My first step to a safe amp:
The valve buffered amp I use actually, uses a big torroid tranny for the LM3875.
And three print-trafos for the valve: one 6V/450mA for the filament, and two 2x15V/250mA in series, they give 2x17.5V at 10mA.
So, I had the dilemma of an extended primary cabling. Thats why, I made one package out of the three print trannies, casted in epoxy (Astorit, is the brand, some Araldit expanded with slate flour).
Here, the fotos:
Franz
The valve buffered amp I use actually, uses a big torroid tranny for the LM3875.
And three print-trafos for the valve: one 6V/450mA for the filament, and two 2x15V/250mA in series, they give 2x17.5V at 10mA.
So, I had the dilemma of an extended primary cabling. Thats why, I made one package out of the three print trannies, casted in epoxy (Astorit, is the brand, some Araldit expanded with slate flour).
Here, the fotos:
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Franz
I try now to continue this thread myself, even, if my english is not so good.
I cannot take some responsibility and try to resume here, what I learned by now.
Look at security issues in this sense: what goes in, can come out of the box, in the worst case.
So: the most important thing is to realize all primary cabling in a safe way. Of course, there are some norms regarding in- and output terminals of an amp. But with the voltages within a chipclone, there is imho no big risk at this terminals, when you make at the primary/mains side everything correct.
Let's beginn with the protection earth (PE) connection of the box (hoping, your enclosure is out of metal).
This is very essential, specially with alu cases: as alu is always a little bit corrodized what is an isolation, a very good contact must be ensured. The basic rules are:
- use star washers between the case and a dish washer to ensure contact (the star washer bites into the corrodized skin of the alu and ensures contact)
- use yellow/green LITZ wire for this connection
- dont solder, use cable feets, crimped (in case of a shortage, soldering can melt quickly; massive wire can break within the use of the device)
This is, how I did the primary connection with my best knowledge.
What I describe here, is just true, when you use the same style of mains connectors. They include:
- ICE-connector
- RF-filter (min. 6A)
- Fuses
- mains switch
The rules here:
- everything at the primary side double isolated
- every open contact (double) isolated
- No soldering, crimping
Please, dont shrink tubes with the lighter, as it is to hot. Use a heatgun!
Here you see the 100-220 Ohm resistor, connecting the star ground to the enclosures PE-point.
In this VBIGC are several trannies included in one block, as I showed before. So, just ONE double isolated cable leads to this block.
Last but not least: should you use an input-poti or an -selector, the they must be grounded very well to (to avoid hum or shock when touching), using star washers or separate grounding cables.
Happy and safe gaincloning!
Franz
I cannot take some responsibility and try to resume here, what I learned by now.
Look at security issues in this sense: what goes in, can come out of the box, in the worst case.
So: the most important thing is to realize all primary cabling in a safe way. Of course, there are some norms regarding in- and output terminals of an amp. But with the voltages within a chipclone, there is imho no big risk at this terminals, when you make at the primary/mains side everything correct.
Let's beginn with the protection earth (PE) connection of the box (hoping, your enclosure is out of metal).
An externally hosted image should be here but it was not working when we last tested it.
This is very essential, specially with alu cases: as alu is always a little bit corrodized what is an isolation, a very good contact must be ensured. The basic rules are:
- use star washers between the case and a dish washer to ensure contact (the star washer bites into the corrodized skin of the alu and ensures contact)
- use yellow/green LITZ wire for this connection
- dont solder, use cable feets, crimped (in case of a shortage, soldering can melt quickly; massive wire can break within the use of the device)
An externally hosted image should be here but it was not working when we last tested it.
This is, how I did the primary connection with my best knowledge.
What I describe here, is just true, when you use the same style of mains connectors. They include:
- ICE-connector
- RF-filter (min. 6A)
- Fuses
- mains switch
The rules here:
- everything at the primary side double isolated
- every open contact (double) isolated
- No soldering, crimping
Please, dont shrink tubes with the lighter, as it is to hot. Use a heatgun!
An externally hosted image should be here but it was not working when we last tested it.
Here you see the 100-220 Ohm resistor, connecting the star ground to the enclosures PE-point.
An externally hosted image should be here but it was not working when we last tested it.
In this VBIGC are several trannies included in one block, as I showed before. So, just ONE double isolated cable leads to this block.
Last but not least: should you use an input-poti or an -selector, the they must be grounded very well to (to avoid hum or shock when touching), using star washers or separate grounding cables.
Happy and safe gaincloning!
