 1N4007s (or whatever you used before) or is this a fresh build or a repair?*[size=78%] [/size]Hum could be new-defective e-caps. Pull the PI tube. Your output section should be quiet and stable.  Why do you suppose the FREDs req'd so much heat? Usually diodes have beautifully-tinned leads and suck up solder with gusto. The other day I found a bag of unused imported 800 v 1.5 amp diodes I pulled out of the garbage from a place I worked in 1982 and soldered into something with no problem. Good for tubes. Yes, I scraped the leads lightly w/a knife blade. Diodes are awfully tough..maybe you did lunch them, though.

Doubt it's the heaters.

RF pickup could be an oxidized tube pin or an inadequate resistor on a preamp grid somewhere. For a radio signal to show up in your output section, it has to be picked up and amplified way earlier in the amp.

You have 2 different issues with 2 different causes.

[Gainzilla]

Thanks for the reply, Eleventeen!


Great info about the scope! The one I have is a Tektronix T922 15Mhz scope. Looks like the voltage range at 120v is 90-130v. It does have probes, so I'll get a cap. It's not an e-cap so it's not zapping myself I should be worried about, but zapping other components?


This isn't a new build, but a new Jet City 20 watt jcm800 style amp (20HV). The reason I had trouble with the diodes is that it was my first encounter with non-lead solder (I'm just a hobbyist, and have been working off the same big spool I've had for like 10 years). I was able to remove the old diodes easily, but struggled to install the new ones ( couldn't get my solder braid to pick up enough solder to make a suitable hole). I learned a trick where you immediately ream the wet solder hole out with a wet toothpick. I'm wondering if I should swap those out for NEW new ones, now that I have a method that won't overheat them. Also, I didn't replace the bias supply diode. I probably won't, unless there is some reason why I should.


Honestly, I can't remember if the problem existed prior to the diode swap. It sounds really good with the master below 10:00, if that helps.


Anyway, if I remove the PI tube, and it's quiet, does that imply a preamp issue? If so, can I also imply a power amp/ power issue if it's still noisy?


Thanks again!

[eleventeen]

Thanks for the reply, Eleventeen!Great info about the scope! The one I have is a Tektronix T922 15Mhz scope. Looks like the voltage range at 120v is 90-130v. It does have probes, so I'll get a cap. It's not an e-cap so it's not zapping myself I should be worried about, but zapping other components?

<<< I had one of those, in the small blue molded case! Mine had an intermittent connection to the input connector and I had a much better scope so I tended not to use it. It's 10x the scope you need for audio and there isn't a microsmidgen of anything wrong with it. A .02 [input voltage-blocking] cap charged up to 400 volts will sting you real good and that charge will remain on it for quite a while. I doubt it would damage anything else in tubeland but it would devastate any kind of solid state anything you might move to....*including the scope itself*. It's not so much the danger of getting shocked...it's the danger of holding a nice new 6L6 in your hand and reflexively flinging it across the room as you are getting ready to install it into your amp. You clear away some of the crud on your bench, forget that your blocking cap is charged, and zap!

Mechanically speaking, the blocking cap is quite the irritation esp if you have a decent and convenient probe with clip-on tip, but is VERY important. I just clamp onto one side of the blocking cap and use a normal alligator-clip jumper to measure, for example, ripple or hum on B+. You may very well have to use it or similar to measure signal on a preamp tube plate running at 180 volts. I have a pretty nice scope, certainly nothing special, but it has a "300 v" warning on the front panel connector so I stay within that limitation by using such a cap.


This isn't a new build, but a new Jet City 20 watt jcm800 style amp (20HV). The reason I had trouble with the diodes is that it was my first encounter with non-lead solder (I'm just a hobbyist, and have been working off the same big spool I've had for like 10 years). I was able to remove the old diodes easily, but struggled to install the new ones ( couldn't get my solder braid to pick up enough solder to make a suitable hole). I learned a trick where you immediately ream the wet solder hole out with a wet toothpick. I'm wondering if I should swap those out for NEW new ones, now that I have a method that won't overheat them. Also, I didn't replace the bias supply diode. I probably won't, unless there is some reason why I should.Honestly, I can't remember if the problem existed prior to the diode swap. It sounds really good with the master below 10:00, if that helps. Anyway, if I remove the PI tube, and it's quiet, does that imply a preamp issue? If so, can I also imply a power amp/ power issue if it's still noisy?

