Lin topology current driver rebuild - Page 3 - diyAudio
Go Back   Home > Forums > Amplifiers > Solid State

Solid State Talk all about solid state amplification.

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 16th June 2006, 05:33 PM   #21
Shredly is offline Shredly  United States
diyAudio Member
 
Join Date: May 2006
Quote:
Originally posted by AndrewT
Hi,
there is absolutely nothing wrong with running multiple grounds to an audio Central Star Ground. For the moment I shall assume the designer got his multiple grounds right.
I'm not so sure about that, particularly with the cap from the heatsink to the audio (i.e., preamp) ground.

Quote:
Originally posted by AndrewT
What is much more likely is that the safety earth and the audio ground have been connected. This is probably the source of your hum.
The strings of all electric guitars are connected to the audio ground. This is true whether one uses a guitar that puts out hi-Z, where the negative sides of the pickup coils are connected to the audio ground, intended for connection to a traditional preamp running (usually) A-class, or lo-Z, where there are three wires: audio ground, positive output, and negative output, intended for input to a diff amp. This absolutely is a safety issue, as you note below. I will not do anything that compromises this arrangement, no matter how quiet it might be alleged to make things; it will certainly make ME quiet, because it will make me dead.

Early guitar amp designs had problems with the way this stuff was done sometimes. I got shocked by an incorrectly wired amp once, it was an old tube amp, and I have a practice I follow religiously: I always check the voltage on the strings against safety ground before I put my hands on the axe when I hook up to a new or modified amp, and I also always check the voltage on the negative speaker output against safety ground. It's just good sense.

Quote:
Originally posted by AndrewT
The finger pointing at Zeners etc may have accounted for wideband noise, but you have now discounted them.
We completely agree on this.

Quote:
Originally posted by AndrewT
There is a big problem here.
Guitarists get killed if the grounding is unsafe.

You need to be absolutely sure of the safety side of your grounding and how it interacts with your metal guitar strings BEFORE you contemplate ANY changes to your pre or power amp.
I am very aware of the risks. I like breathing. Your concern is heartening. It is always clear in my mind, and I don't trust- I test.

I was trained on high voltage in the Navy (and I mean HIGH voltage- klystrons and like that- stuff you put a safety lock on that you're the only holder of the key to, and post a Marine next to with clear orders to allow no one to even look at it, much less touch it, before you even think of opening the door, and when you open it, you reach in with a grounded six-inch thick copper probe on the end of a fiberglas rod and ground every dang thing you can see or might touch before you go inside), and I have an extremely clear idea about the importance of safety ground. I don't point at a live circuit, having had my hand slapped until I knew better by habit.

Quote:
Originally posted by AndrewT
I could advise in a non professional way but at either end of a forum I cannot guarantee some misunderstanding will not lead to a wiring error that could be unsafe.
So, for your safety I bow out.
I don't necessarily see that as a problem- believe me, if you make a suggestion and I don't think it's safe, I won't hesitate to speak up about it, and in any case, like I say, I don't trust- I test. I would value your experience; please reconsider.

Quote:
Originally posted by AndrewT
Your Email is still off.
That shouldn't be a problem any more.
  Reply With Quote
Old 16th June 2006, 06:06 PM   #22
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
Hi,
I think you have persuaded me that you understand and know how to check the safety of your Guitar/Pre/Power system.

Let's restate my belief.

The grounds (or returns) for safety, signal, power, speaker, PSU, relays etc. and digital should all have their own dedicated connection to the Central Star Ground.

The safety ground must be connected to all external conductive chassis parts permanantly. This connection should never be removed nor need to be removed to do any other maintenance.

The Central Star Ground (audio ground) should be connected to the safety ground with a disconnecting network consisting of any combination of one or more of
1. side by side (inverse parallel) power diodes,
2. hi frequency capacitor,
3. resistor,
4. ground lift switch. The ground lift switch is an optional ADDITIONAL item to 1, 2, and/or3.

All the other returns should come to a Central Star Ground.
Arrange the lugs/terminals, so that you group high current together, separate from PSU (grouped), separate from digital, relay, separate from signal (at the end).

The heatsink is capacitively coupled to the output devices. The designer has recognised this and decoupled the heatsink taking account of the series link he was creating. You have destroyed his design philosophy by shorting out the second half of the series capacitive coupling. Maybe the designer was wrong and you know better!