Franz
Oh my god! I thought I had seen some crazy stuff in this business, but molding stuff in a food package, simply great Franz! 
You seem to have done a proper work here. The use of an integrated inlet/fuse holder/switch/mains filter is a convenient and safe start. But being a besserwisser I can add/point out a few things which might be useful.
1. Protective earth connection point.
You are right about using star washers here. Also, like you did the PE screw should be a dedicated PE connection and not also be used as a mounting detail for any component (such as the IEC inlet).
For each PE connection you should use a separate nut and washer.
If you are using anodized aluminium for the case, remember that the anodized surface is a perfect insulator and there might not be electrical contact between different panel segments. If in doubt about the contact integrity, use separate PE connections from each segment to the PE point.
2. Mains cable color coding.
In Europe, the standard colors are:
HOT - Brown
NEUTRAL - Blue
PE - Green/Yellow
It is advisable to stick to these colors. Also, make sure you use wire intended for carrying mains voltage. The heavy-gauge wire used for speaker connections and power supply wiring is generally NOT ok to use for mains voltage. Since you don't know which of the HOT and NEUTRAL wires that really is the "hot" one both should of course be treated as HOT!
Generally, make the PE wire from an IEC inlet longer than the other two. If someone applies force and the IEC inlet should become loose the last wire to "leave the case" should be PE!
3. Do not think you are safe just because you have used protective earth!
In many buildings (especially old ones) the PE is generally not very good. It might actually be more dangerous to use PE rather than avoid it altogether! PE is only effective with a PE current sensing circuit breaker and the new "5-lead" system that is the standard for new installations in Europe today.
But whenever you build something with a metal case - connect the case to PE!
I have to rush now, but later I will tell you a funny thing with those wirewound resistors in alu housing...
Cheers
/Magnus
You seem to have done a proper work here. The use of an integrated inlet/fuse holder/switch/mains filter is a convenient and safe start. But being a besserwisser I can add/point out a few things which might be useful.
1. Protective earth connection point.
You are right about using star washers here. Also, like you did the PE screw should be a dedicated PE connection and not also be used as a mounting detail for any component (such as the IEC inlet).
For each PE connection you should use a separate nut and washer.
If you are using anodized aluminium for the case, remember that the anodized surface is a perfect insulator and there might not be electrical contact between different panel segments. If in doubt about the contact integrity, use separate PE connections from each segment to the PE point.
2. Mains cable color coding.
In Europe, the standard colors are:
HOT - Brown
NEUTRAL - Blue
PE - Green/Yellow
It is advisable to stick to these colors. Also, make sure you use wire intended for carrying mains voltage. The heavy-gauge wire used for speaker connections and power supply wiring is generally NOT ok to use for mains voltage. Since you don't know which of the HOT and NEUTRAL wires that really is the "hot" one both should of course be treated as HOT!
Generally, make the PE wire from an IEC inlet longer than the other two. If someone applies force and the IEC inlet should become loose the last wire to "leave the case" should be PE!
3. Do not think you are safe just because you have used protective earth!
In many buildings (especially old ones) the PE is generally not very good. It might actually be more dangerous to use PE rather than avoid it altogether! PE is only effective with a PE current sensing circuit breaker and the new "5-lead" system that is the standard for new installations in Europe today.
But whenever you build something with a metal case - connect the case to PE!
I have to rush now, but later I will tell you a funny thing with those wirewound resistors in alu housing...
Cheers
/Magnus
Regarding molding, I must add some advices, to avoid loss of money:
- Test your trafos very well before molding. Make very good soldering joints, inside!
- Every contact with mains voltage, should be more than 5 mm away from the outside, or you use a plastic box to mold, wich you dont remove afterwards.
- Dont use commercial resin, wich gets hot at hardening! It will destroy your circuit. Just cool hardening, not too much shrinking material!
Franz
BTW:
Before I made the trafo-package, I tested the trafos with an old 4K headphone and and a detector receiver, to learn about the humming magnetic field of the trafos and to choose the best orientation.
Unfortunately some people won't follow safety procedures even if their life depended on it. (pun intended)
Here goes for the rules;
1 Only use hook up wire for the mains that is rated for the voltage. i.e. UL1007, 300V insulation rating. Available anywhere. There is also a 600 V version, UL1015.
2 The spacing between two points with voltage differential of 40 to 400 volts is to be 4 mm. For voltage between 400 and 1200 the spacing is to be 8 mm.
3 For mains connections in the pictures I have seen here use INSULATED spade connectors. For extra safety, heat shrink can be placed over the connectors. This will give double insulation. But, the heatshrink is also to be UL recognized.