Philosophically, you are trying to isolate and confirm that the building blocks in your block diagram are working right. If you have nominal B+ to the output tubes and they are not redplating and the speaker output is quiet...then you have powerful evidence that your output section is working right; especially if you can hear a tiny, tiny bit of toob noise from the speaker. Probably I should have said to pull the tube *before* the PI. The PI+output tubes are kind of one block. With no input from prior stages (because you have pulled tubes PRIOR to the PI, there's not much you will learn from how it (the PI) affects the amp. But the point is, most people seem to like to establish proper operation of the output section before troubleshooting the preamp, even if it seems backwards.


Except: In the case of a PPIMV with problems...this is where you will have to spend some attention if you have this hum issue.

The point is...you have to figure out a way or ways to interrupt the signal path through the amp and subdivide the chain into segments whose operation you kind of understand. Pulling out tubes is an easy way to do this. For the most part, the preamp section of your amp doesn't know if it is driving the output section--its job is to amplify and tone-shape a tiny guitar signal. If you can probe with your scope a few stages into the amp and confirm that the signal is getting bigger, without massive hum and noise or RF (which will appear as a filled-in "cloud" on your scope----this is a possible sign for the radio issue you described) then it is probably working. Likewise, with no signal input, if the output section isn't blowing fuses or redplating and you can hear tiny, tiny tube noise on the output..then it is working right. This is the process of troubleshooting. Isolate > narrow down > kill.

 
 

[stratele52]

No need scope , use you voltmeter to read AC signal when hum

[Gainzilla]

I had one of those, in the small blue molded case!

That's it! It's a loaner from a friend. Prior to that I made an audio probe (nail/capacitor/shielded cable to 1/4 jack). I use a tone generator on my iPhone plugged into the amp, then probe around for signal. The other end of the probe is plugged into the battery powered micro Marshall half stack I got for Christmas. Lol

It's not so much the danger of getting shocked...it's the danger of holding a nice new 6L6 in your hand and reflexively flinging it across the room as you are getting ready to install it into your amp. You clear away some of the crud on your bench, forget that your blocking cap is charged, and zap!

That was the best explanation I've ever heard.   Ok, duely noted!


Alright, thanks for the help. I'll start with the tubes and try and isolate a general building block, then go from there.


Stratele52, what are you looking for when using a meter to locate the source of the hum?

[stratele52]

Stratele52, what are you looking for when using a meter to locate the source of the hum?

1- First the hum is louder when you crank Master = hum is coming before Master circuit  .
Any schematic fot me ?
2- Your meter on AC scale , take some reading from Master vol and go back to the amp  input


You should read AC voltage ( hum ) on the signal path  where you should not read something if no guitar plugged


I did not read all the post , too long and my english is poor . Do you do some mods ? What Is FRED circuit ?
If the amp working fine before , check your mods

[Gainzilla]

Adding schematics...


Sorry, FREDs are Fast Recovery Epitaxial Diodes.


The only mods I did we're to replace crappy Chinese signal caps with Mallory's, and swap out the rectifier diodes with The UF4007s.


Thx for the troubleshooting info!

[stratele52]

Adding schematics...


Sorry, FREDs are Fast Recovery Epitaxial Diodes.


The only mods I did we're to replace crappy Chinese signal caps with Mallory's, and swap out the rectifier diodes with The UF4007s.

[/size]
[/size][size=78%]You had no hum before that caps replacement  and diodes ?[/size]
[/size]
[/size][size=78%]Or you have hum ?[/size]

[Gainzilla]

UPDATE:

I pulled the tube right before the PI: Dead quiet
Put that back in and pulled the 1st: a little hiss, no hum
Replaced the Tung Sol from 1 with another Tung Sol: HUM!
Replaced THAT Tung Sol with a NOS RFT: some hiss, but NO HUM!

What gives? Amp just doesn't like Tung Sol? The RFTs construction in some way minimizes hum in this amp? Or just 2 random bad tubes that behave well enough in other circuits?

Anyway, a mighty facepalm is in order!

Thanks for the help

[HotBluePlates]

Replaced THAT Tung Sol with a NOS RFT: some hiss, but NO HUM!

What gives? Amp just doesn't like Tung Sol? The RFTs construction in some way minimizes hum in this amp?

It is probably a case of heater-to-cathode leakage in the Tung-Sol-skis.

