paralleled first gain stage for noise reduction

[Tone Junkie]

Would this be a good way to go about paralleling your first gain stage for a high gain amp for noise reduction the only schematic I could find that does this is the vht deliverance.
Thanks Bill

[tubeswell]

The plate bypass resistor in that case (i.e. at the V1 position) may only make it noisier (by shunting unwanted HT noise onto the plate). Remember that the goal in this regard at V1 is to keep the noise:signal ration as low as possible.   Also the cathode resistor should be half the normal value of what it is in a single stage, in order to get the equivalent bias point.

[Tone Junkie]

So the bypass cap on the plate instead passing unwanted hash out of the signal could actuall bring it back in off of the ht winding. I ran the numbers if I figure it at 4.4k and .47 cap it came out to  71hrtz if ran it at 2.2k it came out to 153hrtz so that would be half of a really cold biased side any suggestions on what you would do to get 2 to have the power of one normal gain stage. thanks Bill
                                                                   

[terminalgs]

So the bypass cap on the plate instead passing unwanted hash out of the signal could actuall bring it back in off of the ht winding.

you could run C1 to ground instead of to HT, since that's where you ultimately want those high frequencies to go anyhow,  right?

[jjasilli]

paralleled first gain stage for noise reduction:  not getting this.  Noise reduction as compared to what?

[kagliostro]

the only schematic I could find that does this is the vht deliverance

http://www.el34world.com/charts/Schematics/files/matchless/matchless_chieftan.pdf

http://www.el34world.com/charts/Schematics/files/matchless/matchless_clubman.pdf

http://www.el34world.com/charts/Schematics/files/matchless/matchless_dc30_old.pdf

http://www.el34world.com/charts/Schematics/files/matchless/matchless_spitfire.pdf


on this tread http://www.el34world.com/Forum/index.php?topic=11610.0 this:

http://www.el34world.com/Forum/index.php?action=dlattach;topic=11610.0;attach=23501;image

http://www.el34world.com/Forum/index.php?action=dlattach;topic=11610.0;attach=23503;image

From Mr. Merlin Preamp book:

Anode resistance is halved, and other circuit resistances will usually be lower than for a single valve too, meaning reduced noise. This is the main reason for using parallel preamp valves in hifi. This also makes them most useful as an input stage in guitar amp, where the signal-to-noise ratio is very important.

Kagliostro

p.s.: Have you a link for the entire VHT schematic ? ........ thanks

[Tone Junkie]

jjasilli earlier I had made a post about reducing the noise floor on a high gain amp and have gotten many  good replies you guys always teach me alot even though this is not a high gain amp site the knowledge in all areas of amp building is second to none. There have been so many good things to work on this is just one of the answers brought up. so i asked specifics. I think I have actually found one area that will be great for me to start in from reading on another forum they said not to use the miniature alphas in a high gain amp because of noise, my noisy section started with one of those it just happened to be what i had laying around Im waiting on some good 2 watt condustive plastic pots then Im going to impliment the other things discussed in that earlier post like the 2 or 3 watt metal film resisters for plate resisters of course I will do one at a time to see which gives me the best noise floor for the amp.
 Oh lets not forget the dc referanced hum dinger thats also on the board for this amp also.
Thanks Bill                 
ps. kagliostro I emailed you what I had the person that gave it to me asked to only share it that way not put it up on the net so i must honor his request

[jjasilli]

Got it. Thanks!

[PRR]

What noise? random hiss?

For low voltage noise on an already well-designed amp, you need More Current.

The combined plate resistor should be no higher than two of the original in parallel: 50K. 47K or 39K is reasonable for 12AX7.

Leo Fender's (G.E.'s) 100K:1.5K bias is a bit cool, use 1K or less per triode, or 470 for two combined.

All of this is moot as long as you have the 68K in front. Leo's value was actually 34K (two 68K). Even that adds thermal resistance hiss to a 12AX7. Take it down to 10K, but stay away from transmitters (AM or CB).

