Does using shielded grid wires really help much?

[William S.]

I have noticed that some people use shielded wires on all the grids of the preamp tubes. Does this really make much of a difference? Should the PI tube also have shielded wires on the grids? I am building a BF Super Reverb head and would like to keep it as quiet as possible.

Thanks,
William

[phsyconoodler]

I only use shielded wire on input grids.Lead dress takes care of unwanted noises.

[OlAndrew]

it can make considerable difference. Also, some recommend attaching the shield to the plate, rather than ground. Makes a very very low value cap out of the wire, helps cut RF interference and prevent parasitic oscillations. I've always just used shield to ground, myself.

[GearHeadFred]

I found it quieted my BF Super build when I used shielded wire from the input jacks to the first grids. Only ground 1 end of this -- not both -- or you can create a ground loop. I experimented with this and found that grounding the jack side was quieter.

[Tonic Amps]

BF Supers do not require shielded cable for the preamp tubes unless you have noise issues. It's quite critical on high gain amps to keep the signal leads isolated from the rest of the amp, so you'll often see shielded cable in these types of amps. Some builders use it as a matter of eliminating issues before you have them as well. If you keep the signal leads as short as possible and route them as you see in photos of old BF Supers, you should be fine. Straying too far from a known good layout and you are asking for trouble.

[jellewelagen]

I agree with TONIC. The Blackface layout is pretty smart about this. Just look at the two grid wires running to the first preamptube. The signals on these wires are opposite in phase yet they are running parallel a good stretch. What do you think is going on here? That is NFB into the grid wire of the first gain stage.

Yes that will sound different than a high capacitance shielded wire like the silverface amps have.

Take a good look at pics of Blackface leaddress. Note how wires run, and experiment.

Have fun!

Jelle

[mooreamps]

I have noticed that some people use shielded wires on all the grids of the preamp tubes. Does this really make much of a difference? Should the PI tube also have shielded wires on the grids? I am building a BF Super Reverb head and would like to keep it as quiet as possible.

Thanks,
William

Does it make a difference? Absolutely, because the input impedance to most gain stages are very high, and thus sensitive to noise ; including the PI if you are using the standard value of 1 meg on the grid resistors. This sounds like a great opportunity for you do do a real good first class job when using all shielded cable for the signal path.




-g

[HipKitty]

to answer the orignal question...yes!

[VacuumVoodoo]

it can make considerable difference. Also, some recommend attaching the shield to the plate, rather than ground. Makes a very very low value cap out of the wire, helps cut RF interference and prevent parasitic oscillations. I've always just used shield to ground, myself.

In this case cable capacitance is in parallel with the grid-plate capacitance and is effectively multiplied by the Miller effect. 10pF cable capacitance thus becomes a virtual 100-300pF and DOES affect treble response.

One should also have in mind that there will be 100 to 250V DC between shield and signal wire and many shielded cables designed for audio signal routing are not designed to withstand this. With time and elevated temperature the isolation between signal wire and shield may deteriorate with nasty crackling noises as results, and in the worst case a tube failure.
PTFE isolated shielded wire is the minimum required if you want to put the shield at high DC potential.

If you really need to tame RF/high treble with Miller capacitance use 5pF to 25pF ceramic or mica capacitor mounted directly on the tube socket and ground the cable shield in the traditional way. A compulsive tweaker may want to use a variable trimmer capacitor here and tune the high end response to his hearts delight.

[hasserl]

Good advice as usual Alex. Besides I hate the idea of applying HV to a shield that someone else down the road may reasonably be expecting to be at ground potential. Seems like a possible accident waiting to happen. May just be me, but I don't like it.

[OlAndrew]

Wasn't recommending the idea, and don't do it myself, just ground on end of the shield. It's mentioned on some of the books, so I thought I'd include it.

[mooreamps]

In this case cable capacitance is in parallel with the grid-plate capacitance and is effectively multiplied by the Miller effect. 10pF cable capacitance thus becomes a virtual 100-300pF and DOES affect treble response.

