My apologies. This is a somewhat truncated version that may be more accessible to Third World countries like England.
Most grounds are not grounds. They are actually the return route for the current that came from the Source.
First and more important than all other rules:
Return the current to the Source and use a close coupled pair to minimise the "aerial" effect.
AFTER that has been wired, then look at what parts of the circuit need a voltage reference to allow it to work with another part of the circuit.
This last is the common voltage reference. It passes virtually no current between circuits and between modules. It can be a star ground, or a star of stars, or a buss, or a plane.
But what is most important is that no attachment to this voltage reference should inject any significant CURRENT into the reference. The current MUST be Returned to the Source.
Last edited:
I had to laugh when I finished reading the thread because I had to go back to the first post to read what it was about when it started.
I have played some great sounding point to point wired amps. I have also played some PCB amps that sounded great. I have played some of each type that sounded really bad. The fact that they were P to P or PCB was pretty far down on the list of likely causes for them being good or bad in my mind.
I will say that certain brands with PCBs like Mesa Boogie I really dislike when I need to work on them. Too much stuff crammed in to a very small space put together with a type of solder that I find very hard to deal with.
For the most part I agree with Struth about the tube issues. My experience indicates there is a break in time, but if there is no break in time and nothing changes what would cause the bias setting to change after a few hours. I could could see how my ears could lie to me but not my Fluke..lol
Anything metal that goes from room temp to red hot is going to move. Will it come back to the exact same place when it cools off? Perhaps but not likely. I think that if one had really well made tubes they would last a very long time.And, in fact they do. I just rebuilt a 1937 radio with what I think were the original tubes and two of the tubes tested good as new. On the other hand I just had to replace two 12AX7 tubes that were less than two years old that were in a 18 Fender reissue. BTW I was not impressed with the P to P wiring in that amp.
I only replace tubes for clients when they go bad or when they demand that I change them. There much better ways to make money than ripping off clients selling them stuff they don't need.
So, back to the original thread topic. If I were making a few hundred amp of very high quality I would use PCBs. If I was making a few high quality amps I would use P to P. I see no reason why both would not work equally well.
Just my two cents worth from a old guy who don't know much.
Cheers,
Billy
I have played some great sounding point to point wired amps. I have also played some PCB amps that sounded great. I have played some of each type that sounded really bad. The fact that they were P to P or PCB was pretty far down on the list of likely causes for them being good or bad in my mind.
I will say that certain brands with PCBs like Mesa Boogie I really dislike when I need to work on them. Too much stuff crammed in to a very small space put together with a type of solder that I find very hard to deal with.
For the most part I agree with Struth about the tube issues. My experience indicates there is a break in time, but if there is no break in time and nothing changes what would cause the bias setting to change after a few hours. I could could see how my ears could lie to me but not my Fluke..lol
Anything metal that goes from room temp to red hot is going to move. Will it come back to the exact same place when it cools off? Perhaps but not likely. I think that if one had really well made tubes they would last a very long time.And, in fact they do. I just rebuilt a 1937 radio with what I think were the original tubes and two of the tubes tested good as new. On the other hand I just had to replace two 12AX7 tubes that were less than two years old that were in a 18 Fender reissue. BTW I was not impressed with the P to P wiring in that amp.
I only replace tubes for clients when they go bad or when they demand that I change them. There much better ways to make money than ripping off clients selling them stuff they don't need.
So, back to the original thread topic. If I were making a few hundred amp of very high quality I would use PCBs. If I was making a few high quality amps I would use P to P. I see no reason why both would not work equally well.
Just my two cents worth from a old guy who don't know much.
Cheers,
Billy
I agree with virtually everything above. I don't doubt that tube gains and bias points will change slightly during a burn in period (first month?). I check and readjust the bias in my tube amps from time to time knowing this. But many people seem to think that a burn in period mellows out the sound. It seems random to me whether or not burn-in improves the sound in any way, if it makes any audible difference at all. To me, it just means you should check and readjust the bias from time to time during the first month (at least) or so.