There are a couple different mechanisms by which hum can be coupled from the heater to the cathode. There are also a couple of different approaches to fixing it, depending on which exact cause you're encountering.

Look at the schematic of the Fender 5F6-A Bassman. See the 250uF cathode bypass cap on the input 12AY7? It's not really 250uF because it gives full gain down to the bottom of a bass' range (it does, but the -3dB point of that circuit is 0.77 Hz). The 12AY7 was marketed at the time as a low-hum tube meant for the input stage of audio amps. Fender then hedged their bet by applying a huge cathode bypass cap, which would then bypass any heater-to-cathode leakage hum to ground.

Looking at the schematics you posted, the input stage has a 1uF cathode bypass cap, while the 2nd stage has no bypass cap. Either of those would be susceptible to hum due to leakage, if you had a leaky tube. That means some tubes will result in audible hum, others won't (as you found out).

The tubes with leakage hum may work just fine in later stages of the amp, where their hum is not amplified so much.

It's not clear if referencing the heaters to a positive d.c. voltage instead of ground would mitigate hum. Or if that's even feasible in your amp (meaning whether you can physically wire the amp that way, or if the heaters are on board traces.

[Gainzilla]

Thanks HotBluePlates!


Re: bypassing hum to ground. Are you saying that only a large value cap would push audible hum to ground, or just that any cap fully bypassing the cathode resistor? I feel like 250uf is a pretty big cap (The Bassman is a bass amp originally, after all). I would likely then need to find somewhere else in the circuit to trim the bass.



Sadly, the heaters are on traces on the board, although I've heard of some who have cut the heater traces and wired twisted pairs away from the board. Regarding the actual noise floor, I'm starting to believe that adding a choke and replacing the pots with higher quality ones would quiet things down considerable (at least as far as the hiss)... Probably replacing some of the signal path resistors with higher quality metal film too...

[eleventeen]

Before you throw time and parts money into this, just ponder the notion that hum and hiss are completely different phenoms. If you just group them together as "noise I don't want"---logical and experienced-based measures you would take to reduce either one are unlikely to affect or help the other. For example, I don't think changing pots or adding a choke will affect hiss in the slightest.


In older Fenders, metal film plate resistors in place of carbon comps are remarkably effective at reducing hiss. I have not tried replacing signal resistors with MFs---undoubtedly, they cannot make things worse. BUT in general, the newer the (plate) carbon comps, the less this effect will be apparent. I suppose you could have some really noisy new carbons. OTOH, this is a 20 or 30 cents-per-resistor affair versus maybe as much as $35-$50 for a choke and a full round of pots.


After a recent lengthy thread on heaters, as well as several others over the past past year or two, I have come to believe that heaters are generally low contributors to hum. Oh sure, one could deliberately route signal wires so that they picked up heater hum. Here's an idea for an experiment: Suppose you get your amp going, into a speaker and everything and you find you have hum which you suspect is heater-induced. Suppose you now come along with a pair of dykes and cut one heater wire-ideally where it feeds the preamp tubes---leaving the power tubes lit up. You should have a second or two of sufficient heater temperature in the preamp tubes to run the amp. If when you do that, your (you suspect) heater hum falls off a cliff, really drops in level dramatically, then OK, I could buy that you have  a heater hum issue. Otherwise....meh. I would pursue the lines HBP suggested if you can rule out power supply hum.


If all your parts are on one board and you can't cut anything but a "master" wire to ALL the tubes without taking out the board and cutting a trace and then remaking the trace, I would still consider this a worthy experiment. If you have a heater-induced hum problem, the hum will drop off a cliff when you cut those wires. Either way, turn the amp OFF quickly after you cut the fil wires. The same kind of thing applies if you suspect AC-induced hum--shut off power (but not standby) and the AC hum should just cease right away if you have such an issue. Easy.

Edit; shortened excess blank space at bottom of post.  Willabe

[Gainzilla]

Right, perhaps I'm jumping the gun, but I was operating under the assumption that the hum was isolated to the tubes. I do hear a very high oscillation when I chopstick some of the leads around, but I think I got those minimized. The Soldano version of this amp uses a choke, presumably to smooth out ripple in the B+ (i.e. hum). I can avoid the choke with the large cathode bypass caps, but would I need to shave off some bass somewhere else in the circuit to avoid blocking distortion (smaller coupling caps maybe)?