I do not know how a guitar amp is wired to guarantee absolute safety to the player in the event of two co-incidental errors. Firstly, the safety ground becoming faulty. Secondly, faulty internal mains wiring making some internal low voltage component live.
Can you inform me how this is done?
__________________
regards Andrew T.
  Reply With Quote
Old 16th June 2006, 06:42 PM   #23
Shredly is offline Shredly  United States
diyAudio Member
 
Join Date: May 2006
Hi Andrew,
Quote:
Originally posted by AndrewT
Hi Shredly,
the CCS and input filters are all OK.
The NFB DC block and bass roll off are set excessively low but should not cause a problem. Slight performance benefits may accrue from adjustment to the input filter (59Hz) and NFB roll-off (22Hz) frequencies, but due to the specialised nature of the amp may not be worth the effort to alter.
The low E string on a concert pitch-tuned guitar (A2=440) is just above 80Hz; for this application I don't see much point to it. If I was looking to use it for the effects channel on a home theater system, it would be worth doing, or if I was playing a bass guitar through it; but bass guitars are notorious for causing the speakers of guitar amps to exceed the flexibility of their surrounds, and as you probably know once this happens it generally damages the voice coil and rapidly blows the speaker up, so I have a bass amp with bass speakers for my bass guitar.

It's worth mentioning that I run 2 2x12 cabinets, at 4 ohms impedance each, one for each channel (the amp is two channel, and only the left channel is shown in my schematic; you probably caught that, but I'll note it for any bystanders). Note that these are simple 12" high-bass-to-midrange drivers, with no tweeter or bass woofer. Their frequency response rolls off at about 7kHz and about 50Hz, which is relatively traditional for guitar amp speakers.

Quote:
Originally posted by AndrewT
The main high frequency roll of is due to C8, the Miller compensation cap. This is a cheap and cheerful method of ensuring stability in the amplifier and a little bit of stability adjustment from C10 is usually all that is necessary for a well behaved amp in mass production. It works and it is often used.
You're right, and I should have spotted it. Gee, that class was a long time ago! Thanks for pointing out this is a Miller cap- that helps my understanding quite a bit.

Quote:
Originally posted by AndrewT
Prior to the C8 roll off, the input filter, as you have correctly noted, reduces the audio response to -3db @ 15kHz, about -1db @ 7.6kHz. In view of the harmonics the guitar is capable of and your speaker response do you think there is any advantage in opening up the HF filter by say half an octave?
You'd be surprised just how nasty those harmonics can get. Part of what manufacturers found early on was that guitars make a lot of high-frequency sounds that don't necessarily contribute to a subjectively/psychologically pleasing or "musical" sound. The highest note on most guitars is around 2kHz, and musically pleasing sounds at least from a guitar stop around 5 or 6kHz. In addition, using wound strings, when one moves from note to note one gets "string squeak," and attenuating the highs tends to somewhat suppress this. I rediscover these things relatively often when listening to newbies playing through home stereos because they can't afford an amp :P . I have the habit of rolling down my tone knobs, particularly on my acoustics, which I run direct into the board; the high end from a 12-string acoustic will blow the top of your head off if you don't roll it off.

It's worth noting that guitar tone knobs generally are nothing but a cap to ground on one side of a pot, with the input signal to the top of the pot and the output wired to the rotor. Volume knobs are a pot from input to ground, with the output on the rotor. Pickups are coil-and-magnet transducers; the strings vibrate within a magnetic field, and the moving magnetic field cuts the coil, generating the signal. Various pickup combinations and wiring to pickup switches and tone and volume knobs are done; the only tricky thing worth noting is something called "humbucking," which is done by using two pickups wound in opposite directions and connected forward-and-reverse, so that while the desired signal adds, being in-phase, any EMI picked up by the two coils is out-of-phase and at least partially cancels. This arrangement works pretty well. In general, the guitar is unpowered; everything is passive. Some manufacturers put various active electronics into their axes; I have generally found these to have various undesirable attributes probably not worth going into here, although this methodology is much more useful for acoustics, and is how both of mine are set up. Acoustic pickups are a whole different world; the two main methods are piezo-electric contact microphones, stuck to various points inside the body, and "coaxial" pickups that are placed under the bridge saddle where the strings go over. The coaxials don't feedback, but have tonal problems; the piezos have awesome tone, but feed back at the drop of a hat. Various notch filtering and phasing schemes are used to avoid these problems with the piezos, with varying degrees of success.

