This project has taken bloody forever (and cost over twice my original estimate), but finally got it powered up last night for the first time. Sounds quite good, and so far I'm impressed, but I have two issues to solve... (#1 is easy, #2 I need help with, so for those who are easily bored with rambling, skip to #2 
)
1) Apparently my heatsinks, which I thought were pretty large, are not large enough. Output devices get too hot to put your finger on after its been on for an hour or so, and bias keeps slowly rising. No big deal...I'm going to get a pair of 120mm 12V Vantec Stealth fans and power them from a wall-wart, and bolt them close to the heatsinks, blowing air across them. In the meantime, I have bias reduced to about 1.2A to at least listen for a while.
2) This is the issue I need help with. I hear a faint (but loud enough to be annoying) 'buzz' through the output, especially with headphones (the amp has a couple of headphone jacks). I poked and prodded all over the amp, and finally I unbolted the transformer (an AVEL toroidal from Parts Express) and swung it away from the driver boards. The buzz immediately went away. The chassis is not large enough for me to move the transformer anywhere else...
How do I tell if this is electrical noise, or EMI? I suspect that it is EMI. If so, and the solution is to use either mu-metal or perhaps the TI metal from Michael Percy, how do I go about using the stuff? Does the entire transformer need to be sealed in the stuff? Or just wrap the outer circumference?
Or could you take strips of mu-metal, and wrap them through the toroid, working your way around the thing? And does this new shield need an electrical connection to chassis ground?(the amp is powered by only hot and neutral...no earth ground).
I lean towards trying the TI metal from Percy, since the claim is that it works for both electrical and EMI shielding, but it is awfully expensive...
What works, what doesn't? Your thoughts are greatly appreciated...
)1) Apparently my heatsinks, which I thought were pretty large, are not large enough. Output devices get too hot to put your finger on after its been on for an hour or so, and bias keeps slowly rising. No big deal...I'm going to get a pair of 120mm 12V Vantec Stealth fans and power them from a wall-wart, and bolt them close to the heatsinks, blowing air across them. In the meantime, I have bias reduced to about 1.2A to at least listen for a while.
2) This is the issue I need help with. I hear a faint (but loud enough to be annoying) 'buzz' through the output, especially with headphones (the amp has a couple of headphone jacks). I poked and prodded all over the amp, and finally I unbolted the transformer (an AVEL toroidal from Parts Express) and swung it away from the driver boards. The buzz immediately went away. The chassis is not large enough for me to move the transformer anywhere else...
How do I tell if this is electrical noise, or EMI? I suspect that it is EMI. If so, and the solution is to use either mu-metal or perhaps the TI metal from Michael Percy, how do I go about using the stuff? Does the entire transformer need to be sealed in the stuff? Or just wrap the outer circumference?
Or could you take strips of mu-metal, and wrap them through the toroid, working your way around the thing? And does this new shield need an electrical connection to chassis ground?(the amp is powered by only hot and neutral...no earth ground).
I lean towards trying the TI metal from Percy, since the claim is that it works for both electrical and EMI shielding, but it is awfully expensive...
What works, what doesn't? Your thoughts are greatly appreciated...
This seems to be such a common problem that I don't understand why no one makes toroid shields. Simply a round steel sheet metal case that the torroid sits in and is grounded to the chassis. Seems pretty straight forward. Hmm, maybe there;s a market there and I should just shut up and make them...
Hmm.
Like an upside-down metal bowl, replaces the upper washer, bolt goes through it, throught the torroid and to the chassis. Any takers?
-Brian
Hmm.
Like an upside-down metal bowl, replaces the upper washer, bolt goes through it, throught the torroid and to the chassis. Any takers?
-Brian
Thanks guys!!
I can't rotate the transformer...it sits on top of a Hammond enclosure, and most of the electronics (including the bridge) are down below, and I have a circular cutout with a rubber grommet feeding the wires through to the inside of the chassis. I can rotate it a cm or two...not enough to matter.
I wish someone DID make enclosures for toroids. I'd be in line to buy one right about now.
I believe that I'm going to buy a sheet of the TI shield from Percy. Supposedly you don't have to worry about deformation of the metal reducing its effectiveness.
If I do this (and I believe I have a method that may work to enclose it) does the shield need to be run to chassis ground?
I can't rotate the transformer...it sits on top of a Hammond enclosure, and most of the electronics (including the bridge) are down below, and I have a circular cutout with a rubber grommet feeding the wires through to the inside of the chassis. I can rotate it a cm or two...not enough to matter.
I wish someone DID make enclosures for toroids. I'd be in line to buy one right about now.
I believe that I'm going to buy a sheet of the TI shield from Percy. Supposedly you don't have to worry about deformation of the metal reducing its effectiveness.
If I do this (and I believe I have a method that may work to enclose it) does the shield need to be run to chassis ground?
No, not in this case.
