|
|
Post by watchtower on Feb 14, 2014 21:43:57 GMT -6
This will be a giant post. Hopefully I can get some much needed help from you guys, though. Last summer I built an AMI/Tab-Funkenwerk U47. I am now deviating from (i.e. modding) the original design, but I have a ton of questions because I am still pretty new to all of this. Let me try to lay it all out before I start with the questions. The schematic of my mic can be found here: www.tab-funkenwerk.com/sitebuildercontent/sitebuilderpictures/U47ArchutAltTubewithRemote.jpgSome of the values are not quite right because my PSU is 120V, not 105V, but that is the basic layout of the circuit. Goals-Flexibility/versatility -Ability to engage a "True" cardioid mode -Use more "vintage" values for resistors to discover sound difference How I intend to reach goals-Install a relay in the mic that will allow me to switch out the rear diaphragm of the capsule from the circuit. -Install a switch in/on the PSU that allows me to activate or deactivate that relay -Install a switch in/on the PSU that allows me to "shunt" the rear diaphragm polarization voltage to ground -Install a mod in the PSU to be able to adjust heater voltage. It is currently set at about 5.6V measured at tube. -Install a mod in the mic to be able to adjust tube plate voltage -Install a mod in the mic to be able to adjust cathode voltage -Replace 1G resistors with more historic values QuestionsI did not design this PSU, so I have some questions that I think you guys will be able to help me with. I want to be able to adjust the heater voltage. The heater is on the yellow wire on Pin 6 of the Binder, and you can see where it attached to the board. Is there somewhere where I could install a potentiometer to physically adjust the heater voltage? Or perhaps is that white phillips screw driver right next to the heater connection on the PCB a potentiometer for adjustment? I plan to switch the 1G resistors to 60M, 100M, and 150M. These values coincide with what was used in the U47, U48, and M49. 150M is for the rear diaphragm, 100M for the backplate, and 60M for the "grid bias." Will this cause any problems inside the mic with the supplied voltages from the PSU, or is it safe to try these values? I am not positive how these resistors work in the circuit, so it is hard for me to know if messing with their values will cause problems. The fact that we can use 1G tells me the PSU is not really supplying many amps, as we are not using these as dropping resistors. I noticed that the M49 uses a 100M then the 150M on the rear polarization, while the U48 uses a 150M and then another 150M. However, the U48 is tapping from the 105V B+, whereas the M49 has its own remote 120V polarization supply. Is this what accounts for the difference? A relay can be used to disconnect the rear diaphragm from the mic amp. It is my understanding that I can use a relay that is powered by the tube heater voltage, or I can use a latching relay, which I have heard would be better for noise. With a 5V single pole Relay, it seems I could attach the tube heater supply to this relay, and "switch" 5V temporarily into the relay to activate it, and initiating true cardioid mode. Does this sound right? Is is simpler to go with the non-latching relay? In either case, do I need to worry about the Relay drawing too many amps and causing a fluctuation in the tube heater? What sort of specs do I need to look for in the relay? Tube heater is anywhere from 4.8V to 6.3V if I can successfully get it variable as described earlier. Does it matter if the C4 capacitor that blocks DC from the rear diaphragm to the front is "before" or "after" the relay? Some designs have the internal "cardioid only switch" disconnect the cap from the front diaphragm, but I have seen some designs where it's the other way, and the cap remains connected at all times to the front diaphragm, even when the rear gets disconnected by a relay or switch. My current Gotham GAC-7 cable uses the thick blue wire for the heater return. This heater return is tied to Ground in the PSU on Pin 5 as you can see in the attached picture. If I use a non-latching Relay, I will need to use this blue wire in the cable for the Relay return, and connect the heater return to ground. Would I be better served using this thick blue Wire for Ground instead of the Relay return, or does it not matter much? The Omni setting on the PSU is supposed to supply 0V to the rear diaphragm. However, this 0V does not have electrical continuity with ground, so I figured it would be best to install a switch that shunts the polar pattern to ground, so that I get a true 0V on the rear diaphragm. Can I just use a 2-way switch connected to green wire Pin 7 and the 0V-120V rotary switch to either pass the voltage from the rotary, or to pass it to ground? Is it safe to connect 120V to Ground through this switch if I happened to have the setting on Bi-Directional and I flipped the switch to shunt it? Do I need to runt he signal through a resistor before it hits ground? I am having trouble figuring out how one calculates the value of R8 to get his/her desired plate voltage. Is it possible to replace R8 with a potentiometer to be able to have simple control over plate voltage? I ask because even though R4 and R8 are in series, it seems like people leave R4 at 100K, and change only R8. R2 and R3 form a voltage divider for getting a positive voltage on the cathode. Can I replace R2 with a potentiometer to "dial in" my desired cathode bias/voltage? What sort of specs would that potentiometer need? Every time I turn off the mic, there is a large transient in audio. This doesn't happen in my C12. Is there a way to "fix" this, or is it expected behavior? I don't particularly care, as long as it is not damaging the internal components of the mic. Thanks a lot! Attachments:
|
|
|
|
Post by svart on Feb 15, 2014 15:29:21 GMT -6
This will be a giant post. Hopefully I can get some much needed help from you guys, though. Last summer I built an AMI/Tab-Funkenwerk U47. I am now deviating from (i.e. modding) the original design, but I have a ton of questions because I am still pretty new to all of this. Let me try to lay it all out before I start with the questions. The schematic of my mic can be found here: www.tab-funkenwerk.com/sitebuildercontent/sitebuilderpictures/U47ArchutAltTubewithRemote.jpgSome of the values are not quite right because my PSU is 120V, not 105V, but that is the basic layout of the circuit. Goals-Flexibility/versatility -Ability to engage a "True" cardioid mode -Use more "vintage" values for resistors to discover sound difference How I intend to reach goals-Install a relay in the mic that will allow me to switch out the rear diaphragm of the capsule from the circuit. -Install a switch in/on the PSU that allows me to activate or deactivate that relay -Install a switch in/on the PSU that allows me to "shunt" the rear diaphragm polarization voltage to ground -Install a mod in the PSU to be able to adjust heater voltage. It is currently set at about 5.6V measured at tube. -Install a mod in the mic to be able to adjust tube plate voltage -Install a mod in the mic to be able to adjust cathode voltage -Replace 1G resistors with more historic values QuestionsI did not design this PSU, so I have some questions that I think you guys will be able to help me with. I want to be able to adjust the heater voltage. The heater is on the yellow wire on Pin 6 of the Binder, and you can see where it attached to the board. Is there somewhere where I could install a potentiometer to physically adjust the heater voltage? Or perhaps is that white phillips screw driver right next to the heater connection on the PCB a potentiometer for adjustment? I plan to switch the 1G resistors to 60M, 100M, and 150M. These values coincide with what was used in the U47, U48, and M49. 