The preamp in this system is A-class and opamps; the A-class can be overdriven to give musically pleasing distortion; only this A-class has gain control, the remainder are fixed-gain setups with volume being controlled by a pot wired to input and ground, with the output to the amplifier circuit wired to the rotor. Tone is a three-band Baxandall-topology-derived network, with about 12dB boost-and-cut. A coil reverb (a rather interesting circuit in and of itself- there are some very esoteric tricks to making a coil reverb sound good, apparently) and a bucket-brigade chorus (mostly just a delay line, but with a time-varying phase shift) may be switched in or out of the signal path. A single channel effects loop is available on the front panel prior to insertion to the reverb and chorus board. I generally run a noise gate (a device that shuts off the output when the input goes below a threshold) and a 6-band graphic EQ in that loop.

Power for the preamp and effects boards is derived from the +/-40V power amp rails, using a 7815/7915 regulator pair; it ain't efficient and it ain't pretty, but it works and it's easy to design and easy to implement. This arrangement is very quiet, other than some pink noise from the opamps and the A-class; there is only a very little bit of hum perceptible, off the top of my head down by about 60-70dB from the signal level, and while the pink noise is increased by higher settings of the treble sections of the tone networks, and by turning up the gain on the A-class, it isn't objectionably high, as such things go. If I do anything to the preamp, it will be to beef up the filter caps on the +/-15V regulation circuit, double-check that it's properly bypassed, and perhaps modify the very first input buffer amp to accept lo-Z balanced TRS cables from my guitars, by running it in differential mode rather than the single-ended mode it's in now. But that would be a heck of a lot of work, because I have an effects chain I run in front of it that's all hi-Z unbalanced, as well as rewiring the axes (and I have three that I run through this amp).

Quote:
Originally posted by AndrewT
Do you use the mix out or the pre out? They both have medium to high source impedance. Very short cables and low input capacitance on the next stage are vital.
Actually, low-impedance three-wire balanced outputs are available, and I use those; since the board has differential inputs, this is the lowest noise option. They come from the chorus board, derived from the signal that winds up on the amp board at the mix out J4 and pre out J1 jacks you're looking at. The output of the board goes to the amp in J2 jack.
  Reply With Quote
Old 16th June 2006, 09:10 PM   #24
Shredly is offline Shredly  United States
diyAudio Member
 
Join Date: May 2006
Thanks for the grounding advice. I'll look it over with that idea in mind; thinking about it, I believe that what you're describing is a lot like what I'm looking at here, with the exception that the safety ground lug on the chassis is hard-wired to the star point, rather than with dual opposing power diodes or any of the other arrangements you've suggested. This actually happens through one of the aluminum standoffs that the amp board is mounted to the chassis with, for reasons that are now relatively obvious to me the one closest to the star point, and there's a short jumper wire, so I could put one of those arrangements in place there; but I'll have to think about it to make sure it's safe.

The star goes to the appropriate points on the two big filter caps, which are on either side of the star point; these are C20 and C21, not shown on the schematic, and they are 6m8F 50V. They are very near the edge of the board, and the PSU ground enters at the edge of the board closest to the star point, at J7.

A very short jumper connects the star to the ground points of R32 and R33, and C17, C15, and R29; this ground area is quite large, and a trace from it also goes to the ground connected to D6 and R19, and to R24, and their counterparts in the right channel.

A trace emerges from the star, and splits in two; these two run all the way around the outside of the board side-by-side to J3, where one of them connects to J3-6. In the original schematic, this has been labeled "dirty ground." This "dirty" ground is also connected to R7 and C3, and C7, but to nothing else on the amp board. The other goes to J3-3, and also to D1, R3, C2, and R10; this is labeled "preamp ground," and true to its name it goes to the preamp, and is the only ground present there. This trace also connects to the grounds of J1 and J2, and this is the ground that the cap to the heatsink was connected to. The "dirty ground" is used a couple places on the chorus board, and for the LEDs that signal which channel is in use and whether the chorus and reverb are on or off. By and large, the ground used on both the chorus board and the preamp board is the "preamp ground;" and the +/-15 supply is derived right where +/-40 comes onto the board. +40 is also sent to the preamp board, where it is used only for the control switch circuits and for the amplifier circuit that sends signal to the reverb coil. I have noticed that there is hum in the reverb, and now I know why.

