In/out grounding is *situational*.
In large school/factory PA systems, there is often an option to run the output full-floating, un-grounded, because this can improve system stability facing many hundreds of feet of wire. (The MC3500 has a jumper for this.)
W.E. was the Phone Company. Much of their stuff drove long-long lines. Grounding was defined on the *system* diagram, not the *amplifier*. Most W.E. stuff is full-floating in and out; also the power supply and chassis are not connected inside the amplifier. These details are given in the *system* documentation. Normally the supply and the rack the chassis is bolted to are bonded at the building ground portal (a concept essential to large exposed audio systems but unknown in small systems). In long-line work the line might be run Floating (relying on Carbon Blocks to absorb dangerous unintended voltages), or in high-RF environments there might be a ground on the CT of the receiving amplifier's input transformer (sometimes cap-grounded to shunt RF yet allow audio to float to a self-balanced condition).
None of which applies to *domestic* work, where you do not need a truck to roll-out the wire. I can't see any reason not to tie secondary to chassis and to electric PE/ground. It is nearly universal even in the cheapest radios, suggesting that UL required it in systems not subject to "professional oversight".
Beyond the new-gear objectives, here we have "unknown transformers", possibly old. All transformer insulation WILL fail over time. It can even be plotted, predicted, and in serious work, guaranteed. Electric Utilities pay close attention to power, temperature, and time, because they judge first-cost against transformer working life. "unknown transformers" may not be so carefully designed and wound, so insulation breakdown is a very real possibility, especially as the tranny ages.
I won't say I've never run speakers un-grounded. I recall some Concert In The Park with very long lines and a generator on wet ground. But certainly in domestic work I've always grounded.
In large school/factory PA systems, there is often an option to run the output full-floating, un-grounded, because this can improve system stability facing many hundreds of feet of wire. (The MC3500 has a jumper for this.)
W.E. was the Phone Company. Much of their stuff drove long-long lines. Grounding was defined on the *system* diagram, not the *amplifier*. Most W.E. stuff is full-floating in and out; also the power supply and chassis are not connected inside the amplifier. These details are given in the *system* documentation. Normally the supply and the rack the chassis is bolted to are bonded at the building ground portal (a concept essential to large exposed audio systems but unknown in small systems). In long-line work the line might be run Floating (relying on Carbon Blocks to absorb dangerous unintended voltages), or in high-RF environments there might be a ground on the CT of the receiving amplifier's input transformer (sometimes cap-grounded to shunt RF yet allow audio to float to a self-balanced condition).
None of which applies to *domestic* work, where you do not need a truck to roll-out the wire. I can't see any reason not to tie secondary to chassis and to electric PE/ground. It is nearly universal even in the cheapest radios, suggesting that UL required it in systems not subject to "professional oversight".
Beyond the new-gear objectives, here we have "unknown transformers", possibly old. All transformer insulation WILL fail over time. It can even be plotted, predicted, and in serious work, guaranteed. Electric Utilities pay close attention to power, temperature, and time, because they judge first-cost against transformer working life. "unknown transformers" may not be so carefully designed and wound, so insulation breakdown is a very real possibility, especially as the tranny ages.
I won't say I've never run speakers un-grounded. I recall some Concert In The Park with very long lines and a generator on wet ground. But certainly in domestic work I've always grounded.
Let me offer another scenario: measure or estimate your output transformer's primary to secondary parasitic capacitance. Mount a capacitor of that size from your B+ to an accessible spot. Touch the chassis with one hand, say "Hey guys, hold my beer. Watch this!" and grab the free end of the capacitor with your other hand.
Does anyone really want to do that to a child, pet, or other loved one? Then don't be that guy.
All good fortune,
Chris
Does anyone really want to do that to a child, pet, or other loved one? Then don't be that guy.
All good fortune,
Chris
Bet your beer the C is 300p to 2,000pFd.