150M is for the rear diaphragm, 100M for the backplate, and 60M for the "grid bias." Will this cause any problems inside the mic with the supplied voltages from the PSU, or is it safe to try these values? I am not positive how these resistors work in the circuit, so it is hard for me to know if messing with their values will cause problems. The fact that we can use 1G tells me the PSU is not really supplying many amps, as we are not using these as dropping resistors. I noticed that the M49 uses a 100M then the 150M on the rear polarization, while the U48 uses a 150M and then another 150M. However, the U48 is tapping from the 105V B+, whereas the M49 has its own remote 120V polarization supply. Is this what accounts for the difference? A relay can be used to disconnect the rear diaphragm from the mic amp. It is my understanding that I can use a relay that is powered by the tube heater voltage, or I can use a latching relay, which I have heard would be better for noise. With a 5V single pole Relay, it seems I could attach the tube heater supply to this relay, and "switch" 5V temporarily into the relay to activate it, and initiating true cardioid mode. Does this sound right? Is is simpler to go with the non-latching relay? In either case, do I need to worry about the Relay drawing too many amps and causing a fluctuation in the tube heater? What sort of specs do I need to look for in the relay? Tube heater is anywhere from 4.8V to 6.3V if I can successfully get it variable as described earlier. Does it matter if the C4 capacitor that blocks DC from the rear diaphragm to the front is "before" or "after" the relay? Some designs have the internal "cardioid only switch" disconnect the cap from the front diaphragm, but I have seen some designs where it's the other way, and the cap remains connected at all times to the front diaphragm, even when the rear gets disconnected by a relay or switch. My current Gotham GAC-7 cable uses the thick blue wire for the heater return. This heater return is tied to Ground in the PSU on Pin 5 as you can see in the attached picture. If I use a non-latching Relay, I will need to use this blue wire in the cable for the Relay return, and connect the heater return to ground. Would I be better served using this thick blue Wire for Ground instead of the Relay return, or does it not matter much? The Omni setting on the PSU is supposed to supply 0V to the rear diaphragm. However, this 0V does not have electrical continuity with ground, so I figured it would be best to install a switch that shunts the polar pattern to ground, so that I get a true 0V on the rear diaphragm. Can I just use a 2-way switch connected to green wire Pin 7 and the 0V-120V rotary switch to either pass the voltage from the rotary, or to pass it to ground? Is it safe to connect 120V to Ground through this switch if I happened to have the setting on Bi-Directional and I flipped the switch to shunt it? Do I need to runt he signal through a resistor before it hits ground? I am having trouble figuring out how one calculates the value of R8 to get his/her desired plate voltage. Is it possible to replace R8 with a potentiometer to be able to have simple control over plate voltage? I ask because even though R4 and R8 are in series, it seems like people leave R4 at 100K, and change only R8. R2 and R3 form a voltage divider for getting a positive voltage on the cathode. Can I replace R2 with a potentiometer to "dial in" my desired cathode bias/voltage? What sort of specs would that potentiometer need? Every time I turn off the mic, there is a large transient in audio. This doesn't happen in my C12. Is there a way to "fix" this, or is it expected behavior? I don't particularly care, as long as it is not damaging the internal components of the mic. Thanks a lot! PSU pot: You can do this, but remember that any resistance you add will become a heater, so use a much higher wattage rating pot for this because it'll get hot. 1G resistors: All resistors are dropping resistors, but it depends on the load. Since the 1G resistors are supplying essentially unloaded current to the capsule, you get a voltage but very little current draw. What R10 and R5 do is create an electrostatic charge between the film and the backplate. As the film vibrates, the charge changes and therefore the voltage changes. Depending on the capacitance of the capsule used and the input bias current of the tube/FET used, the value of the bias resistor needs to be changed, thus the difference in the resistors in the different models of mic. With values in the Megohms and Gigohms, I don't think you'll hurt anything, but your audio might change for better or worse. Usually with lower values in these places you'll lose low end. Relay: Not sure why you'd want to do this when the variable voltages for tuning pickup pattern work fine.. AND I'd worry about making/breaking high voltage on the capsule skin through the relay due to voltage transients, which might cause issues like arcing or the skin sticking to the backplate.. But anyway, sure, you can use a relay on DC heaters. Just make sure your relay doesn't use so much current that the heater rail starts to sag. You'll also want to use an anti-kickback diode so that you don't get voltage transients on the heater rails when the relay coil collapses. There are a couple different latching type relays. The most prevalent one needs a voltage on two pins, one to set one way and the other to make it go back. I'd just use a regular relay for the ease of use. C4: Should be after the relay, as in always connected between the skins. Relay return: Don't bother. Hook both grounds together in the mic and use the thickest wire as ground return. Onmi: You are dealing with CHARGE.. so as long as it sounds like it's working, it's probably fine as-is. R8: It's a dropping resistor. You need to know how much current is being used by the circuit after R8 to calculate how much it drops. Either you can put a DMM in series with it and measure current, or you can measure the voltage drop across it and calculate it that way. Or you can just put a 50K pot in rheostat hookup and use that. R2: Yep, you can do it. I don't know how much current that tube needs on the cathode, but I don't think it's much. The schematic shows 1/2W resistors, so use a 1W pot if replacing both, or 1/2w pot if only replacing R2 with a rheostat.. Transient: Unknown. You could try tracking it down by removing certain sections of the circuit and powering up/down. |
|
|
|
Post by jazznoise on Feb 15, 2014 18:52:47 GMT -6
I'm not sure you should load any capsule with that much more resistance. 1G to 150Meg is about 4.5 octaves of a difference, to speak in musical terms. Since we can see B+ as being ground to AC signals, it would follow that the bias resistor + capsule form a hi pass filter, so moving your -3dB point will move up quite a bit. If I'm right, and the U47's internal amplifier was set to -3dB at 21Hz, then your new cut off would be about 400 Hz.