-40 goes straight from the entry point on the power amp board to the big power transistors, with smaller traces leading away from them down into their support circuitry; it also runs under the back edge of the heatsink. +40 runs along the front of the board, outside the split preamp and dirty grounds, and jumpers into the support circuitry, with a big trace going up to the power transistors; both +40 and -40 wind up on the far side of the board, going to the chorus board connector, along with the incoming signals and the two grounds.

If I were designing the board, I'd have a power plane and a ground plane internal, and two trace layers on the outsides; and little shield traces around the signal path traces; and the dirty ground run pretty much where it is now. For a single-side design, it's fairly well laid out, and the power and ground are run nicely. And single side boards are, of course, cheap.

The power comes in to the transformer through a board with the fuse and a "ground select" switch, that is connected to a big poly cap to safety ground from the power cord. You can select to put the top of this cap to hot, neutral, or nothing. There is a convenience outlet wired directly to the incoming power, no fuses, no switches, nothing, just straight wired. After traversing the safety fuse and power switch, hot goes to the primary of the main transformer; this transformer is wound 2:1. Neutral is direct wired to the transformer. There is a bypass and anti-turnon-surge cap across the primary. The safety ground from the cord goes directly to a chassis lug, and the convenience outlet's safety ground goes to that same lug. The transformer is center-tapped, with a bridge across the legs. The center tap goes to the star point via a spade connector, J7, and the outputs of the rectifier to the +40 and -40 spade connectors at the power input end of the power amp at J5 and J6. Since the star point is connected to the chassis through the standoff, at J10, this is the safety ground entry point.

The grounds on the connectors are all wired to preamp ground, and that includes the two input connectors and the front-panel effects loop, as well as all the pre-out, mix-out, and amp-in connectors, as well as the footswitch connectors. The only exception to this rule is that J8 and J9, the speaker outputs, whose grounds are wired to the tops of R32 and R33; given that these are 0.68 ohms and in parallel, and can handle 10W, I do not see this as a safety concern, but I am careful with the ground sides of the speaker cables nevertheless, just in case. I've used insulated plugs here.

Overall, it appears that this amp is wired much the way you have recommended. The power amp ground, the ground for the chorus and preamp, the ground for the chassis, and the PSU ground all go separately to the star; the safety ground is wired to the chassis, not the star, and according to you noise could be reduced by adding a disconnecting network between the star and the chassis lug on the power amp board. This would not compromise the safety of the chassis, which would still be direct-wired to the safety ground from the power cord, nor would it compromise the safety of the jacks, which all go to preamp ground and then to the chassis through lugs on the preamp and chorus boards. I'll certainly try that. I won't mess with the preamp and chorus lugs, because they're not having much noise. Which precise option would you recommend for the disconnecting network? The ground lift switch I can work out for myself; I'll have to think a bit to see if I think that's a good idea from a safety standpoint. I think it might be OK, but I want to be sure.

As far as the cap to ground from the heatsink, I tried with it and with a jumper, and it was quieter with the jumper. It's worth noting that the jumper is to the preamp ground, and that the heatsink connects to a larger finned heatsink that is connected to the chassis, but there is heatsink grease involved and there might be insulation as a result, and even capacitance.

I'm not entirely certain I understood your point about the series capacitances; could you elaborate a bit on that?

But in the end, I have to tell you, I still feel strongly that regulating this supply will not only provide more consistent power to the system, but also will eliminate much of the noise I'm currently fighting. I'll try these two mods you've suggested,

Quote:
Originally posted by AndrewT
I do not know how a guitar amp is wired to guarantee absolute safety to the player in the event of two co-incidental errors. Firstly, the safety ground becoming faulty. Secondly, faulty internal mains wiring making some internal low voltage component live.
Can you inform me how this is done?
The safety ground is THE safety device. It is required by UL that the safety ground pin of the power plug be wired directly to a lug on the chassis immediately upon entry to the chassis, and that the ground pins of any convenience outlets be wired to the same lug; no electrical device is permitted between the ground pin and the lug, except wire and connectors, and the wire must be at least the gauge of the largest of the hot and neutral at their largest point(s). It is required that the ground on all external connectors with the exception of speakers be wired to the chassis as well, but this can be indirect, i.e. through a circuit board, although it must be a hard-wired connection; ground lugs like the one at J10 on the power amp board exist on the preamp and chorus boards to ensure this. This is the only guarantee that a dangerous voltage cannot exist on the metal parts of the chassis.

