HPF Solution... help needed

[johnheath]

Hi all...

I have been fiddling with a simple preamp based on the V72 in the first stage and a SRPP stage as the driver.

I would like to add a simple passive RC HPF between the two stages but I seem unable to get it to work as wanted... or do I need a hearing aid??? .  What I want is the HPF to have a calculated roll-off of 106HZ and 55Hz (the two 15nF caps with the 100k resistor) but it seems like the HPF is affected by other caps and resistors and I cannot see how.

I am also concerned if the Pot with 100k actually would create a "variable" HPF in which I have no intrest.

Anyhow I post a schematic of the two different solutions that I have been working on.

In the A I wonder how the suggested R1 and R2 would isolate the HPF from the EF86 coupling cap? And also if I could use the R3 to add some resistance to get 100k in total with the parallell 1M (R4)?

I also wonder how C1 and C2 would affect the HPF... is the C1 just in series with the two 15nF?


In the B schematic I think if the C3 and R4 would be isolated from the HPF as a single RC filter and the HPF with R3 as a separate RC filter?

Any advise would be very much appreciated.

Cheers

/John

 

[kambo]

not exact answer to ur question but,
if u lower the value of C1 you will be cutting bass anyway...

try 0.01 uf...

 

[PRR]

What response DO you get?

But-- your LPF is *inside* a NFB loop. The NFB will correct-out as much LPF HPF effect as it can. As your NFB encloses the whole preamp, there isn't an obvious way to do what you want.

Same for your proposed gain pot. The amp NFB wants to hold gain to 57.6. Turning-down is just spoiling the NFB. NFB will be gone before gain drops.

If you toss the NFB-- the pentode's output impedance is near 220K, and you show a 100K pot load (alternate 100K pot || 100K R). Plus some no-value resistors so I dunno what you really have. The pole must be computed on ALL this resistance. And in general a simple C-R cutoff is much easier designed from a low source impedance into a high load impedance.

 

[PRR]

What is 68K at V2B plate??

Why are two stacked identically-biased triodes sitting at 38V and 52V p-k?

 

[johnheath]

Thanks for your input both of you. Kambo: I do not want to cut the bass in general I just would like a HPF to swith when I want to cut som low frequencies.

PRR: First the 68k on the anode is for lowering the voltage to desired ~95V... according to the data sheet of the 6922? Maybe I am way out of control here?

And the NFB is not needed for frequency response but maybe for gain?

The "no-value" resistors were shown for discussion purpose. I guess that the R1 could be rather low for not affecting the gain (10k would not change much in gain I suppose)

R3  is there if it could be used as a resistor adjusted with the pot value (which is not always spot on)

Regards

/John

 

[kambo]

Thanks for your input both of you. Kambo: I do not want to cut the bass in general I just would like a HPF to swith when I want to cut som low frequencies.

i wonder if there is any reason why wouldnt you be able to switch that 0.001/0.002/1uf ?

 

[johnheath]

Well, that is what I thought too but when doing so I get not difference. I have done the same HPF in another design but in that preamp I am using three double triodes and use the HPF in between two triodes and the pot is after that tube and it works just fine... but in this preamp it does not...

I have been soldering and testing like a maniac but the problem still remains so that is why I turn to this forum for som advise.

I do not have a computing design program so maybe that is why I can't find the answer?

Regards

/John

 

[kambo]

interesting!
just added a NFB to my WIP bench preamp, and lowering C1 worked for me...
i am curious too now, i wonder whats the trick with this design !

 

[johnheath]

Yes, lowering C1 would alter the frequency but do you calculate  the value of C1 in series with the caps in the HPF?

This is one of my questions how that would affect the frequency with the pot or a static resistor?

Maybe it is my poor english that affects my ability to understand what is being said?

Regards

/John

 

[ruffrecords]

Your HPF is currently inside the NFB loop. The NFB will do its best to ignore it. For a HPF to work it needs to be outside the NFB loop or part of the NFB itself.

Cheers

Ian

 

[johnheath]

Thanks Ian...