I'm not trying to imply anything here, but I'm just not seeing this one. The signal cable, lets say an a couple inches in length from the input jack to the card, less than one puff if that, and in parallel with the tube. I just don't see how that plays into the miller effect of the tube.




One should also have in mind that there will be 100 to 250V DC between shield and signal wire.

I'm not seeing this one either. That would imply there is voltage potential of 100 to 250v present on the input grid, and that just is not the case.






-g

[Swarty]

Do shielded leads help noise? Yes.
Do they noticably impact the top end? I'd think most ears would say yes.
I'd recommend using them only after standard wire was used and the noise was an issue, and then as few as possible, hopefully only on the first gain stage.
Good lead dress should really be all that is needed in a Super Reverb. Fender added a sheilded lead to the first gain stage in the early '70s to overcome inconsistent/sloppy lead dress. This along with the top end robbing suppressor caps on the output tubes really muffled the amps (they lost a lot of the openess).

FWIW, the plate connected shield was stock on lots of Marshall 2203/2204s.

[VacuumVoodoo]

I'm not trying to imply anything here, but I'm just not seeing this one. The signal cable, lets say an a couple inches in length from the input jack to the card, less than one puff if that, and in parallel with the tube. I just don't see how that plays into the miller effect of the tube.

It doesn't matter where a capacitance between grid and plate is located physically or what form or shape it has. Any external capacitance applied between grid and plate pins is in parallel with the internal grid-plate capacitance. Miller effect applies to any capacitance between grid and plate in a common cathode amplifying stage. In cathode follower or common grid stage Miller effect is negligible. It's all in the old books and many are available for download from the web.

In HF applications we used to solder a piece of sleeved wire to the grid and plate each, then twist them tighter or looser to trim HF response. Ask any old HAMster.



I'm not seeing this one either. That would imply there is voltage potential of 100 to 250v present on the input grid, and that just is not the case.
Signal wire on the grid is at 0V (usually), plate is at whatever DC operating point the stage is designed for. Cable shield connected directly to plate is at the same DC potential as the plate. Therefore there is a potential difference (otherwise known as voltage) between signal wire and shield. It's obvious, isn't it?

The answer OPs original question if shielded wire helps much is both yes and no. It can help a lot or screw thing up, it depends on how and where it is applied. Miller and Maxwell rule unthreatened.

[Jade]

Do shielded leads help noise? Yes.
Do they noticably impact the top end? I'd think most ears would say yes.
I'd recommend using them only after standard wire was used and the noise was an issue, and then as few as possible, hopefully only on the first gain stage.
Good lead dress should really be all that is needed in a Super Reverb. Fender added a sheilded lead to the first gain stage in the early '70s to overcome inconsistent/sloppy lead dress. This along with the top end robbing suppressor caps on the output tubes really muffled the amps (they lost a lot of the openess).



:agree

With proper lead dress, you shouldnt need any shielded wire in a Super.

In many cases, using shielded wire further down the signal path than the first grid can cause more problems than it fixes.

Jade

[mooreamps]

no. any cable capacitance is in parallel to the tube. The miller effect is only those components, either resistive or reactive which is in "series" with control grid.


-g

[VacuumVoodoo]

no. any cable capacitance is in parallel to the tube. The miller effect is only those components, either resistive or reactive which is in "series" with control grid.


-g

Gary, for reasons known only to yourself you seem to refuse to accept my explanations. I could be rude and say you don't have the faintest idea what Miller effect is. However, I'd rather blame myself for being unable to convey the idea clearly enough for an esteemed rocket scientist to grasp.

Perhaps you'll find Mr. Millers original analysis acceptable?

It's here:
http://web.mit.edu/klund/www/papers/jmiller.pdf

If that is too hard to swallow there's a simpler explanation courtesy of Randall Aiken:
http://www.aikenamps.com/MillerCapacitance.html

If it is still too chewy here is a very basic one with pretty colorful pictures:
http://www.zen118213.zen.co.uk/RFIC_Theory_Files/Miller_Effect.pdf
(httphttp://www.zen118213.zen.co.uk/RFIC_Theory_Files/Miller_Effect.pdf://)

[mooreamps]

yes, all very good articles. But, I hope this is something you would consider spending some time to ponder.
The miller effect is only those components, either resistive or reactive which is in "series" with control grid.