Sorry, off subject again. My only major thought on PCB's is don't put power tubes or hot power resistors on the board, and consider shock mounting the board to minimize input tube microphonics if you think you should.
With an amp turned up to max gain, with the input shorted, you should be able to smack the amp fairly hard, and hear very little ringing out the speaker, and no sustained feedback at all due to conducted radiation/vibration from the speaker to the input tube.
Sorry, off subject again. My only major thought on PCB's is don't put power tubes or hot power resistors on the board, and consider shock mounting the board to minimize input tube microphonics if you think you should.
With an amp turned up to max gain, with the input shorted, you should be able to smack the amp fairly hard, and hear very little ringing out the speaker, and no sustained feedback at all due to conducted radiation/vibration from the speaker to the input tube.
Hi Guys
Infant tube sound is harsh and noisy compared to the burned-in tone. This is universal not random. Those 12AX7WBs ElectroHarmonix was flogging years ago seemed to be the rejected noisy tubes - but even those mellowed out with burn-in. Unless you do an accelerated burn-in, it happens gradually and you won't notice it much unless you have a higher gain amp.
Tubes on PCBs do not present a heat issue if the heat of the tube is generally removed from the are and in nearly all tube amps this is the case - Vox and Laney would be the exceptions. The tube is usually outside and the chassis provides a heat shield for the board.
As I mention in my FAQ and books, the PCB should use metal stand-offs to be less microphonic. Nylon or plastic is super microphonic.
Everything MI is smoother when broken in: tubes, speakers, guitar strings...
Have fun
Infant tube sound is harsh and noisy compared to the burned-in tone. This is universal not random. Those 12AX7WBs ElectroHarmonix was flogging years ago seemed to be the rejected noisy tubes - but even those mellowed out with burn-in. Unless you do an accelerated burn-in, it happens gradually and you won't notice it much unless you have a higher gain amp.
Tubes on PCBs do not present a heat issue if the heat of the tube is generally removed from the are and in nearly all tube amps this is the case - Vox and Laney would be the exceptions. The tube is usually outside and the chassis provides a heat shield for the board.
As I mention in my FAQ and books, the PCB should use metal stand-offs to be less microphonic. Nylon or plastic is super microphonic.
Everything MI is smoother when broken in: tubes, speakers, guitar strings...
Have fun
I attempted to repair an 800 series Marshall amp head, which had 4 EL34's mounted directly to the rather thick circuit board. The EL34's 9operated upside down, so the heat would rise into the circuit board area (even though the chassis blocked much of the rising heat energy). The board showed discoloration from the heat where the EL34 tube sockets were soldered to the board. The board had become conductive to the point of causing the EL34 plates to glow red. I put many hours into trying to get that board to work, even grinding out areas of the PCB around grid wiring. I checked all coupling caps and replaced some of them anyway. I went on the web researching this amp, and found that many other people had had the same experience with this amp model. For a while Marshall was offering new circuit boards (loaded) for about $300, but not any more... I had to tell my customer the amp is toast. Dead. Land fill. When tubes are mounted upside down, as in MANY guitar amps, thermals are a big issue. One of the benefits of running input stage heaters on DC (derived from the 5VAC tranny tap), is that you can use that same 6.3VDC to power a 12 volt computer fan, which run at the lower voltage, will still be substantially effective and also relatively quiet.
Last edited:
Hi Guys
Bob, that's definitely an example of how NOT to use PCBs !
When I was servicing amps, I found that most Marshall OT failures were caused by bias failures of either the tubes or the bias path. EL-34s with their wire screen grid is simply not as reliable a structure as the beam-forming plates in a 6L6 or true 6CA7, so the EL-34s are also more prone to thermal runaway caused by vibration and thermal cycling. Compound those issues by putting the tubes upside down on a poorly soldered PCB and disaster is the only result.
