My viewpoint is not based on aesthetics but on the actual circuits function, signal integrity and EMC. Point to point wiring is like LPs its from a bygone era things have moved on and so have PCBs, no one does commercial P2P these days except for archaic audiophile companies that want to scam their customers out of thousands of dollars... P2P is just a load of dipoles waving around in air with no real control of return current paths, it was great in the day and for very simple valve designs and DIY prototyping is adequate, but for anything more complex a PCB is best...
Hi Guys
Marce: My mistake. I was actually commenting about the quote in your post from gabdx which does seem to be based a lot on aesthetics.
I agree with you that PTP is often presented as something that it is not, but it does cost more time and labour to assemble so commands higher money even when the perfomrance is actually compromised due to out-dated layout methods being followed.
One detail about the tube aging and burn-in and how it relates to negative feedback loops: Guitar power amps follow two basic patterns but 99% are built to just one of these and that incorporates feedback. The power of that loop is quite lkow as feedback theory will illustrate, so the vaguaries of tube performance are still quite evident and one can hear the difference of tube swaps and aging of the tube samples.
That ubiquitous guitar PA is a hifi circuit with a Schmitt splitter front-end and push pull output stage. Hifi amps built to this pattern exhibit all the same defects from ideal performance - defects which are in fact the attributes of the tube sound and why some people choose them. More elaborate circuits endeavour to be truer to the engineering ideal, so basic linearity is improved and feedback is made more effective. However, it is still a tube circuit and the aging effects are still audible.
In both the MI and hifi cases, one can measure the THD of the infant tubes and compare it to the broken-in tube THD and see a change in the raw number and in the profile of the distortion harmonics. These changes are more apparent in free-running gain stages that comprise all production guitar preamplifiers and many hifi preamps.
Have fun
Marce: My mistake. I was actually commenting about the quote in your post from gabdx which does seem to be based a lot on aesthetics.
I agree with you that PTP is often presented as something that it is not, but it does cost more time and labour to assemble so commands higher money even when the perfomrance is actually compromised due to out-dated layout methods being followed.
One detail about the tube aging and burn-in and how it relates to negative feedback loops: Guitar power amps follow two basic patterns but 99% are built to just one of these and that incorporates feedback. The power of that loop is quite lkow as feedback theory will illustrate, so the vaguaries of tube performance are still quite evident and one can hear the difference of tube swaps and aging of the tube samples.
That ubiquitous guitar PA is a hifi circuit with a Schmitt splitter front-end and push pull output stage. Hifi amps built to this pattern exhibit all the same defects from ideal performance - defects which are in fact the attributes of the tube sound and why some people choose them. More elaborate circuits endeavour to be truer to the engineering ideal, so basic linearity is improved and feedback is made more effective. However, it is still a tube circuit and the aging effects are still audible.
In both the MI and hifi cases, one can measure the THD of the infant tubes and compare it to the broken-in tube THD and see a change in the raw number and in the profile of the distortion harmonics. These changes are more apparent in free-running gain stages that comprise all production guitar preamplifiers and many hifi preamps.
Have fun
You said that tubes lose some treble gain after about 100 hours of use. I said I don't know of a mechanism can cause this alleged loss of treble gain, yet apparently leaving unaffected the gain at lower and higher frequencies. How is this being argumentative?Struth said:
I referred to RF as presumably any mechanism which reduces gain at, say, 10kHz would have a bigger effect at, say, 10MHz. I am not aware of any reduction in RF gain after 100 hours of use; certainly the tube receivers and transmitters I have built don't seem to suffer from this alleged problem. After 1000's of hours gain reduces, but at all frequencies.
Why bring in THD? You mentioned gain and frequency response.
Has it? Affecting just treble?
Hi Guys
DF96: In reality it is the THD that changes much more than the gain, but because the sound moves from a bright one to a less-bright one, the impression players and hifi listeners have is that some gain is lost. It is a fact that humans perceive brighter sounds as "louder", so this effect is often described as a loss of gain even by tube sellers.
