Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => AmpTools/Tech Tips => Topic started by: Mike_J on April 30, 2016, 04:43:58 pm
-
Working on the AB763 one-channel build. Have increased the amps gain by changing the mid pot to 100K, removing the 47K resistor and adding a switch to remove the feedback circuit from the amp. Like the tonal changes but unfortunately increasing the gain injects some hum into the amp.
Want to create a test probe to see if I can identify where the hum is located. However, have some rudimentary questions as I am not sure how to do same.
What I was hoping to do is use a stereo to create a signal. Maybe connect a guitar cable from the stereo's headphone jack and plug it into the input of the AB763 build. First question, have I blown anything up yet? Need to do something a little differently to make that work?
Saw a thread where Sluckey made a probe using a pill bottle. His probe had the negative side of the probe tied to an alligator clip (which would be clipped to the AB763 build chassis as I understand) and the tip was connected to a male banana connector. Guessing I could replace the tip wire with an old guitar cable?
Have a 15 watt solid state Fender Frontman. Do I just plug the probe into the input of the Frontman?
If everything is good up to this point have to touch the probe to the circuit at certain points. Not sure what the process is at this point. Any pointers would be helpful. Also, what am I listening for?
Thanks
Mike
-
1: onlinetonegenerator.com for a signal source, computer audio out to regular amp input. Got cable?
2: As far as a listening amp, I still believe the simplest and cheapest solution is to find a pair of powered computer speakers at a garage sale or thrift store. Should be $1-3. Usually the little amp is in one speaker and the other speaker is a "slave", eg, just a speaker. The really critical thing is that you have to place a "blocking" capacitor in series with the input lead of the computer speakers, because you hit the input of that little transistor amp with 350 volts and you'll send it into the stratosphere. (but only cost yourself $1-3) Such a cap can be enclosed in a pill bottle, that's the traditional cheapskate way to contain the cap that works fine. You'll need a ground wire with an alligator or other clip on the input wire to the speaker set. You'll have to figure out which of the two possible "hot" input wires goes to the powered member of the duo and probably toss out the slave speaker.
3: The big bonus arises if the comp speakers have a headphone out. The lets you hear your probe audio much better than a little speaker.
I would do the computer speakers before risking blowing up your Frontman or pretty much anything else of value. Sure, you can use the Frontman. But you MUST have that blocking cap, or you'll risk obliterating anything solid state working on tube gear. Ideally you want a .047/600 volt cap; the most important thing being the high voltage rating. Could be any ufd from say .01 to .1
-
1: onlinetonegenerator.com for a signal source, computer audio out to regular amp input. Got cable?
Don't have a cable but could probably come up with a converter that would work. Have some old speakers from a dead PC but don't see a headphone jack. A little concerned with using the headphone jack from my computer to the input of the AB763 build. Losing the computer would be a bigger loss than either the Frontman or old stereo by a considerable margin. I do like the onlinetonegererator.com recommendation. Appreciate that reference. That would make using the computer preferable as long as I don't destroy the computer while using it this way.
Not sure how I would use the probe. Want to use it to find the source of the hum I am getting. Have a hand held oscilloscope that I bought so I could find the outside foil on capacitors. Would that be a better tool for finding the source of the hum?
Thanks
Mike
-
In my estimation, yes...provided you are working within its max input voltage ratings.
"Hum" really has little or nothing to do with an input signal....right? You don't get hum when you hit the guitar really hard or play a particular note. Hum is produced for the most part by AC getting into signal wires. And it is constant; The AC might come from the heater string, in which case we would expect 60 Hz hum. Sometimes, the power supply itself (or a wire running too close to it) produces the hum and that generally produces 120 Hz hum with a full wave or a full-wave bridge rectifier...an octave higher than 60 Hz hum. The two are not that easy to distinguish but I believe there is YouTube you can search for titled something like "60 Hz vs 120 Hz hum".
Hum *might* occur with *anything* plugged into the amp and disappear when you yank the input plug; in most amps, removal of the input plug shorts the hot terminal of the input jack to ground. This, if it is the case, gives an immediate clue that your issue is in or at the very first preamp stage. Whatever that stage does is amplified the most of anything in the amp, so anything wrong there gets magnified, a lot.
Hum can also occur in a perfectly working amp *especially* if it is upside down and open, out of the cabinet, under fluorescent lights, or if you are in a room with old noisy dimmers. Or if the cabinet you built doesn't have a shield covering the open side of the chassis. If that's the case, you can spend an eternity chasing down the hum inside the amp and you'll never find it because it isn't there. Just saying.