The 12V fan operating from 8Vdc derived from the 6Vac heaters is something we've promoted since 1990. Added it to lots of amps regardless of tube orientation, but some amps cry out for it to be installed by the builder - like that Marshall you worked on, or any Vox amp, and really any amp with upside-down tubes. Hiwatts need fans since their ventilation is not so great and they often run the tubes at the maximum dissipation.
The loading of the fan is about 160mA or less, so it can be added without worry to any stock heater line. No need to move tubes off the winding.
Burned PCBs should just be thrown out. As you say, the carbonised part is conductive. It is bad enough that an epoxy fibreglass board is a bit hydroscopic, which means it can absorb moisture and that leads to leakage issues (conductivity). If the edges of the board are sealed, along with the edges made by drilling mounting holes through it, then moisture is no longer an issue.
Note that modern turret boards use fibreglass epoxy boards and are thus prone to the same issues of raw PCBs.
Bob wrote:
"With an amp turned up to max gain, with the input shorted, you should be able to smack the amp fairly hard, and hear very little ringing out the speaker, and no sustained feedback at all due to conducted radiation/vibration from the speaker to the input tube. "
That is reminiscent of Peter Traynor throwing an amp off the roof of the music store, then replacing the damaged tubes and playing through the amp. He wanted to show that his amps were rugged enough to fall off a festival stage and keep on working.
Have fun
Bob, that's definitely an example of how NOT to use PCBs !
When I was servicing amps, I found that most Marshall OT failures were caused by bias failures of either the tubes or the bias path. EL-34s with their wire screen grid is simply not as reliable a structure as the beam-forming plates in a 6L6 or true 6CA7, so the EL-34s are also more prone to thermal runaway caused by vibration and thermal cycling. Compound those issues by putting the tubes upside down on a poorly soldered PCB and disaster is the only result.
The 12V fan operating from 8Vdc derived from the 6Vac heaters is something we've promoted since 1990. Added it to lots of amps regardless of tube orientation, but some amps cry out for it to be installed by the builder - like that Marshall you worked on, or any Vox amp, and really any amp with upside-down tubes. Hiwatts need fans since their ventilation is not so great and they often run the tubes at the maximum dissipation.
The loading of the fan is about 160mA or less, so it can be added without worry to any stock heater line. No need to move tubes off the winding.
Burned PCBs should just be thrown out. As you say, the carbonised part is conductive. It is bad enough that an epoxy fibreglass board is a bit hydroscopic, which means it can absorb moisture and that leads to leakage issues (conductivity). If the edges of the board are sealed, along with the edges made by drilling mounting holes through it, then moisture is no longer an issue.
Note that modern turret boards use fibreglass epoxy boards and are thus prone to the same issues of raw PCBs.
Bob wrote:
"With an amp turned up to max gain, with the input shorted, you should be able to smack the amp fairly hard, and hear very little ringing out the speaker, and no sustained feedback at all due to conducted radiation/vibration from the speaker to the input tube. "
That is reminiscent of Peter Traynor throwing an amp off the roof of the music store, then replacing the damaged tubes and playing through the amp. He wanted to show that his amps were rugged enough to fall off a festival stage and keep on working.
Have fun
Last edited:
Not a problem really these days, there are also many grades of FR4 as well as more exotic materials. Do quite a few high voltage PCBs and moisture absorbance hasn't been an issue... including one recently with rf at 1000+ volts...
Heat can be an issue but since lead free the Tg and Td temperatures of epoxy glass based materials has increased, as with most things you will get what you pay for to an extent.... there are choices out there that are perfect for valve based designs, one of our customers at work does them and we did a study on materials etc. for a range of gear, the basic went with a high Tg FR4.
On hot components you can use the PCB to help dissipate the heat (look at the huge amount of bottom terminated components these days), but as always PTH resistors etc that get HOT should be mounted so there is space under them.