-g

[ClinchFX]

yes, all very good articles. But, I hope this is something you would consider spending some time to ponder.
The miller effect is only those components, either resistive or reactive which is in "series" with control grid.
-g

:huh

[VacuumVoodoo]

yes, all very good articles. But, I hope this is something you would consider spending some time to ponder.
The miller effect is only those components, either resistive or reactive which is in "series" with control grid.


-g

Perhaps you would be so kind as to enlighten us by providing a mathematical analysis of how Miller effect operates on a grid stopper resistor, it being in series with the grid as per your statement. A simple schematic drawing illustrating this will help. I am sure this is childishly simple for you to do.

[reaiken]

yes, all very good articles. But, I hope this is something you would consider spending some time to ponder.
The miller effect is only those components, either resistive or reactive which is in "series" with control grid.


-g


:messedup


:munch

[Fuchsaudio]

:messedup


:munch


:munch

Input grid leads from the input jacks, and the wipers of volume controls can (sometimes) benefit from a shielded cable. Alot depends on what's nearby (AC, transformers etc). In a Fender, there are some pretty long lengths on the gird leads in some models. In some amps, if the low level circuits have a short enough path, no shielding is required. Beyond the input channels, the signals get high enough for it not to matter. I've used shielded wires to and from power tube grids (in some Fenders), where there's alot of high voltage AC line or transformer wiring, to wring the last tid-bit of hum out of an amp at idle. That's really nit-picking, but Alexander is right: The effect of too much cable capcitance will suck the high end and life out of otherwise stable circuits, not requiring the shielding...

[mooreamps]

Perhaps you would be so kind as to enlighten us by providing a mathematical analysis of how Miller effect operates on a grid stopper resistor, it being in series with the grid as per your statement. A simple schematic drawing illustrating this will help. I am sure this is childishly simple for you to do.


Talk to brian.


You may post new threads
You may post replies
You may not post attachments
You may edit your posts

[Miles19]

Shielding the input wire in my Fender-style amp made a noticeable improvement in hum/buzz pickup, but my first attempt (using skinny, so high-capacitance, cable) also resulted in a noticeable drop in high frequencies. If you use a short length of reasonably low-cap cable you should be fine. Remember, you've probably got 10 times that capacitance in the cable connected to that input jack! You could shield the following gain stages as well, but I doubt you'd hear any improvement, as the signal level is way higher there, so any induced hum or buzz will have proportionally smaller effect.

And yeah, don't connect the shield to the plate. In the event of a cable failure you do NOT want 250V applied to your guitar's components or yourself!

Miles

[mooreamps]

Remember, you've probably got 10 times that capacitance in the cable connected to that input jack!

Miles

Actually, any capacitance in parallel to the input grid is in parallel with the miller effect. I notice the valve wizard has a pretty good page already on this subject ; along with another page of using a SGR resistor for the EF-86. Something I started writing about, about a year ago. I wonder what's next , someone on one of the other forums going to post a reversed engineered verison of my electronic volume control ? What a F*****G screw job this is turning out to be.




-g

[reaiken]

Actually, any capacitance in parallel to the input grid is in parallel with the miller effect.


No, it isn't. Miller capacitance works on the capacitance between the grid and plate. Cable capacitance appears between grid and ground, so it is not subject to the capacitance multiplication of the Miller effect. Therefore, the two capacitances are not in parallel unless you connect the cable shield to the plate.

The total input capacitance is the grid to plate capacitance multiplied by the gain of the stage, plus the grid to ground capacitance. If you connect the shield to the plate, then the total input capacitance is the tube's grid to plate capacitance plus the cable capacitance multiplied by the gain of the stage. As you can see, this can result in quite a bit more high-frequency rolloff.

