12BH7/12BH7a subs?

Hi People,

I just changed the miniwatt 12ax7wa`s and 12at7 in the dem*t*r VTCL-2a to a matched pair of siemens ECC803S and a telefunken ECC82, wow! now the S/N ratio is 10 dB better!
The question is that the compressor uses a pair of EI 12BH7´s that I`d like to change to a matched pair of better tubes: Do you know what brand of 12BH7 tubes or substitute would be the best? I`m told that the ECC99 is a direct sub for the 12BH7, and also I remember someone telling that would be changed to a 12AU7/ECC81?

You can check the d*m*t*r VTCL-2a schemo here:

http://img.photobucket.com/albums/v321/synthi/DemeterVCTL-2ALineamp.png

http://img.photobucket.com/albums/v321/synthi/DemeterVCTL-2ASidechain.png

Thanks!

Synthi.

YYYUUUUMMMMMMMM!!!!!!! :green:

Peace,
Al.

Dang, where did you get that Amperex!

12AU7a will drop right in (base 9a as with bh7)

Or, if you want to hack up the box and get good sound, punch in a couple of 12SN7's and amuse your friends.

(bh7 is based on sn7)

The 12AU7 is not a satisfactory sub for the 12BH7 in that circuit. To put it in layman's terms, it doesn't have enough "oomph."

I'm not sure about the ECC99. Are 12BH7s that hard to come by in Europe? They're pretty common here.

The White cathode follower operates with a lot of negative feedback, so I don't think you're going to hear much difference between true 12BH7s from different manufacturers.

[quote author="CJ"]Dang, where did you get that Amperex![/quote]

Picked it up on the bay... Two for about 15 bucks, shipping included.

Oh yeah, they're NOS, not pulls.

:twisted:

Peace,
Al.

Hi synthi,
You could have a look on a E80CC really close to a 12BH7 and pretty easy to find ... in "newrope" :razz:

Fly

Well.....

Max Plate Watts:
12AU7 2.75
12BH7 3.5

Max Plate Volts:
12AU7 330
12BH7 450 (notice only 240 is used in the Demeter)

Plate Current:
12AU7 10.5 ma
12BH7 11.5 ma

Rp:
12AU7 7700
12BH7 5300

Gmho:
12AU7 2200
12BH7 3100

Amp Factor (û)
12AU7 17
12BH7 16.5

You make the call.

CJ, Build the circuit, try both tubes and get back to us. Pay special attention to how it behaves with a low-Z load. I've been down this road already, myself... Hey, it was worth a try, since AU7s are cheaper, but there ain't no free lunch.

In the output stage of my line amp, which uses a WCF not totally unlike D*m*ter's, the AU7 will fold up like a card table when asked to drive low-Z loads that the BH7 will handle with relative ease--and that's after optimizing the DC conditions for the different tube types.

Not only does the BH7 have a bigger plate, but it also has a hotter heater (600mA at 6.3V, twice the current of the 12AU7). It does make a difference when you're demanding high current from a tube, as in an output driver stage.

I grant you that if you're only ever driving "easy" loads (>10K), it may not make a huge difference. You would need to adjust the resistor values if you changed tube type, though.

I would not use the au in the fairchild with all that plate voltage, but it looks like the bh is hardly working in the demeter circuit.

i figure about 1 watt per plate, which is just loafing along.

who makes a low z input these days?

heater current don't mean squat.

I'll drop an au7 in my LA2 tonite and see what I hear.

in my la2 clon I use 12au7 in place of the 12bh7 an work fine.by now don`t have many for buy 12bh7.

who makes a low z input these days?

Click to expand...

How do we know he's not using it with old equipment with 600-ohm inputs?

heater current don't mean squat.

Click to expand...

Wrong. The heater temperature has a direct bearing on the cathode emission, which in turn determines how much cathode current can flow before reaching the limit of the cathode's ability to emit electrons. It's basic physics. If heater current didn't mean squat, then why not make all tubes with 20-microamp heaters so that we could run off AA batteries? Sheesh...

I'll drop an au7 in my LA2 tonite and see what I hear.

Click to expand...

Apples and oranges. The WCF in the LA2, besides being quite different from the one in the D*m*ter (compare resistor values and currents--I estimate about 12mA plate current in the D*m*ter), is seeing a pretty "easy" load through that stepdown transformer. Plugging an AU7 into your LA2 will only tell you if an AU7 will work in your LA2; it doesn't have much bearing on Synthi's VTCL.

Like I said before, breadboard the D*m*ter circuit as shown and try it with BH7 and AU7 into low-Z loads, and then see if you still think I'm full of shit.

Wrong.
Tubes require enough heater current to get the cathode up to the tempeture required for that particular tube. Heaters are all different. Obviously the bigger heater will most likely require more current. The au7 is a smaller tube with a phisically smaller cathode, and therefor a physically smaller heater. Therfore, it requires less current to get it up to the same cathode temp as the bh7. This is Kindergarten math.