Wire it to B+. DC. Discharged to a 10k skin resistance, the 40mA discharge is fading in 10micro Seconds, say 100uS to get far down the exponential, which is far-far below the dangerous line. Assuming skin resistance 1k to 100k trades peak current for shock duration
The danger is in an insulation breakdown.
PRR,
Yes, insulation breakdown. But I am not going to grab it. What if the amp is playing electronic music, and an SE output tube is swinging +/- 200V at 3kHz?
2000pF
Xc = 27k.
200V/27k = 7.4 mA.
Sustained note.
I believe there was a documented case where a TV repairman was shocked by the High Voltage on the CRT. The current was not enough to kill. However, the reaction to the shock threw the repairman across the room, and he broke his neck.
Safety first.
Yes, insulation breakdown. But I am not going to grab it. What if the amp is playing electronic music, and an SE output tube is swinging +/- 200V at 3kHz?
2000pF
Xc = 27k.
200V/27k = 7.4 mA.
Sustained note.
I believe there was a documented case where a TV repairman was shocked by the High Voltage on the CRT. The current was not enough to kill. However, the reaction to the shock threw the repairman across the room, and he broke his neck.
Safety first.
Last edited:
All the transformers i have tested sound better (to me) when the secondary is not grounded. Still waiting for my amp to kill me
Please correct that to:
.... WHEN I know about it .
Specially if I wired it (builder´s pride anybody?)
I am very accurate even in double blind tests; I am RIGHT 50% of the time
OK, so it sounds better.
Research why it is better.
I was doing Vector Network Analysis on SE tube Output Transformers. The test had the frame (lams and end bells) of the transformer at test ground. The test had the B+ lead of the primary at test ground. The test had the Common lead of the secondary at test ground. There was an 8 Ohm non-inductive resistor from the secondary common to the 8 ohm tap. So . . . I found one transformer that had better frequency response when the plate and B+ leads were swapped. I wrote it up as probably the way the different capacitances of the windings are (distributed across each winding, distributed from each winding to frame, and distributed from each winding to another winding). The only real problem with this is that you need to remember to reverse the B+ and plate leads on the other channel's transformer, so they are also in phase. Left and Right out of phase does never sounded pretty to me, and now I only have one ear, and out of phase still does not sound pretty to me.
Maybe your secondary has too much capacitance to the plate lead of the primary, so it sounds better when you float the secondary.
Research why it sounds better.
Then fix it so you can ground the secondary.
Curiosity killed the cat. High Voltage killed the person.
"Most things have an explanation, we just do not always know what the explanation is" - Me
Research why it is better.
I was doing Vector Network Analysis on SE tube Output Transformers. The test had the frame (lams and end bells) of the transformer at test ground. The test had the B+ lead of the primary at test ground. The test had the Common lead of the secondary at test ground. There was an 8 Ohm non-inductive resistor from the secondary common to the 8 ohm tap. So . . . I found one transformer that had better frequency response when the plate and B+ leads were swapped. I wrote it up as probably the way the different capacitances of the windings are (distributed across each winding, distributed from each winding to frame, and distributed from each winding to another winding). The only real problem with this is that you need to remember to reverse the B+ and plate leads on the other channel's transformer, so they are also in phase. Left and Right out of phase does never sounded pretty to me, and now I only have one ear, and out of phase still does not sound pretty to me.
Maybe your secondary has too much capacitance to the plate lead of the primary, so it sounds better when you float the secondary.
Research why it sounds better.
Then fix it so you can ground the secondary.
Curiosity killed the cat. High Voltage killed the person.
"Most things have an explanation, we just do not always know what the explanation is" - Me
Last edited:
Well, the schematic tells me that the loudspeaker outputs may be grounded, but they don't need to be. Honestly, I feel much safer if the speaker winding is grounded.
Best regards!
I'd just like to point out in a non moderating capacity that advocating or dismissing/making light of the potential safety hazard involved with ungrounded secondaries is both irresponsible and a violation of forum safety policies.
Fortunately there is a strong safety orientation amongst most of our members and hopefully the hazard is clear.
