Nova/Groove Tubes/Sterling mod questions

[Amled87]

For the most part, yeah

Not sure that 500pF-larger-than-stock input capacitance REALLY works as you might expect, though. Confirmation bias is a fickle mistress...

1nF into 1G gives a cutoff frequency of 0.16Hz - that's zero-point-sixteen Hertz, waaaaaaaaay below what you can expect to hear (not to mention, what you might want to pick up / record).

1.5nF into 1G pushes that even further down, to 0.106Hz. If you claim you can hear THAT difference, you're the most advanced whale / elephant (what other animals deal with infrasound / subsonic frequencies?) i've heard of And "golden ears" like that belong to "audiophool" environments anyway, not in this blue-collar world But i merely jest...

Jokes aside though, you'll also want to add a 100-200k resistor between the C2 "free end", and ground, and probe the signal at the connection between those two (in order to also see the effect of that output capacitor has on the low frequencies, and the load that a preamp that you would plug this mic into, would have on the circuit).

Other than that, yeah, looks about right indeed And if you click on the name of the signal, in the plot window, you get a cursor you can grab the vertical line of, and drag side to side & see a detailed readout of the signal level at the respective frequency.

Although you might want to consider a JFET there, if that's what your real circuit is using..? Or rather, do indeed recreate THE actual circuit you're trying to analyze? But i do understand if / that this was a sanity-check first-attempt, just to make sure the "protocol" is right

The added 500pf is there on the input because I liked what it did to the sound (the original 1n is a cheap polyester cap and I added a silver mica of 500pf) I'll swap the npn transistor out for a jfet which is what is in the circuit lol. Quick run through so just grabbed components from the list lol

 

[Khron]

I liked what it did to the sound

https://en.wikipedia.org/wiki/Confirmation_bias

Just saying...

If you expect a difference, you WILL hear a difference.

 

[Amled87]

https://en.wikipedia.org/wiki/Confirmation_bias

Just saying...

If you expect a difference, you WILL hear a difference.

Hmm, I can't explain it to be honest. It just seemed like a good idea at the time (before making these new changes, I added the silver mica cap to the input and also added a 0.1uF wima polypropylene to the 10uF output electrolytic cap). I've since changed the the capsule coupling cap, output cap to both philips polyester caps though. I'll remove the 500pF because mathematically it's not doing anything as you stated, because no one can hear where it extends the low-end. Perhaps replace the capsule coupling cap with something better all together of same value if I'm going to do anything there.

Quick question, on this simulation should I put the capsule coupling cap between the two 1G as in the schematic or before/after it. Also, should I be adding the 57pF of the capsule capacitance to the input as well? I measured it out of circuit with a meter.

 

[Amled87]

https://en.wikipedia.org/wiki/Confirmation_bias

Just saying...

If you expect a difference, you WILL hear a difference.

How would I go about simulating this cardiod only u87 style schematic? Every time I do, I get a 10dB or so drop at 1kHz that extends on up. The only way I get any of that back is by changing the 470nF capacitor to no cap! I know the 470nF, 120pF/1m is part of the HF nfb being fed back into the Jfet, but it seems to me when I sim it, I see none of that and the only component doing ANYTHING is the 470nF cap because as I said, without it it's a uniform frequency resonse and with it, it dips about 1kHz and up?

 

[Khron]

How would I go about simulating this cardiod only u87 style schematic? Every time I do, I get a 10dB or so drop at 1kHz that extends on up.

How about a screenshot of the schematic in your simulation software? "The devil's in the details"...

 

[Amled87]

How about a screenshot of the schematic in your simulation software? "The devil's in the details"...

So, i started with just the jfet and the 1n across the drain resistor that you first mentioned, and provided the sim and frequency response of that.

Then, i looked at some different u87 style NFB de-emphassis and changed it to a 1M/470pF across the drain resistor. Both ways sound different, with the 1nF cap seemingly killing a bunch of high-end and the 1m/470pF leaving more but it's kind of "peaky" still. I know i'm not going to turn this into a u87, but trying to make it more "smooth" and not as sibilant/harsh on sources while keeping "some" high-end is my goal. My question is, did I do the 1m/470pF thing correctly and is that the correct way to apply it?

 

[Khron]

changed it to a 1M/470pF across the drain resistor. Both ways sound different, with the 1nF cap seemingly killing a bunch of high-end and the 1m/470pF leaving more but it's kind of "peaky" still

Well, that's a "trap for young players" there - in Ltspice, "1m" is the same as "1M", and it's 1 milliohm (0.001 ohm). If you want 1000000 ohms there, you wanna write "1meg".

But that's nowhere near "U87 style NFB", just so you know...

did I do the 1m/470pF thing correctly and is that the correct way to apply it

That depends what exactly was your goal..?

 

[Amled87]

Well, that's a "trap for young players" there - in Ltspice, "1m" is the same as "1M", and it's 1 milliohm (0.001 ohm). If you want 1000000 ohms there, you wanna write "1meg".

But that's nowhere near "U87 style NFB", just so you know...[

QUOTE="Amled87, post: 1175138, member: 129988"]
did I do the 1m/470pF thing correctly and is that the correct way to apply it

That depends what exactly was your goal..?
[/QUOTE]

Well, my goal was to apply nfb to reduce the high-end? Just experimenting and trying to find which way I like most that keeps some high-end but also gets rid of the horrible sibilance/harsh HF.

 

[Khron]

Well, my goal was to apply nfb to reduce the high-end?

Pretty sure "capacitor across the drain resistor" doesn't qualify as negative feedback.

And if i hadn't made it clear in the previous reply, the difference between in the two simulated schematics in your previous post is 1nF vs 470pF. That "1m", as far as LTspice is concerned, might as well not be there, since it's one ***milliohm***.

 

[Amled87]

Pretty sure "capacitor across the drain resistor" doesn't qualify as negative feedback.

And if i hadn't made it clear in the previous reply, the difference between in the two simulated schematics in your previous post is 1nF vs 470pF. That "1m", as far as LTspice is concerned, might as well not be there, since it's one ***milliohm***.

I'm not saying the capacitor across the drain resistor is negative feedback. I meant I got the idea (it's now a 2.2nF cap from the drain to a 10k resistor parallel with 47k drain resistor) from looking at HF de-emphassis in the u87, which is negative feedback, no?

 

[Khron]

I'm not saying the capacitor across the drain resistor is negative feedback.

Well, certain words / phrases have certain meanings - otherwise communication no longer makes sense or has a purpose, right?

it's now a 2.2nF cap from the drain to a 10k resistor parallel with 47k drain resistor

Adding a series resistor to the capacitor shifts the effect from being a low-pass filter to a high-shelf attenuation.

 

[Amled87]

Well, certain words / phrases have certain meanings - otherwise communication no longer makes sense or has a purpose, right?



Adding a series resistor to the capacitor shifts the effect from being a low-pass filter to a high-shelf attenuation.

I'll try to be more clear and use my terminology correct Khron, I apologize!

I do believe I like the cap/resistor better than the capacitor only. The capacitor seemingly killed all high-end above a point while the resistor/cap seemed to retain some.

Thank you for all of the help Khron!

 

[Johnny1234]

This may be of interest to some owners of these mics.




I bought 2 GT-55s cheap. Tried removing the resonator disk but doing so damaged gold on capsule.
Replaced with Ari capsules and night and day difference for the better IMO.