Simple stepped Attenuator values @ 600R line

[schlamn]

Hi,
May I ask for an eye to be cast over these attenuator values.. I am not asking for a check on any maths - I hope its correct .. more just the ballpark values -> Do they seem reasonable (or too large maybe ?)?

Its got 8 steps ~6dB down to ~41dB.. Input is from a Balanced 600R line level source. Output will be to Balanced 600R I/P.

It is going to be built on the cheap 4 wafer type Diy Attenuators that are available and will be at the I/P and also O/P of a Rca BA6B Lim..

Btw.. new poster - so Hi, Been registered and reading for a while..  Invaluable resource - Thanks.

Anyways thoughts will be appreciated.

 

[schlamn]

Ok ..no response .. This is a little odd.  I am sure this is not a complex issue - so either I am asking a question so dumb it doesnt warrant an answer or it has been done to death elswhere and I havent found it !


I will expand slightly...

I built the Redd.47 pre-amp and utilised a Balanced attenuator (Input) design shown by John Hinson which was for a Z in of 200R.

It works of course.

I want the same simple design but to work for an input Z of 600R (then the same on the output - of a RCA BA6B soon to be built)

I changed all resistor values to suit and did some maths.. It seems correct.

But I have not done this before and just wanted a quick "that'll work fine " or otherwise.

I have seen the ladder calculators too, but not one for this particular design. The balanced ones incorporate 2 extra resistors on the O/P. These seem to be able to utilise much lower resistor values.

Are my larger resistor values (despite the fact that they will give the desired attenuation) a bad idea ?

I am really hoping someone could just chime in briefly..

Thanks in advance.

Keith

 

[gary o]

Hi keith I cant answer but there is lots here about attenuattors try some searches bridged t u attenuattors word like that I no Ive asked questions and got good answers it should all be here good luck

 

[schlamn]

Hi Gary, "Bridged T" - I will give that a go cheers.
Just getting bits together so far and have been thinking about taming all this gain.
The switching stuff is a bit complex too !!

 

[gary o]

I used daul 500K pot as daul ladder on input, think I did fixed pad on output followed by a 1K pot suggested by PRR & new york dave....yeah swithing is fiddly I left out the tube test part

 

[PRR]

> Do they seem reasonable....?

No.

You want: constant 600 ohms input, large (>20dB) attenuation, 600 ohm load, do not care what source the load sees?

Set your plan to the "-41" position. The input is at least 40K or 60K, with R1 R2 R3 R4 facing the source. That is not correct.

Start simple. Calculate un-balanced; you can easily convert to balanced after you know the unbalanced version.

To prevent errors such as the above, try extreme values.

For "infinite" attenuation and 600 ohm input, the only possible configuration is shorted output, 600 ohms series. (For 40dB attenuation, the output is about 1/100th the total, so about 6 ohms shunt and about 596 ohms series.)

For zero attenuation, the shunt must be infinite and the series resistor is zero.

Check 6dB because it is an easy 2:1 ratio yet easy to get wrong. The load is 600. If we did not care what the input was, we would use shunt=infinite and series=600. However that gives input=1200, which is twice as high as we want. Therefore change series to half value, 300. And change shunt to give the correct attenuation. 600 shunt in parallel with 600 load is 300 ohms total shunt, with the 300 series gives 2:1, which is the correct value.

> Redd.47 pre-amp

A microphone input? The microphone does not need a specific load, just "high" (over 1K). The transformer inside the preamp does not like to see a low-low source. So the REDD shows a few-hundred ohms to the preamp at any setting, and lets the microphone face high to very-high impedance.

> it has been done to death elswhere

Well, it has been done to death. Although many of the men who did it to death are now dead. Basic attenuator topologies and calculations are very well covered in standard "books".

 

[PRR]

me> The transformer inside the preamp does not like to see a low-low source.

This is the drawback of the Balanced L-pad. At high attenuations the output impedance approaches zero. When driving a grid, this is not a problem. When driving transformers it "might" be a problem.

There is a simple variation: the T-pad. At high attenuations the output Z approaches the nominal Z; at low attenuations it approaches the source Z (which is often nominal Z). The math is quite easy:

http://www.fmsystems-inc.com/index.cfm?tdc=dsp&page=engineers_detail&pid=7

That page computes several unbalanced 75-ohm pads, but clearly shows the equations and modifications for other Z and for balanced operation.

A mono balanced T-pad can be built with three switch-poles, just like the mono balanced L-pad you show.

 

[Aardman]

I built the Redd.47 pre-amp and utilised a Balanced attenuator (Input) design shown by John Hinson which was for a Z in of 200R.

Sorry for the off-topic but is there a link for this anywhere? I looked but couldn't see one,
Thanks

 

[schlamn]

Hi,
I am so sorry I did not answer and thank you for your replies. I had assumed this post had drifted way back into lost posterville. I just saw it by accidentally clicking the "new replies to your posts link" !!

Anyways thank you PRR and Gary O.. I did find alot of stuff under Bridge T attenuators. I have gone for the 10dB stepped half of NYD's 600R design.

I have decided to put this in front of the Input Tx of the BA6B. The same will go after the Output Tx too.. I am hoping that it will work ok.

Everything is going into the box now, so fingers crossed it wont blow up.

Aardman here is the John Hinson stepped att for the R47 if its of use still..