The shielding mechanism in materials like mu-metal is very different from "normal" EMI shields (which needs to be grounded), the later are protecting you from electromagnetic waves (like radio-waves) but mu-metal and similar materials "block" (cancels) low-frequency magnetic fields.
The shielding mechanism in materials like mu-metal is very different from "normal" EMI shields (which needs to be grounded), the later are protecting you from electromagnetic waves (like radio-waves) but mu-metal and similar materials "block" (cancels) low-frequency magnetic fields.
first the ventillation issue -- a 12V fan will run very quietly at around 6 or 7 volts, of course the CFM are a lot lower, but much better than ambient convection. a fan with larger diameter blades is going to be quieter than those mini-donut variety jobbies.
second, the noise of which you complain -- you have to let the transformer radiate heat -- transformers have L, R and C as well as coupling losses -- you could use a screen mesh if it is really problematic, but you have to give the trafo some breathing room.
second, the noise of which you complain -- you have to let the transformer radiate heat -- transformers have L, R and C as well as coupling losses -- you could use a screen mesh if it is really problematic, but you have to give the trafo some breathing room.
It may be a BIG PITA, but moving it is the only practical solution. Making/buying a new enclosure for the transformer....one that will let you move it further away...is the most pragmatic solution.
Trust me........some of us have made that mistake. Only took once.
Your amp will sound better the further away you get that transfomer.
Jocko
The transformer is pretty large in comparison to the rest of the setup...the whole amp is indeed 10lbs of s*!# in a 5lb bag. The only way I could get the transformer further away would be to make it remote from the rest of the amp via some sort of AMP connector. I don't like that solution at all.
Jocko Homo, you seem to be suggesting that even with shielding that I won't be able to solve the issue. Regardless, I have to try. A remote transformer just doesn't do it for me...
Jocko Homo, you seem to be suggesting that even with shielding that I won't be able to solve the issue. Regardless, I have to try. A remote transformer just doesn't do it for me...
Signal wires
Echo,
Do the signal wires run in close proximity to the toroid? If so re-route the signal wiring or use shielded cable with the drain wire connected to ground. Interference from the toroid will be higher top and bottom, so running signal wiring over the top of a toroid will result in hum.
Try to keep all AC wiring and transformers separated from signal wires and the amp pcb. If you can manage this, hum shouldn't be a problem.
Hope this helps,
m
Echo,
Do the signal wires run in close proximity to the toroid? If so re-route the signal wiring or use shielded cable with the drain wire connected to ground. Interference from the toroid will be higher top and bottom, so running signal wiring over the top of a toroid will result in hum.
Try to keep all AC wiring and transformers separated from signal wires and the amp pcb. If you can manage this, hum shouldn't be a problem.
Hope this helps,
m
Signal wires never really run close to the transformer...at least they are as far away as the enclosure allows. But, the design includes a volume control which the input signal wires run to right after they enter the enclosure. Turning the volume up and down makes very very little difference in the level of the 'buzz', which leads me to believe that it is being induced into the driver boards, which are about 5" away from the transformer.
One more thought on the fans:
I haven't installed many fans in amps, but check these out:
http://www.newegg.com/app/ViewProductDesc.asp?description=35-118-212&depa=0
I have used these many times in computer enclosures. Very thin, very very quiet. Highly recommended.
-Brian
I haven't installed many fans in amps, but check these out:
http://www.newegg.com/app/ViewProductDesc.asp?description=35-118-212&depa=0
I have used these many times in computer enclosures. Very thin, very very quiet. Highly recommended.
-Brian
The connector on that fan, I think, is allowing you to use it at either 12V or 7V (by connecting to the 12V and 5V line). At full 12V is will be noiser than I want. Going to go with the Vantec 80mm Stealth, and feed it about 10V or so.
RE: The transformer hum. I got a replay back from Rod...
I had thought that copper would do no good for EMI as such. However, I can buy a cheap piece at the local hardware store and see what it does.
What confuses me is his statement about 'touching the chassis or the top plate'. The toroid is wrapped with whatever it is they wrap toroids with...stuff is an insulator AFAIK...
RE: The transformer hum. I got a replay back from Rod...
If you read this Rod, I mean no disrespect my posting your reply...I'm not the only one fighting this problem.
I had thought that copper would do no good for EMI as such. However, I can buy a cheap piece at the local hardware store and see what it does.
What confuses me is his statement about 'touching the chassis or the top plate'. The toroid is wrapped with whatever it is they wrap toroids with...stuff is an insulator AFAIK...
hmmmm maybe I need to revisit my lead case I built for the same problem..... it completely surrounds the transformer (top and sides) and has the bolt going through it and connected to the chassis!!!! never smoked, and definitely reduces the noise (but didn't eliminate it). I'm now wondering if some huge amount of the transformers rated current is being wasted as some sort of short!!!!
Tony.
Tony.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.