new build troubleshoot

[yunger]

i finally finished my first amplifier project. i gutted my pro junior and built a circuit to suite my needs. it sounds unbelievably good, however it hums. does anyone see any obvious mistakes in my wiring that would contribute to the hum?
btw, the filaments have 2x100 ohm resistors to ground acting as artificial ct.

[Enzo]

Hum comes from a million things, it is not one monolithic problem. You need to determine its source.

DO any of the controls affect the hum in any way?

Is it 60Hz or 120Hz hum?

Pull the first tube, does the hum go away?

If not, pull the second tube, I assume the phase inverter. Does the hum go away?

If not, pull one power tube at a time to see if either one affects the hum more than the other, or at all.

[yunger]

enzo, taking out the first preamp tube completely killed the hum. otherwise it is present even at 0 volume and amplified significantly at around the last 20% of the 250k a volume pot. the tone knob doesn't affect the hum until the volume is at that last 20 or so %. it turns the hum into a hiss at full treble....or so it seems. i can't say i'd recognize 120hz. my hum seems to be about the frequency of when you touch the tip of the guitar cable.

[bdcat]

Hi Nice build you've done there!
You may find that using a screened cable going from the jack socket to the first preamp tube helps-remember only to earth it at the jacks socket end and just insulate the screened cable at the valve base end-this should stop it amplifying any stray signals present as that is the point with highest amplification factor.

Also I have sometimes found that if you can put the heater wires so that they run as close to the chassis as possible (ie. poke the wires as close to the chassis as possible). Although doing this would mean re soldering your other wires on the preamp socket so that they come in directly onto the top of the preamp valve/phase splitter-almost the opposite way to which you have your heater wires and all the other wires going to the valve.
You may find that shortening some of the wires and trying to keep them at right angles from the turret board helps.

And finally I think the best way (provided that your experienced with live working-which obviously I don't recommend) switch the amp on and try moving the wires about lightly with a non conductive plastic pen or drum stick and try to see if the noise/hum changes-remember to keep your left hand in your pocket and only poke about using the stick with your right hand!

Goodluck

Brett

[Enzo]

120Hz is power supply ripple. 60Hz is essentially everything else. If pulling V1 kills the hum, then I doubt power supply is an issue. The two are an octave apart, so are the same note, and sound very similar.

Your schematic would help. then we can refer to specifics.

I think his heater wiring is OK. Two schools of though, both of which make quiet amps - the up and over twisted pair like this, and the down on the metal twisted pair like Brett's.

I agree with exploring the input wiring

[yunger]

thanks brett.
i was under the impression that wire were better off next to the chassis as it provided shielding. however, i've seen so many conflicting wiring and grounding methods. i just picked one and ran with it. hopefully i didn't provide an example of how not to do it. i've chopsticked every wire in there to death with no results. i wonder if i have grounding problems. here's how i did it:
input>tone/volume>*ground@terminal below tone/vol pots*<preamp circuit.....power circuit>filters>bridge rectifier>*ground@pt mounting bolt*.....artificial ct>*ground@indicator lamp*.....output is grounded at the jack.
as for shielded wire, i didn't have any. a wild goose chase around town proved fruitless. i need to order some.

[loudthud]

If there is a ground wire connecting all the filter caps under the board, that's the problem. Cut the wire at the last (preamp) filter cap and let it get it's ground at the tone stack.

[yunger]

here's my schematic.

the 22uF that supplies the preamp/pi is grounded with the preamp and tone stack. i should've mentioned that.

[loudthud]

OK. So all the blue grounds from your schematic go to the same point and not back to the grounds of the other filter caps? Is anything connected to the ground lug to the right of the input jack?

[yunger]

yeah, everything in blue goes to ground at the same point along with the preamp/pi filter cap and the input. the ground lug to the right of the input jack is just there to keep the input wire from sliding down near the eyelet board. the hole was already there from the pro junior circuit board so i just used it.

