Konnichiwa,
You may have to move the transformer off-board. And you may need to use shielding on the valves to kill the hum. Especially the ECC83 is very prone to picking up hum.
Sayonara
Kofi Annan said:
You may have to move the transformer off-board. And you may need to use shielding on the valves to kill the hum. Especially the ECC83 is very prone to picking up hum.
Sayonara
kOFI:
You have DC filiments, BUT the ground of your DC filiment supply should be grounded to your signal ground at some point (star ground). If it was not it would be a good source for your humm. ALSO...a preamp transformer is vibrating your chassie!?! That does not sound right to me. Can you measure and see if your tranny is drawing as much power as the circuit says it should. MelB
You have DC filiments, BUT the ground of your DC filiment supply should be grounded to your signal ground at some point (star ground). If it was not it would be a good source for your humm. ALSO...a preamp transformer is vibrating your chassie!?! That does not sound right to me. Can you measure and see if your tranny is drawing as much power as the circuit says it should. MelB
Seems a shame to rework what I've just done, but if I need to I will. Its a pretty decent hum.
I think I'll double check the grounds first. I should probably own up to the fact that I likely didn't do the greatest job on the grounding. The filament power supply and the B+ power are both independently grounded to the chassis, but their grounds are tied together. Also, the I didn't insulate the RCA jacks, but they are also tied directly to the single point ground.
I'm guessing this could be responsible for some of the hum, right?
Its grounded.
Also, I was shooting for a single point ground here, but would it be better to have an independent signal ground and power ground?
Well, I checked and the chassisi is not vibrating like I thought. I used an old power transformer from a Pilot tube amp. It puts out about 290VAC at the center tap which made it perfect for this project.
OK-- you can admonish away now.
Thanks for all the advice.
Kofi
Konnichiwa,
Based on your notes, yes that seems a good idea.
1) Insulate all jacks from the case. Connect grounds to respective stages SPG.
2) Only ONE connection from the chassis to +B PSU Single Point Ground (SPG)
3) +B 0V connects to the Input Stage SPG and Heater 0V connects there too.
4) Use multiple SPG's (or Stargrounds) per stage where each stage's local current loops are closed to the SPG of that stage. Then connect the +B's own "0V" to the input stage SPG and "bus" the ground from input stage SPG to output stage SPG.
5) Remove the mains transformer into a seperate chassis (honestly - take my word).
Build correctly the circuit should show << 0.1mV wideband, unweighted noise, somewhat depoendent on the noise levels of the Input valve..
Kofi Annan said:
Based on your notes, yes that seems a good idea.
1) Insulate all jacks from the case. Connect grounds to respective stages SPG.
2) Only ONE connection from the chassis to +B PSU Single Point Ground (SPG)
3) +B 0V connects to the Input Stage SPG and Heater 0V connects there too.
4) Use multiple SPG's (or Stargrounds) per stage where each stage's local current loops are closed to the SPG of that stage. Then connect the +B's own "0V" to the input stage SPG and "bus" the ground from input stage SPG to output stage SPG.
5) Remove the mains transformer into a seperate chassis (honestly - take my word).
Build correctly the circuit should show << 0.1mV wideband, unweighted noise, somewhat depoendent on the noise levels of the Input valve..
OK-- so I've been living with the hum for a while and I JUST CAN'T STANDS IT NO MORE!!!. I need to do a complete rewire of the project including proper grounding and an off-board power supply but, as usual, I need some advice.
Question 1:
When I move the PSU off board, how should I connect the rectified power to the chassis with the phono circuit? Should I use an XLR connector? If so, what connector would you recommend? Any ideas on the best source for such a connector?
Question 2:
Should I also connect the ground from the power supply to the ground for the phono circuit? In other words, should the power ground be one of the lines that is connected passed through the proposed XLR connector to the chassis with the phono circuit?
Question 3:
Can the power ground (for B+ droppers, anode load and related caps) and the signal ground (everything else) in chassis with the phono circuit BOTH be attached to the chassis or do I need to find a gound for them independently and then connect THOSE grounds together?
Question 4:
Since I'm using the center tap on the transformer to power the circuit, I have not grounded the PSU independently, i.e., there is no wire coming off the transformer that goes directly to ground. Is this wrong?
Question 5:
My nanny touched me in a place that's normally covered by my bathing suit. Is this wrong?
Any responses would be greatly appreciated. Any responses that lead to a solution would be even more appreciated than that.
Kofi
Question 1:
When I move the PSU off board, how should I connect the rectified power to the chassis with the phono circuit? Should I use an XLR connector? If so, what connector would you recommend? Any ideas on the best source for such a connector?
