APEX 460 - Formulas for calculating HPF and LPF.

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

vertiges

Well-known member
Joined
Oct 25, 2007
Messages
766
Location
Montreal, Canada
Hi all,

I'm looking for modding this microphone with a Peluso, a CM-2480 and by replacing some caps.

I'd like know which part of the circuit of the APEX-460 act as filters. Could someone help me ?

Here is the schematics of a 460 without the cathode follower. (I don't remember who did draw it)

a460schemCFremoved.jpg


I know for example that if you increase the value of C8 you'll get more bottom. So I imagine it's a HiPass Filter. The Cinemag 2480 I will use is 16.5k/150.

The formula is : Fc=1/(2.Pi.R.C) But what is R ? It's not only the resistance of the Primary, is it ?

There is also C7, R7 which influence a lot the audio. But I don't know to make the calculations. :?:

Thank you,

eD)))
 
> It's not only the resistance of the Primary, is it?

No. But pretend it is. And pretend the transformer is really a resistor. 1uFd and 16.5K comes out to, 9.65Hz. That is the -3dB frequency. -1dB will be near 19Hz, -0.5dB near 38Hz.

The full math will include the 100K in parallel with the plate resistance of the tube. Not indicated on your drawing. I'm going to WAG 40K. Then 16.5K+40K gives -0.5dB at 11Hz.

There's no 11Hz in music. Unless you do large pipe organs, 19Hz is not happening (or is all truck rumble). A piano has a string at 28hz but unless it is a 9-foot Grand, there is about zero 28Hz coming out. The lowest common musical tones are 40Hz.

Meanwhile your speakers are +/-10dB at 20Hz (if they do anywhere near 20Hz). Finding an honest +/-3dB at 50Hz is rare. And your DAW has a virtual knob for +/-20dB at 20Hz. Half-dB "down" at 40Hz is not a problem you should fix.

Actually, the transformer is an inductor. And L-C networks resonate. I don't see an inductance number on the spec-sheet, but their 20K plot suggests 600H. And 600H against 1uFd is 6.5Hz. Is it up or down at 6.5Hz? Depends on the tube plate resistance again; also loading. But if Q is anywhere near "1", then whether up or down at 6.5Hz it will be dead-flat by 13Hz, or 19Hz even if Q is as high as 9. My SWAG for Q is 0.6, so I expect dead-flat by 13Hz.

C7 shelves bass and, again, we wanna know what the tube and voltage is. However it is unlikely that Gm is higher than 1,000uMho, with 100K in plate and 2K7 at cathode it will reflect as more like 2K ohms. Response will drop a bit by 0.8Hz.

The capsule capacitance in series with 1nFd, the 1Gig, and the 200Meg are a low-cut. Dunno what that capsule capacitance is. Can't readily change it. But probably 60pFd. In series with 1nFd or 1,000pFd, still 60pFd (maybe 57pFd but our uncertainty is larger than 57-60pFd). The two resistors sum to 167Meg. (Yeah, I moved the 1Gig over after the 1nFd because I "know" neither really matters.) Then 60pFd and 167Meg is -3dB at 16Hz. This is your main electrical roll-off. Electrically, you want the 200Meg bigger BUT the tube's grid current noise in the larger resistor gives a larger noise voltage and more hiss.

The electronics, as-shown, are flat for any musical bass. You want more, turn a knob and boost it.

The tube is a fancy-number 12AY7 running 0.6mA and Rp is near 60K.
 
PRR,

Thank you for this very instructive and complete answer. :thumb: :shock:

[quote author="PRR"]> It's not only the resistance of the Primary, is it?

No. But pretend it is. And pretend the transformer is really a resistor. 1uFd and 16.5K comes out to, 9.65Hz. That is the -3dB frequency. -1dB will be near 19Hz, -0.5dB near 38Hz.


