I finally finished one of my 6L6 monoblocks. Aluminum was tough to polish, but at last I was able to wire the things. The other one is still sitting on the bench, not quite done.
First, some photos. Then, the bad news
Here's my problem. The amp started up fine, fed by a discman into a decent speaker. Upon playing the music, though, this horrible distorted bass rumble started blaring. I shut down the amp and it "killed" the home-theater voltage regulator thing it was plugged into, resetting the TV, cable box, etc. Surprisingly, the fuse didn't blow.
The amp, ignoring the noise (which I suspect is motorboating), sounded pretty good otherwise. The cymbal hits on Miles Davis' Nefertiti album came through pretty decently.
Back at my bench, I put on some gloves, tipped the amp onto its side, and measured some voltages. Or tried to. Voltage on the first motor run cap, the 100uF main B+ reservoir, started at 450V and rapidly rose to above my crappy tester's 500V limit. I then stupidly blew up said DVM when I put the leads from one side of B+ AC to the otherwords, probably around 1000V. Damn.
I'd had the bias pots at around half way. Without being able to measure tube current, I turned the 20-turn 25K trimpots a few times in one direction, then a few times in another, without getting much difference in sound.
One last thing. Turning the amp on and off back in my workshop didn't blow any fuses. I left the CD player off, but still plugged in, and turned on the amp. It was quiet, just the faintest bit of hiss and B+ ripple noise. I tapped the side of the chassis firmly (as it was vertical). The rumbling hum noise started up again, building pretty quickly to a steady loudness. What the hell? I made sure that not too much stuff was dangling around, even super-gluing a stiff piece of wire to the top capacitor to keep it from vibrating.
My questions - if it is motorboating, is there anything I can do? I made sure not to mess up the wiring of the OPT, but those long runs of DC between those last 22.5uF motor run caps and the input stages might be a problem. My main issue though is that I've never had this problem of crazy noise, and don't know what could be causing it. Someone suggested that tapping the bias voltage off the B+ could lead to a feedback loop, could this be it?
First, some photos. Then, the bad news
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Here's my problem. The amp started up fine, fed by a discman into a decent speaker. Upon playing the music, though, this horrible distorted bass rumble started blaring. I shut down the amp and it "killed" the home-theater voltage regulator thing it was plugged into, resetting the TV, cable box, etc. Surprisingly, the fuse didn't blow.
The amp, ignoring the noise (which I suspect is motorboating), sounded pretty good otherwise. The cymbal hits on Miles Davis' Nefertiti album came through pretty decently.
Back at my bench, I put on some gloves, tipped the amp onto its side, and measured some voltages. Or tried to. Voltage on the first motor run cap, the 100uF main B+ reservoir, started at 450V and rapidly rose to above my crappy tester's 500V limit. I then stupidly blew up said DVM when I put the leads from one side of B+ AC to the otherwords, probably around 1000V. Damn.
I'd had the bias pots at around half way. Without being able to measure tube current, I turned the 20-turn 25K trimpots a few times in one direction, then a few times in another, without getting much difference in sound.
One last thing. Turning the amp on and off back in my workshop didn't blow any fuses. I left the CD player off, but still plugged in, and turned on the amp. It was quiet, just the faintest bit of hiss and B+ ripple noise. I tapped the side of the chassis firmly (as it was vertical). The rumbling hum noise started up again, building pretty quickly to a steady loudness. What the hell? I made sure that not too much stuff was dangling around, even super-gluing a stiff piece of wire to the top capacitor to keep it from vibrating.
My questions - if it is motorboating, is there anything I can do? I made sure not to mess up the wiring of the OPT, but those long runs of DC between those last 22.5uF motor run caps and the input stages might be a problem. My main issue though is that I've never had this problem of crazy noise, and don't know what could be causing it. Someone suggested that tapping the bias voltage off the B+ could lead to a feedback loop, could this be it?
I doubt that it's motorboating because you don't have any feedback loop; and you also say that it is quiet when your cheapie CD player is not hooked up to it.