Smart musicians check outlets before they use them and refuse to use any that are unsafe; I carry an Isobar which has a built-in tester, not only for ground faults but for hot-neutral reversal, because you never know when you'll get a power hit, and I won't plug my equipent in if it shows a ground fault or hot-neutral reversal. This last is important because the fuse is in the hot incoming line. I also try to use GFCI (you know them as "RCD") outlets whenever they are available, and most bars (which is generally where I play) must have them by the electrical code.

There is no provision for the double-fault condition you have brought to light; it is not required in American equipment. Basically, the integrity of the safety ground is your only guarantee. By code, all new wiring for the last several decades requires a ground pin; neutral is not considered safe, even though it's grounded at the feeder transformer, and sometimes at the building entry point. Of course, provided the safety ground is integral, any connection of hot to the chassis via any low-resistance path will blow the power fuse. Manufacturers put all sorts of warning labels on equipment with 3-wire power plugs warning people not to tamper with the plug or use an adapter that disables the safety ground.

It is not permitted to use the safety ground as a return, for obvious reasons; neutral must always be used.

I have seen amps that were wired with two-wire plugs; the one I mentioned earlier that shocked me was one of these. I avoid them. No safety ground == no safety.
  Reply With Quote
Old 17th June 2006, 09:10 AM   #25
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
Hi Shredly,
we do this differently.
Quote:
the wire must be at least the gauge of the largest of the hot and neutral at their largest point
I cannot remember the exact numbers but our logic for an earth fault tries to establish the ratios of resistance from mains distribution board out to the equipment and back through the earth grounding wire. We calculate the voltage on the equipment chassis for the fuse ratings used and limits the chassis voltage to about 50V if the fuse does not blow. This usually results in an earth wire of smaller guage than the power wires but can also be of larger guage. It's not a calculation that most electricians would normally carry out for a domestic installation but the designer of a commercial/industrial system would almost certainly do it. I suspect thay have a set of tables/graphs to make this selection quick and easy.
Quote:
I have seen amps that were wired with two-wire plugs
double insulated with the double concentric square symbol.

This does not necessarily follow
Quote:
No safety ground == no safety.
if the equipment is correctly designed, the possible fault conditions are thought through and risk eliminated/managed.

If guitar /preamp/poweramp systems are designed along the same lines as normal domestic equipment then connecting the RCA/TRS/XLR grounds to chassis introdues a risk of carrying fault voltages back through the system to the guitar.
Quote:
It is required that the ground on all external connectors with the exception of speakers be wired to the chassis as well, but this can be indirect, i.e. through a circuit board,
I think this compromises the separation of mains isolation from the low voltage stages.

Quote:
I'm not entirely certain I understood your point about the series capacitances; could you elaborate a bit on that
No, because I don't know enough about it . Having read it a few times by respected designers I believe it and I think Borbely discusses it in one of his articles.


I don't think this applies
Quote:
regulating this supply will not only provide more consistent power to the system, but also will eliminate much of the noise
you can have a perfectly smooth supply (no ripple and infinitely low output impedance) going to your supply rails and a grounding error/loop will still cause hum at your speakers. It is not ripple on the supply rails that is causing the hum.

I have fitted the parallel combination of 10r and 10nF1000V ceramic from safety earth to audio ground. I have tried the switch but always removed it because in my tests it never helped. I believe ground lift is more common in commercial gear to help eliminate awkward grounding problems and the preferred position is open. It is accessible on the back of a PA amp I recently bought to hear/see how commercial gear compares to domestic. I have only become aware of the diodes since joining this Forum and have never experimented with it.
__________________
regards Andrew T.
  Reply With Quote
Old 21st June 2006, 10:35 AM   #26
Shredly is offline Shredly  United States
diyAudio Member
 
Join Date: May 2006
Quote:
Originally posted by AndrewT
Hi Shredly,
we do this differently.
First, I was unclear: the safety ground wire to the service outlet must be at least as large as the hot and neutral at the largest points within the chassis or in the line cord. The intent is to ensure that the chassis is at ground potential, to the limit of the ability of current to enter the device. If the ground wire is greater or equal to the maximum size of the input wire, then it will not melt until at least the time the input wire does. This is assuming the fuse fails to blow; it ensures that no dangerous voltage can exist on the chassis, from any source, because such a voltage's maximum current capability is determined, at maximum, by the wiring of the hot and neutral inputs. No multiple failure that does not include the ground connection can render it unsafe; the safety ground prevents it. Thus my statement, "no ground == no safety." I'll point out that I wasn't talking about Class II double-insulated equipment, because almost no professional or serious amateur sound equipment in the US is Class II. Quite frankly, I suspect that any such equipment would be noisy, but not having ever seen any that was of sufficient grade to have a power supply that wasn't already noisy as heck, I can't say it for sure.