I am at work at the moment but I will skip the NFB and try again. I guess that it was what PRR also mentioned so I will go for it.

Tricky stuff with NFB

Best regards

/John

 

[kambo]

haha, thats what was my understanding too but, wanted to make sure.
PRR made me confused with "LPF"

made further tests, it is not effective as without the NFB!

u might as well move it front end
http://ethanwiner.com/gadgets.html

 

[johnheath]

I guess that (after reading my first post again) I saw that I wrote RC-filter and I guess that would be a LPF???

My intention was a CR-filter

Regards

/John

 

[abbey road d enfer]

Two reasons why it doesn't work as you thought:
1) As others have noted before, the NFB loop tends to linearize the frequency response.
2) You have not included the output impedance of the 1st stage in your calculations. The output Z of a pentode is very high so the actual output Z of the stage is more or less equal to the plate resistor, so the R value in your calculations should be 320k, not 100k.
I think you could insert a cap between the input xfmr's secondary and the grid resistor of the EF86. That would be "outside the NFB loop" and prevent any possible overload of the 1st stage.

 

[johnheath]

Thank you sir

That is one of my problems... my lack of knowledge because I do not know how to calculate the impedance of a tube… pentode or triode for that matter.

So am I right if I think that you mean calculate 320k for the CR-filter?

Best regards

/John

 

[PRR]

> PRR made me confused with "LPF"

I only have 10 fingers and not all off them work guud. Corrected.

> I do not have a computing design program so maybe that is why I can't find the answer?

*ALL* classic tube circuits were designed without computer programs. The designer usually had a slide-rule (and a good sense of proportion to set the decimal point). But an engineer would not cross the room to fetch his slide-rule if a martini-napkin were handy. Today a calculator is a little easier (and far more common) than a slide-rule. "EE" and "1/x" keys are very handy.

> I do not know how to calculate the impedance of a tube... pentode or triode

As Abbey said: Pentode is normally "equal" to plate resistor.

For fine figuring, assume internal rp is about 10 times Rp. Then output resistance is near 0.9*Rp.

For triodes in most sane conditions, pencil 0.3*Rp. Lower for low-Mu tubes, higher for high-Mu tubes.

 

[PRR]

me> 38V and 52V p-k?

Because the NFB tap sucks-off much of the top tube's current. Spoils the intended push-pull action a lot.

> the 68k on the anode is for lowering the voltage to desired ~95V

But each section has, by your notes, 38V and 52V. Half of this supposed "95V".

Where is it written "use 95V"? The first set of 90-100V conditions are really for "TV Tuner" work, with high current for low hiss; voltage held down because output will be small. And page 12 shows a stacked-up 2-triode configuration taking 180V supply.

Philips 1968 page 6 clearly shows 200V and 5 to 9mA for "power output" use, which is what would normally be wanted at an Output section.

The series 68K with no bypass *seriously* increases output impedance and reduces power output. If you WANT a "easily over-loaded" output stage, with strong 2nd harmonic due to 68K plate feed and 68K NFB tap, then fine.

 

[kambo]

> PRR made me confused with "LPF"

I only have 10 fingers and not all off them work guud. Corrected.

my humble apologies,
i didnt mean to put a blame on you, with my little knowledge 

every single post from you is super educational....
typo once in a while is even more educational  8)

you rock 
along with Abey, John, Ian and others, i wish NYD was still around too

 

[johnheath]

PRR: Super thank you sir for bringing some lights on this for me… I really appreciate it.

About the voltage it is not (as you say) written I just saw the page 12 in the data sheet I had in hand. I have seen some designs using a lot higher plate voltage on the 6922 then I read a thread here were some people were critical to higher plate voltage so I tried to reduce the voltage with the resistor but missed to bypass it… so thank for the info on that.

And about the impedance?… well, saying it again - thanks for the info on how to calculate it even if I guess it is the "simple" way to calculate it

The NFB is gone… voltage back to where I "started" with some fine adjustments and bypassed.

Now I just need to get the HPF back in business.

Thanks again

Regards

/John

 

[johnheath]

And the 1/x button on the calculator is great