Chances are that the output of the CD player has a DC component on it and your amp does not like that. Insert a coupling capacitor between the CD's output and your amps input. That will probably solve the problem. Otherwise use a different source that has a true line out (you will probably need a volume control then).
Hope this helps...
Good luck, Daniel
Chances are that the output of the CD player has a DC component on it and your amp does not like that. Insert a coupling capacitor between the CD's output and your amps input. That will probably solve the problem. Otherwise use a different source that has a true line out (you will probably need a volume control then).
Hope this helps...
Good luck, Daniel
One more thing... I didn't read your entire post the first time...
Try the tapping test again with the amp's input shorted to ground. See if you can get the hum thing started again with the input shorted to ground. If the hum starts up again, it's probably a bad / microphonic tube, or one with bad cathode to heater insulation.
If the hum doesn't start up with the input shorted, then it's most likely what I said in my original post.
Motorboating is a feedback problem, and by your schematic, it looks like the amp currently has no feedback.
Daniel
Try the tapping test again with the amp's input shorted to ground. See if you can get the hum thing started again with the input shorted to ground. If the hum starts up again, it's probably a bad / microphonic tube, or one with bad cathode to heater insulation.
If the hum doesn't start up with the input shorted, then it's most likely what I said in my original post.
Motorboating is a feedback problem, and by your schematic, it looks like the amp currently has no feedback.
Daniel
Hi sorenj07,
A quick look at your wiring and description of your fault tells me you are indeed suffering from motorboating. The culprit is the way you ran the grounds. Look at your schematic. The circuit ground should be near the local filter cap. Imagine each ground connection as a resistance and the problem should become clear to you.
You blew the fuse on your supply for your home theater. By chance was that a computer type UPS? If so, you can't use it for audio gear.
-Chris
A quick look at your wiring and description of your fault tells me you are indeed suffering from motorboating. The culprit is the way you ran the grounds. Look at your schematic. The circuit ground should be near the local filter cap. Imagine each ground connection as a resistance and the problem should become clear to you.
You blew the fuse on your supply for your home theater. By chance was that a computer type UPS? If so, you can't use it for audio gear.
-Chris
Thanks! I got the chassis made up for me at a place down in Dorchester, I forget the name but they use an industrial ultra-high-pressure water jet with sand in it to cut through even the thickest stuff. I gave a plan to the guy outlining just the basic stuff - all the bigger cuts that'd be hard to make on .1" material. After that, I just followed my plan and drilled the screwholes etc.Sal Brisindi said:
Hi, thanks a lot for the schematics. Why would adding feedback be so necessary though? I suspect that I have around the right values for capacitors that i need, but shouldn't I try getting the amp making decent music first before slapping some NFB in there? Also, what's the electrical difference between the two versions?the_manta said:
Well, it doesn't hum until I tap it with the ground shorted. This makes sense because I'm using some old 6SN7's along with new - these 6SN7's date back to the very very first linestage I ever built and scrapped once I realized things like keeping heater wiring separate, how to build a good PSU, and how not to put PT and choke right next to each other in linedanFrank said:
Should I just try different 6SN7's? I recall tipping one of my 6L6WXT+'s over onto a table and it making a bit more rattling but I suspect that it's OK. I don't know whether it was a computer type, but I suspect so.anatech said:
As for motorboating, I just recently read somewhere that it only is possible if there is a negative feedback connection - what is your take on this? The grounding job can probably use some cleaning up, but unfortunately I don't quite see the solution to this problem - would running a copper connector from the main B+ cap (closer to the PT) up to near the middle of the signal bar be better? could I try lifting signal ground with a 10 ohm 1/2w resistor?
Hi sorenj07,
Don't assume wire resistance is negligible. "When is a ground, not a ground? in other words. It's a hard lession to learn sometimes.
Does that help?
Also, I should have said this earlier. Nice Job!
-Chris
Edit: I should have stated that you created a feedbak loop via the supply, which includes your ground.