Question 2:
Should I also connect the ground from the power supply to the ground for the phono circuit? In other words, should the power ground be one of the lines that is connected passed through the proposed XLR connector to the chassis with the phono circuit?
Question 3:
Can the power ground (for B+ droppers, anode load and related caps) and the signal ground (everything else) in chassis with the phono circuit BOTH be attached to the chassis or do I need to find a gound for them independently and then connect THOSE grounds together?
Question 4:
Since I'm using the center tap on the transformer to power the circuit, I have not grounded the PSU independently, i.e., there is no wire coming off the transformer that goes directly to ground. Is this wrong?
Question 5:
My nanny touched me in a place that's normally covered by my bathing suit. Is this wrong?
Any responses would be greatly appreciated. Any responses that lead to a solution would be even more appreciated than that.
Kofi
Konnichiwa,
Considering the voltages involved, do not use XLR. I use the Neutrik Speakon connectors in similar applications.
Usually - yes. It may be worthwhile to connect the earth to PSU ground using a few antiparallel diodes to "virtually disconnect" the earth unless fault conditions arise. Note that this is not fully compliant with electrical code in the letter, though it is in spirit.
Connect power ground to signal-star ground and the chassis to the input starground.
I am not quite clear what you refer to. More details please?
Depends, is she young, pretty and blonde?
Sayonara
Kofi Annan said:
Considering the voltages involved, do not use XLR. I use the Neutrik Speakon connectors in similar applications.
Kofi Annan said:
Usually - yes. It may be worthwhile to connect the earth to PSU ground using a few antiparallel diodes to "virtually disconnect" the earth unless fault conditions arise. Note that this is not fully compliant with electrical code in the letter, though it is in spirit.
Kofi Annan said:
Connect power ground to signal-star ground and the chassis to the input starground.
Kofi Annan said:
I am not quite clear what you refer to. More details please?
Kofi Annan said:
Depends, is she young, pretty and blonde?
Sayonara
Kuei Yang Wang said:
Sounds good. So I will need three Speakons, right? One for B+, regulated filament supply and PSU ground?
Kuei Yang Wang said:
What is the input star ground? I'm thinking there will be two star grounds in the phono circuit box-- one for power (grounded here: the B+ dropper and anode load will ground through the final smoothing cap; the PSU ground from the PSU box, the filament and tube pin 9 ground) and one for signal (everything else).
The question is, since I'm grounding to the chassis, should each individual star ground be grounded to the chassis and then connected together by a ground wire, or should they be insulated from the chassis, then tied together at the chassis?
Kuei Yang Wang said:
Normally, I think you would use the secondary end taps for the transofrmer to power the circuit, then ground the center tap to the PSU ground. Since I'm using the center tap to power the circuit, there's no center tap to ground. As such, the transofrmer has no direct connection to ground. Hopefully I'm explaining this right... I'm trying. I really am.
Kuei Yang Wang said:
Nah... old and wrinkled, but she gives good touch. I can put in a good word for you.
Seriously, thanks a million for all your help. You have no idea what joy it gives me to complete a project and learn so much along the way. You're a superhero.
Konnichiwa,
Speakon is 4-pole, rated at 500V and many Amp's, so one connector is enough.
Not quiet. As I wrote a little further up, you should have a single point ground PER STAGE, build the PSU using a bus-ground system and finally tie all these grounds together at the output stage signle point gtound.
Yes, everything should be insulated from the chassis except for one wire connection.
You mean you use a full bridge rectifier? In that case you are fine with no direct connection of ground to transformer.
In that case it is NOT okay. Change the nanny to young and pretty and try again!
Sayonara
Kofi Annan said:
Speakon is 4-pole, rated at 500V and many Amp's, so one connector is enough.
Kofi Annan said:
Not quiet. As I wrote a little further up, you should have a single point ground PER STAGE, build the PSU using a bus-ground system and finally tie all these grounds together at the output stage signle point gtound.
Kofi Annan said:
Yes, everything should be insulated from the chassis except for one wire connection.
Kofi Annan said:
You mean you use a full bridge rectifier? In that case you are fine with no direct connection of ground to transformer.
Kofi Annan said:
In that case it is NOT okay. Change the nanny to young and pretty and try again!
Sayonara
OKOKOKOK!! Got it, I think. So, to clarify, the 10u and 22u caps will ground to the PSU star ground along with the incoming PSU ground from the outboard power box. The 51K1 and 619 will be tied together in the 1st STAGE ground and the 10n, 1n, 10M and 1000K will be tied together in the 2nd STAGE ground.
These three independent star points that are shielded from the chassis will then be tied together and then tied to the chassis at a single point, probably an RCA jack ground.
