Surprised by the result

[daschnoz]

I literally had all of the parts needed to build another mic. BM800 body, transistor and assorted other parts, capsule... so why not. I took the C12 topology and replaced the tube with a 2N5457. This is a transistor based clone of the tube mic that I made a month or so ago. The capsule here is the same one that I used in that tube mic. Same NTE10/3 x-fmr. So really the only difference is the active component (and the surrounding resistor values for bias and proper operation).

The tube version and the FET version have the same frequency response. I was expecting that the FET circuit would have a bit more top end than the tube, so this was a bit of a surprise.

What I really didn't expect... the FET version seems to have a bit more pop in the upper-mids... it responds faster - the guitar was there more quickly than the tube version, but at no greater amplitude than the tube.

Is this what people are talking about when they say that a tube mic has a more smooth or more mellow sound?

 

[rock soderstrom]

I literally had all of the parts needed to build another mic. BM800 body, transistor and assorted other parts, capsule... so why not. I took the C12 topology and replaced the tube with a 2N5457. This is a transistor based clone of the tube mic that I made a month or so ago. The capsule here is the same one that I used in that tube mic. Same NTE10/3 x-fmr. So really the only difference is the active component (and the surrounding resistor values for bias and proper operation).

That is very interesting, I have made similar attempts to replace tubes with JFETs. On this page something similar is done in a guitar amp environment.

Can you draw a small schematic sketch with your resistor values and adjustments please? I think each FET has to be biased individually, but I'm still interested in your values. Have you also tried other JFETS such as J201?

Is this what people are talking about when they say that a tube mic has a more smooth or more mellow sound?

My experience is that JFETs sound more direct and yes, also faster. I like the harmonic overtones/distortions of tubes better than those of JFETs, I think these are also responsible for the “larger than life” sound of good tube microphones.

Of course, this is all highly subjective so YMMV...

 

[mrgrooves666]

I literally had all of the parts needed to build another mic. BM800 body, transistor and assorted other parts, capsule... so why not. I took the C12 topology and replaced the tube with a 2N5457. This is a transistor based clone of the tube mic that I made a month or so ago. The capsule here is the same one that I used in that tube mic. Same NTE10/3 x-fmr. So really the only difference is the active component (and the surrounding resistor values for bias and proper operation).

The tube version and the FET version have the same frequency response. I was expecting that the FET circuit would have a bit more top end than the tube, so this was a bit of a surprise.

What I really didn't expect... the FET version seems to have a bit more pop in the upper-mids... it responds faster - the guitar was there more quickly than the tube version, but at no greater amplitude than the tube.

Is this what people are talking about when they say that a tube mic has a more smooth or more mellow sound?

Would you be kind to share the guitar tracks?
But, YMMV indeed... also similar with my own personal observations after rolling many of the tube and fet and opa mic projects available here, and trying them out....and why I stopped diy'ing tube mics. Different? yes, nice? yes, although not necessarily better.... worth the extra hassle of PSU, cable, and higher cost?....mmm not really... or at least not for me ..

 

[daschnoz]

It was a sand box project, so I deleted it. I can re-record something though. And yes, I'll be happy to share the schematic. But right now, I'm off to the annual Halloween parade through town. 1 kid is playing guitar on a float, the other kid is in the high school marching band.

 

[kingkorg]

Have you done the test with exact same source material, recorded in the same exact way, same exact conditions? When i did this kind of test, under normal operating conditions with no clipping, linear operating region, i couldn't hear any difference and the tracks canceled in the null test. Even the pink noise track. I did this multiple times, with multiple circuits and capsules as i have modular circuit/capsule system.

Even humidity and temperature on day to day basis can make capsules sound slightly different. Which might sway the result. This could be just one of many reasons why you got the impression you describe.

Imho, tubes vs fets start sounding different when pushed. Gross oversimplification as there are tons of different topologies for each, so even this might not be the case.

 

[daschnoz]

As promised...
Schematics and sound clips. It's the basic acoustic guitar rhythm track that you might hear under any number of 80s hair metal songs, including "Sweet Child of Mine" (if you listen close enough, it's in there, and the mix would sound like it was missing something without it). Block chords, nothing fancy.

