RCA preamp build

[dspruill]

I am working on a copy of a RCA BA-2C preamp. My question is, the connection to the grid via the connection on the top of the tube (1620) I assume this needs to be shielded cable? I have the grid cap but I don't have the metal shields that go over the cap that I see in pictures of original units. Finally does the shield from the cable actually connect to the tube shields and then go to ground?

Thanks,
David

 

[emrr]

Yes, shielded low capacitance wire.  originals have what was sometimes referred to as 'crystal microphone wire'.  Single solid strand with a lot of insulator before you get to the shield.  Ground cable at the chassis end.  The shield caps are tough to find, and usually not necessary.  I've rarely experienced a difference, with or without.  They provide a full shield around the exposed grid cap, and contact the tube body, which is itself grounded at pin 1. 

 

[PRR]

> I assume this needs to be shielded cable?

Nah. It's your amp, you do what you want.

For your own edumacation, tack it up with plain hook-up wire. The amp will work. But if you fondle the grid wire, you hear room-buzz. If you touch the grid-cap, you get ROOM-BUZZ.

This is even useful in diagnostics. If grid won't buzz, the problem is between here and the speaker. Low-level stages buzz more than high-level; it is useful to learn the touch-buzz of each stage while the amp is healthy, so you can instantly check it when it seems ill. The absolute buzz is affected by how nasty your room is, the relative buzz in 1st 2nd 3rd.. stage is consistent.

Many 6J5 and 6J7 in high-level stages just used plain stranded un-shielded wire. Where you have 20V of signal thumping power tubes, and you don't fondle the grid wires while using the gear, the stray buzz is negligible.

In electrically quiet situations, even mike-amps will work fine without shielded grid-lead. But you run the chance that some "minor" change, such as moving the mikeamp over nearer a power supply, will raise the buzz level more than you expect.

If you sell to any fool, who may have a lamp-dimmer in the same rack, you must be pretty obsessive to reduce customer complaints. As Doug says, there were several "standard ways to do it", and often over-kill by DIY standard.

Capacitance per se is not a prime issue. The 6J7 is capable of VERY low C and very high impedance. But in most audio we'd have 500pFd crystal-mike or 100pFd of stray transformer capacitance. The 10pFd of 4 inches of plain shielded cable is not going to be a big difference. And since many mike transformers are designed for the ~~50pFd input of a triode, the <5pFd input to a pentode may be lower than optimum, and a few more pFd is not bad. (OTOH, many mikes-amps worked the first 6J7/1620 pentode in triode mode, so you should not carelessly add lots of pFd.)

You often find 6J7 wired with plain fat-rubber hookup wire, plus a thin strand of hookup wire wound around this lead, grounded at the source. It's like 50% shield, which may be 100 times better than no shield, yet the C is low and the production cost is far less than stripping a little run of shielded cable.

Build it. Use it. Does it buzz around the grid lead? If so, add the minimum shielding which makes the buzz negligible.

 

[PRR]

Oh: don't miss the 6SJ7. It is the same tube as 6J7/1620, but without the grid-cap. 1pFd more capacitance, but with transformer and tube adjacent, the short direct run may be 2pFd less C than an up-and-over grid-cap lead. The under-chassis space "should" be already shielded and quiet, so no wrapping wire or stripping braid.

6SJ7 "is" 6J7/1620, with the grid cap moved to a pin, and without the extra weed-out that 1620 got. 6SJ7 is probably the more practical Audio tube. Broadcasters stuck with top-cap 1620 because some radio circuits do benefit from low-low-low C, and stocking one good pentode for all functions saved inventory complexity.

While there is not a 162x weeded-out version of 6SJ7, costs of old-production 6SJ7 are still so low that you can buy several and do your own weeding. Try three tubes, listen for hiss, crackle, dirt. Use the best for 1st stage, the next best for 2nd stage. Unless you buy a heap of rejects, odds are that <$40 will get you two excellent tubes and a not-bad spare.

 

[emrr]

I can't remember if the 5693 is supposed to be lower noise than a stock 6SJ7, or simply more shock-proof and long life.  I'd use 6SJ7 myself if starting from scratch with a pentode design.  Or, in the BA-2, try 6C5's.  We have that rumor about which says 6C5's are simply triode connected 6J7's internally.  The one I cracked open sure looks like it's a possibility.  The few RCA 6J7/1620 preamps I have that lack a shielded grid wire do have more hum.  They all originally had external metal shields that went over the entire tube, and I've not found the correct fit so far.  I can't say if hum will go down or not, once I get a shield in place.    The hum isn't enough to injure anything, just notably higher. 

 

[tubologic]

I am working on a copy of a RCA BA-2C preamp. My question is, the connection to the grid via the connection on the top of the tube (1620) I assume this needs to be shielded cable? I have the grid cap but I don't have the metal shields that go over the cap that I see in pictures of original units. Finally does the shield from the cable actually connect to the tube shields and then go to ground?

Thanks,
David

The 1st grid of the 1st stage of a high gain preamp is the most sensitive point in the circuit and also the most prone to hum pickup. Unless you're trying to build a 60 Hz noise generator,using shielded cable at this place is an absolute  MUST.  All vintage broadcast tube equipment used shielded cable and top grid shield caps for this same (good) reason.
To avoid possible ground loops the shield must be connected only to the source side (in the preamp chassis) and not to the grid shield caps. Please follow and adhere strictly to the good old wiring techniques, RCA (and others) surely knew what they did then, and were not beginner D.I.Yers. AFAIK, nobody makes top grid shield caps today and they are quite hard to find. With some perseverance you may find them on the well-know auction site where I bought two boxes full of N.O.S shield caps a couple of years ago. Good luck with your project.

 

[PRR]

> if the 5693

Ah! Good catch!!

> is supposed to be lower noise

"For applications where extreme dependability and uniformity are paramount."

So not necessarily more shock-proof. Or lower noise. Mostly: you can put a matched-pair above the arctic circle, and not have to dog-sled back out there all winter. But of course the only way they could claim high uniformity for 10,000 hours is to be VERY CAREFUL in the making of the tubes. And since many audio troubles are just borderline "bad" tubes, using 5693 improves your odds. OTOH, 6SN7 are still cheap enough to hand-pick your tube the way you hand-pick your mike or guitar.

My 1963 Harrison catalog shows these retail prices:
6SJ7Y $1.95
1620 $7.55
5693 $6.60

The "Special Red" jobbies cost three times more than the run-of-mill bottle. That means your BEST worker, paid to work very slow and careful, long run on the pump, extended burn-in, many tests with narrow limits.

6J7 is not listed. I suspect everyone who ever maintained 6J7 gear 1937-1957 already had a closet-full of the things, and the spares market was saturated.

 

[buildafriend]

What's a fair price for the shielded grid cable? are they specific or will any shielded proper gauge cable do?

I have a mached pair of RCA 1620's that need grid caps and cables.

Now I wish I got 6sj7's!!!

Thanks for the info!

 

[letterbeacon]

I just used normal shielded cable, and that seemed to work for me.

I have a mached pair of RCA 1620's that need grid caps and cables.

Now I wish I got 6sj7's!!

Yes but the grid caps look cool!