Right? Well, wrong, porbably, but hopefully not that far off.
Thanks an additional million for all your help.
Kofi
Attachments
Konnichiwa,
No, each of these capaictors connect locally at the cathode/cathode resistor of the respective stage they decoupled.
Nope.
10uF, 51K1, 619R, 10n & 1n and input RCA jack ground get tied to the first stage SPG (one SPG for both channels!!!).
ECC88 cathode, 22uF, 10M & 1M (1000K) and output RCA jack ground get tied to the second stage SPG (one SPG for both channels!!!).
PSU ground from external PSU gets tied to 2nd stage SPG.
Chassis gets tied to 1st stage SPG.
1st stage SPG gets tied to 2nd stage SPG preferably with a piece of flat silver or laquered copper.
Heater supply negative line (which should be DC and well filtered) gets tied to the 2nd stage SPG.
All signal/heater/psu circuit section parts are completely insulated from the chassis except for the connection of the Phonostage Chassis to the 1st stage SPG and the connection of the PSU Chassis to Earth and to PSU (+B) Ground via a pair of anti-parallel diodes.
The anti-parallel diodes (need to be high current!!!) normally break the connection between chassi and earth, UNLESS the potential between chassis and earth becomes > 0.5V, in which case they clamp the fault voltage to < 2V untill they short out from overcurrent.
Sayonara
Kofi Annan said:
No, each of these capaictors connect locally at the cathode/cathode resistor of the respective stage they decoupled.
Kofi Annan said:
Nope.
10uF, 51K1, 619R, 10n & 1n and input RCA jack ground get tied to the first stage SPG (one SPG for both channels!!!).
ECC88 cathode, 22uF, 10M & 1M (1000K) and output RCA jack ground get tied to the second stage SPG (one SPG for both channels!!!).
PSU ground from external PSU gets tied to 2nd stage SPG.
Chassis gets tied to 1st stage SPG.
1st stage SPG gets tied to 2nd stage SPG preferably with a piece of flat silver or laquered copper.
Heater supply negative line (which should be DC and well filtered) gets tied to the 2nd stage SPG.
All signal/heater/psu circuit section parts are completely insulated from the chassis except for the connection of the Phonostage Chassis to the 1st stage SPG and the connection of the PSU Chassis to Earth and to PSU (+B) Ground via a pair of anti-parallel diodes.
The anti-parallel diodes (need to be high current!!!) normally break the connection between chassi and earth, UNLESS the potential between chassis and earth becomes > 0.5V, in which case they clamp the fault voltage to < 2V untill they short out from overcurrent.
Sayonara
Kuei Yang Wang said:
Uh... yeah... that's what I meant.
Steeeee-rike two!
OK. Got it.
Makin' sense!
Time!! Here's the deal-- I have a regulated DC supply for the filaments with a trimpot sos I can set the voltage wheres I wants to. When you you say the "negative supply", I'm thinking "ground", which I'm sure you're going to say is really the negative supply, right?
That would make sense to me, but I just wanted to confirm it.
OK. Getting blown away here with the anti-parallel diodes thingee. On the one hand, I'm thinking I could do without these and still have a perfectly acceptable grounding scheme. On the other hand, "anti-parallel diodes" sounds really Star Trekky and it would be cool to insert them into a conversation:
"Yeah: just your standard grounding scheme using anti-parallel diodes. Huh? Yeah anti-parallel diodes are pretty cool... Huh? Yeah... sure you can touch me where my bathing suit covers..."
So, in an attempt to keep this simple, are these diodes a requirement or can I make this work OK without them. Sorry to seem lazy, but I'm working on a pretty steep learning curve as it is, as you are aware...
Also, should the decoupling caps / anode load / B+ dropper be some distance away from the tube socket as they are carrying some high voltage or is closer to the socket the way to go?
One more thing-- will these work for power speakon interconnects? They seem relatively cheap and I wonder if their performance will suck or be unsafe for a power application.
Add an addiitonal 1.5 million thanks to your already bloated "thank you" coffers.
Kofi
Konnichiwa,
Yup, I tend to avoid the term "ground" it is a convenient fiction but in reality there is no "ground".
That would make sense to me, but I just wanted to confirm it.
Sweet and simple. The earth connection is needed for safety, but is the chief source of hum-loops (often called "Earth Loops"). So ideally you have no connection at all between the signal circuit and mains earth.
By connecting a pair of diodes (which require at least 0.5V to conduct any current) you more or less disconnect signal ground and earth, except when a fault condition exists, which is what the earth connection is intended to protect against.