So if R1 and R3 are set much higher than an original U47, then it must be to compensate for capsule differences, bias differences or amplifier differences. The easiest way is to check, of course. You'll know it when you hear it!
|
|
|
|
Post by tonycamphd on Feb 15, 2014 21:51:25 GMT -6
This is a cool thread fellas, following thanx
|
|
|
|
Post by watchtower on Feb 16, 2014 14:04:08 GMT -6
PSU pot: You can do this, but remember that any resistance you add will become a heater, so use a much higher wattage rating pot for this because it'll get hot. 1G resistors: All resistors are dropping resistors, but it depends on the load. Since the 1G resistors are supplying essentially unloaded current to the capsule, you get a voltage but very little current draw. What R10 and R5 do is create an electrostatic charge between the film and the backplate. As the film vibrates, the charge changes and therefore the voltage changes. Depending on the capacitance of the capsule used and the input bias current of the tube/FET used, the value of the bias resistor needs to be changed, thus the difference in the resistors in the different models of mic. With values in the Megohms and Gigohms, I don't think you'll hurt anything, but your audio might change for better or worse. Usually with lower values in these places you'll lose low end. Relay: Not sure why you'd want to do this when the variable voltages for tuning pickup pattern work fine.. AND I'd worry about making/breaking high voltage on the capsule skin through the relay due to voltage transients, which might cause issues like arcing or the skin sticking to the backplate.. But anyway, sure, you can use a relay on DC heaters. Just make sure your relay doesn't use so much current that the heater rail starts to sag. You'll also want to use an anti-kickback diode so that you don't get voltage transients on the heater rails when the relay coil collapses. There are a couple different latching type relays. The most prevalent one needs a voltage on two pins, one to set one way and the other to make it go back. I'd just use a regular relay for the ease of use. C4: Should be after the relay, as in always connected between the skins. Relay return: Don't bother. Hook both grounds together in the mic and use the thickest wire as ground return. Onmi: You are dealing with CHARGE.. so as long as it sounds like it's working, it's probably fine as-is. R8: It's a dropping resistor. You need to know how much current is being used by the circuit after R8 to calculate how much it drops. Either you can put a DMM in series with it and measure current, or you can measure the voltage drop across it and calculate it that way. Or you can just put a 50K pot in rheostat hookup and use that. R2: Yep, you can do it. I don't know how much current that tube needs on the cathode, but I don't think it's much. The schematic shows 1/2W resistors, so use a 1W pot if replacing both, or 1/2w pot if only replacing R2 with a rheostat.. Transient: Unknown. You could try tracking it down by removing certain sections of the circuit and powering up/down. Thanks a lot for your post! PSU Pot Thanks for the tip. Do you know where I would insert the potentiometer? Do you know what the phillips head white plastic circle is right next to the heater source on the PCB? Is that an adjustment spot? 1G Resistors Can you explain how R5 and R10 work exactly? I know we are polarizing the backplate with 60V, and varying the voltage on the rear diaphragm to create whatever polar pattern we want. And I know the difference/changes in voltage is caused by sound pressure level. But I don't understand how R5 and R10 work. In other words, I understand the idea of how it works in general, but not how the electronics/passive components come into play really. Relay Can you explain more about the relay causing voltage transients on the capsule? You're right; the current remote polar pattern switching works fine. The reason I want to mod the mic to give myself the option of "true" cardioid is for the sake of U47 authenticity. U47 has hard-switching between cardioid and omni, which cut off the rear diaphragm from the circuit in cardioid mode and actually increased the mic output by 5dB. Many people claim true cardioid sounds different in general. I'm not sure if I believe it. However, even Neumann says true cardioid has a better S/N ratio than remote pattern cardioid, and they installed an internal S2 switch in the M49 to allow you to put the mic in cardioid-only mode. C4 This cap will always be connected between the skins. The question is if you use the relay to disconnect the rear diaphragm from C4 (and thus the front), or if you disconnect C4 from the front of the capsule. Thick Blue Wire So you suggest making this Ground? I read it was good to use the thick red for the tube heater, as it draws the most amps of any part of the circuit, but I never knew where to use the thick blue. Does it really matter if I use the thick blue for ground or not? It might be too cumbersome to rewire the ground connection of the 7-pin cable. Omni Yes, Omni is working fine as is. Here is where I'm going with this... I want a way to control the polarization voltage even when I have the rear diaphragm out of circuit via the relay. I have read thatt he voltage on the rear affects the sound of the capsule even if the rear is not contributing any audio. By this logic, a U48 and U47 would sound different in cardioid (assuming all other aspects are the same) because a U47 has no charge on the rear, and a U48 gets about 105V on the rear whether it's in Cardioid or Bi-Directional Mode. So If I install a relay to control "true" cardioid, as well as a switch to shunt the polarization voltage, I can get "U47 cardioid," "U48 cardioid," "M49 cardioid," or really anything I want. I read that when you switch an M49 to cardioid-only internally, you're supposed to put the PSU on Omni so you get 0V on the rear, but that even this doesn't sound the exact same as the U47 cardioid due to "capacitive leak via buffer resistors to ground." R8 I have measured the voltage drop across this before actually. I'll have to check again, but I think it was dropping about 20V or possibly 30V. Then the 100k resistor after drops another 40V or 50V. What would the calculation be? R2 The U47 puts 1.1V on the cathode. I think it's the same idea here, except that we use this voltage divider tapped from the heater. I measure about 1.3V on the cathode in my mic with R2 as 100ohm. I figured rather than swapping R2 to 110ohm instead of 100ohm, I could just install a pot and tweak it however I like and see if I hear a different at 1.1V. I'm not sure you should load any capsule with that much more resistance. 