[loudthud]

Yunger,

One of the most difficult things to get right on a guitar amp is the ground. Opinions vary as it seems everyone has their own way of doing it. I for one follow Kevin O'Connors method to a large degree. This is contrary to the practice of many people on this and other boards. The first deviation of the two methods is the grounding of the raw unfiltered HV supply to a transformer bolt. Second is where the main filters get their ground and where the output stage ground is connected. I call this more common method the Fender method because it follows early Fender amps pretty closely. There is a good tutorial on the Hoffman board. In my understanding of how the Fender method works as well as it does, you have followed that method and there is still hum. What has not been described so far is exactly where and how the HV, filter and output stage connections are made to the chassis. A description of these connections should help others here to assist you in correcting the problem. To follow my method you will need to re-do much of you work and your amps looks so nice as it is, I understand why you wouldn't want to do that.

[Diablo]

Maybe I spotted a problem, maybe not.
Above the circuit board main filter caps, you have a terminal strip with a green wire soldered to one side of a terminal, and the other side of the terminal you have a capacitor connected. It looks like the terminal is also the mounting lug for the terminal strip and you are grounding out to the chassis there. Did you intend that to be a ground connection? From the schematic, I think not, but I'm not sure.

[yunger]

here's a map i quickly drew of the grounding points.

[loudthud]

Your grounding looks like I would expect it to following the Fender method.

2x 47uF is quite a bit of filtering for a 2xEL84 amp. If you removed one of those caps, the hum will probably go down. The wire from the transformer bolt to the board should be a nice fat low resistance wire. You can probably measure a DC voltage across it of 10mV. Double or triple it up. Make sure connections are tight.

I'm assuming if you listen to the hum you will hear kind of a low organ note with a buzz on top of it one octave up. To reduce the low note part, move the heater CT from ground to the cathodes of the EL84s.

[yunger]

2x 47uF is quite a bit of filtering for a 2xEL84 amp. If you removed one of those caps, the hum will probably go down. The wire from the transformer bolt to the board should be a nice fat low resistance wire. You can probably measure a DC voltage across it of 10mV. Double or triple it up. Make sure connections are tight.

I'm assuming if you listen to the hum you will hear kind of a low organ note with a buzz on top of it one octave up. To reduce the low note part, move the heater CT from ground to the cathodes of the EL84s.

that's interesting. i didn't know you could over-filter. i have 33uF's...would they be more appropriate? the wire i used was 18 awg...i'll add another one next to it. i'm going to try to narrow the source of noise as per the geofex article also. i'll report back.

[loudthud]

The filter caps draw current in little bursts from the PT. Because the caps just temporarily store the charge and then pass it along to the load, the intergal of the little bursts must equal the DC consumed by the load. What this means is the as the capacitance is increased, the bursts get shorter but higher. It's an area under the curve kind of thing. When these little bursts get into the signal path, you hear a 2x line frequency buzz. I'd try just lifting one of the 47uF caps and see what happends.

These little bursts flow in the transformer , the bridge rectifier and the wires to first filter cap. There are small voltages drops in each wire of several millivolts. Because the chassis ground is midway along the ground side wire, there is a buzzy voltage difference between ground on the board and the chassis. It would be better if the ground side wire went first to the cap and them to ground, but that's another can of worms.

[yunger]

i've been dividing and i think i'm close to conquering here.
i've narrowed the loudest hum to be coming from between the two triodes by grounding grids. the volume and tone knobs are microphonic, r8 & r9 (voltage divider/attenuator) are SUPER- MICROPHONIC. the amp buzzes like crazy when my finger gets within 2" of them. does this lead you to any conclusions about grounding or layout in that vicinity? possibly a bad mechanical connection? i'm questioning the placement of both r10 & r7 but i'm not experienced enough to know for sure.
as a side note, w/ volume knob turned all the way down, volume is not completely off. maybe this is a clue to grounding issues.
-tried many different tubes to no avail.
-lifting a 47uF cap didn't kill much hum that i could tell.
-i still need to try shielded wire and thicker ground wire when the order arrives.

anyway, i labeled the schematic and drew a picture of the layout so we can know for sure what we're referencing.

thanks for sticking w/ me on this. i'm determined to be a success story.

