Unfortunately, my tube preamp and tube power amp are separated by about 20 feet of cable. I can not shorten that distance.
My first thought was to use a balanced line between the two components as it would solve the cable length issue and avoid any ground loops.
I have a Jensen PC-2XR transformer (actually, I have two!) that could be placed near the power amp, but what can I use at the preamp's output?
The PC-2XR is a pro to consumer level transformer, so that may not be a good choice unless I add a gain stage to the preamp?
Is there a better way to do this or what should I do?
My first thought was to use a balanced line between the two components as it would solve the cable length issue and avoid any ground loops.
I have a Jensen PC-2XR transformer (actually, I have two!) that could be placed near the power amp, but what can I use at the preamp's output?
The PC-2XR is a pro to consumer level transformer, so that may not be a good choice unless I add a gain stage to the preamp?
Is there a better way to do this or what should I do?
Cables of that length MUST be shielded. Long shielded cables have significant capacitance. It is required that the circuitry driving the cables have both a low O/P impedance and substantial current capability.
You don't need a balanced setup. Shielded "twinax" wire, with the shield grounded only at the "upstream" end is quite suitable. However, the driving electronics must have elephant sized cojones. What does the circuitry driving your preamp's O/Ps look like? You probably are going to need a buffer and FETs do the buffering job quite well. If the thought of "sand" in the signal path makes you break out in hives, build the buffer around 1 of the "super" twin triodes: 5687, ECC99, or 6H30П (6n30p).
You don't need a balanced setup. Shielded "twinax" wire, with the shield grounded only at the "upstream" end is quite suitable. However, the driving electronics must have elephant sized cojones. What does the circuitry driving your preamp's O/Ps look like? You probably are going to need a buffer and FETs do the buffering job quite well. If the thought of "sand" in the signal path makes you break out in hives, build the buffer around 1 of the "super" twin triodes: 5687, ECC99, or 6H30П (6n30p).
Three issues for long signal runs (as I see it):
Ground loop.
Signal loss due to line capacitance.
Potential for EMI...pickup of unwanted signals
Ground loop can be reduced significantly with a transformer or with balanced feed.
Signal loss due to cable capacitance can be compensated with inverse roll off filtering at either end or by reducing the feed (source) driving impedance.
EMI can be reduced with shielded lines, twisted balanced feed, ferrite clamp-on sleeves or low pass filtering at the load (amp) end.
There are many other ways but it depends on the issues that actually crop up in your installation.
Ground loop.
Signal loss due to line capacitance.
Potential for EMI...pickup of unwanted signals
Ground loop can be reduced significantly with a transformer or with balanced feed.
Signal loss due to cable capacitance can be compensated with inverse roll off filtering at either end or by reducing the feed (source) driving impedance.
EMI can be reduced with shielded lines, twisted balanced feed, ferrite clamp-on sleeves or low pass filtering at the load (amp) end.
There are many other ways but it depends on the issues that actually crop up in your installation.
The driving amp is a PAS 2 tube amp.
I know that is pretty junky, but at some point I will roll my own preamp and that should allow me to address the drive issue in a more eloquent way.
The power amp is my own design and has a 10K Ohm input impedance.
Using a total run of 20 feet of LC-1 Blue Jeans cable (with a capacitance of 12.5 pf) at 10K impedance at my power amp yields an F3 of 64 kHz, which sould be fine.
The caveats are that the run needs to go through a wall, so I would need to add connectors and a wall plate. Not sure what that does to my calculations.
The beauty of the balanced line is no ground loops, which is an issue that I need to contend with.
My choices there are to either buy a set of transformers or build a balanced out drive circuit.
Using tubes to make a balanced out circuit seems excessive and just adds too many components in the chain.
I can use something like a DRV134 diff amp, but the mystique of tubes is kind of broken, but more importantly, it requires an additional ± power supply for those chips.
I have two pairs of Jensen PC-2XRs that I could use one at the power amp side, but they require a +14 dB higher drive than a simple 1:1 transformer.
I am thinking I should just sell those and look into buying two pairs of something else. Not sure what.
Eli, I assume that the 5687, ECC99, or 6N30P would be a cathode follower circuit to get the low impedance drive?
Chris, The preamps and all of the associated inputs run off of a circuit that is unique to the cabinet they reside in. The power amp must use a neighboring wall socket, so it would not be easy or possible to simply plug the amp in at the same outlet strip as the preamp.
Lastly, if the room layout is changed, the amp may need to be moved another 15 feet away, so the balanced line becomes even more attractive.
I know that is pretty junky, but at some point I will roll my own preamp and that should allow me to address the drive issue in a more eloquent way.
The power amp is my own design and has a 10K Ohm input impedance.
Using a total run of 20 feet of LC-1 Blue Jeans cable (with a capacitance of 12.5 pf) at 10K impedance at my power amp yields an F3 of 64 kHz, which sould be fine.
The caveats are that the run needs to go through a wall, so I would need to add connectors and a wall plate. Not sure what that does to my calculations.
The beauty of the balanced line is no ground loops, which is an issue that I need to contend with.
My choices there are to either buy a set of transformers or build a balanced out drive circuit.
Using tubes to make a balanced out circuit seems excessive and just adds too many components in the chain.
I can use something like a DRV134 diff amp, but the mystique of tubes is kind of broken, but more importantly, it requires an additional ± power supply for those chips.
I have two pairs of Jensen PC-2XRs that I could use one at the power amp side, but they require a +14 dB higher drive than a simple 1:1 transformer.
I am thinking I should just sell those and look into buying two pairs of something else. Not sure what.
Eli, I assume that the 5687, ECC99, or 6N30P would be a cathode follower circuit to get the low impedance drive?