1G to 150Meg is about 4.5 octaves of a difference, to speak in musical terms. Since we can see B+ as being ground to AC signals, it would follow that the bias resistor + capsule form a hi pass filter, so moving your -3dB point will move up quite a bit. If I'm right, and the U47's internal amplifier was set to -3dB at 21Hz, then your new cut off would be about 400 Hz. So if R1 and R3 are set much higher than an original U47, then it must be to compensate for capsule differences, bias differences or amplifier differences. The easiest way is to check, of course. You'll know it when you hear it! I have been told that 1G is used because it gives you more low end and it's a more modern choice. The original U47 used 60M or 100M, as they didn't have access to 1G back in the day. I don't think the 1G is used to compensate for something else, but the thought did cross my mind. There's really no way to no unless Oliver (the designer of the circuit I link to in my post) specifically told us, and so far, all he's told me is 1G provides more bass, and that even in a real U4, increasing that resistor to 500M or 1G is almost always a sound improvement. Regardless, I don't think we're talking about 21Hz versus 400Hz. More like shifting the -3dB point to 40Hz. It can't possibly be to compensate for the capsule. No specific brand of capsule is specified for the build, and Neumann still makes the K47 anyway. Perhaps you could compensate for a thin-sounding tube, but I'm not sure if the EF800 or EF802 or EF14 sounds thin. In fact, even among those that say the EF14 is not a good VF14 replacement, they still say the low-end is correct; it's more the mids and highs that sound different. |
|
|
|
Post by jazznoise on Feb 16, 2014 15:08:16 GMT -6
Neumann specified the K47 preamp to be used with the K47 capsule, so the capacitance of the capsule would have been fixed. But I see in the schematic I was referring I saw, in error, 1 G's when in fact the values you're describing. The guy on R/E/P says these values give you a nominally flat to 21Hz - so I'm guessing the "modern" values used extend into infra-sonics. Have a read of it here yourself; repforums.prosoundweb.com/index.php?topic=35368.0Regardless: since it's a passive filter, a load resistor shunting to ground (both B+ and ground are ground to an AC signal)with series capacitance can only be taken to mean a standard passive HPF with a 6dB/Octave slope. Q aside, the cutoff is a simple F=1/2piRC equation. Regardless of what C is, halving R doubles F. Inversely, a capacitor shunting to ground with a series resistive load forms a LPF with the same function as before. The more you know... |
|
|
|
Post by svart on Feb 16, 2014 17:24:26 GMT -6
PSU pot: You can do this, but remember that any resistance you add will become a heater, so use a much higher wattage rating pot for this because it'll get hot. 1G resistors: All resistors are dropping resistors, but it depends on the load. Since the 1G resistors are supplying essentially unloaded current to the capsule, you get a voltage but very little current draw. What R10 and R5 do is create an electrostatic charge between the film and the backplate. As the film vibrates, the charge changes and therefore the voltage changes. Depending on the capacitance of the capsule used and the input bias current of the tube/FET used, the value of the bias resistor needs to be changed, thus the difference in the resistors in the different models of mic. With values in the Megohms and Gigohms, I don't think you'll hurt anything, but your audio might change for better or worse. Usually with lower values in these places you'll lose low end. Relay: Not sure why you'd want to do this when the variable voltages for tuning pickup pattern work fine.. AND I'd worry about making/breaking high voltage on the capsule skin through the relay due to voltage transients, which might cause issues like arcing or the skin sticking to the backplate.. But anyway, sure, you can use a relay on DC heaters. Just make sure your relay doesn't use so much current that the heater rail starts to sag. You'll also want to use an anti-kickback diode so that you don't get voltage transients on the heater rails when the relay coil collapses. There are a couple different latching type relays. The most prevalent one needs a voltage on two pins, one to set one way and the other to make it go back. I'd just use a regular relay for the ease of use. C4: Should be after the relay, as in always connected between the skins. Relay return: Don't bother. Hook both grounds together in the mic and use the thickest wire as ground return. Onmi: You are dealing with CHARGE.. so as long as it sounds like it's working, it's probably fine as-is. R8: It's a dropping resistor. You need to know how much current is being used by the circuit after R8 to calculate how much it drops. Either you can put a DMM in series with it and measure current, or you can measure the voltage drop across it and calculate it that way. Or you can just put a 50K pot in rheostat hookup and use that. R2: Yep, you can do it. I don't know how much current that tube needs on the cathode, but I don't think it's much. The schematic shows 1/2W resistors, so use a 1W pot if replacing both, or 1/2w pot if only replacing R2 with a rheostat.. Transient: Unknown. You could try tracking it down by removing certain sections of the circuit and powering up/down. Thanks a lot for your post! PSU Pot Thanks for the tip. Do you know where I would insert the potentiometer? Do you know what the phillips head white plastic circle is right next to the heater source on the PCB? Is that an adjustment spot? 1G Resistors Can you explain how R5 and R10 work exactly? I know we are polarizing the backplate with 60V, and varying the voltage on the rear diaphragm to create whatever polar pattern we want. And I know the difference/changes in voltage is caused by sound pressure level. But I don't understand how R5 and R10 work. In other words, I understand the idea of how it works in general, but not how the electronics/passive components come into play really. Relay Can you explain more about the relay causing voltage transients on the capsule? You're right; the current remote polar pattern switching works fine. The reason I want to mod the mic to give myself the option of "true" cardioid is for the sake of U47 authenticity. U47 has hard-switching between cardioid and omni, which cut off the rear diaphragm from the circuit in cardioid mode and actually increased the mic output by 5dB. Many people claim true cardioid sounds different in general. I'm not sure if I believe it. However, even Neumann says true cardioid has a better S/N ratio than remote pattern cardioid, and they installed an internal S2 switch in the M49 to allow you to put the mic in cardioid-only mode. C4 This cap will always be connected between the skins. The question is if you use the relay to disconnect the rear diaphragm from C4 (and thus the front), or if you disconnect C4 from the front of the capsule. Thick Blue Wire So you suggest making this Ground? I read it was good to use the thick red for the tube heater, as it draws the most amps of any part of the circuit, but I never knew where to use the thick blue. Does it really matter if I use the thick blue for ground or not? It might be too cumbersome to rewire the ground connection of the 7-pin