C again

[Gus]

Had anyone had the chance to put values to the schematic?

http://www.omnipressor.com/MicSchLib/Bphantomc.GIF

I don't think it is a great design but I am curious about the use of a NS process 52 type fet at the front without a source resistor bypass cap(higher noise lower gain) and a process 50 as the SF.

The process 52 has a higher noise rating and input cap than a 50 reading the specs in my data book. That does not make sense

People seem to like this microphone. I have not heard one. Maybe it is because it has a very good MBHO capsule and OK body grill design people like it.

FWIW I have seen a picture of inside a horsh tube microphone from a site on the web. Again if you find it look close it looks like a MBHO capsule(very good) but the leads on the tube look to be short the heat of soldering and mechanical stress can hurt the vacuum seal I believe.

 

[chrissugar]

Hey Gus

I totally forgot about it. If I scan a paper with the drawing of the components and resistor values attached as colour codes it is ok?I know it is a shame that after more than 25 years of electronics I'm still uncapable to learn the colour codes so please don't laugh.
I wrote down on a paper the colours from the resistors to have the values decoded later :grin: .

chris

p.s. by the way, how do you know at a 1% metal film resistors that has five colour rings which one is the first ring. :grin:
I'm not joking this is a serious question.

 

[bcarso]

Chris, typically there is a little separation between the tolerance band (brown for 1%) and the four color bands. The other clue is to decode from each end if it is not clear and then compare to a chart of standard 1% values.

Once in a while you will get a palindrome which is correct regardless of the starting end. 100 ohms is an example (brown black black black brown).

If you can see the four bands as a group, distinct from the tolerance band, then have the four bands on your left and read from left to right: the first three are the significant figures and the last is the decimal point code (or exponent of ten code).

Example: from left to right, brown black black red decodes as 1, 0, 0, 2. So this is 100 with two (red) zeros to the left of the decimal point, or 10,000 ohms.

A few of the standard 1% values are coincident with 5% values and to make matters worse a number of manufacturers now offer 1% tolerance parts with nominal 5% values.

A handy way to learn to color codes is to get a small piece of 10-conductor color-coded ribbon cable and carry it around for a while. All you have to remember initially is that black is zero and brown is one.

The only thing beyond this is to know that black in the fourth position means no zeros (it can be thought of as the decimal point, or a mulitplier of 10^0 = 1), gold in the fourth means the decimal point is to the left of the third band code (or a multipler of 10^-1 = 0.1), and silver means the decimal point is to the left of the second band code (a multiplier of 10^-2 = .01).

 

[chrissugar]

[quote author="bcarso"]Chris, typically there is a little separation between the tolerance band (brown for 1%) and the four color bands.[/quote]

Problem number one is that I never see there is a separation between the first four bands and the last one.

[quote author="bcarso"] The other clue is to decode from each end if it is not clear and then compare to a chart of standard 1% values.
[/quote]

I was afraid someone will say something like this. No I'm too lazy to do this.

[quote author="bcarso"]
If you can see the four bands as a group, distinct from the tolerance band, then have the four bands on your left and read from left to right: the first three are the significant figures and the last is the decimal point code (or exponent of ten code).

Example: from left to right, brown black black red decodes as 1, 0, 0, 2. So this is 100 with two (red) zeros to the left of the decimal point, or 10,000 ohms.
[/quote]

Thank you Brad, but I know the procedure, the problem is elsewhere.
I have three problems that stop me from reading colour codes:
1- never see the space between the first four rings and the fifth.
2-can't memorise which is 1 and which is 0, black or brown.
3-can't memorise the powers of ten.

[quote author="bcarso"]
A few of the standard 1% values are coincident with 5% values and to make matters worse a number of manufacturers now offer 1% tolerance parts with nominal 5% values.

A handy way to learn to color codes is to get a small piece of 10-conductor color-coded ribbon cable and carry it around for a while. All you have to remember initially is that black is zero and brown is one.

The only thing beyond this is to know that black in the fourth position means no zeros (it can be thought of as the decimal point, or a mulitplier of 10^0 = 1), gold in the fourth means the decimal point is to the left of the third band code (or a multipler of 10^-1 = 0.1), and silver means the decimal point is to the left of the second band code (a multiplier of 10^-2 = .01).[/quote]

Oh, no I knew it is more complicated than I thought. Too many variables. :grin:


I know it is hilarious, I can memorise formulas, even things I learned long time ago like trigonometry formulas, thermodinamics and mechanics formulas I never used for more than 10 years and still remember, but I'm not capable to learn this. :grin:
Or maybe I'm too lazy for this.
OK now you can all laugh at me.

chrissugar

 

[bcarso]

Sounds like a blockage. I've had similar problems when certain material (matrix algebra for me) was introduced during a traumatic period, and it can be difficult but not impossible to eradicate. Sometimes re-living the events in meditation and then "editing the film" can work well.

