Stepped Input Attenuator Rin Value

[gg85]

You realize the position you have marked 1 is actually OFF, so there is an infinite step there; is it what you want?
Now the 39.2k resistor should be feeding the string of resistors.

Yeah I thought that my 39.2k sits in between 1 and ground. Should the 150 ohm resistor sitting between 2 and 1 on the switch actually be between 1 and ground? Essentially moving all my resistors down 1 position, and then the 39.2k between my input signal and position 24 on switch?

 

[abbey road d enfer]

Yeah I thought that my 39.2k sits in between 1 and ground. Should the 150 ohm resistor sitting between 2 and 1 on the switch actually be between 1 and ground? Essentially moving all my resistors down 1 position, and then the 39.2k between my input signal and position 24 on switch?

No. I'm too lazy for redoing the whole calculation but you want 24 positions, the last one being -50, thus the 1st one is -38.5, correct? You also want the total resistors totalling 100k, so the last resistor is 50dB smaller than 100k, or 100k/316=316ohms.
The second resistor should total 333.6 ohms, so actually be 17.6 ohms, and so on. The total resistor string (at the 24th position should be 1159 ohms (11.5 dB greater than 316 ohms), so the series resitor (between source and position 1) should be 100k-1159=98841 ohms
That's my perception, in view of the basic data.

 

[gg85]

No. I'm too lazy for redoing the whole calculation but you want 24 positions, the last one being -50, thus the 1st one is -38.5, correct? You also want the total resistors totalling 100k, so the last resistor is 50dB smaller than 100k, or 100k/316=316ohms.
The second resistor should total 333.6 ohms, so actually be 17.6 ohms, and so on. The total resistor string (at the 24th position should be 1159 ohms (11.5 dB greater than 316 ohms), so the series resitor (between source and position 1) should be 100k-1159=98841 ohms
That's my perception, in view of the basic data.

Got it, thanks so much for taking to time to spoon feed me with that one. The attenuator I'm making will be 47k but that's easy enough for me to figure out now. I had missed the step of resistance required vs resistor required to achieve that total resistance.

Back to Mouser to place an order, and I'll let you know how I go

 

[gg85]

Then you must know that the actual xfmr response will vary with the source impedance, which may necessit tuning the secondary loading. You would need to input a square wave from a generator with an impedance matching that of your intended source, and tun ethe loading for minimum overshoot, as seen on an oscilloscope.

Realised I missed this and it seems important. Does this mean with no h-pad, if I am coming out of a mic pre with a 50 ohms output impedance, that's going to drastically change the transformer's response? And by having a h-pad of any attenuation, so long as it's 600 ohms, is going to give the transformer what it needs?

So I could in theory have a 1dB h-pad (600 ohms) and keep my 47k attenuator which will load the secondaries correctly according to Lundahl's datasheet, regardless of the source impedance?

 

[abbey road d enfer]

This is becoming more and more cryptic. What do you want to build?
Attenuating signal by 60dB does not make much sense.

 

[gg85]

This is becoming more and more cryptic. What do you want to build?
Attenuating signal by 60dB does not make much sense.

Sorry for that! My goal is to make a 0.5dB input attenuator for my sta level build. I'm not 100% sure what level of attenuation I'm going to need so I am trying to understand the different ways that I can adjust the attenuation while keeping the same 0.5dB steps. The current configuration of the input is 1:8, so there is a lot of level coming in. I can wire it 1:4, which means I won't have to attenuate the signal as much, but if I've already built up my attenuator, I want to know if I can adjust the start/finish range of attenuation while keeping the correct loading on the secondary.

Along with that, Lundahl is saying that with 600-ohm source and 47k loading on secondaries, that is the ideal response for the transformer I have. So I am wondering if the use of a 600-ohm h-pad on input will mean the transformer is always seeing 600 ohms as the source, or does it not work like that? If it does mean that, then I will use a pad on input, and my attenuator will start at a higher level as I won't need to knock off as much level.

Apologies that my questions are scattered, just the more I understand each approach the better I can decide which way to go and can be sure I know what I'm doing.

It's starting to click but there are some definite gaps in my knowledge that I'm trying to fill so I really appreciate your patience and answers

 

[abbey road d enfer]

Sorry for that! My goal is to make a 0.5dB input attenuator for my sta level build. I'm not 100% sure what level of attenuation I'm going to need so I am trying to understand the different ways that I can adjust the attenuation while keeping the same 0.5dB steps. The current configuration of the input is 1:8, so there is a lot of level coming in.

I wonder why you have a step-up xfmr with such a high ratio for a line level input.

I can wire it 1:4, which means I won't have to attenuate the signal as much, but if I've already built up my attenuator, I want to know if I can adjust the start/finish range of attenuation while keeping the correct loading on the secondary.

You can adjust the start by modifying the large resistor before, but not the others and certainly not the one at the foot.

Along with that, Lundahl is saying that with 600-ohm source and 47k loading on secondaries, that is the ideal response for the transformer I have. So I am wondering if the use of a 600-ohm h-pad on input will mean the transformer is always seeing 600 ohms as the source, or does it not work like that?

It will, as long as the source impedance is also 600 ohms.

If it does mean that, then I will use a pad on input, and my attenuator will start at a higher level as I won't need to knock off as much level.

Whether you lose level via an input attenuator or an output attenuator, it seems a bit wrong to me. If you wire the xfmr as a 1:4, you'll have to analyze the consequences on loading. The optimum load at the secondary will probably be around 12k.
Now I'm not sure the loading of the xfmr matters that much. Some xfmrs are quite tolerant to loading, some are retive.

 

[radardoug]

Your idea of a .5 dB per step attenuator is not smart. On a 24 position switch you would have 12 dB of range. Thats why people use 2 dB or more steps, and higher steps at lower levels.
If your Lundahl can handle line level, then just do the pot type attenuator after it before the grids. The oldies who built these things knew the ways to do it.
Why do you think you are smarter than them? Build based on their knowledge and you wont go wrong.

 

[gg85]

Your idea of a .5 dB per step attenuator is not smart. On a 24 position switch you would have 12 dB of range. Thats why people use 2 dB or more steps, and higher steps at lower levels.
If your Lundahl can handle line level, then just do the pot type attenuator after it before the grids. The oldies who built these things knew the ways to do it.
Why do you think you are smarter than them? Build based on their knowledge and you wont go wrong.

Haha, I definitely am not smarter than them! I'd say I am at the bottom of the ranks with this kind of thing

In the schematic I'm using there is both a 20dB h-pad before the transformer and the pot before the grid. I've built mine without the h-pad and was just unsure if that was the way to go, or if it's better to have it in there and accommodate with less post transformer attenuation...or if it even matters.

The reason I want 0.5dB steps is that I want to be able to move in that accuracy. I know the original design isn't a super accurate unit, but I am wanting to see if I can mod the design to be more precise so I can use it for mixing and mastering (dual unit). I don't plan to be doing 20+ dB of compression, and I've redesigned the metering to show me what's happening from 0-10dB GR. I've been using it for a while with an attenuator built the way you mentioned, with higher steps at lower levels, and it works great. But I seem to always be around the same position on the switch and I realised I don't need that much range on it.

If I was to build the attenuator starting at too low a range, the unit wouldn't compress enough, or too high a range and it would over compress even at the lowest input position, so I have just been thinking of the best way to change that range without having to order more resistors and rebuild the switch from scratch - hence the many questions about varying the secondary loading, and whether it's best to attenuate with a h-pad before my switch.

It's all experimental for me and has been a great way to deep dive and learn how to design or redesign something...and when to leave it alone!