Hum can occur with bad filter caps. With your scope, provided it can handle the B+ volts inside the amp, you should be able to read AC ripple on the power supply nodes and/or on the preamp tube plates. Be careful of the reveb driver tube, that tube has the highest of the high B+ volts on it. On Fenders, these nodes are the "A", "B", and "C" points on the power supply. You should be able with NO input signal to read AC on any tube plates and that number should be very very low. Luck is kind of in your favor here because you test the lower-voltage preamp tubes before wandering over to the power section.
Hum can occur with a heater-to-cathode short in a tube. These are very rare but they do happen. The only way to prove that is the issue is to substitute tubes. You can swap any of the 12A_7 tubes with any other for this test.
Hum can happen if a signal wire is running too close the a heater wire. AC radiates. We twist the heater wires not just to keep them together, but to reduce the amount of radiation coming from them.
In general, you search for and try to isolate hum by pulling one of the small tubes after another. If you start with the PI tube, pull it, and the hum stops, your hum is absolutely in the PI or the power section which suggest a bad e-cap. If that does not do it, go to the preamp side of the amp and start pulling tubes one at a time. The one that kills your hum either IS the source of the hum or is AFTER the source of the hum.
And finally, the particular way you have increased gain and opened up the feedback loop definitely opens up possibilities in terms of having a wire that was going *somewhere* (and thus creating an impedance or load for some part of the circuit) now connected ONLY to the back of a switch and thus acting like an antenna. Where is that switch? Wherever it is, physically in the chassis, it now can act as an AC field receiver and can send that induced AC to someplace the amp does not like. Maybe you need to place a high-value resistor or a very small cap across that switch so that instead of "opening up" the feedback loop you are "defeating" the feedback effect by 90%. Maybe you need to run shielded wire to that switch.
I can only throw out ideas, the usual, and the bizarre. Some hum problems can be very difficult to chase down and some layout choices can make it almost impossible. That's why among the grizzled veterans on this forum there is an enormous preference towards using proven layouts and doing things quite the way Leo did, with only some improvements here and there. Because without even asking, I know that the two dozen or so folks who have built multiple amps have built amps they could never get the hum out of and they either ripped them apart or rebuilt them using known-good configs.
-
And finally, the particular way you have increased gain and opened up the feedback loop definitely opens up possibilities in terms of having a wire that was going *somewhere* (and thus creating an impedance or load for some part of the circuit) now connected ONLY to the back of a switch and thus acting like an antenna. Where is that switch? Wherever it is, physically in the chassis, it now can act as an AC field receiver and can send that induced AC to someplace the amp does not like. Maybe you need to place a high-value resistor or a very small cap across that switch so that instead of "opening up" the feedback loop you are "defeating" the feedback effect by 90%. Maybe you need to run shielded wire to that switch.
Thought I would start with the wheel that appears to be squeaking the most. Took a picture of the switch and its location so you could see it. You hit the nail on the head with the wire from the switch acting like an antenna. Wiggled the wire going from the switch to the pot and could change the hum significantly. Got rid of a lot of it but some is still there. Running shielded wire makes a lot of sense. Moving the pot closer to the switch also makes sense as far as layout is concerned (put the pot in first, had I put the pot and switch in at the same time don't think the pot would be as far away from the switch as it is now. Also, when I built my Dumble amps I noticed he placed a 22Meg pot across the common leads on the switch. Is this to reduce the impedance or load that you were mentioning? Created a revised layout so you could visually see what I am talking about.
Appreciate your help it is definitely on point.
Thanks
Mike
-
Hard to say. I can tell you that getting rid of hum can be very frustrating, obviously, because you have to build the whole amp (almost) to find out you have it and then all of your parts are nailed into their positions and you are more or less committed.
I think the next thing you might consider as a test to isolate and verify that this is or could be your problem is to disconnect the neg feedback loop WIRE from the place it re-enters the amp. Not at the speaker-connection end, at the re-injection point. The idea is that if that wire is indeed acting as an antenna, then you have to take whatever measures are necessary. You have to realize that the NFB feed from the speaker out is across the OT secondary, one side of which is grounded, so this is a pretty large signal held off ground by 4 or 10 or 12 ohms (the speaker winding) which is pretty close to a short circuit in tubeland. With that loop opened up and thus with no "short" across it, the stock circuit is really asking to be an antenna.
What I am saying is, if you open up the NFB loop it could make quite a bit of difference whether you cut that connection at the speaker jack (leaving the rest of the wire intact) or at the re-injection point, leaving the wire to the speaker jack intact. I myself have never performed this experiment because I tend to play pretty clean and I want the NFB and all its glory, but it certainly makes sense to me that if you ramp the gain of the amp and connect an antenna to the input, you ought to expect to pick up noise and maybe a lot of noise.