For the black brown distinction, think perhaps of black holes in association. Then realize that the hole looks like a zero.

The central part of the color sequence is like a rainbow of primary colors starting with red, through violet. That fills in 2 through 7, leaving only the somewhat counterintuitive grey and white for 8 and 9.

 

[chrissugar]

Yes, I also think it is a blockage but can't identify the real source of the problem. I think it is from the fact that I can't memorise the black, brown, white and grey.

chrissugar

 

[chrissugar]

Hey Gus,

Here are the values I decoded from the colour codes.
Because I'm not sure I decoded them correctly I will also add a list with the colour codes I wrote down when I disected the mic.
http://groupdiy.twin-x.com/displayimage.php?pos=-356

270k=red/violet/black/orange/brown

2k7=red/violet/black/brown/brown

19k1=brown/white/brown/red/brown

30k=orange/bleack/black/red/brown

10k=brown/black/black/red/brown

check the colour codes, maybe they are reversed. :grin:

chrissugar

 

[PRR]

> 2-can't memorise which is 1 and which is 0, black or brown.
> 3-can't memorise the powers of ten.

{sigh....}

I hate to repeat this because it is grossly offensive, but:

0 = Black
1 = Boys
2 = Rape
3 = Our
4 = Young
5 = Girls
6 = But
7 = Violet
8 = Gives
9 = Willingly

The phrase "Black Boys" was offensive when I learned it, which was before LBJ and all that. BUT it does tell you "black from brown". (Assuming you remember to count from zero.) (Ignoring the obvious fact that, although most cultures call the darkest skin tones "black", "black" skin is closer to brown.)

OTOH, the suggestion about Violet's willingness is not the scandal that it once was.

Alternatively, mutter "Bad Boys...." and remember that Brown-Black-Brown is 1-0-1 which is 10 plus one zero or 100 ohms.

Blue/"But" is less of an issue because there are very few blue stripes in 10% values. Likewise you see a lot of Green/"Girls" and not so many Grey/"Gives".

Whoops... I see you are from Romania, so naughty chants in english may not stick in your mind.

 

[Crusty2]

[quote author="PRR"]>
0 = Black
1 = Boys
2 = Rape
3 = Our
4 = Young
5 = Girls
6 = But
7 = Violet
8 = Gives
9 = Willingly[/quote]

I remember my freind's 80 year old grandma rattling that off to us. She learned it when working at an aircraft plant during the war. Such is the power of a good mnemonic!

 

[chrissugar]

Thank you PRR but that thing will not help, but this morning I wake up and the first thing that came to my mind was the colour code and the fact that black and white are the oposite colours, (ying/yang thing), and they are at the extremes of the colour code. It starts with black and ends with white. I realised that I will never forget that black is zero and white is nine. If I know this I also know brown and grey. The rest from red to violet was never a problem.
Thank you all for this thread because now one problem is solved, I know the right order in the colour code. :grin:

One problem still remains, the powers of ten. The problem with me is that if it is not something that comes to my mind in a second like 8x8=64, and I have to do asociations (colour-number-power of ten) then I leave it in a second. Now that would be interesting to find a way to know instantly the powers of ten.

chrissugar

 

[drpat]

[quote author="PRR"]> 2-can't memorise which is 1 and which is 0, black or brown.
> 3-can't memorise the powers of ten.

{sigh....}

I hate to repeat this because it is grossly offensive, but:

0 = Black
1 = Boys
2 = Rape
3 = Our
4 = Young
5 = Girls
6 = But
7 = Violet
8 = Gives
9 = Willingly

The phrase "Black Boys" was offensive when I learned it, which was before LBJ and all that. BUT it does tell you "black from brown". (Assuming you remember to count from zero.) (Ignoring the obvious fact that, although most cultures call the darkest skin tones "black", "black" skin is closer to brown.)

OTOH, the suggestion about Violet's willingness is not the scandal that it once was.

Alternatively, mutter "Bad Boys...." and remember that Brown-Black-Brown is 1-0-1 which is 10 plus one zero or 100 ohms.

Blue/"But" is less of an issue because there are very few blue stripes in 10% values. Likewise you see a lot of Green/"Girls" and not so many Grey/"Gives".

Whoops... I see you are from Romania, so naughty chants in english may not stick in your mind.[/quote]

Am I the only black person on this forum???

 

[bcarso]

I'm sure you're not----and also please understand that I think we all find this appalling and offensive. I'm sure it was presented as, at best, an embarrassing historical artifact. May it serve to remind us of how recently something like this could be in circulation.