Chris, The preamps and all of the associated inputs run off of a circuit that is unique to the cabinet they reside in. The power amp must use a neighboring wall socket, so it would not be easy or possible to simply plug the amp in at the same outlet strip as the preamp.
Lastly, if the room layout is changed, the amp may need to be moved another 15 feet away, so the balanced line becomes even more attractive.
I have used a 5687 as a Long Tail Pair driving an output transformer, for differential drive, with step down for about 2x gain overall. This involves a voltage steep down of about 9:1 which means that the ouput impedance is very low from an already low impedance tube and the current capability must be in the 100's of mA.
This was not a difficult build and only comprises 1x5687 which is transformer loaded (Push Pull) to the output. I happen to use an input phase splitting choke - but this is not essential. All in all its about the best thing I have built and is so simple its laughable.
A valve differential preamp can be very simple and relatively efficient to boot.
Shoog
This was not a difficult build and only comprises 1x5687 which is transformer loaded (Push Pull) to the output. I happen to use an input phase splitting choke - but this is not essential. All in all its about the best thing I have built and is so simple its laughable.
A valve differential preamp can be very simple and relatively efficient to boot.
Shoog
Yes, I think I need to either add a buffer or bite the bullet and simply build my proposed preamp circuit.
Yes. Hum from a ground loop. The cable ground is also disconnected at the power amp.
I am looking for a long term robust solution.
I had some similar problem from the phono amp to preamp (4m in single-ended).
Had a pair of Lundahl LL1592 and some balanced cables laying around, and as my pre has balanced input...
hum and noise disapeared and sound is great. So great I'm thinking of getting some more of these trafos for balanced to SE input on the amps
Hope this helps
You have two options for the ground loop, to break the "galvanic path" of the safety grounds plus signal grounds, or to shorten the ground loop.
The first option requires transformers, which are great if properly applied, but don't like high impedances and especially along with high external capacitances.
The second option is to shorten the ground loop by bringing AC line power along with signal from the source end. A simple DIY and "code" solution exists:
Amazon.com: Datacomm 50-3323-WH-KIT Flat Panel TV Cable Organizer Kit with Power Solution - White: Electronics
All good fortune,
Chris
Last edited:
Loren42,
I've done some thinking about your long cable issue. Blue Jeans, IIRC, uses a simple coaxial construction, which is not suited to the task at hand. Shielded "twinax" in a differential, if not true balanced, configuration seems correct. That length of shield, when carrying signal, can all too easily pick up noise.
The shield should not carry signal and only be connected to the upstream end chassis, not signal, ground.
You previously mentioned wall plates. 2X Neutrik 203Ms will do the job for you. Online Components carries the Neutrik plates.
No ifs, buts, or maybes, you will have to buffer the PAS preamp. I'm giving the buffer some thought. More about that to to follow soon. However, this much is clear. You will need trafos to break ground loops up. Jensen can be pricey. CineMag offers competetive products, at lower cost. CineMag's CMOB-2L looks like a candidate for your situation. A DC blocking cap. is needed between the buffering voltage follower and the trafo. Contact CineMag for assistance in sizing that part.
I've done some thinking about your long cable issue. Blue Jeans, IIRC, uses a simple coaxial construction, which is not suited to the task at hand. Shielded "twinax" in a differential, if not true balanced, configuration seems correct. That length of shield, when carrying signal, can all too easily pick up noise.
The shield should not carry signal and only be connected to the upstream end chassis, not signal, ground.You previously mentioned wall plates. 2X Neutrik 203Ms will do the job for you. Online Components carries the Neutrik plates.
No ifs, buts, or maybes, you will have to buffer the PAS preamp. I'm giving the buffer some thought. More about that to to follow soon. However, this much is clear. You will need trafos to break ground loops up. Jensen can be pricey. CineMag offers competetive products, at lower cost. CineMag's CMOB-2L looks like a candidate for your situation. A DC blocking cap. is needed between the buffering voltage follower and the trafo. Contact CineMag for assistance in sizing that part.
Loren42,
I've done some thinking about your long cable issue. Blue Jeans, IIRC, uses a simple coaxial construction, which is not suited to the task at hand. Shielded "twinax" in a differential, if not true balanced, configuration seems correct. That length of shield, when carrying signal, can all too easily pick up noise.
You previously mentioned wall plates. 2X Neutrik 203Ms will do the job for you. Online Components carries the Neutrik plates.
No ifs, buts, or maybes, you will have to buffer the PAS preamp. I'm giving the buffer some thought. More about that to to follow soon. However, this much is clear. You will need trafos to break ground loops up. Jensen can be pricey. CineMag offers competetive products, at lower cost. CineMag's CMOB-2L looks like a candidate for your situation. A DC blocking cap. is needed between the buffering voltage follower and the trafo. Contact CineMag for assistance in sizing that part.
Eli,
Thank you for the information. Would a Jensen CI-2RR work as the isolation transformer if I can get it at a competitive price?
Long term I would like to roll my own preamp, but in the interim, does it make sense to upgrade the PAS 3 with one of those Tubes4HiFi PC boards?
I would like to retain the tone controls, so I was considering the TCLA here:
Dynaco PAS3 upgrades & modifications & parts
Unfortunately, I can't find a schematic nor a specification for the output impedance (uses a 12AX7), so I have no idea if it would be a solution or a waste of money. Your thoughts?
If I understand your wiring scheme, you would use a high quality shielded twin lead cable. Connect the two inner leads to the preamp out hot and ground. Do not connect the shield at the preamp end.
The other end goes to the isolation transformer where the shield is connected to the isolation transformer box, the hot to hot, and the other lead to the return side of the transformer.
The cable from the transformer box would connect that shield to the amp chassis, while the output of the transformer connects to the input of the power amp.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.