If you measure mu, that will not have changed perceptibly if at all, and RF response remains unchanged. I would not be surprised if the THD at RF is improved after burn-in just as it does in the audio band.
Speakers usually are slightly more efficient once broken-in since their mechanical parts are more flexible to begin with and become moreso with initial use.I'm not certain about the mechanism of change in the tube but suspect it has to do with the attributes of the element surfaces and thier micro-alignment. The macro alignment does not change except with extreme mechanical jarring, so the raw characteristics stay within spec.
Have fun
DF96: In reality it is the THD that changes much more than the gain, but because the sound moves from a bright one to a less-bright one, the impression players and hifi listeners have is that some gain is lost. It is a fact that humans perceive brighter sounds as "louder", so this effect is often described as a loss of gain even by tube sellers.
If you measure mu, that will not have changed perceptibly if at all, and RF response remains unchanged. I would not be surprised if the THD at RF is improved after burn-in just as it does in the audio band.
Speakers usually are slightly more efficient once broken-in since their mechanical parts are more flexible to begin with and become moreso with initial use.I'm not certain about the mechanism of change in the tube but suspect it has to do with the attributes of the element surfaces and thier micro-alignment. The macro alignment does not change except with extreme mechanical jarring, so the raw characteristics stay within spec.
Have fun
When you write 'one can measure ... ' do you mean that you have measured it yourself (or have access to someone else's measurements)? If so, can you present these measurements to us, or point us where to find them?
One set of actual measurements is worth more than any number of subjective impressions.
Hi Guys
I did measure this decades ago. I was pressed for proof and this is the are where a skeptic will find it - in THD measurements.
The subjective impressions are described succinctly and any player should be able to understand them. Musicians get beaten over the head with all kinds of sales hype and completely wrong BS that uses a few techy buzz-words to make it sound true.
The fact is, tubes will last for decades if not mechanically upset. Their tone does not change perceptibly after they have been broken-in, which for most players takes about six months to a year. The sound change at that point is obvious to most players - at least to the ones I've dealt with - and once they learn that the tubes are not actually shot, they are very pleased.If the player misses some of that the brighter tone of the infant tubes, it is a simple mod to most amps to restore some brightness albeit in a different manner.
As for the PTP vs PCB "superiority": you have to make a fair comparison for it to be valid, as I've described at length above.
Have fun
I did measure this decades ago. I was pressed for proof and this is the are where a skeptic will find it - in THD measurements.
The subjective impressions are described succinctly and any player should be able to understand them. Musicians get beaten over the head with all kinds of sales hype and completely wrong BS that uses a few techy buzz-words to make it sound true.
The fact is, tubes will last for decades if not mechanically upset. Their tone does not change perceptibly after they have been broken-in, which for most players takes about six months to a year. The sound change at that point is obvious to most players - at least to the ones I've dealt with - and once they learn that the tubes are not actually shot, they are very pleased.If the player misses some of that the brighter tone of the infant tubes, it is a simple mod to most amps to restore some brightness albeit in a different manner.
As for the PTP vs PCB "superiority": you have to make a fair comparison for it to be valid, as I've described at length above.
Have fun
I beg to differ....there are several NOT-'archaic' companies making very fine
amplifiers using P-to-P wiring that are definitely NOT scams. This company:
Amplifiers – Milkman Sound is very well respected and produces some of the very best guitar amplifiers made; certainly WAY better than the Fender 'reissue' models that are PCB based.
Are the output transformers the same? No, well they have greatest effect on how a valve amplifier behaves when clipping, which is what guitar amplifiers do.
Its an archaic old fashioned way of working, just QAing P2P is a nightmare etc...
Why I dont know, but ultimate hi fidelity is not one reason... How many guitarist want the ultimate fidelity, they want to create there own sound and the amplification is a big part of that with electronic instruments, its probably the extra noise p2p provides that makes them liked so much.....
