Question about input sensitivity?

[mr mojo]

Howdy folks,

I had a question about input sensitivity-the 7591 PP amp I built uses a stereo 500k attenutator for volume control connected via 15k resistor to the grid of the pentode section of a 6AN8 for the gain stage. Schematic lists .475v input for max volume.

My questions:

With just .475v needed for full volume, am I overdriving the amp when I use a CD player? Aren't most CD players 1v-2v output?

Next, I checked the schematic for the Scott 299C which uses a stereo 500k volume pot connected via 150k resistor to the grid of the pentode section of a 6GH8, slightly different pin-out, but very similar characteristics otherwise.

I'm thinking about replacing the 15k resistor in my amp with a 150k in hopes of avoiding overdriving the amp into clipping, but I have to admit I really have no idea if this is a problem looking for a solution or a solution looking for a problem.

Anybody got any insights? I'd sure appreciate any and all help.

Best,
mr mojo

[ray_moth]

If the 500k pot is at the input, then there's no risk of overdriving anything. 475mv is a very suitable sensitivity, IMHO.

[janneman]

Originally posted by mr mojo
Howdy folks,

I had a question about input sensitivity-the 7591 PP amp I built uses a stereo 500k attenutator for volume control connected via 15k resistor to the grid of the pentode section of a 6AN8 for the gain stage. Schematic lists .475v input for max volume.

My questions:

With just .475v needed for full volume, am I overdriving the amp when I use a CD player? Aren't most CD players 1v-2v output?

Next, I checked the schematic for the Scott 299C which uses a stereo 500k volume pot connected via 150k resistor to the grid of the pentode section of a 6GH8, slightly different pin-out, but very similar characteristics otherwise.

I'm thinking about replacing the 15k resistor in my amp with a 150k in hopes of avoiding overdriving the amp into clipping, but I have to admit I really have no idea if this is a problem looking for a solution or a solution looking for a problem.

Anybody got any insights? I'd sure appreciate any and all help.

Best,
mr mojo

Well, isn't the pot there so you can control the level going into the amp to prevent max level that split your ears, and also to prevent overdrive?

Jan Didden

[kevinkr]

Slightly off topic, but I would really recommend replacing that volume control pot with a 50K to 100K pot. High value pots exhibit fairly poor high frequency performance and depending on the miller and stray capacitance this pot is driving you might find that at certain volume settings (worst at -6dB) that your frequency response does not even make it to 20kHz.

Kevin

[Giaime]

Originally posted by kevinkr
Slightly off topic, but I would really recommend replacing that volume control pot with a 50K to 100K pot. High value pots exhibit fairly poor high frequency performance and depending on the miller and stray capacitance this pot is driving you might find that at certain volume settings (worst at -6dB) that your frequency response does not even make it to 20kHz.

Kevin

Sooooooo true Kevin, thank you for have pointed that out. Most CD players can even drive a 5k pot if that's the problem

Stepped attenuators are still the best

[Joel]

Most CD players can even drive a 5k pot if that's the problem.

Are you kidding? Have you measured the output impedance of your player? Mine is almost 1.1k ohms at 1kHz.

The concept that cd players can drive any impedance and/or capacitance is simply false, in my experience.

Joel

[Giaime]

Originally posted by Joel


Are you kidding? Have you measured the output impedance of your player? Mine is almost 1.1k ohms at 1kHz.

The concept that cd players can drive any impedance and/or capacitance is simply false, in my experience.

Joel

My NAD 514 is flat 200 ohm from 3Hz (-3dB) to 20kHz (-3dB)

But I won't feel right if it has to drive 5k. Mine was just an exhageration. 50k is good.

[mr mojo]

Folks I really appreciate the info, but I'm getting the feeling there's more to this situation than I thought...

If anyone is so inclined I could use some illumination...

Before I built my 7591 PP amp I used a Scott 299C that I restored. I didn't hear any loss of HF using a CD player into the aux input of the Scott. Didn't check with a scope or anything like that, but the Scott was my first introduction to the difference in sound quality between "black boxes" and good tube amps-in fact I still think that Scott is a fine sounding amp!

I figured since the Scott and the amp I built were so similar-comparable pentode/triode gain and cathodyne phase splitters driving 7591s-that if the Scott used a dual 500k pot for volume and the article from which I built my amp suggested a dual 500k pot for volume control, then I would be fine using a dual 500k stepped attenuator for volume control. In a lot of ways, other than plate and screen voltage on the finals, these two amps are nearly identical if you ignore the tone controls and phono pre-amp sections of the Scott.

Since the attenuator is mounted in the chassis and connected to the 6AN8s by about 3 inches of wire and the gain stage is a pentode, wouldn't the pentode minimize any Miller's capacitance effects? With short wire connections wouldn't cable capacitance be minimized as well?

What other considerations am I missing here?

Thanks again for taking the time to help a slow-learner!

Best,
mr mojo

[SY]

What's saving you here is that the input tube is a pentode. Grid to plate is 0.04pF, so Miller effect is pretty miniscule. Grid to cathode is 7 pF, which in conjunction with a 250K source (middle volume setting on a 500K pot), gives a rolloff at over 90kHz.

A smaller pot won't hurt, but it's not necessary. If you go to a triode input stage, a smaller pot will be mandatory.

[kevinkr]

Hi Sy, Mr. Mojo
Actually the thevenin equivalent resistance of a 500K pot at mid point is 125K which should give a roll off in the region of 180KHz (probably lower in practice.) given the estimated input capacitances.

The real issue is noise, and the equivalent pot noise at mid point (which is worst case) is 46nVrtHz which over a 20kHz bandwidth is about 6.5uVrms of additional noise at the input which doesn't need to be there.

Another issue, one which I can't rightly explain is that these high value pots in many instances exhibit surprisingly high levels of thd and other non-linearity as compared to a lower value. The effects of small amounts of signal level dependent grid current from the wiper to the element might be the main culprit here, but I'm not sure of all possible causes, and dc current flowing in the element will also increase the overall noise level significantly.

Channel to channel tracking is not usually that great either..

As Giaime pointed out a stepped attenuator is a good investment, and these are what I use in all of my equipment.. There are actually some quite economical ones on eBay made in Hong Kong I think.. I haven't tried these yet but plan to given the modest asking price.

Kevin

edit: fix typo