Well, not a negative feedback connection. By definition, it's positive feedback (also when negative feedback goes wrong and becomes positive). But, current flowing through resistive connections (wire is not 0 ohms) creates a small voltage. Current increases making these bigger in amplitude. That is what I meant by looking at your layout. If you drew it and replaced the positive supply wires with small value resistors and the same for your ground traces, I think you may see the answer.
Don't assume wire resistance is negligible. "When is a ground, not a ground? in other words. It's a hard lession to learn sometimes.
Does that help?
Also, I should have said this earlier. Nice Job!
-Chris
Edit: I should have stated that you created a feedbak loop via the supply, which includes your ground.
Sorenj07,
You do not have a high gain circuit, so motorboating is also rather funny to me. The usual weird first question, with respect, in detective "where were you at the time of the murder" style. You are sure of all the connections, etc.?
After that I would support Chris. From what I see (I did not enlarge) your earth (common) goes from your loudspeaker terminals picking up everything on the way to the input socket and then with a length of wire to the power supply caps. The first thing I would try is common of output and 6L6s directly to the reservoir caps' negative. All the other commons (drivers last) with a separate wire to the same reservoir caps' negative point. You could also earth the chassis there.
Forget about mechanically fixing caps etc. at this stage. There is neither your problem nor microphony there (urban legends exist thataway). Then also as others suggested a bad tube, though I doubt that; still. (Hope we are not all overlooking something silly.)
Good luck!
You do not have a high gain circuit, so motorboating is also rather funny to me. The usual weird first question, with respect, in detective "where were you at the time of the murder" style. You are sure of all the connections, etc.?
After that I would support Chris. From what I see (I did not enlarge) your earth (common) goes from your loudspeaker terminals picking up everything on the way to the input socket and then with a length of wire to the power supply caps. The first thing I would try is common of output and 6L6s directly to the reservoir caps' negative. All the other commons (drivers last) with a separate wire to the same reservoir caps' negative point. You could also earth the chassis there.
Forget about mechanically fixing caps etc. at this stage. There is neither your problem nor microphony there (urban legends exist thataway). Then also as others suggested a bad tube, though I doubt that; still. (Hope we are not all overlooking something silly.)
Good luck!
Motorboating and NFB
Oh yes....
Motorboating as a problem with NFB is meant as a stability problem. Yes, it is eventually positive fb, but then because of poorly proportioned l.f. time constants all phasing together to create positive fb from the mid-band nfb in the end (below the audio band).
That statement is also conditionally true. Insufficient decoupling can also cause motorboating without any signal nfb as we know it. It is then as above; the wrong time constants (as in all co-inciding in frequency) again being the problem. But the possibility is less with a low-gain circuit such as this.
Oh yes....
Motorboating as a problem with NFB is meant as a stability problem. Yes, it is eventually positive fb, but then because of poorly proportioned l.f. time constants all phasing together to create positive fb from the mid-band nfb in the end (below the audio band).
That statement is also conditionally true. Insufficient decoupling can also cause motorboating without any signal nfb as we know it. It is then as above; the wrong time constants (as in all co-inciding in frequency) again being the problem. But the possibility is less with a low-gain circuit such as this.
Hi Soren,
Nice amp. I've seen your others on facebook and they always look good. I had a few things to add/ask.
Check your cathode voltage on your second 6SN7 triode section. They are only rated for 100V htr-cthd, and since your plate resistor value isn't shown, I'm not sure what your plate voltage on the first section would be. Sounds like when you tap it, something mechanical inside one of the preamp tubes might be shorting (mainly the grid of the second triode stage contacting the heater, and when it cools it retracts back into place.) Are they both russian 6H8C?
Also, how do you like the Edcor output transformers? I was thinking about getting some for another project.
good luck,
-micah
Nice amp. I've seen your others on facebook and they always look good. I had a few things to add/ask.
Check your cathode voltage on your second 6SN7 triode section. They are only rated for 100V htr-cthd, and since your plate resistor value isn't shown, I'm not sure what your plate voltage on the first section would be. Sounds like when you tap it, something mechanical inside one of the preamp tubes might be shorting (mainly the grid of the second triode stage contacting the heater, and when it cools it retracts back into place.) Are they both russian 6H8C?