cable. Omni Yes, Omni is working fine as is. Here is where I'm going with this... I want a way to control the polarization voltage even when I have the rear diaphragm out of circuit via the relay. I have read thatt he voltage on the rear affects the sound of the capsule even if the rear is not contributing any audio. By this logic, a U48 and U47 would sound different in cardioid (assuming all other aspects are the same) because a U47 has no charge on the rear, and a U48 gets about 105V on the rear whether it's in Cardioid or Bi-Directional Mode. So If I install a relay to control "true" cardioid, as well as a switch to shunt the polarization voltage, I can get "U47 cardioid," "U48 cardioid," "M49 cardioid," or really anything I want. I read that when you switch an M49 to cardioid-only internally, you're supposed to put the PSU on Omni so you get 0V on the rear, but that even this doesn't sound the exact same as the U47 cardioid due to "capacitive leak via buffer resistors to ground." R8 I have measured the voltage drop across this before actually. I'll have to check again, but I think it was dropping about 20V or possibly 30V. Then the 100k resistor after drops another 40V or 50V. What would the calculation be? R2 The U47 puts 1.1V on the cathode. I think it's the same idea here, except that we use this voltage divider tapped from the heater. I measure about 1.3V on the cathode in my mic with R2 as 100ohm. I figured rather than swapping R2 to 110ohm instead of 100ohm, I could just install a pot and tweak it however I like and see if I hear a different at 1.1V. I'm not sure you should load any capsule with that much more resistance. 1G to 150Meg is about 4.5 octaves of a difference, to speak in musical terms. Since we can see B+ as being ground to AC signals, it would follow that the bias resistor + capsule form a hi pass filter, so moving your -3dB point will move up quite a bit. If I'm right, and the U47's internal amplifier was set to -3dB at 21Hz, then your new cut off would be about 400 Hz. So if R1 and R3 are set much higher than an original U47, then it must be to compensate for capsule differences, bias differences or amplifier differences. The easiest way is to check, of course. You'll know it when you hear it! I have been told that 1G is used because it gives you more low end and it's a more modern choice. The original U47 used 60M or 100M, as they didn't have access to 1G back in the day. I don't think the 1G is used to compensate for something else, but the thought did cross my mind. There's really no way to no unless Oliver (the designer of the circuit I link to in my post) specifically told us, and so far, all he's told me is 1G provides more bass, and that even in a real U4, increasing that resistor to 500M or 1G is almost always a sound improvement. Regardless, I don't think we're talking about 21Hz versus 400Hz. More like shifting the -3dB point to 40Hz. It can't possibly be to compensate for the capsule. No specific brand of capsule is specified for the build, and Neumann still makes the K47 anyway. Perhaps you could compensate for a thin-sounding tube, but I'm not sure if the EF800 or EF802 or EF14 sounds thin. In fact, even among those that say the EF14 is not a good VF14 replacement, they still say the low-end is correct; it's more the mids and highs that sound different. PSU pot: It looks like it already has one, if the schematics I found are correct. Just adjust it for what you want. R5/R10 They restrict extreme amounts of current. Since the capsule needs to be biased, but a lot of current would be noisy, you need very large value resistors to drop this current to extremely small amounts. R5 biases the backplates up to whatever voltage the divider R6/R7 makes (63V for 105V input). R10 biases either to 0V for omni, to 63V or 2x 63V for figure 8, Relay transient: If you power and unpower an inductor of any sort, the energy in the field will build or collapse and will attempt to keep it's current steady to the load so V=IR means that the voltage will change because current and resistance stay the same. Look up "flyback diode" or "inductive kickback" and you'll see what I'm talking about for the circulation diode for the relay. For the capsule transient, the same thing applies, but with the capsule capacitance instead. As the relay contacts open or close, there will be a small arc as the charge from the capsule is made or broken. C4 Leave it connected and only relay the lead to the head amplifier. C4 will bleed charge off of one capsule skin to the other. Blue wire: You always want the thickest wire to be your ground. Regardless of what kind of current is on your red wire(or whatever wire), the ground has to return ALL currents from all wires back to the source. Yes, ground has currents on it! Ground is just a name, it is a current carrying conductor just like all others. Ground should always have the least impedance of all conductors. Personally, I would hook both my shield and my thickest wire together as my ground. Omni: Omni should be 0V front, 63V backplate, 0V backskin. Cardioid should be 0V front, 63V backplate, 63V backskin and figure 8 should be 0V Front, 63V backplate, 126V backskin, according to your schematic anyway. That would make perfect omni, perfect cardioid and perfect figure 8.. If you are referring to the alternate funkenwerk schematic with just the switch in place of C4, you'll get CHARGE bleeding from the backplate to the back skin since it's not connected to a load. It'll essentially charge up to the 63V that the backplate is biased to. Then it will modulate with the amount of charge. Personally I'd prefer just using the polarity switch on the PSU instead. R8: The formula is (Vin-Vout)/R=I. So (105-25V)/130K=.0061A R2: The values shown in the schematic for R2 and R3 should put 1.13V on the cathode. If you show higher, then your heater voltage is higher than 5.05V. Adjust that downwards like you had planned and you should be OK. |
|
|
|
Post by watchtower on Feb 17, 2014 21:07:47 GMT -6