Going from one lean 12AX7 with 68K in front, to two fat 12AX7 with <<68K, should reduce input noise *on shorted input* to half (6db).

However pickup impedance runs 10K to 200K in the hiss spectrum and may make all this insignificant.

Overall this still looks conflicted. The gain you got in the first stage may be partly wasted by R5 and whatever comes after. C1 cuts highs, but C4 seems to be there to boost highs.

I assume you already have your axe turned FULL-up when max signal is needed? Any turn-down on guitar makes amp-diddles rather silly.

hiss in 10KHz bandwidth:
68K: 2.6uV
10K: 1uV
lean 12AX7: 2uV
two fat AX7: 1uV
guitar: 2.5uV-3.5uV

> unwanted hash

Power supply should have near-zero HISS. It has some buzz, this can be reduced with power supply techniques. An AC-heated preamp will have humm; this can also be reduced. Buzz and hum are NOT reduced by paralleling triodes.

[tubeswell]

I would've commented on the 100k shared plate resistor, but the title was 'hi-gain', so I assumed the OP would've already thought about putting a 220k resistor on one triode, and was halving that already. But 47k will be even quieter.

[Tone Junkie]

See guys this is what i mean there is such a wealth of knowledge here thank you PRR you gave me some very good ways to go about what I am trying to do I cant wait to try these out. Tubeswell that front end just happened to be the only one i could find with parallel front tubes untill kagliostro posted the matchless amps so I was just putting up what I had to go off from.
 I hadnt even thought about lowering the 100k plate because I just didnt understand what they were doing untill this great explination by PRR  I understand it now, Thanks to you also tubeswell for your help. Thanks guys Bill
                                        
                                                
These still remind me of my kids(LOL)

[firemedic]

Noise cancellation is nice but I like paralleled triodes because they sound awesome.

[Fresh_Start]

As PRR pointed out, both the plate resistor and cathode resistor need to be halved when you parallel triodes to achieve the same bias point.  47K and 680 ohm would be close to the "standard" Fender 100K and 1.5K.  However, none of those Matchless circuits use a similar bias point.  The Lightning & DC-30 both use 220K and 1.5K, equivalent to 440K and 3K for a single triode.  That's a peculiar bias point.  Very popular circuits though.

FWIW, 220K over 2.7K is a nice alternative to the "standard" triode resistors.  Paralleled, that'd be 100K over 1.2K or 1.5K.

Geezer used 82K over 1.2K in his "Little Wing" circuit, corresponding to 164K over 2.4K for a single triode.  Just food for thought.

Also, putting the snubbing cap between the anode and the cathode might be the best approach, especially for this first gain stage.  With the cathode bypass cap, the cathode is an AC ground. (see Merlin's preamp book)

HTH

Chip

[Jonas]

I am working on lowering noise floor on an amp.  I am considering changing the design (see attached) to a paralleled connection for V1. 

I have heard that attaching the input shield to the plate instead of GND will reduce noise.  I think Marshall might have done this, but I could not confirm......

Could someone look at this preamp and help me with recommendations for lower noise?

[PRR]

> shield to the plate

No.

Also you show paralleled on V2 not V1.

What "noise" do you have? Shielding will not reduce hiss. Paralleling has little effect on hiss without significant circuit changes. Shotgunning old-spouse remedies is not efficient.

[Jonas]

thank you PRR, the noise is "hiss".  V1 and V2 are DC heated inducing no hum.  The amp is grounded per hoffman recommendations.

The hiss increasing as gain knob turned up.  I get that this is normal, but I want to investigate ways to minimize this hiss.

Also you are spot on!!!  :laugh:I am guilty of shotgunning old spouse remedies so my thoughts were to rewire V1 and V2 so that V1 is paralleled (gain stage 1) and put gain stages 2 and 3 on V2.  I am open to significant circuit changes if required.

Thanks!!!!!