• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Choke Load for a 6922/ECC88

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi All,


I'm planning out a new build and I think I'll be using my first choke loaded gain stage. I'm going that way because I need more gain from a single stage, I want to use a very linear tube that doesn't have huge mu, and I need to use a lower than normal B+ (about 130vDC). I'm probably going to use an ECC88 or ECC99 and I'm looking for recommendations for plate chokes that would give wide bandwidth. Idle current will probably be around 10mA. Electraprint AC9037 look like the best bet for me. Are there better options in a similar price range?



Brian
 
Or try an IXYS current source, instead of a choke.

The downside of current sources (much tho' I love them) is that the operating point of the anode tends to drift quite a bit away from the design spec over time. Not so with a large-ish inductor. At 30 Hz, a 150 H choke presents an effective impedance of 2πFL = 6.28 × 30 × 150 ⇒ 28.3 kΩ. And obviously, at all higher frequencies, higher impedance above 28.3 kΩ.

The point tho' is this: say the operating point of the valve is "2 ma" thru. And say the series resistance of the choke is 3.7 kΩ. (Hammond 156-C).

Well

E = IR
E = 0.002 × 3700
E = 7.4 V
This would be the nominal voltage drop of B+, at the anode (assuming direct connect to the choke, to B+). However, the buck EMF of the choke is enormous, and it will react … at 30 Hz … just like the 28 kΩ load.

Moreover if the valve quiescent current changes, the anode voltage hardly changes at all. A few volts.

Anyway, that's the tires-on-the-road view. I've been through the whole IXYS current source thing, and when it can be tuned to working "naked, like a choke" at about ½ B+, well … its fine. Then things go whacky, and it swings pretty far from ½ B+ as the valve ages.

Just saying,
GoatGuy
 
Choke-coupled, tube voltage hardly changes.
CCS loaded, tube current hardly changes.

Which gives the "best" for all wall-outlet and tube drift, who knows?

CCS load self-adjusts (fairly well) for line voltage changes. If its very sensitive operating point is selected to float anode at say ⅔B+, then it'll do so relative to the particular B+ designed for. For instance
B+NOM = 200 V
VANODE = ⅔ ( 200 V ) = 133 V (design point)
VHEADROOM = 200 - 133 (simplistic) = 67 V … 33% of B+
… but line is 10% less …
B+ACTUAL = 180 V
VANODE = 133 V
VHEADROOM = 180 - 133 (simplistic) = 47 V … 26% of B+

Whereas the VANODE of a choke loaded stage will be what, 7 to 20 volts (depending on choke) less than B+ … and will track whatever line voltage variations percent-to-percent.

As PRR says, who's to say? You lose some headroom with the CCS on low line voltage conditions, but that's about all.

However, as the valve ages, with lower mu (in general), the quiescent current for VANODE will drop. This in turn results in CCS reducing its voltage drop, increasing the VANODE to drive more current thru tube. Again, VANODE ends up substantially drifting toward B+, reducing headroom. C'est La Vie, I guess.

By comparison, the choke load remains doggedly close to IQUIESCENT • RCHOKE voltage drop below B+, regardless of line voltage drift and valve drift. The reverse ("buck") EMF of the choke's inductance goes on to expand the headroom that the valve sees and controls almost as if by magic.

I've done amplifiers both ways, and I appreciate the “drop in and forget for the life of the amplifier, and almost without regard for people 'tube rolling' later o” aspect of choke loads.

Just saying,
GoatGuy
 
GoatGu By comparison said:
choke load[/b] remains doggedly close to IQUIESCENT • RCHOKE voltage drop below B+, regardless of line voltage drift and valve drift. The reverse ("buck") EMF of the choke's inductance goes on to expand the headroom that the valve sees and controls almost as if by magic.

I've done amplifiers both ways, and I appreciate the “drop in and forget for the life of the amplifier, and almost without regard for people 'tube rolling' later o” aspect of choke loads.

Just saying,
GoatGuy
Say GG, can you show a schematic with the type of choke load you are referring to?
 
This is what I was thinking... GoatGuy
 

Attachments

  • 190115a-choke-loaded-triode-line-stage.jpg
    190115a-choke-loaded-triode-line-stage.jpg
    71.1 KB · Views: 1,406
Chokes, cheap or otherwise, have considerations that need to be taken:

A. They need to be wound properly, or the distributed capacitance will reduce the high frequency load, high frequency distortion, and high frequency response.

B. They need to be properly oriented and have enough distance to other magnetics that are on the amplifier . . . power transformer, power supply chokes, and output transformers (or you get . . . Hum for the first two, feedback for the output transformer).

I have used both solid state CCS, and choke current sources, both work well but each have tradeoffs.