Neumann specified the K47 preamp to be used with the K47 capsule, so the capacitance of the capsule would have been fixed. But I see in the schematic I was referring I saw, in error, 1 G's when in fact the values you're describing. The guy on R/E/P says these values give you a nominally flat to 21Hz - so I'm guessing the "modern" values used extend into infra-sonics. Have a read of it here yourself; repforums.prosoundweb.com/index.php?topic=35368.0Regardless: since it's a passive filter, a load resistor shunting to ground (both B+ and ground are ground to an AC signal)with series capacitance can only be taken to mean a standard passive HPF with a 6dB/Octave slope. Q aside, the cutoff is a simple F=1/2piRC equation. Regardless of what C is, halving R doubles F. Inversely, a capacitor shunting to ground with a series resistive load forms a LPF with the same function as before. The more you know... Ah yes, I've referenced that thread many times. I've somehow been researching the U47 for over a year at this point haha. There are folks on both sides of the argument as to whether 1G is better or not in a U47. PSU pot: It looks like it already has one, if the schematics I found are correct. Just adjust it for what you want. R5/R10 They restrict extreme amounts of current. Since the capsule needs to be biased, but a lot of current would be noisy, you need very large value resistors to drop this current to extremely small amounts. R5 biases the backplates up to whatever voltage the divider R6/R7 makes (63V for 105V input). R10 biases either to 0V for omni, to 63V or 2x 63V for figure 8, Relay transient: If you power and unpower an inductor of any sort, the energy in the field will build or collapse and will attempt to keep it's current steady to the load so V=IR means that the voltage will change because current and resistance stay the same. Look up "flyback diode" or "inductive kickback" and you'll see what I'm talking about for the circulation diode for the relay. For the capsule transient, the same thing applies, but with the capsule capacitance instead. As the relay contacts open or close, there will be a small arc as the charge from the capsule is made or broken. C4 Leave it connected and only relay the lead to the head amplifier. C4 will bleed charge off of one capsule skin to the other. Blue wire: You always want the thickest wire to be your ground. Regardless of what kind of current is on your red wire(or whatever wire), the ground has to return ALL currents from all wires back to the source. Yes, ground has currents on it! Ground is just a name, it is a current carrying conductor just like all others. Ground should always have the least impedance of all conductors. Personally, I would hook both my shield and my thickest wire together as my ground. Omni: Omni should be 0V front, 63V backplate, 0V backskin. Cardioid should be 0V front, 63V backplate, 63V backskin and figure 8 should be 0V Front, 63V backplate, 126V backskin, according to your schematic anyway. That would make perfect omni, perfect cardioid and perfect figure 8.. If you are referring to the alternate funkenwerk schematic with just the switch in place of C4, you'll get CHARGE bleeding from the backplate to the back skin since it's not connected to a load. It'll essentially charge up to the 63V that the backplate is biased to. Then it will modulate with the amount of charge. Personally I'd prefer just using the polarity switch on the PSU instead. R8: The formula is (Vin-Vout)/R=I. So (105-25V)/130K=.0061A R2: The values shown in the schematic for R2 and R3 should put 1.13V on the cathode. If you show higher, then your heater voltage is higher than 5.05V. Adjust that downwards like you had planned and you should be OK. PSU Did you find a schematic for the AMI PSU? R5/R10 Ok, makes more sense now. We are dropping as muchc urrent as possible, but still passing through a voltage. Btw, the schematic I link to is not 100% correct. The PSU is 120V, not 105V, so the R6/R7 divider is actually two 1Meg resistors. Relay transient Do you think this is a real danger? Wouldn't C4 block the DC, at least to the front skin and tube? In the schematics I've seen with a relay, they did not involve a kickback inductor. Maybe I'll PM one of them to you. C4 This will definitely still be there even if I use a relay. I have to block DC from the rear from getting to the front. I still don't understand where you're saying to put the relay. I can put it between C4 and the Rear, or between C4 and the Front. Blue Wire I have the shield of the cable tied to the connector housing of the cable, which is ultimately attached to ground at some point in the PSU. Actually, I've always wondered why when I screw the Binder connector into the mic body, I get small signal level in my preamp. Omni Yes, all polar patterns are working correctly. The idea is to add an additional option of "true cardioid," which will at the very least increase output level and decrease noise. The relay/switch would not replace C4, as I still have to block DC between the back and front so I can get all patterns. 60V at the front would probably kill the tube. R2 Yes, for a heater of 5V you would get 1.1V on the cathode. However, my PSU came stock at 5.67V heater, not 5.05V. So I'm getting 1.3V cathode. I do not know if the PSU has a built in potentiometer for adjusting the heater, or if Oliver ultimately decided that 5.6V was the optimum point for the tube to heat at. Thanks! |
|
|
|
Post by jazznoise on Feb 18, 2014 6:16:24 GMT -6
If the original value gives you your roll of at 21Hz, then the only differences I'd worry about are the AC impedance seen by the capsule and the capsule voltage drop it creates. These mics don't respond much down at 21Hz, so I can't see moving the roll off into the infrasonic territory helping. In fact, in the case of plosives or very fast transints, the spike in IMD might create nasty artifacts. You mightn't hear 10hz, but you don't want to hear Signal +10Hz and Signal-10hz..unless you're building a ring modulator.
|
|
|
|
Post by watchtower on Feb 18, 2014 8:15:52 GMT -6
I agree about the infrasonics. That's why to a degree I'm not sure why they really increased things up to 1G, and why I'm planning to go back down to the vintage values. Unless for some reason the 1G was more like a low shelf instead of just a lower cut-off haha - but I don't think it works like that  Do you have more info about the AC impedance the capsule sees, or that general idea you were mentioning? |
|
|
|
Post by svart on Feb 18, 2014 8:53:52 GMT -6
PSU: I think this is it, but not entirely sure. img.photobucket.com/albums/v283/jazz347/47-inspired%20mic%20build/PSU_Schematic.pngR5/R10: This is not a regulated supply, so the voltage will fluctuate. Don't get too hung up on it being perfect. Besides, most cheap DMM meters have 10-20% tolerance, so reality is that you could be way off in measurement and the mic will still sound OK. Ohm's Law V=IR, means that current, voltage and resistance(load) are all coupled mathematically. If one changes, then the others change. There is no such thing as "passing voltage", heck, there is no such thing as "voltage" either, it's just a number derived from current and resistance to make something quantifiable and easy to understand. What happens is that the resistors "resist" current and due to the load of the capsule, a "voltage" is mathematically created. Actually, since the capsule is a big capacitor, the best measurement would be "Coulombs" to measure the charge on the capsule. However, nobody has a charge counter in their drawer, a voltage can simply be measured by a DMM and get reasonable approximation of charge. C4: Transients look like AC and would pass right through a capacitor. Fast rise and fall times. Look up relay transients, or contact arcing for lengthy papers on the subject. I honestly don't have any idea how bad they'd be. Blue wire: Shields aren't meant to carry current, they are meant to be low impedance pathways for RF and other EMI to follow to earth ground so you don't get radio stations and such on your audio. If the sheilds have heavy return currents from heaters and such on them, then they aren't considered low impedance and won't work as well. Omni: I think you misunderstood what I was saying.. Cardioid NEEDS the back skin to be biased the same as the backplate according to your schematic. Unhooking it via a relay would give you something very strange I think. Besides, Looking at the schematic you posted again, I have no idea how you have multiple patterns in that mic, unless you turned everything around and used the backplate as the grid control and the front skin grounded and back skin variable voltage. R2: Ok. just adjust the value of the resistor using a voltage divider formula to get what you want. Vout=Vin*(R3/R3+R2) So 5.67*(29/100+29)=1.27 Since this is really just a ratio, adjust R2 up a little bit. 5.67*(29/120+29)=1.10V Make your R2 around 120ohms for 1.1V. |