Also, how do you like the Edcor output transformers? I was thinking about getting some for another project.
good luck,
-micah
thanks! i'll still try to puzzle out where the grounds go but the replies below might just have taken the burden off. it makes sense to put the ten ohm 6SN7 cathode caps to the PSU ground since the bulk of current passes through that point. in fact, the large amounts of current run past all the small-signal ground stuff, where i would really want it at one end or another. in this case, re-orienting the ten ohm resistors is kind of a hassle, whereas all I have to do to equalize the ground is make it an I-shape, with PSU caps bottom left and bottom middle, HV and filament CT on bottom middle, chassis ground bottom right, and signal ground top left and power/OPT common on top right. basically move the vertical ground connection to the right between signal ground and the 6L6's. sound good?anatech said:
Johan Potgieter said:
Boris_The_Blade said:
thanks! i'll still try to puzzle out where the grounds go but the replies below might just have taken the burden off. it makes sense to put the ten ohm 6SN7 cathode caps to the PSU ground since the bulk of current passes through that point. in fact, the large amounts of current run past all the small-signal ground stuff, where i would really want it at one end or another. in this case, re-orienting the ten ohm resistors is kind of a hassle, whereas all I have to do to equalize the ground is make it an I-shape, with PSU caps bottom left and bottom middle, HV and filament CT on bottom middle, chassis ground bottom right, and signal ground top left and power/OPT common on top right. basically move the vertical ground connection to the right between signal ground and the 6L6's. sound good?anatech said:
thanks! I'm pretty confident that between me and all of you out there, this problem can get sorted out. as for connections, I can always reflow a few for good measure. I hear what you say about where signal ground is on the buss; the I-shape ground that I mention above is probably how I'm going to try to deal with it, or at least the thing I'm going to try first. It's a quick thing to do.Johan Potgieter said:
Love the Edcor transformers, first off. I've used two 6.6K 60W 4-8-16 OPT's in a stereo push-pull EL34 amp and I really like how it sounds. I've also used 7.6K 25W OPT's in a PP 6BQ5 amp that I have since sold; both I and the client like the sound. Of course it's hard to separate transformer coloration from the rest of a completely scratch build, but at least you know they function to magnetically couple audioBoris_The_Blade said:
My next project, after trying to sell these monoblocks (I'm tired of them already ) will probably be some type of PP 811A monoblocks. A nightmare to bias them, and a 10K a-a 50W OPT is no joke. Should be fun!I just forgot to write in that value, it's 47K 1W. I took it straight from the good Mr. Williamson so I assume that the heater to cathode ratio is all right. If not, should I ditch the CT on the heater line and tie the 6.3V to B+ with a resistor?
It's probably ok with the CT at ground. I plotted the operating point of the first section of the first 6SN7, assuming an approx. B+ there of 310V (using psud). It's about right at 100V, as are other williamson designs (I own a pair of heathkit W4's), but they had slightly lower B+ on the first section, so they may be more around 90V (if i remember test resutls correctly) . The cathode of the next section will be a few volts higher than the plate of the first. You could change your 33K series drop resistor to 47k or 68k and see if that helps. Maybe an older American tube might tolerate it better, that is of course if this is the actual problem... Also note that your first 47k plate resistor is dissipating about 0.9W+, so you may want to change that to a 2W for long term reliability.
Can you put the meter on the plates and cathodes mentioned to see? (do you have a working meter?)
Hi sorenj07,
Great looking chassis, do you remember where in Dorchester you went to get this work done, I live next town over and could use someone that good..
Do you have a scope? One technique to apply is to try and isolate the problem to the culprit stage. I'd remove the output tubes use a variac and check to make sure none of the tubes in this thing are a subsonic noise generator. (Do you have extra 6SN7 to try?)
This should not be making noise even without negative feedback so if you are thinking about adding it later fix this problem first..