4 In the bio-med world the maximum resistance allowed between the earth connection at the end of the power cord and anywhere on the chassis is 10 milliOhm. This should give you an idea as to what to do for connections of chassis panels to chassis panels, to make certain there is a good connection between them.
5 All equipment connected to the mains is to be hi-pot tested for both DC and AC of 1500 volts with a max current rating of 2 mA. Granted we do not have the capabilites to hi-pot as DIYers but keep a couple things in mind.
The DC test verifies that there are no direct connections from the chassis to high voltage. (between output devices and heatsink) The AC test verifies that there isn't any leakage due to capacitive or inductive coupling to the chassis. An example of this would be a defective power transformer, filter caps that are leaky or from an earlier post, a RC connection from common to earth. If an RC connection is going to be used than the C should be a UL recognized component. An X or Y cap. Do NOT use standard film caps. They are not rated for leakage or voltage in a safety application.
6 All switches, wires, transformers, power cords, etc. are to be UL recognized. For you EN people the items are to be VDE or TUV recognized. Any components that are recognized by a safety agency will have the mark of that agency on the component.
7 Do not use hot melt for hi voltage insulating purposes. Hot melt will not stick to items that have any oil on them. Hot melt is also really good at letting loose due to vibration or age. This should also apply to electrical tape. Tape is rated but it likes to let loose over time.
8 When using crimp connections on mains wires do a pull test. Pull on the connector as hard as you can, if the connector comes off the crimp isn't any good. Yes, this means that some of the crimping tools being used should be replaced. Pliers are not good for this application. And yes, a good crimp will pass this test. There is a UL spec for this but I won't go there.
9 The reason IEC connectors are used is because if the power cord is hard wired to the product using a strain relief, the power cord then has to go through a pull test. This test is to have the product sitting on the end of a table and a 20 pound weight is hung from the power cord for 5 minutes. The power cord or strain relief can not fail. The weight required for this test is different for different product lines but 20 pounds is a general rule. It is not an easy test to pass! If an IEC connector is used the product does not have to be subjected to this test because it is assumed that the cord will disconnect from the product removing power.
10 Because a "safety rated" transformer is used in a power amp. At least they are supposed to be. An earth connection to chassis is not required by UL. Many consumer receivers, CD players, etc. do not have an earth connection. But, an earth connection is always a good idea.
The earth connection is to be rated for the amount of current the product CAN pull from the AC mains while under full load. This means that for a 200 watt amp the earth is to be able to carry over 5 amps. UL does test this. One thing that is common for DIYers is to use an earth that is TO SMALL. You can't have an earth that is to big.
11 Soldering the earth and/or mains instead of using crimp connectors is OK. Again, the issue with many DIYers is that the wire and connections are to small. If you are worried about the connections getting to hot and melting the solder... think about that long and hard.
12 If you want really good connections use medical grade wall outlets and power cords. These will have a green dot on them.
The wall outlet is rated for a maximum amount of leakage within itself, a maximum amount of contact resistance and a minimum amount of pull force required to remove the plug from the outlet.
The power cord is rated for leakage and the ends will have clear plastic around the crimps. Clear plastic is used so the techs can see if the crimps are failing over time.
13 Banana connectors are an ugly subject. Banana's were used in the old days because that's what they had. But remember, in those days a power amp might have been 20 watts. At that power level this connector was not an issue. Now move the clock forward to today. They are still used because of being "grandfather'd in." UL hates them big time. A big amp can expose the user to hi voltages as we know. There isn't much we can do for this except use the new connectors mentioned in previous posts.
14 Speakers can get away with using banana's because they are not recognized, listed or even looked at by the agencies. Speakers are not connected to the mains so they are "assumed" to be inherently safe. I'm not saying this is real life but that's the way it is.
15 If you are really nervous about safety then replace your wall outlet used for the stereo with a GFI unit.
16 The agencies don't care if a manufacturer makes the biggest piece of junk on the planet. All they care about is that if the product fails, it fails in a safe manor.
17 EMI/RFI issues are a totally different subject. Power amps (excluding switch mode types) are not tested to these standards because they do not contain a clock. And at $1500/day for lab time I don't know of to many audio companies that have subjected their gear to this. This subject is a whole other post.
Happy and safe DIYing
Mike
Here goes for the rules;
1 Only use hook up wire for the mains that is rated for the voltage. i.e. UL1007, 300V insulation rating. Available anywhere. There is also a 600 V version, UL1015.