|
|
|
Post by jazznoise on Feb 18, 2014 12:32:29 GMT -6
I agree about the infrasonics. That's why to a degree I'm not sure why they really increased things up to 1G, and why I'm planning to go back down to the vintage values. Unless for some reason the 1G was more like a low shelf instead of just a lower cut-off haha - but I don't think it works like that Do you have more info about the AC impedance the capsule sees, or that general idea you were mentioning? If you agree then you might experiment with moving the value up. I've never needed more 30Hz on a vocal. If you brought it up to 50/60Hz (80M for 40Hz, so about 60M for 60Hz, give or take) you'd be just rolling off the fundamental of your kick drums but your plosives would be a lot less aggressive. IMD should drop too, if the mic clips, and your headroom should somewhat increase. Douglas Self takes briefly about it in his Small Signal Audio Design Book. He's a low noise, low distortion type of guy so it's all about good low output impedance with high but quiet input Z. An interesting but useless fact is that because the capacitor microphone is almost purely reactive (No direct resistance as current can't flow across it), the Johnson noise is minimal. In his example he uses a 5pF capsule with a 3.2G Ohm resistor to get a cut off of 10Hz but with the op amp having an input Z of about 9G Ohms. Take this to mean that the input impedance of your valve will load down your input, so using very high value resistors on low Z inputs doesn't really give you much advantage anyway. Not quiet sure what the implications are - I'll dwell on that zen vodoo for a while. Johnson noise should drop slightly from lowering the resistance too, but I don't think it'll break the bank. The big deal is maintaining the same polarity voltage - higher V's increase capsule sensitivity, if I remember right. There's negatives to that, obviously. I recommend googling it! |
|
|
|
Post by Martin John Butler on Feb 18, 2014 12:37:42 GMT -6
This is all way over my head, but I'm curious if the capsule for your Oliver Archut design is the Tab_Funkenwerk 47, or the Thiersch, you can get either one from Oliver if you use the Blackspade U17 or the U17R .
|
|
|
|
Post by watchtower on Feb 19, 2014 9:36:16 GMT -6
If you agree then you might experiment with moving the value up. I've never needed more 30Hz on a vocal. If you brought it up to 50/60Hz (80M for 40Hz, so about 60M for 60Hz, give or take) you'd be just rolling off the fundamental of your kick drums but your plosives would be a lot less aggressive. IMD should drop too, if the mic clips, and your headroom should somewhat increase. Douglas Self takes briefly about it in his Small Signal Audio Design Book. He's a low noise, low distortion type of guy so it's all about good low output impedance with high but quiet input Z. An interesting but useless fact is that because the capacitor microphone is almost purely reactive (No direct resistance as current can't flow across it), the Johnson noise is minimal. In his example he uses a 5pF capsule with a 3.2G Ohm resistor to get a cut off of 10Hz but with the op amp having an input Z of about 9G Ohms. Take this to mean that the input impedance of your valve will load down your input, so using very high value resistors on low Z inputs doesn't really give you much advantage anyway. Not quiet sure what the implications are - I'll dwell on that zen vodoo for a while. Johnson noise should drop slightly from lowering the resistance too, but I don't think it'll break the bank. The big deal is maintaining the same polarity voltage - higher V's increase capsule sensitivity, if I remember right. There's negatives to that, obviously. I recommend googling it! Thanks for the info. I'm definitely planning on putting in 100M, 150M, and 60M. Probably two or three carbon comps in series because you can't find values that high. That Johnson noise stuff you're talking about is a little over my head, I think haha. In terms of the grid bias resistor, ultimately I think it's the tube input impedance that affects the sound from the capsule in this case, not the 1G. The 1G is tied to ground, so as far as I understand, it's not adding resistance, but just raising the audio signal from ground. PSU: I think this is it, but not entirely sure. R5/R10: This is not a regulated supply, so the voltage will fluctuate. Don't get too hung up on it being perfect. Besides, most cheap DMM meters have 10-20% tolerance, so reality is that you could be way off in measurement and the mic will still sound OK. Ohm's Law V=IR, means that current, voltage and resistance(load) are all coupled mathematically. If one changes, then the others change. There is no such thing as "passing voltage", heck, there is no such thing as "voltage" either, it's just a number derived from current and resistance to make something quantifiable and easy to understand. What happens is that the resistors "resist" current and due to the load of the capsule, a "voltage" is mathematically created. Actually, since the capsule is a big capacitor, the best measurement would be "Coulombs" to measure the charge on the capsule. However, nobody has a charge counter in their drawer, a voltage can simply be measured by a DMM and get reasonable approximation of charge. C4: Transients look like AC and would pass right through a capacitor. Fast rise and fall times. Look up relay transients, or contact arcing for lengthy papers on the subject. I honestly don't have any idea how bad they'd be. Blue wire: Shields aren't meant to carry current, they are meant to be low impedance pathways for RF and other EMI to follow to earth ground so you don't get radio stations and such on your audio. If the sheilds have heavy return currents from heaters and such on them, then they aren't considered low impedance and won't work as well. Omni: I think you misunderstood what I was saying.. Cardioid NEEDS the back skin to be biased the same as the backplate according to your schematic. Unhooking it via a relay would give you something very strange I think. Besides, Looking at the schematic you posted again, I have no idea how you have multiple