 

[Gus]

By the time I learned that it was "changed" to bad.

drpat

I find a lot of things I hear/read offensive.
The whole thing is offensive

 

[chrissugar]

Hey Gus,

This hand drawn thing is usable to you? Can you check with your schem if the values are the good ones?

chrissugar

 

[Gus]

I will check it out and let you know.

 

[Gus]

j201 IDSS .2 to 1ma p52 maybe a .7ma operating point as a first guess?


BC264d IDSS 7 to 12ma p50

mark up the

http://www.omnipressor.com/MicSchLib/Bphantomc.GIF

R7 49.1K
R11 49.1K
R12 49.1K
R10 10K
R8 30K
R9 49.1K
R6 2.7K
R2 2.7K
R3 270K
R1,R4,R5 1 gig

C1 .01uf 63V
C6 470uf 25V
C5 470uf 25V
C3 10uf 63V
C4 100uf 63V
C2 560PF

D1 BZX55 need the voltage marking after I will guess 20V because of the 25V caps used

Chris need

the voltages at the
Zener
Sources of the two fets
Drain of the fets
and the center tap of the transformer
R8 R9 node

Sw1 looks to be a sensitivity control by adjusting the capsule voltage

Rant section

This will help to understand the operating points and maybe way it is designed the way it is.

Is this a bad design? or a good design with a sound in mind? looks cheap.

Look at all the electros!

R6 not bypassed with a cap? higher noise?

I would make the 1gigs 3 gigs for low noise with this circuit (AN106 at Vishay)

Why a J201 process 52???? makes no sense, process 50 fets in 1,000 are $0.06 each in the USA for some numbers. (One thing that comes to mind is maybe a better yield of fets that will drop in without changing the source R. This would go along with C3, R3 being used because of the drain voltage shift with different fets)

A BJT could be direct coupled to the first fet for even better drive to the I will guess low ratio transformer 2:1 to maybe 4:1 also gets rid of R3,C3

I think I gave away a better design here for a cardiod only LD(a modded km84 circuit)

Again a good capsule and nice grill/body are more important than the circuit sometimes IMO.

I have a VM1 it sounds very good but I don't like the PCB build inside.

PRR, Brad anyone Ideas about the design??????

 

[bcarso]

[quote author="Gus"]
This will help to understand the operating points and maybe way it is designed the way it is.

Is this a bad design? or a good design with a sound in mind? looks cheap.

Look at all the electros!

R6 not bypassed with a cap? higher noise?

I would make the 1gigs 3 gigs for low noise with this circuit (AN106 at Vishay)

Why a J201 process 52???? makes no sense, process 50 fets in 1,000 are $0.06 each in the USA for some numbers. (One thing that comes to mind is maybe a better yield of fets that will drop in without changing the source R. This would go along with C3, R3 being used because of the drain voltage shift with different fets)

A BJT could be direct coupled to the first fet for even better drive to the I will guess low ratio transformer 2:1 to maybe 4:1 also gets rid of R3,C3

I think I gave away a better design here for a cardiod only LD(a modded km84 circuit)

Again a good capsule and nice grill/body are more important than the circuit sometimes IMO.

I have a VM1 it sounds very good but I don't like the PCB build inside.

PRR, Brad anyone Ideas about the design??????[/quote]

Gus, Chris's 1's look like 4's, hence the 49.1k R's should be 19.1k, a standard 1% value.

From the characteristic curves in the old National transistor book, it looks like the use of process 52 in the input stage is for ensuring low gate leakage. Process 50 gate leakage under normal bias begins to turn upward at around 9-10V Vdg, whereas process 52 does so around 35V, and does so then with a gentle slope.

I would have probably held down the voltage on a process 50, maybe with a folded cascode, and gotten the lower voltage noise. But possibly they also liked the more classical square-law behavior and associated distortion structure of the longer-channel process 52 as well. As you point out the unbypassed R6 doesn't help noise much either. But it may make the gain more consistent.

 

[Gus]

Thanks for the info on the gate leakage part. FWIW the KM84 neumann has the fets drain at 10V to ground.
In the neumann book at the infopool section "Microphones" there is a circuit fragment showing a 3819.

Chris the zener voltage might help some more to understand what might be going on. I never asked how does it sound?

 

[chrissugar]

Sorry Gus but I had access to that mic only for a week to test it.
Didn't measure any voltages, just made some photos and wrote down on a paper the component values.
It is a nice sounding microphone and at that price I think it is a good choice. I had no time to compare to something common like the U87 but I think it is better and cheaper. Although the U87 has more options.

chrissugar

 

[Gus]

Chris

thanks for all the work.

Brad

what do you think of 15V 20V and 24V kind of standard zener voltagess to try to work out the operating points. Do you think it could be a 24V with 25V rated caps at the zener section?