• 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.

suggestions for driver tube for 2A3

Status
Not open for further replies.
Jump to Latest
nomiconxp

nomiconxp

Member
Joined 2009
#1
Greetings!

I bought a pair 2A3's and just finished winding my OPT's and now i want to start making the amplifier!

I am searching for a driver tube and i am thinking for the 12BH7 or 6CG7..
Does someone has experience with them and can share his findings of them or if there is something better i am open to any suggestions...
 
#2
Yes, i have. 12BH7 and ECC99 are intended for this purpose and both excellent. I personally like JJ ECC99 and would use them to drive 2A3. They show very low distortion, are reliable and will be available in the future (or similar tubes like 12BH7).
I used them for driving transformers at line level signals.
 
#5
Are you going RC coupled or are you also winding input and driver transformers? The better designs of yesteryear used push pull drivers in to transformers with push pull drive. RCA advises separate coils (bring out both wires instead of just a center tap) since separate bias adjustment is usually needed for balance. Cathode adjustment is possible too but that requires separate filament windings. To get the full power of vintage designs, the bias feeds to the drive transformer are low impedance and it's a step down transformer so you can get the added power of class A2 or AB2 on peaks without drive distortion or bias shifts.. The smaller vertical output twin triodes and similarly made tubes not marketed for high voltage are all good choices for transformer drive. 6BL7 12BH7 6CG7/6FQ7 6SN7 6GU7. There are a number dissimilar vertical triodes that will work well using the bigger half for the driver and smaller for the input stage. Some of the larger ones have capability near that of the 2A3 (6GF7A etc). If using RC coupled drive you'll get less power and blocking distortion when overdriving, but much lower currents work then with the choice being limited mostly by the miller capacitance of the output. The 6SL7 suggested was no doubt used in that mode, being low current like a 12AX7. If you want Class A2, AB2 capability with RC and direct coupling, look at the drive section of the Heathkit W6A or W6M. A 6SL7 and 6BL7 or 6SN7 will replace the 12AX7 and 12BH7 if you want the more vintage octal look.
 
#6
You might also want to consider the 5842/417A particularly if you are interested in using choke loading or interstage transformers.

Other potentially interesting choices include the D3A and C3G although in most applications the gain will likely be too high.

The Russian 4P1L DHP driving an IT in triode connection or in pentode RC coupled might be worth investigating .. I've not yet gotten around to doing something with the ones I have, but will..

If you need little gain and like American dhts the 26, 30, 31, and 112/12 might also be worth looking at.
 
#8
Currently i possess two D3A and the perfomed very nice!
I know that the gain is two much
I am thinking about CCS load. In all testing i have seen the D3A perfoms very good.

What about a cascade driver?
 
Last edited:
#11
Maybe 6AQ5's triode connected? They don't see too much love as pentodes even if they are basically a tiny 6V6. They work fine, I think the only reason I never liked them that much was burnt fingers as a kid. They get very hot. I kind of hate to see the 12HG7, 12GN7, 12HL7 tubes get used up as triodes when there are so many suitable real ignored triodes like 6GU7s just laying around. The video tubes also are designed for low screen voltages, see the derating curves. The video tubes are so good for RF projects, and if we stick a current mirror or darlington current mirror under them with a horizontal output transistor collector to the plate they'd make a wild output tube. Triode or pentode connected, the boost would track the tube curves, so it should sound like a tube. I suppose when all of the 12HG7,12GN7, 12HL7s are burned up, current mirror boosted 6CB6s could be the rage for output stages. It's one strange way to get tube curves and power too at transistor prices. Maybe a Nuvistor driver with...
 
Last edited:
#15
If you connect a 9 pin socket with 6.3v on 4&5 g1 on pin 2 k on 3 and connect 7 8 9 to a ccs running 10 - 20ma then you can get some d3a 6j9 6j52p ef184 el84 etc etc with a red led on the cathode and try them out for yourself. Most of these tubes are very cheap and worth trying. I,m doing this in a pp 12av5 amp currently and d3a and 6j9p are ahead of the pack,, 6j52p is notable for its bass though. Lift pins 7 & 8 and I think from memory you can try 6c45pi whih is very low distortion. I also like triode connected 6e5p. Different pin out though. Its great. 100% dht with 4p1l i'll def be giving a go but may mean three stage as was noted :-(
Enzo

Enzo
 
#16
Hi!

misok said:
If i use driver that cannot full-swing the power tube, am i getting just less power, or there are some more hidden consequences (like freq-related or so)? thanks..
Click to expand...