I recommend anyone with strong feelings to the contrary not discuss it here.
Fortunately there is a strong safety orientation amongst most of our members and hopefully the hazard is clear.
I recommend anyone with strong feelings to the contrary not discuss it here.
Nobody seems to have suggested that you measure the resistance between secondary and primary, obviously with the amplifier unpowered.
. . . . or such resistance between secondary and amplifier earth - just to make sure there is not another resistive path lurking somewhere.
P.S: Correction: It was just about suggested somewhere, so my post withdrawn. But measure the resistance anyway!
. . . . or such resistance between secondary and amplifier earth - just to make sure there is not another resistive path lurking somewhere.
P.S: Correction: It was just about suggested somewhere, so my post withdrawn. But measure the resistance anyway!
Last edited:
My friend Iain has a picture of something extra dangerous, from the bad old days:
retro vintage modern hi-fi: Freed MEGOHMMETER Test equipment.
500 Volts DC on the front panel. 100K Megohms on the top scale. Yikes.
All good fortune,
Chris
retro vintage modern hi-fi: Freed MEGOHMMETER Test equipment.
500 Volts DC on the front panel. 100K Megohms on the top scale. Yikes.
All good fortune,
Chris
We don't yet know how this Freed machine will work (needs restoration first) but it does have a TV flyback transformer and HV rectifier inside. Seems likely to need a lot of voltage to measure 100K Megohms. I wish I had a schematic (without having to draw one out the hard way). If wishes were horses, beggars would ride.
All good fortune,
Chris
All good fortune,
Chris
i worked 15 years in the oil and gas industry constructing oil and gas facilities,
i have done lots and lots of high pot testing of cables, switchgears and transformers, test voltage at 40kv, in Russia, we used a Kamas truck to test 110 kv volt cables and high pot to 110kv....really scary stuff, but this activity required me to submit a work permit, outlining the steps to be taken and safety precautions taken, and safety office will approve, obviously this can not be a one man job, we had teams of safety officers, technicians, watch men at the other end of the cable which can run over a kilometer long....radios to communicate with the field, and techs along the route of the cable....high pot testing is done before sand backfilling and concrete tiles along the way if successful....yes i have seen and done them all, so i know the danger working with high voltage....even today, the first time livening up a tube amp sends me jitters....
i have done lots and lots of high pot testing of cables, switchgears and transformers, test voltage at 40kv, in Russia, we used a Kamas truck to test 110 kv volt cables and high pot to 110kv....really scary stuff, but this activity required me to submit a work permit, outlining the steps to be taken and safety precautions taken, and safety office will approve, obviously this can not be a one man job, we had teams of safety officers, technicians, watch men at the other end of the cable which can run over a kilometer long....radios to communicate with the field, and techs along the route of the cable....high pot testing is done before sand backfilling and concrete tiles along the way if successful....yes i have seen and done them all, so i know the danger working with high voltage....even today, the first time livening up a tube amp sends me jitters....
McIntosh MC240 and MC275 did not ground the common secondary tap neither, for the reason already mentioned regarding the PA systems. You can ground it with a jumper.
It depends of the design, especially on the feedback side. If your OT has an extra feedback winding, there is no reason why the Common HAS to be grounded. If there is none and the feedback is taken from a speaker tap, then obviously the Common has to be referenced to ground.
In the OP case, there is no feedback therefore no common grounding is necessary. Now if you see a DC presence there, I certainly would suspect an OT (dangerous) leakage, or a wiring problem.
My 2 cents.
It depends of the design, especially on the feedback side. If your OT has an extra feedback winding, there is no reason why the Common HAS to be grounded. If there is none and the feedback is taken from a speaker tap, then obviously the Common has to be referenced to ground.
In the OP case, there is no feedback therefore no common grounding is necessary. Now if you see a DC presence there, I certainly would suspect an OT (dangerous) leakage, or a wiring problem.
My 2 cents.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.