While this is a "Soldano" in spirit, it's still a cheap Chinese amp. The resistors are metal film, but they are cheap. My general goal with this amp was to upgrade the most critical components to improve tone/reliability. So ok, if I swap out plate resistors with high quality metal films, and from what I've been reading, the input grid resistor as well, that should help? keep the hiss down?

The reason I was considering replacing the pots was due to thermal noise. These pots are essentially cheap carbon film resistors, and I've read the gain pot in a high gain amp is a key part in keeping noise down... I've read that even going so far as using conductive plastic... Do you think the general idea warrants merit, or should I leave it alone?

I've also been reading that high value resistors in divider networks can cause noise as well, so a general guideline would be to keep amp designs simple and avoid 1M gain pots. With that in mind, what do you think the impact would be to lower the gain pot to 500k? Negligible or noticeable?

BTW, its amazing to me how different tubes impact this design. I have a NOS RCA from the 50s. In the 2nd slot, its LOUD and bright. I've had similarly distinct and varied observations with other combinations. This doesn't bode well for my inner amp-tweaker. I am really liking the NOS RFTs in both slots though. Tight, crunchy, with a smooth top end.

Anyway, thanks again!

[Colas LeGrippa]

Hi,

as I can see, the filter is badly designed. Disconnect the OT red lead from where it is hooked up now and plug it after the resisor. Or add a double 50uF cap can with a choke in between ( pi filter ) and plug the OT red lead right after the second 50uF. I had exactly this problem and now, I ALWAYS hook up the OT  right AFTER  the choke, and not before like in many schematics that I frquently look at.

Colas

[HotBluePlates]

Re: bypassing hum to ground. Are you saying that only a large value cap would push audible hum to ground, or just that any cap fully bypassing the cathode resistor? I feel like 250uf is a pretty big cap (The Bassman is a bass amp originally, after all). I would likely then need to find somewhere else in the circuit to trim the bass.

But I did not tell you to add large bypass caps...

UPDATE:
...
Put that back in and pulled the 1st: a little hiss, no hum
Replaced the Tung Sol from 1 with another Tung Sol: HUM!
Replaced THAT Tung Sol with a NOS RFT: some hiss, but NO HUM!

With this test, you proved the tube itself is the cause of the hum.

I said there were several possible causes: it could be capacitive coupling of hum from the heater to the cathode across the heater-cathode insulation. It could be emission from the heater to the cathode during the moments the heater is more-negative than the cathode voltage (a diode effect).

The diode-effect case can be fixed by elevating the reference voltage of the heaters from 0vdc to some positive voltage. That's generally done by moving the heater center-tap (or 100Ω resistors, as in this amp) to a positive voltage of 20-50vdc.

True capacitive coupling may not be fixed by this tactic. I once was contacted by an amp maker whose name & products you'd know to figure out why hum showed up when the big 250uF bypass cap was taken off the input stage of some of their amps. They were using smaller bypass caps to adjust bass (but even 25uF caps didn't cure the hum all the time). The problem existed even though they tried a positive d.c. voltage for the heaters (at least that was my understanding at the time; it's been quite a few years since then).

Read what I posted before again... the -3dB point for a 250uF cap and an 820Ω cathode resistor is 0.77Hz! The electric bass only goes down to around 40Hz. The excess sizing of the cap is because it forms a capacitive voltage divider along with the capacitance across which the hum is leaking. A super-huge cap then looks much like 0Ω, and the hum is eliminated.

Read the description towards the top of the G.E. and Tung Sol 12AY7 data sheets. This tube was expressly made for low hum in low-level gain stages, yet Fender still took the belt & suspenders approach of having a huge cathode bypass cap. When Marshall copied the Bassman, they used the 250uF first, then 320-330uF bypass caps for the Bass models.

Each of these forms of hum are dependent on individual tube samples. This is exactly what you found: the amp hummed with a couple tubes, but not with the RFT. You could probably make no changes to the amp, audition various tubes for the input socket, mark the ones that are quiet, use the noisy tubes for the phase inverter (or not-input socket).

Obviously, the big-cap approach isn't appropriate in all amps (because it will contribute to mud, and you may want to use bypass caps for tone-/gain-shaping). You can try the d.c. voltage reference for your heaters... My concern about traces for the heaters boils down to where the 100Ω resistors are located, whether you can cut their connection to ground easily, whether you are able to easily derive a +30-50vdc reference voltage, and how easily you can apply that to the point where the 100Ω resistor were previously connected to ground.