Definitely look at your grounding scheme as well.
Good luck..
Great looking chassis, do you remember where in Dorchester you went to get this work done, I live next town over and could use someone that good..
Do you have a scope? One technique to apply is to try and isolate the problem to the culprit stage. I'd remove the output tubes use a variac and check to make sure none of the tubes in this thing are a subsonic noise generator. (Do you have extra 6SN7 to try?)
This should not be making noise even without negative feedback so if you are thinking about adding it later fix this problem first..
Definitely look at your grounding scheme as well.
Good luck..
Regarding the h-c voltage of the 6SN7, the peak spec. is 200V, so that there should not be a problem with the design. Williamson indicated 105V on the phase splitter cathode, with 450V h.t.
But the solution in case of doubt is easy. Elevate the heater voltage to about (say) +60 - +70V, by using a resistive divider from h.t. The standing current needs not be more than 1mA. Bypass that point to common with an electrolytic. This will also assist in eliminating any hum resulting from a small h-c leakage, where the positive heater cannot radiate to the cathode. This is an often used measure, especially for pre-amps.
Soren07, I still feel uncomfortable - we are overlooking something. This behaviour of yours is uncharacteristic, especially in absence of NFB. Of course, one way to eliminate possible microphony is to simply unplug the loudspeaker. Hope you can find a scope, although a meter (especially analogue) would also show whether the l.f. instability is still there. I would also try different tubes as suggested by Kevinkr, but that may not be practical. Also, while "fishing", does shorting the input make a difference? (Further, while you are testing and this occurs, do not let the condition go on for too long. You could be generating spikes in the output transformer with detrimental effect.)
I am really interested in the solution, and wish you luck.
But the solution in case of doubt is easy. Elevate the heater voltage to about (say) +60 - +70V, by using a resistive divider from h.t. The standing current needs not be more than 1mA. Bypass that point to common with an electrolytic. This will also assist in eliminating any hum resulting from a small h-c leakage, where the positive heater cannot radiate to the cathode. This is an often used measure, especially for pre-amps.
Soren07, I still feel uncomfortable - we are overlooking something. This behaviour of yours is uncharacteristic, especially in absence of NFB. Of course, one way to eliminate possible microphony is to simply unplug the loudspeaker. Hope you can find a scope, although a meter (especially analogue) would also show whether the l.f. instability is still there. I would also try different tubes as suggested by Kevinkr, but that may not be practical. Also, while "fishing", does shorting the input make a difference? (Further, while you are testing and this occurs, do not let the condition go on for too long. You could be generating spikes in the output transformer with detrimental effect.)
I am really interested in the solution, and wish you luck.
Thanks! I'm shipping off to college in less than a week and hope to get these suckers done before then. People might remember that about this time last year I was making noises about building a 6L6 amplifier in the style of the first Hafler/Keroes modification of the traditional Williamson. It's been a long-*** time and I've definitely learned a bit since then, no small part because of people on forums like these. Thanks guys!burnedfingers said:
This makes sense. I have half a mind to give the 6SN7's a break and use 47K 2W and 33K 2W instead of the 33K and 22K resistors. Not to mention that bigger resistors will be easier to mold in 3D into pleasing rectilinear shapes 0_oBoris_The_Blade said:
Will do. The place is called "Metalsmiths" and they're located at:kevinkr said:
15 Banton St.
Dorchester Ctr, MA
(617) 265-4040
http://www.bostonmetalsmiths.com/index.html
The people who work there are really chill, especially the water cutting guy. A bit of advice if you plan on working with the guy. He programs CAD into a terminal to operate the water cutter, so if you can come up with the CAD files yourself, you'll save yourself a lot of money that you'd have to pay him to draft it himself. If you have any doubts about the precision of the machine, ask him about the clock he made out of thick stone cut to the exact outline of Germany that he made for a friend
I do have a smallish scope kicking around but not a Variac. I also have an ancient Eico signal generator that desperately needs some calibration/cap replacement, but seems to be working decently.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.