Allways consider the loop in which the signal current will flow. You want these loops as tight (short) as possible.
They will be fine. Speakon connectors have a inherent design that is safe, making all metal parts carying voltage tough proof and they are locking, so various little house pests cannot dislodge them easily. The cable used to carry the voltages of course also needs siutable insulation.
Sayonara
Kofi Annan said:
Yup, I tend to avoid the term "ground" it is a convenient fiction but in reality there is no "ground".
That would make sense to me, but I just wanted to confirm it.
Kofi Annan said:
Sweet and simple. The earth connection is needed for safety, but is the chief source of hum-loops (often called "Earth Loops"). So ideally you have no connection at all between the signal circuit and mains earth.
By connecting a pair of diodes (which require at least 0.5V to conduct any current) you more or less disconnect signal ground and earth, except when a fault condition exists, which is what the earth connection is intended to protect against.
Kofi Annan said:
Allways consider the loop in which the signal current will flow. You want these loops as tight (short) as possible.
Kofi Annan said:
They will be fine. Speakon connectors have a inherent design that is safe, making all metal parts carying voltage tough proof and they are locking, so various little house pests cannot dislodge them easily. The cable used to carry the voltages of course also needs siutable insulation.
Sayonara
Kuei Yang Wang said:
Seems like a nice solution to isolation. You mentioned earlier in the thread about it not meeting the letter of safety regs. It would seem that the only way to have an unsafe situation would be for a diode to fail open. I would guess that's rare?
Sheldon
My 2 cents
Kofi stay with it man. Today I finished my pre using Kuei's (bow in honour) Valve ElCheapo. It's good. It's very very good. It's quiet and completely blows away my previous 3 phono stages, a Chinese commercial product, JLH ss1540 and Trichord Dino. I'm hearing things on vinyl I never knew were there. It's well balanced and gives an excellent soundstage, heaven knows what his better one sounds like, I suspect this one will hang around for quite some time.
FYI I built mine into a box with S&B Tx103 mc trafos wired straight onto the valve base, then out of the 6922 into the selector switch and onto the TVC I bought in the GB last winter. My cartridge is a very low output Dynavector XX2 so the TX 103's are wired at X20. With the volume turned right up I get a small amount of hum but I daren't drop the stylus at that volume for fear of launching my speakers out of the window. At normal (loud) volume the stage is dead quiet.
I mostly wired mine in the week so I missed Kuei's input. My earthing arrangements are, input stage is taken back to the power star, 2nd stage is also taken back to the power star with a separate conductor, heater _ve is taken to chassis ground, filter caps at the valves are taken back to the power star with their own conductor. Chassis ground has the trafo shields and power trafo body connected as well as the heater supply -ve. Chassis ground is connected to power star by a 100ohm resistor. The mains trafo is fastened onto the chassis so everything is in oneplace. I may try altering the grounding in the light of Kuei's comments but there really is little need.
Kuei
Many thanks for the design
Kev
Kofi stay with it man. Today I finished my pre using Kuei's (bow in honour) Valve ElCheapo. It's good. It's very very good. It's quiet and completely blows away my previous 3 phono stages, a Chinese commercial product, JLH ss1540 and Trichord Dino. I'm hearing things on vinyl I never knew were there. It's well balanced and gives an excellent soundstage, heaven knows what his better one sounds like, I suspect this one will hang around for quite some time.
FYI I built mine into a box with S&B Tx103 mc trafos wired straight onto the valve base, then out of the 6922 into the selector switch and onto the TVC I bought in the GB last winter. My cartridge is a very low output Dynavector XX2 so the TX 103's are wired at X20. With the volume turned right up I get a small amount of hum but I daren't drop the stylus at that volume for fear of launching my speakers out of the window. At normal (loud) volume the stage is dead quiet.
I mostly wired mine in the week so I missed Kuei's input. My earthing arrangements are, input stage is taken back to the power star, 2nd stage is also taken back to the power star with a separate conductor, heater _ve is taken to chassis ground, filter caps at the valves are taken back to the power star with their own conductor. Chassis ground has the trafo shields and power trafo body connected as well as the heater supply -ve. Chassis ground is connected to power star by a 100ohm resistor. The mains trafo is fastened onto the chassis so everything is in oneplace. I may try altering the grounding in the light of Kuei's comments but there really is little need.
Kuei
Many thanks for the design
Kev
Konnichiwa,
Yes, it is rare. In many years I have yet to see to se a semiconductor diode or the like fail open. They invariably fail by shortcircuiting. Using a diode bridge rated at 10A+ should not fail at all untill the mains fuse blows out.
So, while the UL (Underwriter Laboratories) would not certify such a solution, in practice it is perfectly safe.
