This thread was started 18 june. No amp is produced yet.
AT the beginning of juli i spotted a ST35 on tradera. It was in factory config, except for the EL84 it was in good shape,
even the can cap was acceptable thus i did not as planned install EFB (yet). Some measurments confirmed that
hum was ok, distortion was ok and output power 17w. 4 matched EL84 and a power cable was all needed.
IT sound good, nu hum no problems whatsoever.
I personally find no reason to change anything on it, just listen to the music.
AT the beginning of juli i spotted a ST35 on tradera. It was in factory config, except for the EL84 it was in good shape,
even the can cap was acceptable thus i did not as planned install EFB (yet). Some measurments confirmed that
hum was ok, distortion was ok and output power 17w. 4 matched EL84 and a power cable was all needed.
IT sound good, nu hum no problems whatsoever.
I personally find no reason to change anything on it, just listen to the music.
Yes, I know.
- Events in life often take priority over amplifier building.
- Some people are slower than others.
- I'm in no hurry since I just finished another EL84 tube amp kit a few months ago, and I'm enjoying listening to that.
- I enjoy learning, which takes time. For example, I now know a lot more about grounding than I did a few days ago.
I did want to add the front-mounted rocker power switch, the CL-90, and binding posts spaced for dual banana plugs. Those took time because it involved fabrication, metal cutting, and a lot of filing. I'm glad that part is over. The decals for the chassis should arrive tomorrow. They are late for some reason. They will take a few more days to apply and finish.
Last edited:
The drawing in post #101 will work, but best practice would have PE/safety Earth to chassis connection kept separate from signal "star" ground to chassis connection, which would be right at the "star".
An interesting thing to think about is that single-ended stereo interconnections from a stereo common-grounded source will always cause its own loop. The shields of the two interconnecting cables are connected at the source end, so it's unavoidable. Sometimes folk will lift the input jack and its associated grid leak and cathode resistor from signal "ground" by 4R7 or 10R (you would need to cut a circuit board trace). This raises the question of where to attach feedback signal returns (from the OPT secondary 0 tap).
All good fortune,
Chris
An interesting thing to think about is that single-ended stereo interconnections from a stereo common-grounded source will always cause its own loop. The shields of the two interconnecting cables are connected at the source end, so it's unavoidable. Sometimes folk will lift the input jack and its associated grid leak and cathode resistor from signal "ground" by 4R7 or 10R (you would need to cut a circuit board trace). This raises the question of where to attach feedback signal returns (from the OPT secondary 0 tap).
All good fortune,
Chris
Now I am confused. I thought we were trying to avoid that exact situation? I could ground the heavy bare copper wire to the chassis right at the 2-terminal strip. However, we then have two different connections to the chassis, some distance apart. Why would we do that?
I have seen people mention interconnects and digital coaxial cables with the shield connected at only one end. I never have known what to think about that.
Last edited:
Does this text below relate to what you said? This part confuses me, and I don't understand what it means in relation to my situation:
Wouldn't this result in this arrangement?
Or maybe the text refers to this arrangement, which I saw somewhere online?
I have gotten a lot of mileage out of my drawing. I'm glad I made it.
Last edited:
You're right in thinking that that would be non-ideal for signal return paths. In the case of PE/safety Earth though, no current flows except under fault conditions, so the safety issue of mechanical connection to chassis predominates.
This is possible with two conductors inside the shield, the inner conductors carrying signal hot and signal return, and the shield being only a shield. A glaring problem with either twisted pair or simple coax interconnections is that signal return (from a common-grounded source to a common-grounded amplifier input) for each channel is split equally between both channels' signal return conductors, and noise immunity comes from having equal signal current in each pair of conductors per channel. Stereo interconnections violate this, and badly. There's an additional issue of the shield also carrying signal return current, so the whole thing stinks on ice.
If we could somehow stop using the grandfathered-in RCA/"phono" connectors, we could massage a lot of this drama away, but we're stuck with 'em except for totally DIY.
Another ground loop issue to keep honest folk up at night is the path around the PE wiring, which unless measures are taken to break the signal "ground" to PE "ground" path are another path for signal ground current. It's not surprising that folk are confused about "grounding" - there are straightforward rules and those rules (one and only one path) are always broken.
All good fortune,
Chris
This is what "balanced" connectors are for on expensive amps and preamps, right? I can't afford that stuff. 😉
In any case, I just edited my last post above with another question and some reference text about grounding inputs. Perhaps someone can assist with my understanding of what it means.