One example of a choke current source, that usually is not thought of that way, is the primary of a single ended output transformer.
It just happens to have a secondary that passes signal on to the loudspeaker.
 
Last edited:
GoatGuy,

Very interesting, a CCS in the cathode, and a choke current source load.
A good way to control the parameters around the tube.

Hi GG, thanks for the schematic. Does this scheme only apply to the ecc88, or, what other considerations are taken into account for different tubes. For example the ubiquitous 6sn7 ? I forget did you use this to drive a single ended output tube? regards, Dak

It is an excellent way to “control the parameters” in an adjustment-in-the-future free way, I've found. As the footnote on the 2015 drawing remembers, those lil' FETs (as opposed to much more precise IXYS current sources!) simply need to be measured in service.

I tend to put a 50 Ω resistor in series with a 500 Ω 10-turn trimpot … while building … and measure the voltage drop across the 50 Ω resistor as a proxy for current measuring. When I get to the desired quiescent current flow, I turn the whole thing off, wait of course, then measure the combined resistance of the trimpot and 50 Ω resistor.

Then, I go to the Resistor Bin, and pick out a suitable 2% (or convenient 5%) resistor that fills the bill. That's what gets soldered in.

Interestingly, by doing this, you can also "roll" a whole host of not-all-that-similar-but-pin-compatible tubes thru the circuit to see how they perform. Its kind of fun to swap in B329, 6189, 6201, 12AX7, ECC83, 12AY7, 12AZ7, 12BZ7, 12BH7 … and all the rest of the "lil' 9 pin glass double triode" tubes. The circuit self-adapts easily (and if you leave in the 10-turn trimpot, you can adjust operating point according to the individual tube's desires, rather easily.).

In particular, using a choke instead of an anode resistor, also frees you to a much wider choice of operating current(s) in your trial-and-see phase. The choke works at any current up to its dissipation maximum.

Have fun!
That is Rule Zero in this hobby!

Just saying,
GoatGuy
 
Last edited:
Is there a part number you recommend for this 100H 3k4 choke? Supplier?

The part was the $53/ea Electraprint AC9037, referred to by the original poster in message № 1.

However, using (my preference) the Hammond 156 C (150 H, 8 ma max, 3.7 Ω RDC) is also an excellent sounding choice. I've had great luck with it and at $15.60 qty 5 at Mouser… I usually just by the 5-at-a-time and keep 'em in stock.

GoatGuy
 
The 156C has 3.700 ohm Rdc
The quality (iron and coils) seems not so good for the purpose; at this value of H the parasitic at high frequency will be quite high so the Z will decrease rapidly with frequency due the parasitic capacitive coupling.
In my opinion, if the gain is not the problem, using a lower Rp tube can allow the use of a L with lower value (= less complicate)


Walter
 
The part was the $53/ea Electraprint AC9037, referred to by the original poster in message № 1.

However, using (my preference) the Hammond 156 C (150 H, 8 ma max, 3.7 Ω RDC) is also an excellent sounding choice. I've had great luck with it and at $15.60 qty 5 at Mouser… I usually just by the 5-at-a-time and keep 'em in stock.

GoatGuy

I am the original poster. I'm not sure how I was supposed to know what choke you were recommending (or if it was even a recommendation) from context but, thank you, I'll check out the hammond.

The 156C has 3.700 ohm Rdc
The quality (iron and coils) seems not so good for the purpose; at this value of H the parasitic at high frequency will be quite high so the Z will decrease rapidly with frequency due the parasitic capacitive coupling.
In my opinion, if the gain is not the problem, using a lower Rp tube can allow the use of a L with lower value (= less complicate)


Walter

OK! Now we're getting somewhere. I wasn't aware of this trade-off. Thank you.
 
OK! Now we're getting somewhere. I wasn't aware of this trade-off. Thank you.

From experience (having used them) I can tell you that the Hammon 156-C chokes are quite good in practice. Theoretically their interwinding capacitance would allow some HF to pass through. In practice, there is plenty of 7,500+ Hz treble in the output. Very nice, actually.

If you are even modestly worried about that though, there is a really simple solution. To put a relatively small, very low capacitance choke in front (below, in the circuit diagram) and in series with the larger choke. The lil' one has its reactance in the way-high frequency band, and passes the lower frequencies to the larger heavier, higher capacitance main choke.

In a way, this is analogous to why 0.1 µF capacitors are put in parallel with 22-to–470 µF capacitors for power supplies and bypasses: On the theory that larger capacitors have much higher internal inductance, paralleling a small, high performance capacitor allows the HF to tunnel thru it preferentially.

Same idea as the chokes in series.

Just saying,
GoatGuy
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.