Second, I was also unclear about double-insulated, Class 2 equipment, which does not have a safety ground connection. Because dangerous voltages cannot exist on the chassis without a double fault, due to the reinforced insulation, it is permitted (and in fact required) that the chassis not be grounded. This is supposedly safer because there is no ground path in case of a double fault. This is how most kitchen, bath, and outdoor powered appliances and tools are constructed in the US. It is not, however, how most if not all musical instrument amplifiers, televisions, and high-grade audio equipment, are constructed. Nor is it how computers are built. Again, in the US. How you guys do it over there I got no idea.

Quote:
Originally posted by AndrewT
This does not necessarily follow if the equipment is correctly designed, the possible fault conditions are thought through and risk eliminated/managed.
You may be right, if it's low power and correctly built; but amps get hot sometimes, and everything that's a good insulator will generally burn. While I can see Class II equipment's point, I prefer Class I; even if the insulation melts, the chassis is still connected to ground, and there are other paths to ground than the chassis of a Class I device.

Quote:
Originally posted by AndrewT
[B]If guitar /preamp/poweramp systems are designed along the same lines as normal domestic equipment then connecting the RCA/TRS/XLR grounds to chassis introdues a risk of carrying fault voltages back through the system to the guitar.I think this compromises the separation of mains isolation from the low voltage stages.
If it's double-insulated, then the chassis isn't connected to anything and won't be unless there's a double fault, and it's no more risk than touching the chassis. If it's Class I, the chassis is connected to safety ground, and it's no more risk than touching the chassis. Either way, it's no more risk than touching the chassis.

I've never seen a guitar amp where the ground on the plugs wasn't connected to the chassis, nor a mixer where the shield of the XLR wasn't either; that includes the few Class II practice amps and old stereo amps I've taken apart, too. Sorry, but that's just the way it is. Ask anybody who's looked inside, or designed them for the US marketplace. I'll say, though, that I haven't seen a British amp built for use in Britain, although I've played plenty of Marshalls built for the US.

Quote:
Originally posted by AndrewT
[B] No, because I don't know enough about it . Having read it a few times by respected designers I believe it and I think Borbely discusses it in one of his articles.
As far as I can make out, the idea is that the transistors are capacitively coupled to the heatsink. That being the case, you can decouple them to ground by attaching the heatsink to ground. I still don't see the point of adding another capacitor in series, unless it's to reduce the parasitic capacitance (remember, capacitances parallel in series and add in parallel). In general, decoupling capacitances aren't that critical. I'd like to see your references.

I'll also point out that I appear to have reduced the noise by directly coupling it to ground, rather than through a capacitor. I guess I really DO know more than the people who designed it, judging by the results I got.

Quote:
Originally posted by AndrewT
I don't think this applies you can have a perfectly smooth supply (no ripple and infinitely low output impedance) going to your supply rails and a grounding error/loop will still cause hum at your speakers. It is not ripple on the supply rails that is causing the hum.
This ignores the point.

I'll try again. If I connect a shorting plug (that's a 1/4" TS plug with the tip connection soldered to the sleeve) to the power amp input jack, the hum doesn't go away. Now, there ain't no ground loop- the input is directly shorted to ground. And it's still humming. You tell me, sport, where's the hum coming from? Am I beginning to get through here? There is one and only one possible source for this hum, and that's the unregulated power supply, which I have already shown both theoretically and by measuring it is putting a 5V sawtooth on the positive and negative rails. There just ain't anywhere else for it to come from, unless space aliens are injecting it to torment me. Wugga wugga.

Now, this is a Lin amp, so it's got a diff amp at the front end that's supposed to be cancelling that hum. And that's all real nice and like that, but I gotta tell you, IT AIN'T WORKIN. So I got two choices here:
1. Fix the diff amp so it removes the hum. Oh, and by the way, check and make sure it really is the diff amp, and not the class-A amp in the next stage. Great, where do I start? I gotta rattle cages left and right to get anything at all on this site, and friend, I STILL don't have any indication from you people how to bias this freakin thing and I been waitin two weeks here. All I get is, "well, what you suggested is wrong." Neato. Thanks, that was lots of help. I kinda figured that one myself, bein as how it's all biased now and works fine. WHAT'S RIGHT? Is there a clue available here? Hello? Should I care? I'm thinkin NOT.
2. Fix the power supply so it don't hum in the first place. This I know how to do, and I don't need no "expert" to come along and give me any "advice" neither. I think I'll do it. I bet it works. Here's the thinkin: it hums when its quiescent, there ain't no signal comin from the input. We already know that the power supply is a dirty POS, with 14% ripple. I'm removin the ripple. Care to place a wager?