6oz copper clad boards are an interesting choice for valve based designs, costly but fun...
Thread cleaned up and posts removed by moderationWith all due respect, fixating on the direction in which the electrons move is not only short-sighted, it is mathematically inaccurate. Which is why it was thrown out decades ago, except at the lower levels of technical education, where, apparently, a minus sign still causes consternation, for reasons I don't quite understand.
Consider this: inside the battery powering the electronics, we have both positive and negative ions, and the two types of charged particles move in opposite directions when DC current flows. Which direction is the "real" current? Is it the direction in which the negative ions move? Or is it the direction in which positive ions move?
Sticking with the world of vacuum tubes/valves, the same thing happens inside a neon bulb: electrons move one way, positively charged ions move the other way. Which one is the direction of the current?
We have the same problem to deal with in gas-filled regulator tubes, gas-filled rectifiers, in the arc of an arc-welder, in the sun's atmosphere, and many other places. In all those cases, the flow of electric current is the result of moving positive charges, as well as moving negative charges. In all those cases, the two types of charge move in opposite directions. Which one is the direction of the current flow?
The answer has nothing to do with electricity, or electron flow, or "conventional" anything. It merely has to do with basic algebra, namely, the meaning of a minus sign! Let me explain.
Suppose I have two bank accounts, A and B, each account initially containing $100.
Now I move one dollar (+$1) from account A to account B. At the end of the transaction, account A contains $99. Account B contains $101. Which way did the money go? Clearly, from A to B. When a positive dollar moves from A to B, so does the flow of money.
Now let's consider negative dollars. (We sometimes see them in our accounts, representing credits or earned interest.)
Once again, let's say I have $100 in account A, and also $100 in account B. Now I move (-$1) from account B to account A. When the negative dollar arrives at account A, the balance there falls from $100 to $99. When the negative dollar is removed from account B, the balance there rises to $101.
(Remember that a double negative is a positive - subtracting a negative quantity is the same as adding a positive one on.)
Now, which way did the money flow? Clearly, from account A to account B, since the balance fell in A, and grew in B. But which way did the negative dollar go? From B to A.
What did we find out? When a negative dollar moves from B to A, the flow of money is in the opposite direction, from A to B.
In our example, the moving dollars, of course, are stand-ins for moving charges. The flow of money is a stand-in for the flow of electric current. Electric current is defined as the rate at which electrical charge is transported - it is not defined as the direction in which any particular polarity of charges might happen to move.
Now it should all be clear: electric current flows in the same direction as moving positive charges, should there be any. Electric current flows in the opposite direction as moving negative charges, should there be any. Mathematically, both situations are exactly the same. All one has to do is understand the meaning of a double negative sign.
We ignore this basic algebra at our peril; the entire mathematics of electricity becomes wrong if we confuse the direction of electron movement with the direction of current flow! For example, if we fixate on "electron flow", Ohm's Law then becomes I = (-E)/R, rather than I = E/R.
Making this mistake also means electric current flows uphill (from negative E to positive E), which is absurd, and unlike anything else in nature: water doesn't flow uphill, for instance! And this is no accident, it is a general principle of physics that systems will attempt to minimise their total energy; water flowing downhill is a manifestation of that fact.
In short: any confusion that existed between "electron flow" and "current" exists only in the minds of those who do not understand the meaning of a double negative, and have never taken a college-level electromagnetism course. Any such confusion should have been dispelled long ago, but, like beliefs in fairies or a flat earth, it somehow lingers on, and on, and on!
-Gnobuddy
The problem with "looking at the waves" is that some people insist on using full wave theory (which, in most cases, they don't actually understand) when circuit theory (the low frequency approximation of wave theory) is entirely adequate to solve the problem. They have just enough knowledge to confuse themselves and impress others.
...and we digress yet again.
