Mesa Subway Blues some hum/buzz 100Hz, out of ideas

[iefes]

Hi everybody,

I have some hum issues with my Mesa Boogie Subway Blues. I know that these amps are famous for hum after they ran for a while but I'm pretty sure it's not that particular issue, especially since it's the newer version with the switch rather than different input jacks. And it's the non-Reverb version of the amp. Additionally I need to tell you that the hum is not unbearable. It is definitely more noise than my Crate V3112 produces and is unfortunately not well usable at bedroom levels. That's why I want to find the source for the hum and eliminate it.

First of all I need to mention that I have already contacted Chuck H who has been a great help with the issues and helped me understand what exactly is going on in this amp.

Okay, here's the story: I bought this amp used a few weeks back. All good except some hum. I did several tests, swapped tubes etc. --> The hum is mainly 100Hz but some 50Hz and higher harmonics seem to be included as well (I'm in Germany, so european power). Pulling tubes revealed that the hum seems to be related to the circuitry around V2. V2, as you can see from the schematic is located before and after the effects-loop. A former owner of the amp had disconnected the transistor-buffer for the effects-loop and ran a wire from the 0.01u capacitor after V2a directly to the effects-send. I tried to reconnect the buffer but found out that no signal was coming through. So I tried bypassing the whole effects-loop and the hum got a little better by connecting from the 0.01u to the 0.02u cap directly (via a c. 12 cm wire). As I would like to have an effects-loop I rebuild the transistor circuitry on strip grid with onboard send- and return-jacks and power supply. I have connected the unit with shielded wires whose shields are grounded at one end. Now the effect-loop works but the hum got a worse again. Moving the wires around inside the chassis while the amps running doesn't change the hum. The ground-wire for the effects-loop sub-pcb is connected to the negative leg of the bias-supply filter cap, as this was relatively close.

Here's all I know / what I have done so far:

The hum is not affected by any of the pots or the switch.

When switching off the Power-Supply while running the hum faded before I used shielded wire. Now it stops immediately so I suspect something related to the power-supply (+/- Grounding).

Installed new filter-caps (4x 33u/500V).

Checked grounding and made clear that the connection to the chassis is only made at the input-jack. The power-supply-Gnd is connected to the Preamp-Gnd from V1Gnd to the Gnd of the last filter-cap. I think that there shouldn't be any potential ground loops but I have to admit that I don't exactly know how the Gnd-traces run on the PCB.

The heater-wires for the power-tubes are not twisted but are attached to each other and run parallel to the power-tube sub-pcb with the HV wire also attached in parallel. However, moving them around in the Chassis didn't change the hum. The heater supply for the preamp tubes run as traces on the PCB, which I suppose is not great. The heater winding does have a center tap connected to Gnd.

I have additionally included some mods to the amp that have been suggested by Chuck H here on the music-electronics-forum. The mods included some resistor and cap-changes in the preamp which alter the preamp to be very Ab763-like. These mods did not alter the hum issue at all.


I have put together a google-drive folder with a video, pictures, schematic and noise figure (recorded with a smartphone app). There's also a schematic that shows the modifications suggested by Chuck H. I have adapted most of the preamp modifications, but did not change the bright-switch yet. And the input circuitry uses a 1M to Gnd followed by a 10k grid stopper with 150pF to ground (as per Merlins suggestions). But as I said these modifications did not change the hum.

Heres the link: https://drive.google.com/open?id=1gn...OcGpjzBpkxMpmT

I would be very grateful if anybody here can help me hunt down the hum. I don't have a scope here to check the hum/buzz-sines but any other ideas are very much appreciated.
I like this little amp and I already learned a lot by trying to figure out the hum issue.

Thank you very much.

[alexradium]

does it hum if you plug directly in the return jack?

[iefes]

I tried this, yes, but the effects-loop was already wired differently from the original layout. It didn't help. Strangely, with the new sub PCB it is still humming when plugged into the Return jack.

[alexradium]

then it should be a filtering problem or a bad ground connection

[iefes]

What do you mean exactly by "bad ground connection"? A cold solder joint or a missing connection? Or a bad ground-wiring scheme?

[iefes]

For testing purposes I have bypassed the effects-loop and just wired directly from the 0.01uF capacitor after V2a to the 0.02uF capacitor leading to the grid of V2b. I have used shielded wire for this. The hum is still present. Moving the shielded wire around with a chopstick doesn't change anything about the hum. I've tried pulling tubes again to confirm that with tube V1 being pulled the hum remains unchanged. But when V2 is pulled the hum disappears completely. I've tried V1 in the V2 position but this didn't help. So the hum is definitely produced around V2. What parts should I check around V2? Do you have any more suggestions?
Is it possible to pull the grid to ground for testing purposes, to see whether V2a or V2b are catching the hum? Would I need a capacitor for this or can I just connect the grid to ground directly?
Thanks!

[alexradium]

i use alll the time a wire with alligator clips,one attached to a ground point,the other side i can touch any grid pin or a suspect ground point.

[Enzo]

I do that as well, plus some points have DC on them, like plates, so I put a 0.1uf or 0.047uf cap in series with my ground clip wire. That way I can "ground" the point in terns of AC/signal without grounding off the DC.