I find this unnecessarily confusing because it uses the word "ground", which has about a dozen different meanings, interchangeably. Signal "ground" and PE/safety "ground" are connected to chassis for different reasons and would in a perfect world not even be connected to each other. For DIY audio many folk will include a device to keep them separated except in fault conditions, a "ground breaker". PE/safety Earth must still always be connected to chassis with dedicated bolt, lockwasher, nut. Signal "ground" only needs to be connected to chassis under fault, so can be a Volt or two away.
All good fortune,
Chris
The new drawing in post #107 shows multiple signal grounds to chassis, so isn't ideal. Also, PE/safety ground to chassis needs to be separate from signal ground to chassis.
All good fortune,
Chris
All good fortune,
Chris
A note on "star grounds": there really is no such critter; there are only multiple small ground loops. If we try to make a hybrid of the multiple small loops ideal with a circuit board which already has its own internal grounding scheme, we have a jackalope. Also, even an ideal star ground ignores all the other ground paths in a stereo amplifier in a system with amplifier connected to AC power with PE and other stereo components, also with PE. So, in that respect also it's a fiction.
All good fortune,
Chris
All good fortune,
Chris
I am so glad I have that drawing now because I would never achieve an understanding of this without it. It took me a number of hours, but it was well worth what I am learning.
So, I understand these three points above, I think. So, why not do this? It seems to satisfy all criteria.
I should correct this to say that the rules are always broken in conventional home installations, and it's usually good enough for rock'n'roll. Professional installations demand a higher standard, and instrumentation demands a yet higher standard. "We get away, almost every day, with what the girls call, what the girls call, what the girls call murder" -Liz Phair
Post #113 shows one signal ground to chassis connection at one of the input jacks. I'd have two reservations about it: first, the fault currents in a fault condition would be carried on PCB traces of questionable survivability. And second, signal return currents in the interconnecting cables would divide unequally in some complicated way.
The previously quoted article about guitar amplifiers doesn't really apply well here, although it's reasonable for its intended application. For a stereo amplifier with common (to both channels) power supply hard choices must be made. Usually this is to ignore the signal voltage differential between input jacks and to chassis ground both of them. Yes, it's a loop!
Like war, there are no winners, only varying degrees of losing.
All good fortune,
Chris
The previously quoted article about guitar amplifiers doesn't really apply well here, although it's reasonable for its intended application. For a stereo amplifier with common (to both channels) power supply hard choices must be made. Usually this is to ignore the signal voltage differential between input jacks and to chassis ground both of them. Yes, it's a loop!
Like war, there are no winners, only varying degrees of losing.
All good fortune,
Chris
Got it. That makes sense. Fault through circuit board trace = no good. Imbalance, not so good either.
Well, I did that in the second picture in post 107 where I grounded both input jacks right at the jack. In the original ST-35, the jacks were both connected to the chassis there. However, that arrangement also violates a "rule" in that it has two connections to the chassis, one from each jack, and a fault would go through the circuit board traces. So do we resort to what you said in post 105?
That leaves us with this with the "star" connected to the chassis right at that point. I could just put a ring terminal under the 2-terminal strip, correct? If not, there is another bolt a half inch away.
Is this diagram the best we can do under these circumstances?
Last edited:
The transformer center-tap (red/ yellow) should go directly to the filter cap minus terminal - if it doesn't reach, splice some extra wire onto it. Otherwise ripple current will create a voltage drop on your ground bus and a hum you can't seem to fix... While you're at it, twist that wire with the red AC wires for magnetic field cancellation. Connect the output transformers' ground wires to the black speaker posts.
A UL standard says that power cord ground should be on its own stud or bolt, not used for anything else, with star washer. It should have the most slack, so if the cord is yanked out, ground wire breaks last. Black wire in cord should go to rear terminal of fuse holder so fuse won't be "hot" when changing.
A UL standard says that power cord ground should be on its own stud or bolt, not used for anything else, with star washer. It should have the most slack, so if the cord is yanked out, ground wire breaks last. Black wire in cord should go to rear terminal of fuse holder so fuse won't be "hot" when changing.
For 3 wire US power mains, and a 3-wire IEC cord . . .
The first connection of the Hot wire is to the Fuse, then no matter what shorts out, the fuse blows, and Hot is no longer connected to anything.
