Sheldon, thanks for the link, I'll read up on that right now!
Andrew,
I just did the check with disconnecting the amplifier cards from the PSU.
With only the tweeter channel card in place:
Only very faint noise to be heard in the tweeter, no buzz.
Connected the mid- amplifier to power, Hum in midrange driver, additional buzz audible in tweeter
Hooked up subwoofer channel, same result, perhaps a marginal increase in buzz and hum..
So, why is this happening??
One amplifier channel connected to power and the x-over card is just perfect, no hum. (other amplifier cards still connected to X-over card, but not to +/-, 0V power), but connect the other amplifier cards to power, and the hum appears in all channels...
Andrew,
I just did the check with disconnecting the amplifier cards from the PSU.
With only the tweeter channel card in place:
Only very faint noise to be heard in the tweeter, no buzz.
Connected the mid- amplifier to power, Hum in midrange driver, additional buzz audible in tweeter
Hooked up subwoofer channel, same result, perhaps a marginal increase in buzz and hum..
So, why is this happening??
One amplifier channel connected to power and the x-over card is just perfect, no hum. (other amplifier cards still connected to X-over card, but not to +/-, 0V power), but connect the other amplifier cards to power, and the hum appears in all channels...
Last edited:
I didn't quite catch at what stage you adjusted the transformer position and what effect it had, but My impression is that your issue was a bit less than mine as you had to listen fairly close to a horn in order to hear the hum.
Hook up my amp to a proper horn and you would not need to get out of the armchair to hear the hum!
Anyway, my last findings following Andrews test-instructions suggest to me that this is not only a transformer radiation issue, otherwise I would have had the same hum-problem with only the one amplifier card powered up..
Hi Elbert it looks like there are at least two sources of Hum, one radiated and one electrical interaction, I'll leave the electrical to Andrew as he is much more experienced than me in this! At least the radiated one is getting under control. I Will watch with interest as I've never completely eliminated the hum from my integrated amp, I just stopped when it reached a low enough level to not be distracting in quite periods.
Tony.
Tony.
At least the radiated one is getting under control. I Will watch with interest as I've never completely eliminated the hum from my integrated amp, I just stopped when it reached a low enough level to not be distracting in quite periods. Tony.
Tony.
I would tend to agree with you there Tony, Improved magnetic shielding improved the situation, but as the hum went away with two out of three amplifier cards disconnected from power, it seems the root-cause lies somewhere else and that the transformer radiation only influences the magnitude...
If this can be solved, hopefully the answer could be of some value to you as well!
If this can be solved, hopefully the answer could be of some value to you as well!
My guess is a grounding issue. When all your channels are connected, some charging pulses are taking a different route. For three circuits you either need three completely separate supplies (i.e. separate secondaries), or one supply. Remember, current has to flow in loops so make sure they are the loops you determine not ones they find for themselves.
DF69,
My thoughts have started to veer in that direction too..
Now, I'm thinking.. I used three pairs of rectifiers to get three separate feeds to three separate banks of supply capacitors.
That' didn't work out as I ended up with three different grounds, so I connected these grounds at the capacitors. (if I had 3 separate transformers with rectifier pairs and capacitors, wouldn't that create the same problem of three independent grounds??)
Perhaps I must do the same for the +/- voltages as well??
That would effectively leave me with one common power-supply.
As it is now, I have three different and separate +/- feeds returning to a common ground point. Could this be the culprit??
My thoughts have started to veer in that direction too..
Now, I'm thinking.. I used three pairs of rectifiers to get three separate feeds to three separate banks of supply capacitors.
That' didn't work out as I ended up with three different grounds, so I connected these grounds at the capacitors. (if I had 3 separate transformers with rectifier pairs and capacitors, wouldn't that create the same problem of three independent grounds??)
Perhaps I must do the same for the +/- voltages as well??
That would effectively leave me with one common power-supply.
As it is now, I have three different and separate +/- feeds returning to a common ground point. Could this be the culprit??
Last edited:
I didn't quite catch at what stage you adjusted the transformer position and what effect it had, but My impression is that your issue was a bit less than mine as you had to listen fairly close to a horn in order to hear the hum.
Hook up my amp to a proper horn and you would not need to get out of the armchair to hear the hum!
Anyway, my last findings following Andrews test-instructions suggest to me that this is not only a transformer radiation issue, otherwise I would have had the same hum-problem with only the one amplifier card powered up..
This was final tuning after all other issues had been sorted. I have zero tolerance for hum. I view it as a defect, and hunt it down with a vengeance.
Sheldon
I see..
And if you are using horns, I your strong aversion towards hum is quite understandable!
I sure Hope I can get mine a bit lower too, allthough it's approaching the point where it is not of great consequence when listening, having that hum going in the livingroom with the stereo on.. well, it would seem a bit "chaep and nasty"..
And if you are using horns, I your strong aversion towards hum is quite understandable!
I sure Hope I can get mine a bit lower too, allthough it's approaching the point where it is not of great consequence when listening, having that hum going in the livingroom with the stereo on.. well, it would seem a bit "chaep and nasty"..
DF96 makes a lot of sense. BTW, the supplies are not really separate. When the diodes are conducting, they don't provide isolation. So might as well simplify your life and go with a common supply. If you want to split it, I would think it better to split with a final RC section in the power supply for the tweeter, and possibly the mid.
Sheldon
I see..
And if you are using horns, I your strong aversion towards hum is quite understandable!
I sure Hope I can get mine a bit lower too, allthough it's approaching the point where it is not of great consequence when listening, having that hum going in the livingroom with the stereo on.. well, it would seem a bit "chaep and nasty"..