You will get less power and the driver will dominate the sound of the amp since it will contribute most to the distortion. This would be a very poor choice.

I prefer driver tubes which can provide much more voltage swing than is required by the output tube. I shoot for minimum twice better even more.

I make sure the driver tube stays in deep Class A, with their grid voltage staying below -1V for all conditions. Grid current effects and non linearities can kick in in the region between 0V and -1V already.

For that reason I also don't recommend types like D3A or 417 as driver tubes unless the output tube requires only very little voltage swing

Best regards

Thomas
 
#17
I'd agree with Thomas and add that I like to get a driver that will handle slew-limiting ( into the DHT grid ) up to about 50kHz - in this case with 2A3 that means at least 8 or 9mA in the driver. That way you get a clean and extended HF with good tone .

My personal favourite, although again this has more gain than you might need, is C3g pentode with about 8K anode load and the screen set by an OA2 to the cathode.
 
#18
Vinylsavor said:
Hi!



You will get less power and the driver will dominate the sound of the amp since it will contribute most to the distortion. This would be a very poor choice.

I prefer driver tubes which can provide much more voltage swing than is required by the output tube. I shoot for minimum twice better even more.

Thomas
Click to expand...

I look at it differently. Just because a tube is low gain doesn't mean it can't swing higher voltage.

Look at the extremes, a 12ax7 has very high gain, but when it swings the ~90 Vp-p required for the 2A3 its distortion performance isn't that great.

The Ecc99/6n6Pi has lower gan yet can swing 90Vp-p. I think I have simulated every possible single driver for the 2A3 and nothing gives lower overall distortion than a 6n6p/5687/EC99. One would think a super linear tube like the we437 would give less distortion but it doesn't (at least in simulation.) Its almost like the 5687 series was designed to match a 2A3 in 2H cancellation.

For a two stage 2A3 I think one wants a tube with a "highish" Vg. I've measured the heck out of We417's and although very linear swinging 10V's, when the cathode to grid potential gets under .5V's they tend to start squeezing out a lot of higher order distortion. Sims show the same for all the tubes of this type (triode D3a,7788, 6c45pi, etc.) I know a lot of old dht tube guys won't go near these tubes, they always told me "I just don't like their sound" 😉

With todays modern sources (DAC's) the output impedance is plenty low to drive an input stepup transformer if one wants to squeeze more power out of the 2A3. The idea is with a computer/DAC source, take advantage of thier low output impedance and digital volume control.

Another benefit to a lower input sensitivity 2A3 is the driver can be 4P1L, 26, 801, etc if an all DHT amp is desired.

I think the use of high gain drivers like the 12ax7 evolved from a time when the source had lower/weaker output. When it was discovered that the 12ax7had too little current/drive we gravitated toward the 5842-D3A and these are fine for analog sources but for digital we have opportunity to take a different direction.
 
#19
thanks for replies, fact behind my question was more-less a decision 2 vs 3 stage concept, 6N6P -> 6S1P vs 6N6P -> 6N6P -> 6S1P. I'll try the first one, then a second..

(i'd be less off-topic if i had money for 2s3/45 at the moment and didnt have bunch of 6S1Ps here)
 
#20
I've found many of the tubes mentioned work well when their grids are run more positive than -1Vgk, the 6j9p and 6c45pi are both really good well into positive bias which explains why its possible to get such large swings from them even with Va <200v. I had 0.06% @ 188v p-p from a 6j9p, admitedly at an unsustainable bias, but 0.22%/66vrms can be had @ 17ma/163v

Alot of these hi mu/gm tubes imhe also have lowering thd as current rises but as increase in 3rd 4th 5th harmonics so a pot on the ccs could let you test if you prefer low thd with more extended harmonics or vv.

I must try an Ecc99, tried 5687wb with poor results.

enzo
 
Status
Not open for further replies.
Top Bottom