Ciao T
Sheldon said:
Yes, it is rare. In many years I have yet to see to se a semiconductor diode or the like fail open. They invariably fail by shortcircuiting. Using a diode bridge rated at 10A+ should not fail at all untill the mains fuse blows out.
So, while the UL (Underwriter Laboratories) would not certify such a solution, in practice it is perfectly safe.
Ciao T
Kuei Yang Wang said:
Thanks,
I figured as much. I actually tried it on a chip amp I have been using for testing some multi amp speaker designs. The amp was giving an extra mv or so of hum with certain sources. The diode trick eliminated that.
Sheldon
There is no ground. There is no conspiracy. There are no secret police.
So, is there such a thing as an "antiparallel diode" or do you mean that any high-current Schottky diode will do, but connected in a certain magical way? Like:
PSU SPG ==> K - A ==> A - K ==> Chassis
There's probably about nine things wrong with the above statements, but hey-- I'm going for a record. Please correct and admonish as required.
Alright! However, I just looked at the auction and they want $10 for "shipping", which makes the total cost for two 4', 4 conductor SpeakOn connectors $28. I guess that's not bad, but I may try and do better. I really only need one.
Stayin' with it is what I do, man. Just like Mid-East peace, the Iraqi Loya Jirga and the Nagaoka Swans, I'll find a way.
Kofi
Konnichiwa,
There really is no ground, honestly.
But there are (sadly) many conspiracies, most inimical to human life, freedom and self expression. As Jefferson said:
"they believe that any portion of power confided to me, will be exerted in opposition to their schemes. And they believe rightly; for I have sworn upon the altar of God, eternal hostility against every form of tyranny over the mind of man. . ."
Also, from personal experience I know there is a secret police, at least at some times and in some countries. Encounters with said secret people is best avoided....
Schottky diodes are a poor choice in that particular application. Normal "slow" diodes are better. And the connection is not magical, simply Anode of one Diode to cathode of the other on both ends, thus the diodes are in parallel but in opposite directions.
If I interpret this right, yes.
BTW, consider just buying one speakon plug and socket and attach the cable fixed on the powersupply, this may be still cheaper.
Sayonara
Kofi Annan said:
There really is no ground, honestly.
But there are (sadly) many conspiracies, most inimical to human life, freedom and self expression. As Jefferson said:
"they believe that any portion of power confided to me, will be exerted in opposition to their schemes. And they believe rightly; for I have sworn upon the altar of God, eternal hostility against every form of tyranny over the mind of man. . ."
Also, from personal experience I know there is a secret police, at least at some times and in some countries. Encounters with said secret people is best avoided....
Kofi Annan said:
Schottky diodes are a poor choice in that particular application. Normal "slow" diodes are better. And the connection is not magical, simply Anode of one Diode to cathode of the other on both ends, thus the diodes are in parallel but in opposite directions.
Kofi Annan said:
If I interpret this right, yes.
BTW, consider just buying one speakon plug and socket and attach the cable fixed on the powersupply, this may be still cheaper.
Sayonara
Kofi Annan said:
I would think something like the common-as-dirt 1n400x should do fine. If you're passing more than 1 amp to ground, the rectifier will short, so it's just a wire then and your mains fuse will burn. Your chassis should be connected directly to the mains earth.
Looks like what you've drawn is series. You want parallel:
electronics star: anode cathode :chassis ground
: cathode anode :
Sheldon
Stayin' with it is what I do, man. Just like Mid-East peace, the Iraqi Loya Jirga and the Nagaoka Swans, I'll find a way.
challenging times ahead at the UN, good to see that you've still got the Swans to keep you on the straight and narrow.
Interested to hear how quiet you can make the phono stage, if you get it dead quiet then I'll do a rewire on mine to try to get rid of that last bit of noise.
What valves are you using? Mine are NOS 6922 but a chinese 12ax7 and I'm wondering if the bit of white noise I've got maybe the valves.. What do you think?
Keep us posted on the thread for the Swans. They may be my next project, my Maggies are a bit insensitive for my valve amp.
Kev
challenging times ahead at the UN, good to see that you've still got the Swans to keep you on the straight and narrow.
Interested to hear how quiet you can make the phono stage, if you get it dead quiet then I'll do a rewire on mine to try to get rid of that last bit of noise.
What valves are you using? Mine are NOS 6922 but a chinese 12ax7 and I'm wondering if the bit of white noise I've got maybe the valves.. What do you think?
Keep us posted on the thread for the Swans. They may be my next project, my Maggies are a bit insensitive for my valve amp.
Kev
- Status
- Not open for further replies.