Very spcifically:
Do I need to reference DC regulated 6VDC heaters to 1/4 B+ in my Aikido preamp?
Wavebourn Labs in one posting suggested that this wasn't necessary.
The center-tap of the secondaries of my 6.3VAC trafo is referenced to 1/4 B+ via 300K and 100K (the 100K bypassed by a 0.1uF cap). When B+ is applied, the 6.3VAC trafo buzzes.
QUESTION: Can the -ve (GND) connection of the regulator circuit be referenced to 1/4B+? This would move the B+ reference to the DC side of the regulator and may take some of the strain of the trafo.
Thanks,
Charlie
Do I need to reference DC regulated 6VDC heaters to 1/4 B+ in my Aikido preamp?
Wavebourn Labs in one posting suggested that this wasn't necessary.
The center-tap of the secondaries of my 6.3VAC trafo is referenced to 1/4 B+ via 300K and 100K (the 100K bypassed by a 0.1uF cap). When B+ is applied, the 6.3VAC trafo buzzes.
QUESTION: Can the -ve (GND) connection of the regulator circuit be referenced to 1/4B+? This would move the B+ reference to the DC side of the regulator and may take some of the strain of the trafo.
Thanks,
Charlie
The regulator has a GND on it, although I cannot see how it is actually connected to ground, other than via the tube filaments, if at all. Maybe this is the problem, and that I actually need to connect the "GND" on the DC side to IEC GND. If this is the case, then wouldn't the "Alternative Configuration" in my schematic (see previous message) apply?
Charlie
Charlie
cbutterworth said:
Yes, you do need to reference the heaters in the Aikido to prevent heater insulation breakdown resulting in heater to cathode shorts that will kill the tube. This requirement is the same for AC or DC powered heaters for the Aikido. So you are on the right track.
Applying the heater reference on the center tap of the transformer is the usual method for referencing the heaters when running the heaters on AC. But your situation is different because of the regulator.
With the regulator, applying the heater reference on the center tap of your transformer will do nothing for your heaters. All that you accomplished was to apply the heater reference to the input of the regulator. Hopefully the regulator did not fry.
Applying the heater reference to only the -ve terminal as in your second configuration will most likely cause problems too.
You need to create a "virtual center tap" on the output of the regulator by placing two 47R, 0.5W resistors across +ve and -ve. Then connect the heater reference to the node between the 47R resistors. The heater supply +ve and -ve need to "float" so do not connect the +ve or -ve to anything other than the heater circuit. Do not ground the -ve. Ignore the "GND" terminology that Welborne uses in their documentation. There is no ground in this application.
I crudely modified one of your sketches to illustrate the concept
Dave
Attachments
Whew!
I have been running the Aikido for over six months now with the center tap of the transformer connected to 1/4 B+. This likely explains the problem with the buzzing 6.3V trafo. Obviously, it is over-designed and has survived.
What power rating of resistor should I use? I expect that very little actual current will be carried by the resistors. Maybe, I should err on caution and use 2W resistors.
Thank-you for explaining things to me.
Charlie
I have been running the Aikido for over six months now with the center tap of the transformer connected to 1/4 B+. This likely explains the problem with the buzzing 6.3V trafo. Obviously, it is over-designed and has survived.
What power rating of resistor should I use? I expect that very little actual current will be carried by the resistors. Maybe, I should err on caution and use 2W resistors.
Thank-you for explaining things to me.
Charlie
Re: Re: Re: Referencing DC heaters to cathode
A bit less noise based on previous experimentation.
Couldn't find my old notes so I repeated the experiment tonight of measuring the output of my Aikido (quad 12SN7s) applying the two methods of referencing the heaters (dual 47R resistors vs. no resistors and applied to -ve). LM317 as the regulator for the heaters. I observed a bit less noise (a mV or so) on the scope across a resistor on the output with the inputs shorted.
leadbelly said:
A bit less noise based on previous experimentation.