[HotBluePlates]

You know, even the old tube testers said in their manuals, "the best test of any tube is in the socket in which it is intended to operate," which applies here. A tube that hums in that first socket may work without issue in other sockets in that amp, or in any socket in some other amp. And I'm not aware of any brand of tube that will guarantee no hum in that first socket (they'd have to do 100% testing of every tube for that one issue, and that's too much time & money for their wholesale price point).It seems the best option in your case is sorting tubes yourself for which ones perform well in your amp. Maybe re-sell the ones that don't make the grade?

The reason I was considering replacing the pots was due to thermal noise. These pots are essentially cheap carbon film resistors, and I've read the gain pot in a high gain amp is a key part in keeping noise down... I've read that even going so far as using conductive plastic... Do you think the general idea warrants merit, or should I leave it alone?

I wouldn't advocate replacing any part until you go back and use a listening amp (your old jerry-rigged probe) to confirm exactly where the noise is present. You might find the resistors and pots are not really an issue. Or you may find a part you'd never suspect being the culprit.

[alerich]

UPDATE:

I pulled the tube right before the PI: Dead quiet
Put that back in and pulled the 1st: a little hiss, no hum
Replaced the Tung Sol from 1 with another Tung Sol: HUM!
Replaced THAT Tung Sol with a NOS RFT: some hiss, but NO HUM!

What gives? Amp just doesn't like Tung Sol?
The RFTs construction in some way minimizes hum in this amp?
Or just 2 random bad tubes that behave well enough in other circuits?

I had the same situation with a brand new Tung Sol 12AX7. Installed it in V1 of my JCM800 2204 build and it hummed like crazy. Replaced it with a new EH 12AX7 and hum almost gone. The tube behaved fine in another similar circuit. Now that I think about it, I had new Tung Sols in all three positions and still have them in V2 and V3. If I ground the input to the PI all hum stops. I'm going to go swap out V2 for an EH tube and see if that kills the remaining hum... after I finish my coffee.

Here's a desoldering tip for that awful lead free solder. I've also had a tough time getting it to wick up. I heat the joint and add a little conventional solder to the joint. Then desoldering seems to go much easier.

[Gainzilla]

I wouldn't advocate replacing any part until you go back and use a listening amp (your old jerry-rigged probe) to confirm exactly where the noise is present. You might find the resistors and pots are not really an issue. Or you may find a part you'd never suspect being the culprit.

Again, the voice of reason. Good call. I'm assuming at some point I will begin to hear a pronounced increase in noise/hiss? BTW, I DID make an aligator clip with a .02 600v cap (mallory... That should work, right?), so I could use the scope. Is there any advantage? High frequency oscillations that maybe I can't hear? just thinking out loud. All that is assuming that I would be able to recognize noise on the scope. lol

I'm with you on the hum issue. At this point I feel ok with isolating it to the tubes, and am even ok if its picky. Right now it seems pretty good, and if I can quiet the hiss a little I'll be content with this one.

[Gainzilla]

Here's a desoldering tip for that awful lead free solder. I've also had a tough time getting it to wick up. I heat the joint and add a little conventional solder to the joint. Then desoldering seems to go much easier

Nice tip! Can you still get the old stuff anymore? I finally burned through my spool.


I'm really interested to see what the outcome of your tube swap is? I have some old RCAs, RFTs, Sovteks, and a few chinese ones.. In this amp, none of them sound as good as the RFT.


Cheers!

[alerich]

Yeah, NOS solder will be around for a while, I think.

I swapped out the Tung Sol in V2 for a Mullard reissue. It did quiet it a touch. It wasn't a huge amount of hum to start with (while the V1 hum was off the charts) and this build is a 50 watt Marshall stuffed into a little 1x8 combo chassis/cabinet. With everything (tubes, transformers) all packed in so tightly it's gonna hum a little but it's very acceptable. I think the tube swap helped. I'll dig out an EH tube tomorrow and A/B it with the Mullard RI.

[HotBluePlates]

Here's a desoldering tip for that awful lead free solder. I've also had a tough time getting it to wick up. I heat the joint and add a little conventional solder to the joint. Then desoldering seems to go much easier

Nice tip! Can you still get the old stuff anymore? I finally burned through my spool.

Even Radio Shack has leaded solder. Or you can buy a pound of the good stuff from AES (be nice to get other stuff so you split the shipping cost over more than just solder).