2 The spacing between two points with voltage differential of 40 to 400 volts is to be 4 mm. For voltage between 400 and 1200 the spacing is to be 8 mm.
3 For mains connections in the pictures I have seen here use INSULATED spade connectors. For extra safety, heat shrink can be placed over the connectors. This will give double insulation. But, the heatshrink is also to be UL recognized.
4 In the bio-med world the maximum resistance allowed between the earth connection at the end of the power cord and anywhere on the chassis is 10 milliOhm. This should give you an idea as to what to do for connections of chassis panels to chassis panels, to make certain there is a good connection between them.
5 All equipment connected to the mains is to be hi-pot tested for both DC and AC of 1500 volts with a max current rating of 2 mA. Granted we do not have the capabilites to hi-pot as DIYers but keep a couple things in mind.
The DC test verifies that there are no direct connections from the chassis to high voltage. (between output devices and heatsink) The AC test verifies that there isn't any leakage due to capacitive or inductive coupling to the chassis. An example of this would be a defective power transformer, filter caps that are leaky or from an earlier post, a RC connection from common to earth. If an RC connection is going to be used than the C should be a UL recognized component. An X or Y cap. Do NOT use standard film caps. They are not rated for leakage or voltage in a safety application.
6 All switches, wires, transformers, power cords, etc. are to be UL recognized. For you EN people the items are to be VDE or TUV recognized. Any components that are recognized by a safety agency will have the mark of that agency on the component.
7 Do not use hot melt for hi voltage insulating purposes. Hot melt will not stick to items that have any oil on them. Hot melt is also really good at letting loose due to vibration or age. This should also apply to electrical tape. Tape is rated but it likes to let loose over time.
8 When using crimp connections on mains wires do a pull test. Pull on the connector as hard as you can, if the connector comes off the crimp isn't any good. Yes, this means that some of the crimping tools being used should be replaced. Pliers are not good for this application. And yes, a good crimp will pass this test. There is a UL spec for this but I won't go there.
9 The reason IEC connectors are used is because if the power cord is hard wired to the product using a strain relief, the power cord then has to go through a pull test. This test is to have the product sitting on the end of a table and a 20 pound weight is hung from the power cord for 5 minutes. The power cord or strain relief can not fail. The weight required for this test is different for different product lines but 20 pounds is a general rule. It is not an easy test to pass! If an IEC connector is used the product does not have to be subjected to this test because it is assumed that the cord will disconnect from the product removing power.
10 Because a "safety rated" transformer is used in a power amp. At least they are supposed to be. An earth connection to chassis is not required by UL. Many consumer receivers, CD players, etc. do not have an earth connection. But, an earth connection is always a good idea.
The earth connection is to be rated for the amount of current the product CAN pull from the AC mains while under full load. This means that for a 200 watt amp the earth is to be able to carry over 5 amps. UL does test this. One thing that is common for DIYers is to use an earth that is TO SMALL. You can't have an earth that is to big.
11 Soldering the earth and/or mains instead of using crimp connectors is OK. Again, the issue with many DIYers is that the wire and connections are to small. If you are worried about the connections getting to hot and melting the solder... think about that long and hard.
12 If you want really good connections use medical grade wall outlets and power cords. These will have a green dot on them.
The wall outlet is rated for a maximum amount of leakage within itself, a maximum amount of contact resistance and a minimum amount of pull force required to remove the plug from the outlet.
The power cord is rated for leakage and the ends will have clear plastic around the crimps. Clear plastic is used so the techs can see if the crimps are failing over time.
13 Banana connectors are an ugly subject. Banana's were used in the old days because that's what they had. But remember, in those days a power amp might have been 20 watts. At that power level this connector was not an issue. Now move the clock forward to today. They are still used because of being "grandfather'd in." UL hates them big time. A big amp can expose the user to hi voltages as we know. There isn't much we can do for this except use the new connectors mentioned in previous posts.
14 Speakers can get away with using banana's because they are not recognized, listed or even looked at by the agencies. Speakers are not connected to the mains so they are "assumed" to be inherently safe. I'm not saying this is real life but that's the way it is.
15 If you are really nervous about safety then replace your wall outlet used for the stereo with a GFI unit.
16 The agencies don't care if a manufacturer makes the biggest piece of junk on the planet. All they care about is that if the product fails, it fails in a safe manor.
17 EMI/RFI issues are a totally different subject. Power amps (excluding switch mode types) are not tested to these standards because they do not contain a clock. And at $1500/day for lab time I don't know of to many audio companies that have subjected their gear to this. This subject is a whole other post.