This is about as close as I can get to "same source" and not spend all day doing it. The tube mic had at least 30 minutes of warm-up time. The mics were set up one over the top of the other (tube mic is upside down, on top), head baskets less than 1/4" apart. I did my best to align the capsule position; distance from the source, as well as in rotation. The guitar was right about at the vertical center point of the 2 mics, 12th fret, about 2-feet away. Nothing done to the clips except normalizing volume. What you hear is what the mics picked up.

Mic A
View attachment A.wav

Mic B
View attachment B.wav

The text file reveals which mic is which. Enjoy

ETA: Interface - Tascam 4x4, gain at 12:00 for both mics. The FET mic is a tiny bit hotter, but only because I could see the levels. To hear it, the volume difference is meh.

 

[kingkorg]

As promised...
Schematics and sound clips. It's the basic acoustic guitar rhythm track that you might hear under any number of 80s hair metal songs, including "Sweet Child of Mine" (if you listen close enough, it's in there, and the mix would sound like it was missing something without it). Block chords, nothing fancy.

This is about as close as I can get to "same source" and not spend all day doing it. The tube mic had at least 30 minutes of warm-up time. The mics were set up one over the top of the other (tube mic is upside down, on top), head baskets less than 1/4" apart. I did my best to align the capsule position; distance from the source, as well as in rotation. The guitar was right about at the vertical center point of the 2 mics, 12th fret, about 2-feet away. Nothing done to the clips except normalizing volume. What you hear is what the mics picked up.

Mic A
View attachment 137978

Mic B
View attachment 137979

The text file reveals which mic is which. Enjoy

ETA: Interface - Tascam 4x4, gain at 12:00 for both mics. The FET mic is a tiny bit hotter, but only because I could see the levels. To hear it, the volume difference is meh.

So it's two different mics, and two different capsules, two (slightly, but still) different positions.

I do audio measurements all the time. 1mm of positioning difference is enough to give you quite audible difference. Not to mention capsules, and even if they were matched!

To get the idea if any difference comes from circuit, inject identical material through grid/gate and listen, null test at different levels. Music, transients, pink/white noise.... About 390mV rms signal would translate to about 117db spl from a typical LDC capsule, adjust signal level accordingly.

I have color boxes made out of different mic circuits through which i send different material for "vibe". Spoiler alert, unless you're doind something wrong, the circuits really need to be pushed in order to hear the difference. Or mess with biasing, components, values....

 

[rock soderstrom]

As promised...

Many thanks for the schematics and the audio examples!

Kingkorg's comments are realistic but to me the tube mic recording sounds like the tube was starving, your B+ is probably too low. Your tube can cope with relatively low anode voltages, but here it may really be too low. Have you measured the actual anode voltage?

PS: I prefer the FET recording over the tube recording as far as this example is concerned, regardless of what is responsible for this result.

 

[MicMaven]

As promised...
Schematics and sound clips. It's the basic acoustic guitar rhythm track that you might hear under any number of 80s hair metal songs, including "Sweet Child of Mine" (if you listen close enough, it's in there, and the mix would sound like it was missing something without it). Block chords, nothing fancy.

This is about as close as I can get to "same source" and not spend all day doing it.

Golly, Mr. @daschnoz , I do not hear the more forward, up front sound of the FET model you described earlier in the thread. I wonder if you moved about while recording the two tracks, as that can alter the impact of transients, especially when strumming higher pitched steel strings. As others suggest, and you apparently realize, it is difficult to capture identical sounds in separate performances/recordings. (I am NOT criticizing, merely mentioning the point in passing as we duscuss your mic project.)

I am curious, as it affects how I perceive the tone/timber of the instrument, what brand and model guitar, and what gauge strings you played? Again, there is nothing critical in my asking. As a player, and not a recording engineer, it helps to me know what I am listening to, for example a Gibson guitar typically sounds different from a Martin or Taylor or Guild guitar, and different gauge strings sound different even on the same guitar. Crazy, I know, but that helps me understand what was recorded, which, in turn, impacts what I hear. I think.

Parenthetically, while Mr. @kingkorg, Mr. @rock soderstrom, and the others are WAY more qualified to comment than I ... I often A/B test microphones by mounting them right smack dab next to each other on the same stand and record each on a differennt track, and play them back in mono, so they get the exact same pick strokes, chord changes, and percussive punch at the same distance and position. I apologize if you did something like this and I missed it. In any case, that gives me as close to exact comparison as I can manage in my ersatz home "studio" which is really my ham shack with another lablel. Again, forgive me if you did this and I missed it.