The full math will include the 100K in parallel with the plate resistance of the tube. Not indicated on your drawing. I'm going to WAG 40K. Then 16.5K+40K gives -0.5dB at 11Hz.
[/quote]

I'm very surprised. I've read that some people find the microphone to have more basse freq. with a 1.5 µF (6.43 Hz) and 2.2µF (4.38 Hz), that' s why I thought my calculations were wrong. The first schematics of the C12 had a C8=0.5µF (19.29 Hz).
Can a value of the cap change the sound (independently of the influence of its value in the circuit) ? I hope you understand what I mean. I'm just wondering how people can hear a difference... as we talking about frequencies under 20 Hz.

[quote author="PRR"]
Actually, the transformer is an inductor. And L-C networks resonate. I don't see an inductance number on the spec-sheet, but their 20K plot suggests 600H. And 600H against 1uFd is 6.5Hz. Is it up or down at 6.5Hz? Depends on the tube plate resistance again; also loading. But if Q is anywhere near "1", then whether up or down at 6.5Hz it will be dead-flat by 13Hz, or 19Hz even if Q is as high as 9. My SWAG for Q is 0.6, so I expect dead-flat by 13Hz.
[/quote]

Ok, so again with those C8 values, it's quite impossible to hear the roll off.

[quote author="PRR"]>
C7 shelves bass and, again, we wanna know what the tube and voltage is. However it is unlikely that Gm is higher than 1,000uMho, with 100K in plate and 2K7 at cathode it will reflect as more like 2K ohms. Response will drop a bit by 0.8Hz.
[/quote]

Around 160 Volts at B+ and 6.3 Voltsat the heater.
The tube will be a GE6072A.

[quote author="PRR"]
The capsule capacitance in series with 1nFd, the 1Gig, and the 200Meg are a low-cut. Dunno what that capsule capacitance is. Can't readily change it. But probably 60pFd. In series with 1nFd or 1,000pFd, still 60pFd (maybe 57pFd but our uncertainty is larger than 57-60pFd). The two resistors sum to 167Meg. (Yeah, I moved the 1Gig over after the 1nFd because I "know" neither really matters.) Then 60pFd and 167Meg is -3dB at 16Hz. This is your main electrical roll-off. Electrically, you want the 200Meg bigger BUT the tube's grid current noise in the larger resistor gives a larger noise voltage and more hiss.
[/quote]

I don't no about the capacitance of the capsule. No data in the Peluso Web site. (I will probably use a CEK12)

Could you tell me what R7 does in this circuit ?
BTW, is there a good book or articles I could read ? I do like to understand and learn what every part does ?

[quote author="PRR"]>
The electronics, as-shown, are flat for any musical bass. You want more, turn a knob and boost it.

The tube is a fancy-number 12AY7 running 0.6mA and Rp is near 60K.[/quote]

Thanks a lot,

eD
 
I've got another questions. (I'm a noob, but I learn a bit every day... :wink: )

1) How calculate the voltage applied at the plate ?

I mean for knowing the voltage drop by R1 and R2, you have to know the current right ?

Is the following formula is correct ? :

i=1/(R1+ R2 + Rplate + Rcathode + R7) :roll:
And about Rplate and Rcathode, are they given by the manufacturer of the tube ? It's pretty hard to find....

=========================================================================================================

2) I'm a bit confused about how the bias voltage is applied to the grid because I thought it was by DC ; with :roll: C4 it's not possible... Could someone explain ?

There are two different voltages which reach the grid : the AC one from the capsule a the DC negative one (the bias voltage) ?

Thank you,

eD
 
[quote author="PRR"]

I don't see an inductance number on the spec-sheet, but their 20K plot suggests 600H. [/quote]

That 600H figure is somewhat optimistic. On a lowQ I measured it as 104H on my Genrad 1650

Best, M
 
R6 takes the grid to 0V, while the current through the tube and R7, gives a positive voltage across R7.
This gives you a negative voltage(bias) on the grid, compared to the cathode.

Kåre
 
[quote author="lydmann"]R6 takes the grid to 0V, while the current through the tube and R7, gives a positive voltage across R7.
This gives you a negative voltage(bias) on the grid, compared to the cathode.

Kåre[/quote]

It took me the half ot the day to understand... but now it makes sense to me... :?

Thank you !