By the way, is it worth noting that Naim use regulated power supplies exclusively? That there is one of the finer amplifier companies on the planet. I agree with them. Any questions?

Quote:
Originally posted by AndrewT
I have fitted the parallel combination of 10r and 10nF1000V ceramic from safety earth to audio ground. I have tried the switch but always removed it because in my tests it never helped. I believe ground lift is more common in commercial gear to help eliminate awkward grounding problems and the preferred position is open. It is accessible on the back of a PA amp I recently bought to hear/see how commercial gear compares to domestic. I have only become aware of the diodes since joining this Forum and have never experimented with it.
Ground lift in commercial gear is dangerous, IMO. The problem is that the safety ground is the only guarantee you have that you won't get a shock; if you don't have it, then any voltage present on the tip of the 1/4" TS plug is present on the ground of your guitar, there's only the pickup coils in the way, and your hand is touching the ground of the guitar.

And by the way, I've tried ground lifts, and my experience is that they don't work, except to find faults. Of course, I'm anal about grounding my gear, so I don't get ground problems; in my experience, "ground loop" is mythology for improper grounding in the first place, so decoupling it by adding a lift seems to help. If you have a problem that a ground lift seems to cure, my experience is, you're covering up the real problem. I've always found a cable or axe or mike with a problem in it every time a ground lift has seemed to help. I use it as a diagnostic tool, to find and isolate the faulty component. I guess some folks think it's better to just leave it broken and cover it up with something that makes it positively dangerous. No accounting for taste.
  Reply With Quote
Old 21st June 2006, 12:05 PM   #27
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
Hi,
Quote:
Originally posted by AndrewT
I have fitted the parallel combination of 10r and 10nF1000V ceramic from safety earth to audio ground. I have tried the switch but always removed it because in my tests it never helped. I believe ground lift is more common in commercial gear to help eliminate awkward grounding problems and the preferred position is open. It is accessible on the back of a PA amp I recently bought to hear/see how commercial gear compares to domestic. I have only become aware of the diodes since joining this Forum and have never experimented with it.
Ground lift in commercial gear is dangerous, IMO. The problem is that the safety ground is the only guarantee you have that you won't get a shock; if you don't have it, then any voltage present on the tip of the 1/4" TS plug is present on the ground of your guitar, there's only the pickup coils in the way, and your hand is touching the ground of the guitar.

I think we have both misunderstood each other here.

We agree that the safety ground must never be disconnected. no compromise is acceptable.

The (commercial) ground lift I referred to is an additional switch between audio ground and safety ground to make a direct connection between the two grounds or optionally break the direct connection between audio ground and safety ground.

If an example of a retailed piece of gear has a switch that broke the safety ground to chassis connection I would hope the authorities would have identified it as dangerous and prevented it from becoming available to the marketplace. With worldwide sales via the internet this policing may be becoming more difficult.
__________________
regards Andrew T.
  Reply With Quote
Old 21st June 2006, 07:08 PM   #28
Shredly is offline Shredly  United States
diyAudio Member
 
Join Date: May 2006
Quote:
Originally posted by AndrewT
I think we have both misunderstood each other here.
While that is possible, I think it's more likely that you're not familiar with musical equipment; that's easily corrected, and better sooner than later.

Quote:
Originally posted by AndrewT
We agree that the safety ground must never be disconnected. no compromise is acceptable.
On this we agree completely.

Quote:
Originally posted by AndrewT
The (commercial) ground lift I referred to is an additional switch between audio ground and safety ground to make a direct connection between the two grounds or optionally break the direct connection between audio ground and safety ground.
This is true, and IMO dangerous. Remember, the cases of the microphones, and the strings on the guitars, are connected to audio ground. This is universal among electric guitars, though not electric-acoustics, and almost so among mikes.

Note also the existance of "phantom power," which is conveyed as 48VDC along the - and ground of a three-wire low-impedance balanced circuit to power condenser microphones. While the current-handling capability of this circuit is deliberately highly restricted, on the close order of 50mA, it nevertheless shows an important point.