[iefes]

Thank you both. I tried with a wire before reading that I could use a capacitor additionally and checked several ground points and grounded the grids of V1 and V2.
I have recorded the sound during my tests, so anybody who would like to hear what the different results sounded like is very welcome to have a listen (mp3 attached).

Below is the timeline of the recording. I also grounded V1a but you won't hear a difference.

9s : switched on
39s : grid V1b
51s : junction 0.47u 220k
1:00 : grid V2a
1:14 : grid V2b
1:28 : I turn up the volume, some hiss is added
1:42 : grid V1b
1:48 : junction 0.47u 220k
1:53 : grid V2a
2:04 : grid V2b
2:15 : volume turned down (I cut out this part as the pot crackled loudly)
2:20 : stdby off
2:26 : stdby on
2:35 : power off

The best result is achieved when the grid of V2b is grounded but good results are also obtained with V2a grounded. However, you can still hear some 100Hz hum going on even with those grids grounded, but I would consider this acceptable. So, what does this tell me? Which parts should I consider checking? The hum is definitely caught somewhere around V2.

Any more ideas are highly appreciated!
Thanks

grid-grounding.mp3

[alexradium]

it sounds like the reverb part and that potted circuit are the main problem,i would try to isolate those parts.

[iefes]

Thanks for your reply!
Yes, the potted circuitry has some issues to it, but at the moment I have completely bypassed it, except the Bias power supply circuitry which sits in there as well. My amp is the non-Reverb version so no issues with reverb here.

Another thing I tried was measuring the resistance of all gnd-connections in the preamp to the global chassis connection at the input. They all read below 1 Ohm, so they're good. I also tried an additional 33uF cap across the filter cap of the C+ supply, without success. Then I checked the coupling cap between V1b and V2a (47n). It tested well regarding the capacitance.

Yesterday I tried elevating the heater supply to around 65V by using a voltage divider from the A+ supply. This actually helped a tiny bit to reduce the overall amount of hum but didn't really help with the buzzing. I will, however, include this mod eventually.

Is there a reason why the amp would hum even more when I ground the grid of V1b? This is not what I have expected.

Another question: Are the rectifier diodes prone to introduce noise to the circuit when they age? I tested them using a component tester while they were still on the PCB and they tested well, but who knows?

Thanks everybody!

Edit:
I forgot to mention that I got in contact with Mesa and they sent me the attached files to reduce hum produced by the potted circuitry. I'm pretty certain that the hum in my case is not related to that part as I have bypassed it, but the documents might be of interest for somebody else.

[iefes]

Hi again. So, I've worked on the amp again. I've replaced the Bias supply filter cap (220uf @63V) and the rectifier diodes (1N4007) of the main power supply. Unfortunately, the hum got worse, Haha. This makes no sense to me.

I've checked everything over and over again. Polarity is correct, values are fine. While I had the amp open, I changed a capacitor in the NFB path but I'm pretty certain thats not the culprit. Pulling tubes revealed that the amp still is quiet with V2 pulled. Swapping tubes confirmed this.

I've noticed that the tube sockets have more solder on them than they need. There are blobs of solder on some of the socket pins flowing towards the plastic. Could these solder blobs cause some stray capacitances that produce some noise? Does this make sense? If yes, I will go ahead and try to clean the mess, but the socket are hard to get to unfortunately.

The Buzz has all kinds of frequencies included but the major part seems to be 100Hz. However, my smartphone-app also showed peaks at many harmonics especially a large peak at 1250Hz. Sounds like sawtooth buzz to me.

Please, if somebody has more ideas I'd be so grateful. I elevated the heater supply to 64V which helped a little bit and I was almost convinced to leave it like that but now I need to start over again.

Thanks in advance!

[Chuck H]

With the potted board bypassed V2 would be an odd place for noise to manifest. It's also odd that changing the rectifier diodes increased the noise. Some amps are sensitive to diode switching noise. Did you use ultra fast, soft recovery diodes? Is it at all possible that the decoupling cap for the HV node that feeds V2 is not connected properly at either end? Where did you ground the heater elevation and did you use a bypass cap on the bottom of that voltage divider? Just some things to check. Also...

I think I remember a thread here once where elevating the heaters somehow injected diode switching noise. This could have something to do with what I covered above.?.

[iefes]

Hi Chuck, thanks for your reply!
I used the same diodes as those that were in there. The schematic notes they're 1N4007s and that's what I put in. I've checked the decoupling cap going to V2 already. It tested fine and I resoldered it after testing. The heater elevation is grounded directly to the leg of one of the two B+ filter caps. I used a 15u filter cap parallel to the 33k on the voltage divider (33k + 220k) providing the elevated DC.

Before I changed the diodes I had already installed the heater elevation and the hum was way better. So I think that's not the issue.

I will open it again now and will hopefully know better afterwards. Cheers!

[g1]

The heater elevation is grounded directly to the leg of one of the two B+ filter caps. I used a 15u filter cap parallel to the 33k on the voltage divider (33k + 220k) providing the elevated DC.

I'd have to see the schematic of this part but it sounds a little off. Why would heater elevation be grounded?