Couldn't find my old notes so I repeated the experiment tonight of measuring the output of my Aikido (quad 12SN7s) applying the two methods of referencing the heaters (dual 47R resistors vs. no resistors and applied to -ve). LM317 as the regulator for the heaters. I observed a bit less noise (a mV or so) on the scope across a resistor on the output with the inputs shorted.
OK,
I'll get some 47R resistors tomorrow. I can only buy 1W locally, will these do? I can also buy 5W power resistors, I'll get a pair of these at 47R each, the power rating will be way lower than 5W.
Why 47R?
I suppose that a quick calculation will give me an idea:
6VDC across 47R and 47R = 0.19W per resistor with 3VDC output?
So 1W resistors should be more than sufficient?
Is the 3VDC at the far and of the divider )ie. 1/2 Vin) the reason for the 47R?
Thanks,
Charlie
I'll get some 47R resistors tomorrow. I can only buy 1W locally, will these do? I can also buy 5W power resistors, I'll get a pair of these at 47R each, the power rating will be way lower than 5W.
Why 47R?
I suppose that a quick calculation will give me an idea:
6VDC across 47R and 47R = 0.19W per resistor with 3VDC output?
So 1W resistors should be more than sufficient?
Is the 3VDC at the far and of the divider )ie. 1/2 Vin) the reason for the 47R?
Thanks,
Charlie
Well,
I added the 47R resistors to the regulated output. There were no bangs or fizzles. More importantly, I get less hum and great sound, although the amplification factor seems a tad lower, but I don't care about that!
I will try a larger cap on the potential divider. What kind of values should I try? I suppose that the cap acts in some kind of filter capacity, so what is the largest before it becomes detrimental?
Thanks,
Charlie
I added the 47R resistors to the regulated output. There were no bangs or fizzles. More importantly, I get less hum and great sound, although the amplification factor seems a tad lower, but I don't care about that!
I will try a larger cap on the potential divider. What kind of values should I try? I suppose that the cap acts in some kind of filter capacity, so what is the largest before it becomes detrimental?
Thanks,
Charlie
voltage readings can be nebulous!
I have never experienced this before, but I think it makes sense.
When I added the 47R resistors to the o/p of the regulator, I began to measure voltages. I clipped the -ve lead of my meter to the -ve of the regulator. Measured to the center of the 47R/47R divider, I got 3V - GOOD. I then measured back to the potential divider that should read around 80V, based on my 315V'ish B+. Instead of reading 80V, I still read 3V! This puzzled me for a while, but the Aikido sounded great, although I was a little concerned - why did my DMM read 3V and not 80V????
Then, it struck me! The 80V results from the potential divider. The potential divider is between B+ and GND. I moved the DMM -ve clip to my chassis GND and began to re-read the voltages. Regulated filaments now read 6V, while the output of the potential divider read 80V. HaHa - voltage readings can differ based upon their reference point!
Does this make sense, was the proverbial lightbulb above my head justified?
Charlie
I have never experienced this before, but I think it makes sense.
When I added the 47R resistors to the o/p of the regulator, I began to measure voltages. I clipped the -ve lead of my meter to the -ve of the regulator. Measured to the center of the 47R/47R divider, I got 3V - GOOD. I then measured back to the potential divider that should read around 80V, based on my 315V'ish B+. Instead of reading 80V, I still read 3V! This puzzled me for a while, but the Aikido sounded great, although I was a little concerned - why did my DMM read 3V and not 80V????
Then, it struck me! The 80V results from the potential divider. The potential divider is between B+ and GND. I moved the DMM -ve clip to my chassis GND and began to re-read the voltages. Regulated filaments now read 6V, while the output of the potential divider read 80V. HaHa - voltage readings can differ based upon their reference point!
Does this make sense, was the proverbial lightbulb above my head justified?
Charlie
cbutterworth - you got it 
I use 2 watt 100 ohm resistors.
