Continuously variable polar pattern for the G7

[globule_655]

Hi folks

I just finished building my very first G7 in a T.Bone SCT800 body with Tim Campbell's capsule, EF86 NOS tube (tried Telefunken, Mullard and Tesla) and Cinemag 2480 output trafo
It really sounds amazing ! Thumbs up for Jakob's design !

Here's my question though : I was wondering if I could put a variable resistor in place of the 2x100k resistors in the PSU so that I could have a continuously variable polar pattern from omni to figure of eight..?
If I do that does this variable resistor have to be around 200k or the value only depends on how sensitive I want the change in polar pattern to be ?

Another question that's buggin' me : unloaded my psu delivers 150V DC on the B+ but this falls down to approx 125V when loaded. I read in many threads that it happens to many people who built the PSU according to Jakob's schematic. What I'm asking here is : is it really a problem as the sound of the mic is absolutely fine (it outperformed by far my TLM49) ! Should I try to rise the B+ a bit by changing the 15:220V transformer to a higher ratio ? Where does this problem come from ?

My last question is about internal wiring : I used the original PCB of the SCT800 and modified it for my build, cutting traces and used wires to make the connexions I couldnt do otherwise. I stripped an old AES-EBU cable that was laying around for this. I just wondered if my mic could sound better if I used another method for those connexions (I just dont know how this mic SHOULD sound...)

Thanks a lot for your help

Cheers !!

 

[dmp]

Here's my question though : I was wondering if I could put a variable resistor in place of the 2x100k resistors in the PSU so that I could have a continuously variable polar pattern from omni to figure of eight..?
If I do that does this variable resistor have to be around 200k or the value only depends on how sensitive I want the change in polar pattern to be ?

You can use a pot for a continuously variable pattern. Many mics do this. The 2x100k resistors serve as a voltage divider and the change in polarity is due to the voltage being B+, B+/2, or 0v. If you use a pot other than 200k you will load the B+ more or less.

Another question that's buggin' me : unloaded my psu delivers 150V DC on the B+ but this falls down to approx 125V when loaded. I read in many threads that it happens to many people who built the PSU according to Jakob's schematic. What I'm asking here is : is it really a problem as the sound of the mic is absolutely fine (it outperformed by far my TLM49) ! Should I try to rise the B+ a bit by changing the 15:220V transformer to a higher ratio ? Where does this problem come from ?

Your lower voltage is not a "problem". The current drawn by the mic causes a voltage drop across the 10K resistor in the PSU (ohm's law), which causes the voltage drop. The B+ is not regulated so it varies with load. Additionally, the B+ will vary based on the actual wall voltage you are using. You can adjust the transformer ratio to increase the voltage if you want it to, but I think the sound will be the same. Too high a voltage will be more of a problem as it can cause the capsule to not work. If you sit down with the schematic and and Ohm's Law (V=IR) the circuit will start to make a lot of sense.

My last question is about internal wiring : I used the original PCB of the SCT800 and modified it for my build, cutting traces and used wires to make the connexions I couldnt do otherwise. I stripped an old AES-EBU cable that was laying around for this. I just wondered if my mic could sound better if I used another method for those connexions (I just dont know how this mic SHOULD sound...)

You can always try it and see. But probably not.  Problems with wiring generally cause noise / hum / oscillations.

 

[gyraf]

About the voltage - a G7 with Tim Campbell's capsule probably sounds best around 120-130V..

And if it sounds right, it is right

Jakob E.

 

[globule_655]

Okay ! Thanks a lot ! I know this may seem like stupid question but I'm learning and I'd rather ask than doing something wrong in case I'm missing something
I figured there would be a voltage drop but I didnt think it would be 25-30V.

About the pot is it better to place it before or after the 470uF capacitor ? I'd say after but the fact that the pickup for cardioid pattern with the 100k resistors  is made before the cap makes me wondering...

 

[SSLtech]

A few things...

The Gyraf tube mic design was derived from the simple U47 tube design. (although, with a different tube, different transformer, different capsule etc, only the general simple operating principle remains in common).

A few things to bear in mind:

The power supply will ALWAYS rise off load. -It cannot be any other way, since it's a simple resistive feed. -This happens with U47's also.

HOWEVER.

Try raising the on-load voltage with a U47, and watch the rare, precious and expensive tube die. VERY quickly.

The design is set to operate at its WORKING voltage. You may confirm correct operation by measuring a higher voltage off-load, but try running the mic at that higher voltage, and the next time you use the mic might also be the last time you use the mic.

As for variable pattern... certainly you can add a potentiometer, but beware power dissipation, particularly near either end of the wiper... and you know the loud  "BANG!" that you get when you switch patterns at the moment? -Well you'll also get a very similar noise, though 'slewed' when you move the pot... -Pot signal noise is caused by small amounts of DC across the pot. -In this case, we've got HUGE amounts of DC across the pot (about 160VDC) and the output couples to the capsule, after a small amount of filtering. -Also, any dirt or contact variation will induce massive noise.