patterns in that mic, unless you turned everything around and used the backplate as the grid control and the front skin grounded and back skin variable voltage. R2: Ok. just adjust the value of the resistor using a voltage divider formula to get what you want. Vout=Vin*(R3/R3+R2) So 5.67*(29/100+29)=1.27 Since this is really just a ratio, adjust R2 up a little bit. 5.67*(29/120+29)=1.10V Make your R2 around 120ohms for 1.1V. PSU Actually, if memory serves, that is a user named soapfoot's schematic for his own PSU for his U47 build. I'm not even sure if it's the design he settled on. Regardless, unless simply by chance, it's not the same as my AMI-built PSU. R5/R10 Good point about DMM tolerance. I was told my PSU is NG/non-regulated, but at the same time, I've read Oliver state that his Lucas PSUs "monitor the filament for voltage swing, normal it regulates a +/- 10% voltage swing of the mains with no problem, it is more the real hard cases when the sensor kicks in!" I don't know enough about PSU design to know if that's still possibly consistent with a non-regulated supply. Blue Wire So if I have the heater return on the blue wire, the cable shielding wont work as well? Will it work better if I don't tie the cable shielding to the blue wire until the signal has reached the PSU? C4/Omni I think the whole point of C4 is that is allows AC to pass from the rear to the front, but not DC, otherwise the polar patterns wouldn't work. Here's how the polar patterns work. Pin 7 provides anywhere from 0V to 120V depending on your selection. This polarizes the back skin of the capsule in reference to the backplate (60V all the time). The signal from that skin is passed on to the front of the capsule through C4 and is fed to the tube. If you completely disconnect the rear skin from the circuit, you get a true cardioid mic. The rear can't feed any audio, so you don't have omni or bi-directional. Only the front skin provides the sound, so you get cardioid. R2 Yes 120ohm would be about right for a 5.67V heater. But I'm hoping to use potentiometers to more easily experiment with how much of an effect the cathode voltage has on sound. I know almost nothing about potentiometers, so what specs would I need to look for if I replaced R2 with a pot? A 1W pot with a range of 50ohm to 200ohm? As always, sorry for my noobieness, and thanks! Here's two links to schematics of mics that are using relays to have a "true cardioid" mode lucasmicrophone.lefora.com/attach/view/ma/b754e8d8ff40998244e500dac9c26afe9cfc5b73.jpglucasmicrophone.lefora.com/attach/view/ma/e071bc3a026d682c1badeae847c662466fe3c885.jpgThis is all way over my head, but I'm curious if the capsule for your Oliver Archut design is the Tab_Funkenwerk 47, or the Thiersch, you can get either one from Oliver if you use the Blackspade U17 or the U17R . I don't have either capsule. My U47 has a BeesNeez K7 |
|
|
|
Post by Martin John Butler on Feb 19, 2014 10:15:00 GMT -6
ahh.. good luck, I'm sure it's really good.
It's interesting that Oliver releases those schematics, I guess he sells kits for DIY. I'm glad I ordered the Thiersch capsule for my U17, since Oliver uses that in his Lucas mics, I figure it can't be too shabby. My mic's over at Mic Rehab, Shannon's doing the install for me. You guys are all in pretty deep with these projects, but I sure enjoy hearing about them.
|
|
|
|
Post by watchtower on Feb 19, 2014 10:39:33 GMT -6
Yep, the mic sounds fantastic. I am just interested in making my own slight modifications to it to provide extra versatility, while also bringing it closer to "vintage." I will then be eager to compare to the Slate VMS, as I will almost surely purchase that when it's out, even though I love building my own vintage clones.
And yes, Oliver provides kits for U47, C12, and M49 currently. None of them are exact replicas. His U47 Alternate Tube Schematic I linked to in my first post is not correct for the kit he sends, which was incredibly confusing for me as a customer. The other two schematics I linked to in my last post are not readily plastered online, but if you know where to look you can find them (not that you have to now that I posted them haha).
|
|
|
|
Post by watchtower on Feb 26, 2014 12:57:16 GMT -6
Any more comments on my previous post? I haven't made any mods yet
|
|
|
|
Post by svart on Feb 26, 2014 13:35:29 GMT -6
Only that I still don't get how the biasing on the back skin on the schematics works, unless it's solely done for figure 8, which can't work and isn't in the original U47, so it wouldn't be authentic unless you also unhooked the biasing for the backskin with another relay as well as rewiring the capsule. Also: C4 passes electrostatic CHARGE through, and biases the back skin even without DC bias on the backskin. AC signals also pass through this cap. You are still thinking in terms of AC/DC but charge is still the right term for capacitors. So when connected, the back skin will DIScharge the built up charge through the cap giving you "0" volts DC and ultimately an onmi pattern. Without it, the back diaphragm (as with any capacitor) will draw charge across the gap from the backplates and "charge up" to backplate voltage, giving you a roughly cardioid pattern. You see, the way it's hooked up in the schematic could never possibly give you a true figure 8 due to the back diaphragm feeding through C4 to the tube in phase with the front, no matter what bias it's at, and never being truly out of phase with the front. you'd have to feed the grid of the tube with the backplate itself instead to get figure 8, like this: harmoniccycle.com/hc/images/sound/C12/smapex460-1.gif |
|
|
|
Post by watchtower on Feb 26, 2014 14:27:14 GMT -6
|
|
|
|
Post by svart on Feb 26, 2014 14:35:11 GMT -6
Probably getting pseudo figure 8 due to the time delay caused by the distance between the front and back diaphragm through C4. In practical use it's probably close enough, but I suppose it's still not true figure 8 as the time delay causes a phase discrepancy. Who knows. I guess if you like the way it sounds, that's all that matters.
|
|
|
|
Post by watchtower on Feb 26, 2014 14:40:24 GMT -6
|
|
|
|
Post by jazznoise on Feb 26, 2014 16:14:17 GMT -6
Is that cap possibly there to compensate for the group-delay caused by the air gap? It is a very small cap!
|
|
|
|
Post by watchtower on Feb 28, 2014 9:35:20 GMT -6
What air gap, jazznoise?
|
|
|
|
Post by svart on Feb 28, 2014 10:29:56 GMT -6
The one between the capsule diaphragms and the backplate. The capsule is a capacitor, with the dielectric being the air gap.
Honestly, I think it's just to pass AC and charge between the diaphrams. I still think that the way it's designed, that you'd get a sort of figure 8/hypercardioid pattern, with one side a little better at picking up than the other, due to the prop-delay through the cap.
|
|
|
|
Post by jazznoise on Feb 28, 2014 10:35:35 GMT -6
The front and rear plate of the capsule should have a gap. Probably not air, now that I think, but it's still there.
Or have I gone full retard?
|
|