The way of a tube works is fascinating... :thumb:

eD)))
 
> for knowing the voltage drop by R1 and R2, you have to know the current right?

Well, you know at a glace there is at least 10K+100K+2K7= 112K7 in the circuit. If the tube could be a dead-short, the current could be 160V/112K= 1.43mA. Or if the tube were dead-open the current could be zero. Short or open are not likely; the broad-side-of-barn guess would be half-way between, 0.7mA. Or maybe 1/3rd to 2/3rd, 0.5mA-1mA. You rarely need a re-exact answer.

For a more refined answer, you plot 112K7 on the 12AY7/6027 curves, find where the plate current times the 2K7 cathode resistor equals the grid voltage. This is best taught on a chalk-board, but is basic tube-bias stuff and is written-up many places on the web.

You can find Rp by figuring the slope of the plate curves near the bias-point, or the GE 6027 has another plot of Rp versus Ip for a few plate voltages.

> On a lowQ I measured it as 104H on my Genrad 1650

Below 100Hz, or at 1KHz? Signal level also matters.

I was looking at the transformer frequency response for 20K source and 1K5 load. It shows very small roll-off at 20Hz. I agree that 600H seems incredible... but it is not cheap iron and is scaled for very small power.

But taking 100H instead of 600H only moves resonance to 16Hz.

OTOH, the capsule has response limits. As an Omni (does anybody do omni?) the lower limit is controlled by leakage, actually by ability to react to sudden huge pressure change like trunk-slam, and will be no lower than 2Hz-20Hz. The directional patterns may be similar or less (higher Hz) due to the added problems of differential action.
 
[quote author="PRR"]

> On a lowQ I measured it as 104H on my Genrad 1650

Below 100Hz, or at 1KHz? Signal level also matters.
[/quote]

At 1KHz. Way too lazy to connect ext. generator.
BTW, how to measure inductance for different currents with this bridge? Say I have a gapped choke and I want to get inductance @ 8ma? Is there any way?

Best, M
 
I noted you just posted the same question at Klaus's forum.

Time to read some books
 
[quote author="Gus"]I noted you just posted the same question at Klaus's forum.[/quote]

Yes indeed. PRR gave me in details the theoretical explanation, but it still doesn't explain why people can here a difference between 3 values of this cap...
So I guess I need the meta-physical answer... :green:

The more I read about the subject, the more I doubt about the objectivity of the listening tests. I mean for the capsules and tubes changes, it must be obvious, but when you start to play with different caps brands or very close values... I wonder... You have a mic, you record your voice. OK. After that, you make a mod, you record your voice again, even at the same distance it will never be the same voice... So why to give the credit of the change to the mod and not to the voice ? :roll: Do you understand what I mean ? And I'm not even talking about the burning time of the components...
The only way would be to start with to identical stock mics (not easy) and to mod only one of them. After that make A/B tests.

Maybe I'm totally wrong and a change between a 1nF and 2.2 uF is obvious (I'm talking about C8). But if it's the case, I would like to understand why. (I know I'm obsessed :green: ) Because with the calculations results, I don't expect an obvious change but an "inaudible change". So it must exist some unknown factors (to me) who have an influence of the bass response of this cap !!

eD))
 
[quote author="vertiges"]
The more I read about the subject, the more I doubt about the objectivity of the listening tests.....
[/quote]

It is very difficult. When trying to compare objectively, on would usually record an amplified version of something that has already been tracked. So take a vocal or instrument track which has been well recorded, then play out into the world and record again and compare with other known mics. It's not ideal but it is more repeatable.

Unfortunately this doesn't tell you how the mic responds to sssssibilance, plosives, foot noises etc.

It's not easy - our response to sound is inherently subjective!
 
[quote author="vertiges"]
So I guess I need the meta-physical answer... :green:

The more I read about the subject, the more I doubt about the objectivity of the listening tests. I mean for the capsules and tubes changes, it must be obvious, but when you start to play with different caps brands or very close values... I wonder...
[/quote]

Make a search on Klaus forum.
Not long ago Oliver gave an excellent answer to your question.

Best, Mark
 

Latest posts

Back
Top