Hum reflects a design or layout problem. It's not acceptable if it's even close to audible. If you can detect any difference with your amp on or off (no signal), within a foot or so of the speaker, it's too much. Only exception to that might be a DHT amp with AC filaments (just because you want that feature). Otherwise, it's gotta go.
Sheldon
Last edited:
Think about where the return currents go, from the capacitor ground ends back to the bridge/transformer. These ground returns are dirty, as they carry charging pulses, so they must be kept well away from signals. Assume that all the capacitors are different values (as they in fact are because of tolerances) and all the diodes have different forward voltage drops. It is quite difficult to avoid the charging pulse for one set of capacitors going via the ground for another set. It helps if you mentally replace each ground lead with a 1ohm resistor - your grounding scheme must cope with this.
I think you need just one ground connection from the cap ground ends to the amps - one lead needs to fan out to three. This helps ensure that no new loops are created. You can have separate +- connections, as these are assumed to carry ripple anyway. Don't have separate ground connections from each cap set to its own amp.
I think you need just one ground connection from the cap ground ends to the amps - one lead needs to fan out to three. This helps ensure that no new loops are created. You can have separate +- connections, as these are assumed to carry ripple anyway. Don't have separate ground connections from each cap set to its own amp.
Sheldon, a good point there with respect to the diodes not isolating when conducting... My "rationale" for doint this was that one stage could not draw down the supply voltage to another stage, e.g. load on the subwoofer amplifier channel would not affect the power suply going to the tweeter channel..
Anyway, true or not, that will have to go out the window if the only way to get rid of the hum is to connect together the + and - supplies...
Bit difficult to explain this perhaps..
The PSU and connection configuration I have right now, is as per the schematic posted on page one (allso a better one by Tony), except for the ground from the three rectifier pairs are connected together at the capacitors, i.e. the 0v of the three capacitor banks were soldered together.
From this grounding star point, there is a separate 0V wire running to each of the three amplifier cards, and a cable running back to the 0v point of the three bridge rectifier pairs. That should be in line with what you suggest if I read your post right..?
I've seen some put a series diode in a split supply for that purpose. Not sure if it really has much benefit. PSRR for most solid state amps should be pretty good. I'd rather isolate with an RC section.
BTW, I seem to remember a really good article on grounding in the articles section. But I can't find it anymore. Am I tripping?
Sheldon
Yes, an RC section would probably make just as much sense..
Let me know if you come across that article on grounding.
Apart from this thread, I've been searching about for some info, but have found very little that could shed light on my particular case.
I got Bob Cordells amplifier book a while ago, but even there, I found that grounding and hum was covered on a rather general basis...
Will be interresting to hear what Andrew has to say about the tests I did, he usually has some sharp observations..
Let me know if you come across that article on grounding.
Apart from this thread, I've been searching about for some info, but have found very little that could shed light on my particular case.
I got Bob Cordells amplifier book a while ago, but even there, I found that grounding and hum was covered on a rather general basis...
Will be interresting to hear what Andrew has to say about the tests I did, he usually has some sharp observations..
Last edited:
Unless the three ground links to the amps join at precisely the same point (difficult to arrange in practice), there will be a potential difference between them because the cap ground connections are carrying charging pulses. Remember, a star point is a theoretical abstraction. In reality it is a star region, with finite resistance. Also, charging pulses must be kept well away from a signal star point.
As I said, you need one connection to the PSU ground, which can then fan out. This ensures that you are starting from one ground potential. Don't connect the three wires directly to the PSU ground, but a short distance away. It can be as short as possible, but it must be a non-zero distance. This ensures that charging pulses can't stray down here (assuming all other ground arrangements are correct).
If you have access to Doug Self's book on power amps, read the section on the exact negative feedbck pickup point. This is the same issue, in a different context. A piece of wire is a resistor, so if you want a single voltage you have to have a single sample point.
As I said, you need one connection to the PSU ground, which can then fan out. This ensures that you are starting from one ground potential. Don't connect the three wires directly to the PSU ground, but a short distance away. It can be as short as possible, but it must be a non-zero distance. This ensures that charging pulses can't stray down here (assuming all other ground arrangements are correct).
If you have access to Doug Self's book on power amps, read the section on the exact negative feedbck pickup point. This is the same issue, in a different context. A piece of wire is a resistor, so if you want a single voltage you have to have a single sample point.
Just tied the + and - feeds togehter, that didn't make any difference...
DF96,
I must admit that fiddling about with those small distances sounds a bit nit-picky, but it would not be the first time I have underestimated things!
Admittedly, my design is not exactly according to this layout at the moment, but I shall see how I can modify it so that I can try and see what the effect is.
But, if this is really the culprit, wouldn't that materialize with only the one amplifier card connected???
DF96,
I must admit that fiddling about with those small distances sounds a bit nit-picky, but it would not be the first time I have underestimated things!
Admittedly, my design is not exactly according to this layout at the moment, but I shall see how I can modify it so that I can try and see what the effect is.
But, if this is really the culprit, wouldn't that materialize with only the one amplifier card connected???
OK, so I took a separate wire out from the 0v at the capacitor bank, and then fanned this out to the 0V of the three amplifier cards.. No Improvement...
Yes I remember that section about the negative feedback point, and to the best of my understanding, the 0V hook-up I just tried should be in line with that.
Yes I remember that section about the negative feedback point, and to the best of my understanding, the 0V hook-up I just tried should be in line with that.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.