What I also do is match them well. They can be a 96 ohm pair or 48 ohm pair (in your case), just as long as they are well mached. You want the + and - to have an equal reference to their source.
I also 'virtual' center tap my non-floating heaters the same way - referenced to ground. This will reduce humm even further.
I wouldn't use 1 watt resistors though... 2 watt would be a better choice. I use kiwame 2 Watt carbon films - flameproof silicon coated and the price is ok too. 5 watt is overkill.
Also, star to ground using nice big copper wire. And twist that heater wiring well - use a good gauge copper wire for heaters and grounding. Save the small wire for signal and B+ (not as many mA there).
edit: fixed typo
I use 2 watt 100 ohm resistors.
What I also do is match them well. They can be a 96 ohm pair or 48 ohm pair (in your case), just as long as they are well mached. You want the + and - to have an equal reference to their source.
I also 'virtual' center tap my non-floating heaters the same way - referenced to ground. This will reduce humm even further.
I wouldn't use 1 watt resistors though... 2 watt would be a better choice. I use kiwame 2 Watt carbon films - flameproof silicon coated and the price is ok too. 5 watt is overkill.
Also, star to ground using nice big copper wire. And twist that heater wiring well - use a good gauge copper wire for heaters and grounding. Save the small wire for signal and B+ (not as many mA there).
edit: fixed typo
Yes, I matched my 47R resistors.
As for making center taps on the tube heaters, do you mean that you make a center-tap for each tube across the socket pins? Isn't this done in the Bottlehead preamp? Also, to which GND would I reference them? The reason I ask is because my PCB GND is floated by 20R above chassis GND. My guess is that I would float the heaters above the chassis ground.
Charlie
As for making center taps on the tube heaters, do you mean that you make a center-tap for each tube across the socket pins? Isn't this done in the Bottlehead preamp? Also, to which GND would I reference them? The reason I ask is because my PCB GND is floated by 20R above chassis GND. My guess is that I would float the heaters above the chassis ground.
Charlie
Yes, center-tap for each tube at the pins is how i do it.cbutterworth said:
Hmmm. I don't know your amp, but to me there is only ever one ground. 20 R doesnt seem like much to me at the ground level though. It seems to me that it might be to reduce noise coming from a less than optimally grounded circuit...
I center tap all non-floating heaters (in a channel) together and then connect that straight to your grounding screw on the chassis.
My preamp is an Aikido using 6SN7s. It does seem that the 20R lift between PCB GND and chassis GND worked to reduce hum levels. Having said that, the degree of hum is now very, very negligible, and I expect that further tweaks could make is no-existent to an ear next to the speaker. A couple of mods that I plan are:
1. new chassis with aluminum top-plate instead of steel.
2. Using a HT trafo with universal primaries, so that I can use the 120V primaries with my 124V mains rather than at present with a 115V rated trafo.
3. Switching from diode to tube rectification, although it'll probably not reduced hum at all, but the rectifier tubes look good.
4. increase the cap on the potential divider.
5. Center-tap the heater pins, as per your message.
6. More careful routing of signal wiring and GND wiring.
I expect that the largest improvements will be from reducing the stress on the B+ transformer by using one with primaries closer to my 124V mains. The Hammond 300 series are also physically larger than their 200 series 115V counterparts, so they may be less prone to fluxes, or whatever.
Charlie
1. new chassis with aluminum top-plate instead of steel.
2. Using a HT trafo with universal primaries, so that I can use the 120V primaries with my 124V mains rather than at present with a 115V rated trafo.
3. Switching from diode to tube rectification, although it'll probably not reduced hum at all, but the rectifier tubes look good.
4. increase the cap on the potential divider.
5. Center-tap the heater pins, as per your message.
6. More careful routing of signal wiring and GND wiring.
I expect that the largest improvements will be from reducing the stress on the B+ transformer by using one with primaries closer to my 124V mains. The Hammond 300 series are also physically larger than their 200 series 115V counterparts, so they may be less prone to fluxes, or whatever.