Because this is so, to protect the user from dangerous voltages that could appear on the hot lead of a two-wire high-impedance unbalanced circuit-- that is, on the tip of a TS plug-- the sleeve of the TS plug, and the shield in the cable it is connected to, must be connected to safety ground, not merely to audio ground. The same is true if the mike is not double-insulated, and most of them aren't.

While ground lifts are supplied on some equipment, I consider it dangerous and don't use them for this reason. In addition, in 25 years of playing and messing with equipment, I have never seen a ground lift fix a problem that was not better addressed by fixing the real problem rather than compromising the safety of the users; generally, there is a problem with the shield on some cable somewhere, and the one advantage of the ground lift is, if it can be applied channel by channel, the source of the problem can be identifed unambiguously. The fact remains, however, that musicians must perform at the time they say they will, and if a ground lift can fix a problem on stage and allow them to do so, most of them will use it rather than allow their roadies to fix the real problem. This occasionally has tragic consequences. Keep in mind also that many musicians drink (and not necessarily alcohol, although it is common) on stage, and imagine the potential consequences of a drink spilled into an amp that has had the ground lifted.

Quote:
Originally posted by AndrewT
If an example of a retailed piece of gear has a switch that broke the safety ground to chassis connection I would hope the authorities would have identified it as dangerous and prevented it from becoming available to the marketplace. With worldwide sales via the internet this policing may be becoming more difficult.
You can probably see from my points above precisely why the audio ground in a guitar amplifier is and must be the same as the safety ground.

While I cannot accept your suggestion as a permanent arrangement, I could follow it as a diagnostic procedure. However, I should tell you that I am very near to a regulator design that will use the LM338 to regulate, and preserve its current limiting capabilities despite its being used with larger voltages than it is designed to handle. The key to this is a technical paper by National that shows how to use the LM317 (a very similar regulator, also designed to work "floating" as the '338 is) in the context of a 2 to 160V variable supply; the key to the design is a transistor to block the input supply voltage, and a zener to ensure that the voltage the regulator sees is constant and low. I'll publish shortly in the other thread, and then I'll implement it on my system and report the results here.

If I have trouble, your suggestion will be the first thing I do, again, only as a diagnostic; I can't and won't leave it that way. But if it shows a problem, I'll then have to mull over how I wish to solve it.

I apologize for being a little waspish. I'm kinda frustrated over here.
  Reply With Quote
Old 6th July 2006, 05:44 PM   #29
Shredly is offline Shredly  United States
diyAudio Member
 
Join Date: May 2006
I've proven that the problem was fixed by introducing a regulator. So much for ground loops. BTW, it's worth noting that there is a logic error in your earlier statement:
Quote:
Originally posted by AndrewT
you can have a perfectly smooth supply (no ripple and infinitely low output impedance) going to your supply rails and a grounding error/loop will still cause hum at your speakers. It is not ripple on the supply rails that is causing the hum.
The second statement does not necessarily follow from the first, and the fact that I removed the ripple from the power supply and the hum went away shows it conclusively. There were no grounding problems in my amp; only a cheap power supply.

Unfortunately, after allowing for line variations, i.e. making sure that the regulator would not brown out under any likely variation of the line voltage, the regulator was dissipating far too much heat for my liking. I am now trying a capacitor multiplier to see if acceptable performance results; so far, the results in breadboard are looking pretty good. I expect to know whether this will answer my needs this week.
  Reply With Quote
Old 8th July 2006, 08:42 AM   #30
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
Hi,
it difficult for me to be objective in this situation since you have the equipment in front of you.

So, as a result, I still think that the description of the hum was much more likely to be caused by a grounding error.

I am surprised that the ripple in the PSU was causing this, but you have a solution.

I wonder if all those extra resistors off the rails reduced the PSRR?

Keep us informed of how the multiplier performs with some details/schematic of the final solution.

Well done, regarding persistance and debugging.
__________________
regards Andrew T.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Rebuild of old box and driver Moondog55 Multi-Way 1 25th May 2009 05:00 AM
Some topology of current feedback amplifier aht Solid State 7 13th August 2005 03:49 PM
Hawksford wrong on amp topology for current driving loudspeakers? capslock Solid State 5 7th November 2003 03:43 PM


New To Site? Need Help?

All times are GMT. The time now is 02:34 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2