[loudthud]

Microphonic components should respond to tapping with a chop stick or insulated object like the handle of a screwdriver. That's different than a high impedance circuit that hums when you put your finger near it. Shielding will fix that. If a cap does it, try reversing the leads so the outside foil of the cap is connected to the low impedance side of the circuit like ground. Carbon composition resistors can be microphonic, they act like little carbon microphones. Even film resistors can be microphonic. At my work we found some surface mount resistors that were microphonic (panasonic). The volume control issue could just be a bad pot or it could be a bad ground. In the tone/volume circuit keep the leads short and close to ground. Run the ground of the input jack to the board and ground it at 1st tube's cathode resistor.

[guitician]

Could it be the phase of the heater connections? I noticed you used 100 ohm balancing resistors to ground, but are pins 4&5 on V1 connected to 4&5 on V2? Just thought that maybe their is some cancellation not happening.

[yunger]

Could it be the phase of the heater connections? I noticed you used 100 ohm balancing resistors to ground, but are pins 4&5 on V1 connected to 4&5 on V2? Just thought that maybe their is some cancellation not happening.

v1 pins 4/5 go to v2 pin 9. i haven't read voltages on the heaters yet. glad you reminded me.

[guitician]

Yeah, maybe swap you V1 heater connections to see if the hum changes. I remember seeing amps with no center tap on the heater windings using a 100 ohm wire wound pot with the center wiper grounded, and either side connected to the windings. Then it would "Cancel" hum from the filaments as you turned the wiper to a "Null" point.

[yunger]

time for an update for those who care. i've got some interesting things going on here.

-i used shielded wire on input and to/fro the tone/vol pots. (small change)
-grounded filament ct at power tube cathodes. (no change, hum not coming from heater)
-grounded input jack at first triode's cathode ground. (seemed to help)
-all grounds in the tone/vol stack routed individually to ground lug. (seemed to help)

after all this, there is still hum but i know where it is coming from.
grounding the grids of the second preamp tube, the p.i., completely kills the hum. so the noise is coming from the first tubes circuit.
-grounding the grid of the first triode leaves me with a loud, constant hum, uneffected by the tone/vol knobs.
-grounding the grid of the second triode gives a constant hum amplified and colored by the tone/vol knobs.

here's where things get interesting:
putting a 25uF bypass cap on the second triodes cathode resistor completely kills the hum. problem not solved though. with bypass caps on the first 12ax7, the tone changes but not how you would think. gain and volume seems to go down. headroom is actually increased, significantly. the amp's vol/tone knobs become scratchy. the guitar's vol/tone knobs and pickup switch become scratchy as well. my holy grail reverb pedal no longer reverbs but actuall acts like a gain knob and with the reverb knob at fully clockwise, there is no volume. i guess dcv is getting where it shouldn't but i don't know why.

so there you have it. i removed the bypassed caps and now i'm back to a noisy but properly functioning amp. if you suffered through this description and have any ideas please share.

[guitician]

It seems to me like a ground is open, by what you describe happening with the cathode bypass. Could there be a broken conductor inside some wires insulation?

[loudthud]

Voltage and resistance checks can reveal many things that visual inspection will miss. Voltage checks for when the power is on, resistance checks for when the power is off and the power supply caps have been discharged. Using the ohm meter function when the power is on or caps not discharged may damage your meter. It's a good idea to verify the voltage in a circuit is zero before switching to the ohm function.

Are all the points that are supposed to be grounded actually connected? They should show a very low resistance to ground (and a very low voltage when power is on). Pick a ground point and connect the black lead of the meter to it with an alligator clip and ohm to every other ground point. Ohm right to the component.

I suspect a bad 25uF cap (shorted). When installed it shorts the cathode to ground, the plate voltage falls significantly and the tube is in grid leak territory causing scratchy pots. A simple check of plate and cathode voltages (before and after) would confirm this. An ohm check of the capacitor probably will show a short, but different meters respond differently to good capacitors. Measure some unused capacitors to get a feel how your meter responds. Some meters will show a low resistance at first and go higher and higher as the cap charges up eventually showing over-range. See how different value caps read on different resistance ranges. Some meters won't give a stable reading, especially if they auto-range. Turn off the auto ranging.

The meter may also act strangely when a resistor and capacitor are in parallel or there is a cap connected to one end of the resistor. Many meters will correctly read the resistor after a second or two. Plate resistors in the preamp will do this. Remember when measuring across a resistor in circuit, the reading should never be higher than the value of the resistor. You can always unsolder one end of the resistor to verify it's the correct value.

Resistors can be defective when new, damaged by too much heat when soldering, damaged by a circuit problem, or be misread by color code. Many film caps will short if the body is touched by a hot iron. Voltage and resistance checks usually show problems when this happends.

[yunger]

I suspect a bad 25uF cap (shorted).

loudthud, you were right on the money here. i had a shorted cap. good thing i had a solid earth ground connection.

i'm going to send individual wires from each cathode resistor of the first tube to the ground lug. same with the voltage divider, straight to the ground lug.
i'm getting closer and closer to an all out star ground system here. i may as well just go full steam ahead with that but it sure would be ugly. i'll save that for a last resort.

[loudthud]

The grid of the second tube stage after the voltage divider of the 220K and the 270K is very sensitive to noise (hum) pickup. If it were me, I'd solder those resistors between the tube socket and the eyelet board. Keep the leads as short as possible on the tube socket end. If that's not possible, use shielded wire. Same applies to the 68K on the first tube stage but I see you located the resistor at the tube socket.

[yunger]

she hums no more....at least not intolerably. i rearranged grounding points in the preamp and found the best positioning of the grid wires and foil taped them in place. if it gets any better than this, it wouldn't even be worth the trouble.
thanks for the help guys.

[yunger]

i conquered the hum and now i'm fighting constant hiss, unaffected by knobs. it sounds like a tv when the cable goes out. i've located it to be coming from around the second triod (see schem & layout):
-grounding the pot side of r8 does nothing but cut guitar signal.
-grounding the pi side of c6 results in absolute silence.
-lifting r9 from ground makes noise and signal louder and introduces hum.
-bypassing r8 makes the hissing tolerable, probably the normal level of hiss.
the attenuator(r8/r9) seems to be contributing to the noise but i have absolutely no headroom without it. any suggestions?

[martin manning]

I assume you've verified that it isn't the tube. Try changing out the plate resistor R11?

MPM

[yunger]

yeah martin, i've tried a bunch of 12ax7s and 12at7s and the noise is there in all of them. r11 measures the right resistance in the board but i can throw another one in there just in case.

[yunger]

i changed r8,r9,r11 with measured, new, 1w carbon films. i also reflowed solder on all surrounding locations. no change in noise.
is there anything theoretically wrong with having a voltage divider in this location? i thought it was a common location, esp since r8 cuts the head off of really high frequencies.

[yunger]

do i even need to cut signal before the second triod? i don't have a scope nor experience to know where the amp is breaking up. i suspect that i am getting pure power tube distortion after reading about the nature of cathode biased el84s w/ no negative feedback. could i move the voltage divider after the second triode?

[yunger]

don't know if anyone is reading this (long thread) but, i put a 470pf across the plate resistor of v2 and noticed a reduction of what seemed to be a seperate, faint hiss. the noise i'm trying to kill is a much more overbearing wushhhhhhhhh.
maybe i should start a new thread for this issue.

[loudthud]

You don't have a whole lot of gain before the voltage divider. You might try removing the voltage divider but installing a resistor between pin 2 of the tone control and pin 3 of the volume pot. Something between 220K and 470K will get you back to the same amount of attenuation.

[yunger]

thanks loudthud. that sounds cool. i'll giv'er a try.