IMO PTP has several disadvantages, like:
1. The wiring is operator dependent - some will do the wiring better than others.
2. More chances of making mistakes. Most of us have made mistakes in PCB designs; PTP wiring has the same potential for mistakes in every unit.
3. Each wire can act as an antenna - to pick up or radiate RFI. This is true for PCBs as well, but the problem is greatly reduced. The concept of a ground plane has no equivalent in PTP.
The drawbacks of PCBs are mostly related to heat degradation. We still like to keep very hot components off board, even solid state devices, and bend the leads to keep stresses from expansion and contraction from degrading the soldered joints.
IMO PTP has several disadvantages, like:
1. The wiring is operator dependent - some will do the wiring better than others.
2. More chances of making mistakes. Most of us have made mistakes in PCB designs; PTP wiring has the same potential for mistakes in every unit.
3. Each wire can act as an antenna - to pick up or radiate RFI. This is true for PCBs as well, but the problem is greatly reduced. The concept of a ground plane has no equivalent in PTP.
The drawbacks of PCBs are mostly related to heat degradation. We still like to keep very hot components off board, even solid state devices, and bend the leads to keep stresses from expansion and contraction from degrading the soldered joints.
For high speed designs its the best way of looking at them as you have to cater for what the wave travels through...
...and we digress yet again.
IMO PTP has several disadvantages, like:
1. The wiring is operator dependent - some will do the wiring better than others.
2. More chances of making mistakes. Most of us have made mistakes in PCB designs; PTP wiring has the same potential for mistakes in every unit.
3. Each wire can act as an antenna - to pick up or radiate RFI. This is true for PCBs as well, but the problem is greatly reduced. The concept of a ground plane has no equivalent in PTP.
The drawbacks of PCBs are mostly related to heat degradation. We still like to keep very hot components off board, even solid state devices, and bend the leads to keep stresses from expansion and contraction from degrading the soldered joints.
Look at the amount of bottom mounted SMD devices with thermal pads... things have changed a lot, high power and heat density are the norm in many designs as more and more is crammed into finite spaces.... Though the really hot stuff is still best kept of the board.
PTP is better becauseSamAnytime said:
1. Some will do a PCB worse than others.
2. More chances of fixing mistakes with PTP, but a bad PCB means every unit is wrong.
3. The concept of a ground plane can be approximated by the chassis in PTP, but in any case ground planes are not usually relevant to low frequency analogue audio electronics as they make it difficult to control where the currents go.
Of course, but a guitar amp is very low speed design.marce said:
Please, let's not talk about "waves", I'm still trying to get over Alan0354's "lecture" on EMI theory.
Yes that's is true...
What you forget is that P2P is very expensive for commercial equipment, labour intensive.
A good PCB is repeatable every time, and one would hope competently designed for commercial gear (not always the case).
Ground planes and return paths, far to big a subject for this thread But low level (and not so low) analogue can and does benefit from proper layout and proper management of return paths and a ground plane is often the best option, that's a very simplistic statement covering a very complex area electronic systems. But I find the more complex people try to make things (especially return paths) the more problems. There are other advantages, not least proving low impedance paths to get any high frequency noise away from sensitive circuitry.
What you forget is that P2P is very expensive for commercial equipment, labour intensive.
A good PCB is repeatable every time, and one would hope competently designed for commercial gear (not always the case).
Ground planes and return paths, far to big a subject for this thread
But low level (and not so low) analogue can and does benefit from proper layout and proper management of return paths and a ground plane is often the best option, that's a very simplistic statement covering a very complex area electronic systems. But I find the more complex people try to make things (especially return paths) the more problems. There are other advantages, not least proving low impedance paths to get any high frequency noise away from sensitive circuitry.
For mass manufacturing, no doubt PCB's are the way to go. But for boutique amps and most re-issues from the big names, they are usually P2P hand-wired, so the MSRP can be jacked up accordingly to compensate for the increased labor costs. For DIY, I think P2P would still be preferable, given the "one-off" nature of the build.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.