Interesting thread. James

 

[rock soderstrom]

Schematics

I think there is a small mistake in the schematic of the Mini C12 Fet, the two 22K resistors should rather be 2.2K resistors, right?

 

[rock soderstrom]

Which 25mm electret capsules did you use? KK12?

 

[daschnoz]

I think there is a small mistake in the schematic of the Mini C12 Fet, the two 22K resistors should rather be 2.2K resistors, right?

Nope, 22k is correct. The mic input of the board/interface/etc is going to have 2.2k resistors and I wanted something that was high enough to not add parallel AC load to the x-fmr. The B+ on that circuit is about 40V, with about 22V at the drain on the FET.

Same thing with the tube circuit, 22k (lower current draw), B+ is 44V, plate voltage is around 30V. The max plate voltage on the 6021 is 170V, so a 44V B+ looks pretty good on the curves.

The capsule is the 25mm capsule that came in the set of ProAR 414 mics. I upgraded the capsule in those to a 34mm electret, so they are unknown.

 

[rock soderstrom]

Nope, 22k is correct. The mic input of the board/interface/etc is going to have 2.2k resistors and I wanted something that was high enough to not add parallel AC load to the x-fmr.

Okay, I'm a bit confused right now, that's probably because it's the middle of the night here. I think I should go to bed.🥱

Same thing with the tube circuit, 22k (lower current draw), B+ is 44V, plate voltage is around 30V. The max plate voltage on the 6021 is 170V, so a 44V B+ looks pretty good on the curves.

Thanks for the info. I don't think the tube circuit sounds as good as the FET. That reminds me of my own low voltage tube experiments.

The capsule is the 25mm capsule that came in the set of ProAR 414 mics. I upgraded the capsule in those to a 34mm electret, so they are unknown

Thanks for info and feedback.

 

[daschnoz]

My original design of that tube mic was a B+ of ~100V and a red LED for the bias (~2V). The couple of DC-DC converters that I bought from Amazon were dead out of the box. Not wanting to wait, I redid the load line using phantom power. With the high resolution of a pencil line and a datasheet from 1960-something, the bias voltage was around 0.7V, hence the 1N914 diode. On paper, it worked, so I built it. I may still try again with a different DC-DC converter and rework it for 100V.

 

[daschnoz]

Golly, Mr. @daschnoz , I do not hear the more forward, up front sound of the FET model you described earlier in the thread. I wonder if you moved about while recording the two tracks, as that can alter the impact of transients, especially when strumming higher pitched steel strings. As others suggest, and you apparently realize, it is difficult to capture identical sounds in separate performances/recordings. (I am NOT criticizing, merely mentioning the point in passing as we duscuss your mic project.)

I am curious, as it affects how I perceive the tone/timber of the instrument, what brand and model guitar, and what gauge strings you played? Again, there is nothing critical in my asking. As a player, and not a recording engineer, it helps to me know what I am listening to, for example a Gibson guitar typically sounds different from a Martin or Taylor or Guild guitar, and different gauge strings sound different even on the same guitar. Crazy, I know, but that helps me understand what was recorded, which, in turn, impacts what I hear. I think.

Parenthetically, while Mr. @kingkorg, Mr. @rock soderstrom, and the others are WAY more qualified to comment than I ... I often A/B test microphones by mounting them right smack dab next to each other on the same stand and record each on a differennt track, and play them back in mono, so they get the exact same pick strokes, chord changes, and percussive punch at the same distance and position. I apologize if you did something like this and I missed it. In any case, that gives me as close to exact comparison as I can manage in my ersatz home "studio" which is really my ham shack with another lablel. Again, forgive me if you did this and I missed it.

Interesting thread. James

That's exactly how I set up the mics to record the track, as close to each other as possible.

The guitar is a Greg Bennett, Eden Plains model, by Samick (I don't even know if those companies are still a thing). Strings are D'Addario Phosphor Bronze, Custom Light (11-52). The strings were put on sometime in February, and my daughter has been playing it, so they're likely overdue for a change. They do sound kinda dead.

73

 

[k brown]

It's well known that tube go "oooo", and FETs go "ehh" . . .