Hi Guys
marce: You need to relax. Your opinion is correct for you and dotneck335's opinion is correct for him. Obviously there is a communication gap.
dotneck335: Most builders using PTP do so because they can understand it and in most cases do not understand PCB layout or do not have enough production to warrant getting PCBs. As I've said already, the problem of "quality" does not lie in the PTP or the PCB itself, rather in how either is laid out and soldered.
Unfortunately, the bulk of PTP amps built today follow the same layouts as the old PTP amps, and those are flawed layouts. For example, Fender builds single-coil guitars that pick up a tonne of noise from the environment. The amps of the day were quieter than the guitars but not as quiet as they could be made. In our modern time we know how to wire a tube amp to be dead quiet - they did back then too but simply didn't bother. Today we have solid-state amps to compete with tube amps for market share, and those amps are usually really quiet which makes the tube stuff seem noisier by comparison.
Today, there is no excuse for a tube amp to be noisy. Similarly, there is no excuse for a non-MV amp to have its output modulated by hum when clipped. The Komets and other TW amps do that, as do most vintage PTP amps. This is not a fault of PTP, rather of poor wiring and grounding.
The performance you expect from gear depends upon your personal experience.
If you are judging solely on sound, then there will be no difference between a good PTP amp and a good PCB amp - BUT it has to be that fair of a comparison, something you are unlikely to encounter.
Have fun
marce: You need to relax. Your opinion is correct for you and dotneck335's opinion is correct for him. Obviously there is a communication gap.
dotneck335: Most builders using PTP do so because they can understand it and in most cases do not understand PCB layout or do not have enough production to warrant getting PCBs. As I've said already, the problem of "quality" does not lie in the PTP or the PCB itself, rather in how either is laid out and soldered.
Unfortunately, the bulk of PTP amps built today follow the same layouts as the old PTP amps, and those are flawed layouts. For example, Fender builds single-coil guitars that pick up a tonne of noise from the environment. The amps of the day were quieter than the guitars but not as quiet as they could be made. In our modern time we know how to wire a tube amp to be dead quiet - they did back then too but simply didn't bother. Today we have solid-state amps to compete with tube amps for market share, and those amps are usually really quiet which makes the tube stuff seem noisier by comparison.
Today, there is no excuse for a tube amp to be noisy. Similarly, there is no excuse for a non-MV amp to have its output modulated by hum when clipped. The Komets and other TW amps do that, as do most vintage PTP amps. This is not a fault of PTP, rather of poor wiring and grounding.
The performance you expect from gear depends upon your personal experience.
If you are judging solely on sound, then there will be no difference between a good PTP amp and a good PCB amp - BUT it has to be that fair of a comparison, something you are unlikely to encounter.
Have fun
Last edited:
For a "one-off", P to P makes better sense. For a production unit, PCB makes better sense, as long as nothing on the board gets too hot to hold your finger on for say 20 seconds IMO. So no power tubes or power resistors that get hot.
As tubes age, their gain goes down slightly (at all frequencies), and matched tubes may drift apart from each other. In most topologies this drift will increase I.M. distortion, and bring up the 2nd harmonic distortion product (generally a good thing if not taken too far). Output power is likely to go down slightly. In a guitar amp, this increase in distortion is often appreciated, but many guitarists judge an amp primarily by how loud it goes, rather than by how mello and musical the distortion is. If you want the amp to have good sounding distortion when over driven, feedback is not likely to be your friend. I never use it in my guitar amps (which are all tube).
Here's my version of "organized grounding":
Tie all power supply grounds together, and then run one wire from that junction to what will be the "star center" ground post, usually a bolt that connects the ground system to the chassis in only one place (so no ground loops). Each section of circuitry should have a separate ground wire that goes to this star center ground post. All connectors need to have floating grounds, isolated from the chassis. The "earth ground" 3rd wire (green in the U.S.) that comes in on the AC line cord should tie to the chassis right where it comes in, rather than to the star center post. I think that covers everything.
Early amps didn't follow this method, and often hummed a lot. They typically had several "ground loops" due to non-floating connectors, and seemingly random grounding techniques.
If anyone is wanting to build their own guitar amp, you might be interested in the guitar amps I designed and built on my hobby website: Bob's Website
As tubes age, their gain goes down slightly (at all frequencies), and matched tubes may drift apart from each other. In most topologies this drift will increase I.M. distortion, and bring up the 2nd harmonic distortion product (generally a good thing if not taken too far). Output power is likely to go down slightly. In a guitar amp, this increase in distortion is often appreciated, but many guitarists judge an amp primarily by how loud it goes, rather than by how mello and musical the distortion is. If you want the amp to have good sounding distortion when over driven, feedback is not likely to be your friend. I never use it in my guitar amps (which are all tube).
Here's my version of "organized grounding":
Tie all power supply grounds together, and then run one wire from that junction to what will be the "star center" ground post, usually a bolt that connects the ground system to the chassis in only one place (so no ground loops). Each section of circuitry should have a separate ground wire that goes to this star center ground post. All connectors need to have floating grounds, isolated from the chassis. The "earth ground" 3rd wire (green in the U.S.) that comes in on the AC line cord should tie to the chassis right where it comes in, rather than to the star center post. I think that covers everything.
Early amps didn't follow this method, and often hummed a lot. They typically had several "ground loops" due to non-floating connectors, and seemingly random grounding techniques.
If anyone is wanting to build their own guitar amp, you might be interested in the guitar amps I designed and built on my hobby website: Bob's Website
Hi Guys
A "star" is the only organised ground everyone can understand, but it is also the worst one to use in most places. If you take a PTP Fender amp with its random grounding and star ground the amp it will be much noisier than stock. Why? Because you have taken all of the quiet currents and MADE THEM MINGLE with the noisiest currents.
For proper grounding you must have isolated jacks.
Note that the chassis is NOT ground rather it is a shield.
The Galactic Ground is based on circulating currents, which always flow in loops. The rectifier current flows through the PT winding and the first filter cap. This is the noisiest current loop in the amp. At the other end of the circuit, the guitar pickup current flows through the pickup, the input stage grid leak resistor and the cable between the amp and instrument. This is the quietest, cleanest current in the amp. between these two loops are loops of current of intermediate noisiness and amplitude.
Each gain stage should have its filtering local to itself and the power end connections form a star as do the ground-ends of the related elements. For example, the first stage of a Fender amp has the grid-leak resistor, cathode resistor and cathode bypass cap. The filter cap for this stage should also be included as should the ground from the input jack - so five elements form the input stage ground star.
The enxt stage has its own filter cap and its own ground star, and so on. NONE of these stars is tied to anything but itself at this point. We link them together along a wired buss and now have a "galaxy" of stars, hence Galactic grounding.
About midway along this buss, we take ONE wire and tie it to the chassis; NOT from the input jack.
The AC safety ground is also tied to the chassis near where it comes in. if the PT has an electrostatic screen, this goes to chassis too, ear the safety ground. The signal ground should be distant from the safety and ES grounds.Note that the power supply filter caps are now DISTRIBUTED as they should be not clustered with a compromised connection to the circuit.
The above method can be applied to every technology. I've used it for tube amps and solid-state amps, built as PTP and as PCB. if you are laying out a PCB for audio DO NOT use the autorouter nor the ground symbol. It is better to make the software do what you need it to do rather than battle against it.
In a low-gain amp you can commit many sins of wiring and still have "acceptable" noise levels. When you add gain then the instability of traditional layouts scream at you and the noise is like a tsunami. TUT3 shows schematically how the iconic amps evolved and how they should have evolved just a teeny bit more, along with layout drawings to show how they should be wired for lowest noise and best note articulation. Many hobbyists and amp companies follow those layouts and have "the best sounding amps they've heard" (their words).
Have fun
A "star" is the only organised ground everyone can understand, but it is also the worst one to use in most places. If you take a PTP Fender amp with its random grounding and star ground the amp it will be much noisier than stock. Why? Because you have taken all of the quiet currents and MADE THEM MINGLE with the noisiest currents.
For proper grounding you must have isolated jacks.
Note that the chassis is NOT ground rather it is a shield.
The Galactic Ground is based on circulating currents, which always flow in loops. The rectifier current flows through the PT winding and the first filter cap. This is the noisiest current loop in the amp. At the other end of the circuit, the guitar pickup current flows through the pickup, the input stage grid leak resistor and the cable between the amp and instrument. This is the quietest, cleanest current in the amp. between these two loops are loops of current of intermediate noisiness and amplitude.
Each gain stage should have its filtering local to itself and the power end connections form a star as do the ground-ends of the related elements. For example, the first stage of a Fender amp has the grid-leak resistor, cathode resistor and cathode bypass cap. The filter cap for this stage should also be included as should the ground from the input jack - so five elements form the input stage ground star.
The enxt stage has its own filter cap and its own ground star, and so on. NONE of these stars is tied to anything but itself at this point. We link them together along a wired buss and now have a "galaxy" of stars, hence Galactic grounding.
About midway along this buss, we take ONE wire and tie it to the chassis; NOT from the input jack.
The AC safety ground is also tied to the chassis near where it comes in. if the PT has an electrostatic screen, this goes to chassis too, ear the safety ground. The signal ground should be distant from the safety and ES grounds.Note that the power supply filter caps are now DISTRIBUTED as they should be not clustered with a compromised connection to the circuit.
The above method can be applied to every technology. I've used it for tube amps and solid-state amps, built as PTP and as PCB. if you are laying out a PCB for audio DO NOT use the autorouter nor the ground symbol. It is better to make the software do what you need it to do rather than battle against it.
In a low-gain amp you can commit many sins of wiring and still have "acceptable" noise levels. When you add gain then the instability of traditional layouts scream at you and the noise is like a tsunami. TUT3 shows schematically how the iconic amps evolved and how they should have evolved just a teeny bit more, along with layout drawings to show how they should be wired for lowest noise and best note articulation. Many hobbyists and amp companies follow those layouts and have "the best sounding amps they've heard" (their words).
Have fun
Last edited:
I agree with everything Struth says above, but... not sure about the ground buss thing. Each section of circuitry should have it's grounds tied together (each stage), and then run one wire back to the "star center" common reference point ground, because that's the ground connection that is used as a reference for all gain stages. You don't want the output stage to be modulating the ground reference used by the input stage, as can happen with a buss wire. In a feedback amp this can cause instability. A thick buss wire will often work "good enough" for grabbing sub-circuit grounds and taking them back to the power supply ground, daisy chain style, but it's less optimal than all sub-circuit grounds coming to a common point, and having some distance from the power supply cap charging currents.
Forgive me if I misunderstood what you were saying. We might be saying the same thing in different ways.
Forgive me if I misunderstood what you were saying. We might be saying the same thing in different ways.
That's the only safe connection point for the AC ground. That connection must have total integrity.
Last edited:
Hi Guys
Sorry, Bob, you are mistaken in that assertion. Abandon the star ground as it is only going to make things worse.
The Galactic Ground is proven for lowest noise and best note articulation and is universally applicable.
The AC safety ground and the screen in the PT are potentially noisier than the rectifier current loop, so those have to be well away from the signal-ground tie point.
Have fun
Sorry, Bob, you are mistaken in that assertion. Abandon the star ground as it is only going to make things worse.
The Galactic Ground is proven for lowest noise and best note articulation and is universally applicable.
The AC safety ground and the screen in the PT are potentially noisier than the rectifier current loop, so those have to be well away from the signal-ground tie point.
Have fun
Last edited:
I'm not sure I understand why, but an amp hummed significantly less when I tied the green earth ground wire to the chassis separately, right where it came in, after trying it connected to the star center Gnd point. It must have modulated the rest of the grounds somehow.
I was taught the satalite star ground method at Tektronix and again at Dolby Labs, in Engineering groups, and it's been working quite well in both my guitar amps and my Hi-Fi projects. I just don't understand why a daisy chain ground buss would be better.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.