Next, connect the output of the Fuse to the Power switch. Now, with the switch off, Hot is not connected to anything other than the Fuse and Switch input pin.
Connect the IEC socket ground pin Directly to the chassis, make sure to scrape any metal if there is paint, or other substance, and use a good cutting action (integral lockwasher) ground lug. Tighten.
From the power switch output pin, connect that to the power transformer primary (Hot wire if there is any indication of which wire is hot and which wire is Neutral).
Connect the primary Neutral to the Neutral pin on the IEC socket.
Then, check the IEC power cord, from one end to the other, Hot to Hot; ground to ground, and neutral to neutral. Any mixup of the wiring can be fatal -
(I have seen some IEC power cords that were mis-wired by the manufacturer).
Then, check your 3-wire power outlet using an approved plug in test fixture with safe and trouble lights.
Some homes are not wired correctly.
Failure to do all of the above might cause: "The Surviving Spouse Syndrome"
The first connection of the Hot wire is to the Fuse, then no matter what shorts out, the fuse blows, and Hot is no longer connected to anything.
Next, connect the output of the Fuse to the Power switch. Now, with the switch off, Hot is not connected to anything other than the Fuse and Switch input pin.
Connect the IEC socket ground pin Directly to the chassis, make sure to scrape any metal if there is paint, or other substance, and use a good cutting action (integral lockwasher) ground lug. Tighten.
From the power switch output pin, connect that to the power transformer primary (Hot wire if there is any indication of which wire is hot and which wire is Neutral).
Connect the primary Neutral to the Neutral pin on the IEC socket.
Then, check the IEC power cord, from one end to the other, Hot to Hot; ground to ground, and neutral to neutral. Any mixup of the wiring can be fatal -
(I have seen some IEC power cords that were mis-wired by the manufacturer).
Then, check your 3-wire power outlet using an approved plug in test fixture with safe and trouble lights.
Some homes are not wired correctly.
Failure to do all of the above might cause: "The Surviving Spouse Syndrome"
The drawing in post 101 is how I wound up grounding the RCA jacks I'm my ST35 build. It was quieter than just grounding to the chassis like the RCA jacks were grounded in the original ST35. My Dynaco FM-3 had the RCA jacks grounded to the chassis also. I replaced them with new jacks that were isolated from the chassis and ran the grounds to a ground point at the center of the closest tube socket. That took care of the small bit if hum I was getting.
I believe the original Dyna ST35, and the Dynaco FM-3 both had 2-wire power cords.
If possible, the ST35 should use a 3-wire power mains, and 3-wire IEC power cords, and 3-wire IEC input socket (IRC socket ground connected right there with a short wire to the IEC mounting screw with a lock-washer style ground lug).
Then, plug in the Dynaco FM-3 with the plug inserted one way, and then with the plug inserted the other way (180 degrees rotated); for minimum hum.
With the ST35 RCA input connectors return connections insulated from the chassis, but connected to the bottom of the input stages Rg ground, and the bottom of the input stages cathode resistor (the ground end of the resistor, or ground end of the resistor series string).
Then, when you use a CD player with a 2-wire cord, or a turntable with a 2-wire cord, you may also have less hum (try both rotations of the 2-wire plug in the mains outlet).
I analyze schematics pretty well; I do not do a good job of analyzing wiring drawings; I do a little better job with seeing the actual wiring in 3D.
Just my opinions.
"Grounds are Commonly Mis-understood" - me
If possible, the ST35 should use a 3-wire power mains, and 3-wire IEC power cords, and 3-wire IEC input socket (IRC socket ground connected right there with a short wire to the IEC mounting screw with a lock-washer style ground lug).
Then, plug in the Dynaco FM-3 with the plug inserted one way, and then with the plug inserted the other way (180 degrees rotated); for minimum hum.
With the ST35 RCA input connectors return connections insulated from the chassis, but connected to the bottom of the input stages Rg ground, and the bottom of the input stages cathode resistor (the ground end of the resistor, or ground end of the resistor series string).
Then, when you use a CD player with a 2-wire cord, or a turntable with a 2-wire cord, you may also have less hum (try both rotations of the 2-wire plug in the mains outlet).
I analyze schematics pretty well; I do not do a good job of analyzing wiring drawings; I do a little better job with seeing the actual wiring in 3D.
Just my opinions.
"Grounds are Commonly Mis-understood" - me
Last edited: