5 watt SE guitar amp head with sidetone buzzing problem when strings are strummed

[rock soderstrom]

I have modded a China 5 watt SE guitar amp head and I have a constant buzzing background noise sound (when I strum the strings) that I can't get under control. It sounds like something is vibrating with the strings, but it's not a mechanical problem with the speaker or tubes. I've swapped them all to isolate the problem, it does seem to be coming from the amp and I'm running out of ideas now.

Does anyone know the problem or have similar experiences. What could be responsible?

The circuit now is a really good sounding hot Mini Z clone (but with Master Pot) but the problem was already there with the donor amp. Tubes are 1x 12AX7 and 1x EL84.

 

[rock soderstrom]

This is the circuit at the moment

 

[CJ]

Sounds like oscillation, try moving wires unless it is on a PC board, your gain went up with the deleted tone circuit, you could try a divider to attended the signal. Maybe a NFB resistor from the xfmr secondary,

 

[rock soderstrom]

Sounds like oscillation, try moving wires unless it is on a PC board,

I think you're right, I've already opened up all the wiring harnesses but I can't influence the behavior. I think the problem originates on the circuit board.

your gain went up with the deleted tone circuit

That's true, but the problem already existed in the original version. By the way, the tone stack had very little effect, sounded terrible and choked the amp. That was a very good decision to take it out. The voltage divider in front of the gain pot had originally also swallowed a lot of gain.

you could try a divider to attended the signal. Maybe a NFB resistor from the xfmr secondary,

I'm afraid it will change the character of the amp too much, I'm very happy how it sounds now, really great, wild and super loud for 5 watts. This is not my first 5 watt amp, but I really like it now - good recipe!

I'm going to try a grid stopper before V1b, see what that does and I'll contact the transformer primary to the 240V winding to get the secondary voltages down, they're all too high!

Maybe it has something to do with the poorly filtered DC heating?

 

[Alva Goldbook]

Well, there’s a few silly things happening in the circuit, but tell us what the noise sounds like. Go to YouTube and put in “60Hz sine wave”. If you can tune a guitar by ear you should be able to tell if this is the noise the amp is making. If it’s not 60Hz try the same procedure with 120Hz. If it’s neither and sound like a boat engine or a ambulance siren, let me know.

Some obvious bad things with the circuit.
1) no grid stopper on the power tube. Helps prevent oscillation. Value is not critical. 10k. 4k7. 1k. Even 100k. Just put something there.

2) no grid leak on power tube. Check data sheet for upper limit on this. If I knew the power tube being used I’d tell ya.

3) a resistor bypassing the tone stack? I mean, why?

4) A voltage divider on the B+ before the output transformer? Why? Just use the correct power transformer. Iron ain’t cheap. The tube might be able to handle the higher voltage. Dissipation is what matters more.

5) DC heaters with just a single filtering cap? That might be your culprit right there. On such a low gain build, DC heaters are unnecessary. AC heaters would be fine.

 

[ruffrecords]

It is also worth knowing if the level of the noise changes with the setting of the controls, particularly the one before the output tube.

Cheers

Ian

 

[rock soderstrom]

Well, there’s a few silly things happening in the circuit, but tell us what the noise sounds like. Go to YouTube and put in “60Hz sine wave”. If you can tune a guitar by ear you should be able to tell if this is the noise the amp is making. If it’s not 60Hz try the same procedure with 120Hz. If it’s neither and sound like a boat engine or a ambulance siren, let me know.

Yep, I can tune a guitar by ear, but I have no idea what this has to do with my (no solved) problem. I got no 60/120Hz problems because I am in a 50Hz country, but joking aside. There is/was no hum problem, I never mentioned that. It was whirring noise, similar to when parts of the housing vibrate loudly with the strings.

Some obvious bad things with the circuit.
1) no grid stopper on the power tube. Helps prevent oscillation. Value is not critical. 10k. 4k7. 1k. Even 100k. Just put something there.

There is a grid stopper in front of the EL84, 3.5K!

2) no grid leak on power tube. Check data sheet for upper limit on this. If I knew the power tube being used I’d tell ya.

It is clearly described in my first post which power tube is used (12AX7+EL84 inside), just read everything before you post.
There is a grid leak resistor now but that was not the problem.

3) a resistor bypassing the tone stack? I mean, why?

Nonsense, the resistor replaces the tone stack which I completely threw out because it had hardly any effect, sounded bad and stole a lot of level. But I've already written about that in my previous posts. Again, read first, then post.

4) A voltage divider on the B+ before the output transformer? Why? Just use the correct power transformer. Iron ain’t cheap. The tube might be able to handle the higher voltage. Dissipation is what matters more.

What voltage divider? What are you talking about?

5) DC heaters with just a single filtering cap? That might be your culprit right there. On such a low gain build, DC heaters are unnecessary. AC heaters would be fine.

Not my design, but I know where it comes from. This amp is a generic Chinese design that is sold under many brand names. Gretsch, Epiphone, Harley Benton and so on and the first models had the usual AC heater. There were problems with hum, according to internet forums, and the people in charge went with a (rather poorly implemented) DC heater. I wouldn't do it that way, none of my guitar amps have DC heating, I get them quiet without it.

The main problem with these amps is the design of the PCB, where local oscillation occurs due to crosstalk.

The amp sounds really good now and is loud as hell...

 

[rock soderstrom]

It is also worth knowing if the level of the noise changes with the setting of the controls, particularly the one before the output tube.

No, it doesn't change anything (apart from the overall volume). This is an attempt on a master volume so that the amp can also be played at moderate volumes (in rented apartments, for example). This also had no influence on the problem, as the pot was not even present at the beginning.

This works more or less at the start of the potentiometer's control travel, but from a certain position the potentiometer is then ineffective. There are better solutions for this, but it's not worth it for such a low budget amp. They were sold here for 89€ at Thomann.

 

[Khron]

4) A voltage divider on the B+ before the output transformer?

What voltage divider? What are you talking about?

Likely that 200 ohm / 220K combo, which has an... academic effect there. The 220k is likely more of a discharge resistor, than anything else (imho).

 

[rock soderstrom]

Likely that 200 ohm / 220K combo, which has an... academic effect there. The 220k is likely more of a discharge resistor, than anything else (imho).

Yep, just an ordinary bleeder resistor for the PSU caps, a good thing, no doubt.

 

[dai h.]

(I'm just throwing stuff out there, perhaps pictures and video (sound) would help the diagnosis?)

weird buzzing when you have signal (when you play)?

(I have no idea if this would matter in such a low power amp, but..) output transformer magneto-striction--something something. Sometimes when people obtain attenuators (the kind to reduce output going between the amp speaker output and speaker(s)), they freak out since they hear the OT producing sound (which was present previously but they just didn't hear it before since it was drowned out by the speaker(s)).

(Another one--just a WAG really) A problem is described where the stapled alu shield in the cab isn't making good contact with the chassis and causing weird noises:

 

[rock soderstrom]

Thanks Dai, for chiming in. As already written, it was definitely not a mechanical problem but oscillation. The PCB is the problem from my point of view, with some mods and trace hacks on the PCB I was able to solve it so that it is no longer annoying.

Is this now the best amp in the world? No, but honestly, I would have killed for this amp in the late 80s. Back then we only had the cheap transistor amps that sounded really bad. We're talking about a practice amp that's really fun now.

Now this amp really gets down to business, depending on the volume at the guitar you can go from a little clean to nice overdrive to fuzzy distortion, everything is possible. Surprisingly loud for 5 watts, but with the crude master pot solution and the gain control you can make a lot of noise even at lower volumes. Now the amp is much better than when it was shipped. Definitely.

 

[CJ]

Just picked up a Marshall G215R CD and man is it loud for two 8" speakers, made in Korea with Park speakers, stereo chorus, reverb, nice crunch and that Marshall midrange thing going on,

 

[Alva Goldbook]

Likely that 200 ohm / 220K combo, which has an... academic effect there. The 220k is likely more of a discharge resistor, than anything else (imho).

I'm afraid I gotta disagree a bit here. The 220k bleeder is fine, a good safety drain to have, but a 4 watt 220R resistor on the B+ between the reservoir cap and Node A is both pointless and an unnecessary source of heat that could harm the PCB, especially if it's mounted directly on the board. There's just no point in dropping the B+ voltage. The EL84 can take the higher voltage. If it couldn't, then a different power transformer should have been used. As long as the total dissipation is within spec, it should be fine. If the point of the resistor was to fake tube rectifier sag, that's pointless as it's a single ended amp.

 

[capacitorless]

Wouldn't R10 (220 ohm) between C8 and C9 form a CRC stage for the output primary B+?

If so, wouldn't that kind of be a requirement for a low-hum single ended output stage, due to the (relative...I keed) lack of common mode rejection inherent in a single ended vs push pull circuit design? But what do I know.

Edit: @rock soderstrom if this is a screamer, that newly unused 1M pot (and maybe the unused caps) might be able to be repurposed into a bass cut control. I've found it handy in situations like using a Les Paul in the neck position, for higher gain work. Edit again: I mean without messing with the gain.

 

[capacitorless]

Hey, @dai h. thanks for the trip down memory lane! I used to stop into LV's shop on South Broadway here in Denver for tubes and advice. He knew his stuff. What a character!

(edit: his business was called NBS Electronics, of which NBS stood for No B*ll S*it. I just looked at the video further. I can't say I agreed with his political views, but he sure made them known, haha!)

 

[Khron]

There's just no point in dropping the B+ voltage. The EL84 can take the higher voltage.

Ok Mr. / Ms. know-it-all, but you seem to have overlooked an alternate view...

Wouldn't R10 (220 ohm) between C8 and C9 form a CRC stage for the output primary B+?

Yes it would, and yes it does.

Worth keeping in mind this is a "FIVE" watt little amplifier, which is also single-ended and without global feedback, so virtually NO power supply rejection. Why would you NOT want to filter the supply voltage?

PS. Spoilers - a resistor, chunky as it may need to be, is still cheaper than silicon parts...

 

[Alva Goldbook]

Ok Mr. / Ms. know-it-all, but you seem to have overlooked an alternate view...



Yes it would, and yes it does.

Worth keeping in mind this is a "FIVE" watt little amplifier, which is also single-ended and without global feedback, so virtually NO power supply rejection. Why would you NOT want to filter the supply voltage?

PS. Spoilers - a resistor, chunky as it may need to be, is still cheaper than silicon parts...

If I've ruffled some feathers, I apologize. That wasn't my intention. Can you confirm something to me? This is a Harley Benton GA5, correct? Can you tell me if the power transformer bell ends are facing left to right or are they facing front to back? I can't find many pictures online where I can see the power transformer, but the one I saw had the bell ends facing left to right. Is this the case with your amp?

So one rather surprising thing I found when I first learned how to build a linear power supply was just how little filtering it technically needed. If you look at old Fender amps from the tweed days Leo was putting tiny 16uf caps in his amps and not that many of them. In Merlin's book he lays out the math for this. The formula is

C = I / (2 f Voltage ripple)

So, with a push pull amp, Merlin advises no more than 10% voltage ripple, 5% on a single ended design. But let's go crazy and shoot for 1%.

So now we gotta figure out the current. The hell. How do we do that? Well, the datasheet for a EL84 tells ya the plate draws 48mA, the screen draws 5.5mA. 12AX7s max out at 1.2mA, but in the real world it's like 0.7-0.8mA usually. So to make the math easier I just default to 1. So that gives us 55.5mA, (48+5.5+1+1) so we got the number for I.

Your mains is 50Hz, so 2x50 is 100. So how do we find the voltage ripple? Well, we just calculate 1% of the rectified B+ voltage. I'm guessing it's around 320 volts. So 1% is 3.2 x 100 and we're back to 320. 55.5/320 gives us 17.3uF. Your reservoir cap is 22uF. So you got enough filtering. So what happens if you just yank the 220R resistor? Maybe move the 220k bleeder to where C11 is. Just get rid of C11 and replace it with a 220k 1w resistor. Then put a jumper where 220R used to be. What would you effectively have then? Well as parallel caps sum, you'd have a 44uF Reservoir cap. And the reservoir would feed Node A. I gotta tell ya, that makes a whole lot more sense to me than this 220R 4 watt nonsense going on. Especially on a fragile thin PCB like on this amp.

You should check the EL84 dissipation. Many amp companies foolishly bias them too hot. Put your meter across R14. Give me the real actual resistance. It might be like 221.3R or something. Turn the amp on and switch to DC and tell me the voltage drop. Then tell me the B+ voltage. We'll use this to check the output wattage.

Edit: whoops. I just realized Khron isn't the OP.

 

[Khron]

If I've ruffled some feathers, I apologize. That wasn't my intention.

Well, one of your first posts on this whole forum is telling off the "for funsies" project of a(n arguably) long-time and relatively respected member here. We can't know who you are or what your resume' is (as it were), outside of your very brief history here, so... yeah, it can come across as... questionable

Well, the datasheet for a EL84 tells ya the plate draws 48mA, the screen draws 5.5mA.

That'll possibly depend how (hot or not) you bias the thing, but even with 55mA (quiescent), that'll drop about 12V across that 220 ohms, which is a measly 0.66W. On a 4W resistor. Are you gonna argue it's "too" overrated, or..?

And all this, in the context of that resistor having sod-all to do with the issue presented in the original post.

Oh no, this is not designed or assembled as you would have done it. So..?

 

[Alva Goldbook]

Well, one of your first posts on this whole forum is telling off the "for funsies" project of a(n arguably) long-time and relatively respected member here. We can't know who you are or what your resume' is (as it were), outside of your very brief history here, so... yeah, it can come across as... questionable



That'll possibly depend how (hot or not) you bias the thing, but even with 55mA (quiescent), that'll drop about 12V across that 220 ohms, which is a measly 0.66W. On a 4W resistor. Are you gonna argue it's "too" overrated, or..?

And all this, in the context of that resistor having sod-all to do with the issue presented in the original post.

Oh no, this is not designed or assembled as you would have done it. So..?

I imagine they put in a 4W resistor in there for a reason. Regardless, it’s a bad power supply design. Some of these amps have had reported issues due to the power supply. From the accounts I’ve read elsewhere it sounds like a bad power transformer, or a stressed power transformer. From the pictures I’ve seen, it may not even be orientated correctly. Which could possibly have caused the oscillation issue.

I once got a hair brained idea to drop B+ on a Hot Rod Deville I gutted. I wanted to fake a tube rectifier. I checked the dissipation of the resistor, and used a 2 watter. Then I wondered by my preamp tubes weren’t getting any B+ voltage. That extra stout fireproof resistor went from like 100 ohms to about 10 megaohms. So, needless to say I didn’t try that again.