Charlie
Ron,
Take a look at either of the schematics posted earlier in this thread. In my thread starter, there are two scematics in the image file.
The first shows how to make a potential divider for AC. The B+ goes through a 300K, then a 100K paralleled with a 0.1uF cap. The 100K/0.1uF goes to GND. Anyway, these three components form the voltage or potential divider.
For DC regulated heaters, it seems that the potential divider is wired as shown in Dave's message - included in this posting. At least, this is what I have done, and it works, it is also easy to do.
The cap is the only cap in this schematic (0.1uF). At some point, I will try a larger cap, but will wait until I can order a Wima polypropylene cap as part of a larger order. Although I do have a spare 0.1uF 440V cap, which I could parallel with the existing 0.1uF cap to give an extra large 0.2uF of capacitance.
My guess is that things can wait until I re-house the Aikido.
Does this make sens Ron?
Charlie
Take a look at either of the schematics posted earlier in this thread. In my thread starter, there are two scematics in the image file.
The first shows how to make a potential divider for AC. The B+ goes through a 300K, then a 100K paralleled with a 0.1uF cap. The 100K/0.1uF goes to GND. Anyway, these three components form the voltage or potential divider.
For DC regulated heaters, it seems that the potential divider is wired as shown in Dave's message - included in this posting. At least, this is what I have done, and it works, it is also easy to do.
The cap is the only cap in this schematic (0.1uF). At some point, I will try a larger cap, but will wait until I can order a Wima polypropylene cap as part of a larger order. Although I do have a spare 0.1uF 440V cap, which I could parallel with the existing 0.1uF cap to give an extra large
0.2uF of capacitance.My guess is that things can wait until I re-house the Aikido.
Does this make sens Ron?
Charlie
Attachments
Hi Charlie,
I just went through your thread again and realised that you have dc heaters all the way through...
Normal DC heaters (that are not floating) shouldn't be referenced to ground as I described. That is because they (unlike floating DC heaters) are connected to ground at the rectifier.
I was thinking about AC heaters when I replied... I shouldn't answer threads so early in the morning.
Please forgive me. I really hope you get this message, as I can't send you an email through the board...
The way to reference floating DC heaters (as drj759 pictured it) is of course correct.
You could also build a MJ thingy instead of that voltage divider. It will be much quieter...
I just went through your thread again and realised that you have dc heaters all the way through...
Normal DC heaters (that are not floating) shouldn't be referenced to ground as I described. That is because they (unlike floating DC heaters) are connected to ground at the rectifier.
I was thinking about AC heaters when I replied... I shouldn't answer threads so early in the morning.
Please forgive me. I really hope you get this message, as I can't send you an email through the board...
The way to reference floating DC heaters (as drj759 pictured it) is of course correct.
You could also build a MJ thingy instead of that voltage divider. It will be much quieter...
Soulmerchant,
So far, the heaters are wired as per drj759's instructions. The level of hum is now very low, the remainder, I expect will be further improved by going for a B+ trafo that can handle the higher voltage seen at my mains. As I say, when B+ gets switched-on both 6.3V and 250V trafos begin to buzz slightly. My chokes remain quiet.
My B+ supply utilizes an Amperex 10S delay tube to allow for heater pre-warming. When switched, the Amperex trips an Omron relay which simultaneously applies power to the 250V trafo and removes power from the Amperex delay tube. The diodes may also be adding some noise.
Charlie
So far, the heaters are wired as per drj759's instructions. The level of hum is now very low, the remainder, I expect will be further improved by going for a B+ trafo that can handle the higher voltage seen at my mains. As I say, when B+ gets switched-on both 6.3V and 250V trafos begin to buzz slightly. My chokes remain quiet.
My B+ supply utilizes an Amperex 10S delay tube to allow for heater pre-warming. When switched, the Amperex trips an Omron relay which simultaneously applies power to the 250V trafo and removes power from the Amperex delay tube. The diodes may also be adding some noise.
Charlie
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.