Perhaps you should read this paper to gain a better understanding of Miller capacitance and its effect on frequency response:

http://www.aikenamps.com/MillerCapacitance.html

RA

[Fuchsaudio]

Even if the capacitance applied to the grid were part of the miller effect, the effect would be lessened by the value of the series grid resistor, which is usually between the guitar cable the the tube grid. That resistor actually uses the miller effect (to advantage) by causing a HF rolloff (at radio frequencies), albeit at a sacrifice (slightly more noise), from the resistor being present. Try measuring the input stage noise with the resistor shorted and simply sent to ground, as it is with nothing plugged into the amp.

Moore: I'd like to know more about the electronic volume control you "invented".


:munch

[reaiken]

Even if the capacitance applied to the grid were part of the miller effect, the effect would be lessened by the value of the series grid resistor, which is usually between the guitar cable the the tube grid.



True, but I was referring to the topic of this discussion, the use of a shielded cable between the input jack and the tube (not the guitar cable), with the shield connected to the plate. Most of the time the grid resistor is mounted on the input jack rather than on the grid pin where it belongs, so it doesn't come into play. In this case, the shielded cable capacitance, while small, only 5-10 pF or so, would actually get multiplied by the stage gain of 60 or so, resulting in an effective input capacitance of up to 600pF. If the grid resistor is connected to the grid at the tube, then the resistance would indeed swamp out the effect of the cable capacitance as you mentioned.

RA

[reaiken]

Moore: I'd like to know more about the electronic volume control you "invented".


:munch


10 bucks says it is a screen voltage control. Been there, done that.


RA

[Fuchsaudio]

I was referring to the post that the "guitar cable would be ten times that capacitance anyway" or some such fodder.

Randall, you referring to small signal pentodes as variable voltage amps by adjusting the screen supply ? Done-there, been-that too :) .

I think it 's very noble of Mr. Moore to share his wisdom with us.

[JacksonAmpworks]

It really does depend on the amp and the level of gain the preamp has. Your best bet is to add it and see if it makes a difference. I have had some amps that it makes noise worse, which is strange but is to be expected since most design work is 90% experimenting. I always always always shield the input jack to the first grid and in some amps go so far as to shield the phase inverter grids. Again it depends on how much gain the amp has and how quiet you need the amp to be.

Best of luck!

TIP: Try using the Shielded Teflon cable (20AWG) at ApexJr.com I've had great luck with this.

[reaiken]

Randall, you referring to small signal pentodes as variable voltage amps by adjusting the screen supply ? Done-there, been-that too :) .



No, I was referring to varying the power tube screen voltage to control the gain of the output tubes. It works, but can cause horrid crossover distortion in a push-pull amp. You have to vary the bias voltage and the PI drive at the same time, same as if you are varying the plate voltage. The nice thing is that screen current is relatively low, so you don't have to have a high-power MOSFET like you do with plate voltage attenuation. I toyed with the idea of doing this and plate voltage control in my original Invader amplifier back around 1999, but I liked the attenuator design more, it was less complicated, so I went with it instead.

RA

[reaiken]

TIP: Try using the Shielded Teflon cable (20AWG) at ApexJr.com I've had great luck with this.


I like Mogami 2330 - it is very quiet and low capacitance. The center conductor is quite small stranded wire, but when it is soldered the solder goes up into the center conductor a bit and acts like a nice strain relief because the core is so flexible.

RA

[mooreamps]

Maybe others have tried it, and had problems making it work. I don't know. Yes, I did solve certain electrical problems it presented initially, and will use it exculsively on my small signal octal builds. It's also that I don't see anywhere. But then again, I never saw a SGR on small signal pentodes before either, at least not on guitar amps like the AC-4. Maybe Mr. KOC would know more, since he seems to be the noted historian on the subject.


sorry it's off topic. BTW, I just don't recall using any shielded wire that stomped on the high end response.




-g

[JacksonAmpworks]

I like Mogami 2330 - it is very quiet and low capacitance. The center conductor is quite small stranded wire, but when it is soldered the solder goes up into the center conductor a bit and acts like a nice strain relief because the core is so flexible.

RA

Thanks for the tip Randall. I'll check that out.