Happy and safe DIYing
Mike
Mike
I print your text out and will sleep the next week with it under my pillow (or read it on the toilet)!
Really, it needs time to understand everything, because I am not a security expert.
More inputs like this are VERY welcome! Every cloner will be able to build safe amps in future (and it is the best base, for later safe tube/valve experiments).
Franz
I print your text out and will sleep the next week with it under my pillow (or read it on the toilet)!
Really, it needs time to understand everything, because I am not a security expert.
More inputs like this are VERY welcome! Every cloner will be able to build safe amps in future (and it is the best base, for later safe tube/valve experiments).
Franz
Now, this is a bit off topic but I post it here anyway...
DO NOT TRY THIS AT HOME KIDS!
Those wirewound MIL-STD-specified resistors in nice gold anodized aluminium housings have quite a special "failure mode".
This was discovered during some late-night circuit hacking at our student's lab, events that now and then turn into "potential destructive testing sessions...", that is you crank up the juice until something explodes.
Well, these resistors do it in a quite spectacular way. A sharp bang like a pistol shot is followed by the resistor body being fired out of one of the ends like a miniature projectile. With a nice trail of the wire from the spool behind like a harpoon shot!
And the epoxy-filled aluminium housing should be perfectly clean on the inside afterwards. Quite a neat trick but not exactly the kind of failure mode I would want in the fuselage of say a commercial airplane.
I don't remember the brand we tested this on but it might have been ARCOL.
/Magnus
DO NOT TRY THIS AT HOME KIDS!
Those wirewound MIL-STD-specified resistors in nice gold anodized aluminium housings have quite a special "failure mode".
This was discovered during some late-night circuit hacking at our student's lab, events that now and then turn into "potential destructive testing sessions...", that is you crank up the juice until something explodes.
Well, these resistors do it in a quite spectacular way. A sharp bang like a pistol shot is followed by the resistor body being fired out of one of the ends like a miniature projectile. With a nice trail of the wire from the spool behind like a harpoon shot!
And the epoxy-filled aluminium housing should be perfectly clean on the inside afterwards. Quite a neat trick but not exactly the kind of failure mode I would want in the fuselage of say a commercial airplane.
I don't remember the brand we tested this on but it might have been ARCOL.
/Magnus
in case of a failure?
Franz,
I am not sure what you are using that resistor for but I guess you have done a signal ground -> earth connection such as the one described here: http://sound.westhost.com/earthing.htm
Regarding this type of resistor, I think they are quite good. Great thermal properties and easy mounting. For an audio amplifier I would not worry about this "gun shot" type failure (which BTW requires quite a large overvoltage/current). But for more demanding applications like life support systems or aircraft avionics I certainly would not like this kind of behaviour where one blown component could damage other (and perhaps critical) components in the equipment rack. But then, an electrolythic capacitor blows with a pretty large bang too. And leaves quite a mess... Smells bad too. 
Nothing in this world is perfect.
I am not sure what you are using that resistor for but I guess you have done a signal ground -> earth connection such as the one described here: http://sound.westhost.com/earthing.htm
Regarding this type of resistor, I think they are quite good. Great thermal properties and easy mounting. For an audio amplifier I would not worry about this "gun shot" type failure (which BTW requires quite a large overvoltage/current). But for more demanding applications like life support systems or aircraft avionics I certainly would not like this kind of behaviour where one blown component could damage other (and perhaps critical) components in the equipment rack. But then, an electrolythic capacitor blows with a pretty large bang too. And leaves quite a mess... Smells bad too.
Nothing in this world is perfect.This is off subject but one of the most fun times I had was testing a PWM motor controller. One of the tests is to completly encase the controller with tissue paper, turn on the power to the controller and then throw a dead short across the output of the controller. The goal was that the tissue paper can not start on fire. But what a BANG! Smoke everywhere. We did the test outside.
All components come from the factory with smoke installed. When the smoke is released the part has reached the end of it's service life.
All components come from the factory with smoke installed. When the smoke is released the part has reached the end of it's service life.
Someway off topic:
My (hopefully) safe amp is getting its enclosure:
very old walnut wood, now just the first appliance of shellack.
The knob is not definitively: I am dreaming of a custom made, blue anodized alu knob...
And of course, the LED is blue!
Franz
My (hopefully) safe amp is getting its enclosure:
very old walnut wood, now just the first appliance of shellack.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
The knob is not definitively: I am dreaming of a custom made, blue anodized alu knob...
And of course, the LED is blue!
Franz
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.