Don't get me wrong, it IS doable, with some modifications... but I wonder if you appreciate that you'll NEVER be able to adjust it wile listening to the microphone... unless you don't mind replacing monitor speakers (and possibly eardrums) regularly? -If you want to do that, you'd need to build a much more complicated microphone.

Forget the type of wire. -Just don't get me started. -It works. Be happy.

Keef

 

[globule_655]

Okay thanks ! Lots of valuable infos here

Forget the type of wire. -Just don't get me started. -It works. Be happy.

Actually I'd like to know what you have to say about that. As I said I'm learning and eager to know more about things. I already now about the effect of capacitance, resistivity and inductance (not sure about the words in english) in cables. That's why I made my connexions out of AES-EBU wire. First, because that's what I had at hand and second because of its low capacitance. But knowing that this capacitance I'm talking about occurs between conductors and conductors to shield then what about single wires ?? Does only conductivity and inductance count in that case ? If yes then it's no problem what type of wire may use because they only carry DC voltage and inductance in twisted copper wires is almost negligible (with regards to audio frequencies I mean) right ?

 

[SSLtech]

The length of wire in this instance means that it's like wondering if a single drop of milk will make a large lake cloudy...

It's basically FAR too tiny to be significant in this case... though that doesn't stop manufacturers like MXL advertising that their mics are 'wired with Mogami'.

A whole INCH of it.

 

[dmp]

As for variable pattern... certainly you can add a potentiometer, but beware power dissipation, particularly near either end of the wiper... and you know the loud  "BANG!" that you get when you switch patterns at the moment? -Well you'll also get a very similar noise, though 'slewed' when you move the pot... -Pot signal noise is caused by small amounts of DC across the pot. -In this case, we've got HUGE amounts of DC across the pot (about 160VDC) and the output couples to the capsule, after a small amount of filtering. -Also, any dirt or contact variation will induce massive noise.

If one assumes the capsule is drawing zero current (fair assumption with a 1Gohm pulldown resistor), then the current through the pot will be constant while turning and there won't be significant pot noise problems. And again ohm's law... 160VDC across 200k is 0.128 watts power dissipation. No problem. And there isn't worse power dissipation near the ends of the wiper.  Current is constant through a resistance and voltage is linear.

Don't get me wrong, it IS doable, with some modifications... but I wonder if you appreciate that you'll NEVER be able to adjust it wile listening to the microphone... unless you don't mind replacing monitor speakers (and possibly eardrums) regularly? -If you want to do that, you'd need to build a much more complicated microphone.

A stepped switch creates a MUCH louder

"BANG!"

than a pot. Did you change your argument midway through? I have tube power supplies of both types and the pot is much quieter during changing. The only argument for a stepped switch in my opinion is to set a fixed position - i.e. know you have figure 8 exactly and that the pot won't get bumped to change it and wreck your m-s recording. I think that is more important than infinite positions. For a recent pair of instrument mics I built I used 9 position switches to get inbetween patterns, but still have them fixed.

 

[PRR]

> what about single wires ??

A single wire in a totally empty universe: no capacitance.

Also no circuit, and no useful audio system.

A single wire inside a case with other wires and stuff has capacitance. The wire maker can not tell you how much.... it depends how big and close the other stuff is. (The cable-maker can tell how much capacitance to the rest of the cable, and this is often dominant.)

For typical audio gear, this capacitance changes VERY little with insulation or diameter, and is not at all about the type of metal (copper, iron, silver). It does change with spacing (good) and length (bad). That's why the only really capacitance-sensitive wire in a condenser mike, the signal from the capsule, SHOULD be short and kept away from other stuff.

Really, there's way too much puffery about wire. Even from folks who could know better. So much that it has driven "common hook-up wire" off the market. Nothing wrong with 98% copper in a shoe-lace.

Microphone voltage: all power sources sag. I have 124V in my house until I run the oven, then it drops to 118V. Back at the power dam the 5000V sags to 4999.8V. In this case, a large resistor is cheap filtering and also limits the smoke in a dead-short. Bigger tube-stuff often starts from 490V to arrive at 260V after filtering.

Blank slate. What does the tube need? Enough to pump the output voltage. You probably won't have over 1V signal out. Times the transformer ratio, say 8:1, so 8Vrms 12V peak at plate. A resistance-coupled tube "can" make peak signal 20% of supply voltage, so we need at least 12/20%= 60V supply. 120V is generous. I would not expect 140V to sound any different. Like: I have a 6-foot wide car. A 12 foot garage is ample. A 14 foot garage isn't much "better". (Unlike a garage, a microphone can't have two cars squeezed-in at once.)

BUT!! This voltage is also Capsule Bias. If you put enough voltage on a capsule, the diaphragm is sucked to the back-plate! There's some safety-factor, but on the other hand low bias voltage means low signal so it can't have the safety-margin of the Brooklyn Bridge. Do NOT wantonly increase mike voltage unless you KNOW what will kill (and not!-kill) the capsule. Don't even come too close: a surge or transient can throw a barely-stable diaphragm over the edge so it sticks to the backplate.