What did you come up with for plate dissapation?

Yes, I know the higher heater current has to do with "more cathode to heat." But don't you think that "more cathode" and "more plate" count for something when a tube has to provide high peak currents? The amount of heater current is not meaningless.

My quick estimate is about 1.5W per plate in the D* circuit.

> 12AU7a will drop right in
> 12AU7 is not a satisfactory sub

An elephant is like a tree trunk. No, like a wall. No, an elephant is like a rope. Or a snake.

LOOK at the dang curves:


At a 40K load, darn near no difference.

At a 1K load, difference between 30mA and 50mA output, several dB beefier.

So yeah: you will all feel different things as you grope around the 12BH7-12AU7 animal, depending where you get a grip.

> heater current don't mean squat.

Cathode area DOES "mean squat". It says how many electrons can be easily sucked-off (current) and for practical grid construction, cathode area tells your transconductance. The 12BH7 has a bigger cathode (which needs more heat) than a 12AU7.

Thanks for the interesting reading!

Synthi.

Is that 249 cathode resistor correct? Seems a little low to me.

Trying to plot where he is on the load line.

Thanks Paul. Very diplomatic!

It looks low if you're used to thinking of the LA2 output. But the LA2 output is running fairly cold compared to what you would need to get some real "push" going.

Here's my line amp. Notice the R values on the WCF:

http://electronicdave.myhosting.net/miscimages/nydave_lineamp1.gif

Also note that the transformer is only 2:1, which means the WCF sees about a 2400-ohm load. Even going directly into 600 ohms, this circuit can do about +25dBM cleanly.

Here's a Jensen mod for the LA2. It's running the 12BH7 under similar conditions (about 15mA plate current):

http://www.jensentransformers.com/as/as091.pdf

D*m*ter is running the BH7 in about the same region as me and Jensen, which tells me he was going for good drive capability into fairly low-Z loads. That's something you don't always need, but any piece of "pro" gear should be able to do it. Plugging an AU7 in there would mean it would probably work fine with less-demanding loads but would crap out when taxed.

Besides maximum output capability, the higher gm of the BH7 means a lower output Z, which makes it a better transformer/cable/whatever's-in-the-outside-world driver.

Do you draw the load line the same in the follower circuit?

I noticed the D circuit is preloaded with a 10k resistor. Probably keeps the circuit optimized for various loads.

Loadlines are drawn with purley resistive loads. I wonder what happens if the White feeds a reactive load? (transformer)

I bet the au7 would hold up better if fed into a transformer.

> Loadlines are drawn with purely resistive loads.

Only for hasty work. RDH/RDM shows how to draw a reactive load. Pure reactance gives a circle. In most real situations we get an oval. Drawing a non-linear loadline (such as a Class 2 grid) is also easy.

Interesting thing about triodes is: if you can drive the resistance, you can probably drive any likely reactance of similar size. If you draw an ideal resistive loadline up into the knee of a pentode, any reactance is going to give elliptical trouble.

> Do you draw the load line the same in the follower circuit?

Yes, with corrections.

For a simple cathode follower, draw the loadline. Tube Plate-Cathode stream does not care which end the the load is on. But when you read the grid voltages, you have to add the cathode voltages to get the ground-referenced grid voltage. (If Mu is more than 10, you can nearly ignore this correction.)

In the WCF it does get messy. With heavy load, both tube share. So if the actual load is 1K, each tube sees 2K. However with light loads, things go off-balance because each tube not only drives the load but the opposite tube.

In fact, if loads can vary and you need maximum performance in any load, a WCF is severely conflicted. Good lo-Z performance at reasonable cathode area/cost/(heater power) needs low Mu to get the current out, but low output impedance and good hi-Z drive wants high Mu. The universal answer is to get a bigger cathode... 6C33 anyone?

My take on things is: use tube curves to estimate bias current, peak current, dissipation and a good voltage, and then stop playing with theory and BUILD that silly thing. It isn't hard to breadboard, and some good bench-abuse will prove more than all the pencils you can sharpen.

Oh... if you are stuck in a town with no 12BH7 and lots of 12AU7 or similar, double-up on 12AU7. The 12BH7 is very much like two 12AU7 in parallel. And every quasi-civilized land has a 12AU7 in wide use and readily available. You just have to figure out the ECC or Cyrillic for it.

> I bet the au7 would hold up better if fed into a transformer.

Well, some power-guys would see this debate as "a Corgi can pull better than a Yorkie". If you really wanna pull a load, get a Pyrenees or a Mastiff or even a draft-horse. There is no basic difference between a voltage amp and a power amp, but a 12AU7 (or even a 12BH7) is a very small scale power amp, just for fooling around.

This essay was written for much lower-Z loads, but gives some ranking of output power and impedance and heater power.