I have a bunch of cheap 6CB5s from an AES buy I made a couple of years ago. I also have some sets of P-P transformers in the 3k-3.5k impedance range. This sort of whispers "triode P-P" to me. The screen is normally limited to 200V in this tube, but it'll be tied to the plate so that it won't necessarily draw power it shouldn't, especially if I use a resistor from screen to plate to make sure of that. The screen support posts of these 6CB5s also have a sizable radiator tied to the support posts. I've heard this tube looks/sounds good in triode mode. Yea or nay?
Is this the thread?
http://www.diyaudio.com/forums/tubes-valves/179701-se-6cb5a-amp-6n7-driver.html
Colin.
http://www.diyaudio.com/forums/tubes-valves/179701-se-6cb5a-amp-6n7-driver.html
Colin.
Hi!
I wrote about that 6CB5A amp.
More information about amps with this tube can be found on my blog. Click on the 6CB5A label in the sidebar and you'll get all posts about this tube.
I mainly used it in SE but also built a PP (triode strapped)
Link to all 6CB5A related posts on my blog:
VinylSavor: 6CB5A
Feel free to ask any questions about it here on the forum.
best regards
Thomas
I wrote about that 6CB5A amp.
More information about amps with this tube can be found on my blog. Click on the 6CB5A label in the sidebar and you'll get all posts about this tube.
I mainly used it in SE but also built a PP (triode strapped)
Link to all 6CB5A related posts on my blog:
VinylSavor: 6CB5A
Feel free to ask any questions about it here on the forum.
best regards
Thomas
I have a set of 3.5k P-P transformers used in a PPP 6BQ5 amp (Baldwin organ), a set of Electra-Print transformers intended for something like a pair of 6B4Gs, and a couple of Amplimo toroidal output transformers with ~3k impedance. I'll probably try the 6BC5s with the Baldwin transformers, as they are suppossed to be capable of some decent output power. I will also be less worried about unbalanced primary current with the E-core transformers. The Electra-Prints will be reserved for use with some 6S4S triodes and my own take on some isolated switching supplies for the directly heated cathodes.
For the 6CB5s, I may use some kind of single ended interstage transformer or a choke-loaded driver to take advantage of the extra voltage swing afforded by that approach. The trioded 6CB5s have a rather low mu (~3.8), so they will need some volts to drive them to best advantage.
For the 6CB5s, I may use some kind of single ended interstage transformer or a choke-loaded driver to take advantage of the extra voltage swing afforded by that approach. The trioded 6CB5s have a rather low mu (~3.8), so they will need some volts to drive them to best advantage.
That's interesting. I also have some ~6k Fisher P-P iron (Fisher 500B and 500C) that would be more suitable in that case, and certainly higher quality than the Baldwin transformers. If I can indeed run the trioded 6CB5s at 400V B+, there woulldn't be much of a power penalty from running at the higher primary impedance.
More thoughts - I've been leaning toward using a trioded high transconductance pentode like the 12HG7 for the input stage (Stu would say use a D3A, but I want to make this amp completely out of cheap TV tubes to prove a point), with a SE interstage transformer to drive the trioded 6CB5A outputs. I have a pair of Fisher 500B P-P output transformers on hand that shouldn't muddy up the sound too much. The SE interstage will serve both as choke load for the input triode and as a driver transformer. This may be my introduction to winding my own interstage transformer. I have on hand what I think is a suitable bobbin. I need to get out my calipers, but it looks like it's sized for either an EI75 or EI100 lam set. Once that's decided, the next step is to set up the equations and check out the turns needed to support a ~30ma bias current, ~12kG peak flux density, and enough inductance so my lows are ok. No fancy stuff this time around - silicon steel, butt-stacked construction. Another candidate for the input tube might be the big pentode inside the 6AG9, but then I'll have an unemployed triode sulking there beside it looking for something to do...
I wouldn't. Given the 6CB5s and the 3K5 OPTs, I'd use 'em as pentode finals. A PP pair of these should get you right around 60W, and though 3K5 is a bit on the high side, it'll cost you a bit of power (and you can spare some) for lower distortion.
Being that these are TV horizontal deflection types, they'll probably sound pretty good. That's how a project that used the lower power 6BQ6GAs turned out. Even without any gNFB connected, these sounded better than expected: a bit "edgy" or "aggressive", but without the expected pentode harshness.
Anyway, that's what I'd do with 'em.
Many TV horizontal (line) sweep tubes don't like being run in triode. There is a tendency for the tube to run away at idle due to the screen grid voltage being over spec. Some sweep tubes do like triode. The 6AV5 comes to mind. I personally haven't tested the 6CB5, and my recent fascination with a certain big red board (I now have 3 running) tends to favor pentode with Schade mode. It has been stated that the 6CB5 in triode behaves much like a 300B.
One of my favorite amps was a 300B P-P amp. It used some "rusty" 6.6K OPT's and ran 60 mA per tube at 400 volts. Power output was about 25 WPC with the first 10 or so being in class A. It had a very nice punchy sound and it will be rebuilt some day. It didn't use TV tubes for the driver although the 6SN7 and 6CG7 were "TV" tubes, but they don't qualify as cheap though.
I have another project in mind for raw power involving sweep tubes (bigger ones at that). This particular one is an exercise in cheap finesse, using all triodes and low/no GNF.
I've done some back of the envelope calculations as to what I'll need for the interstage XFMR. Given that Pete Millett's musings about the Rp of the trioded 12HG7 (~1.5k) are correct, a primary inductance of ~50H seems to be in order to get the lows without straining too hard (20Hz impedance ~5X Rp). Assuming a stack-and-a-half of EI75 lams, 60ma peak current, and 12kG peak flux, I get about 5000 turns for the primary and the two secondaries (1:1:1 turns ratio). Obviously, I'll need to do a lot of interleaving to get decent bandwidth with that many turns (now do you see why interstage transformers are difficult?), The core needs about 0.3mm gap to get the right inductance. Relaxing the inductance value (downwards) or the peak flux density (upwards) will allow me to use fewer turns, but at the cost of more distortion. Using another driver tube with lower bias current (the 6EJ7 comes to mind) might allow me to also relax the transformer parameters somewhat, especially as I'd be running the tube at about 1/2 the bias current of the 12HG7. However, I'd probably have to give some of the advantage back by having to run with a somewhat higher Lp, as the 6EJ7 is likely to have a higher Rp in triode mode than the 12HG7. Also, the distortion will likely be higher. If I choose this route, the 6JC6A in triode mode might also be worth a look.
I've done some back of the envelope calculations as to what I'll need for the interstage XFMR. Given that Pete Millett's musings about the Rp of the trioded 12HG7 (~1.5k) are correct, a primary inductance of ~50H seems to be in order to get the lows without straining too hard (20Hz impedance ~5X Rp). Assuming a stack-and-a-half of EI75 lams, 60ma peak current, and 12kG peak flux, I get about 5000 turns for the primary and the two secondaries (1:1:1 turns ratio). Obviously, I'll need to do a lot of interleaving to get decent bandwidth with that many turns (now do you see why interstage transformers are difficult?), The core needs about 0.3mm gap to get the right inductance. Relaxing the inductance value (downwards) or the peak flux density (upwards) will allow me to use fewer turns, but at the cost of more distortion. Using another driver tube with lower bias current (the 6EJ7 comes to mind) might allow me to also relax the transformer parameters somewhat, especially as I'd be running the tube at about 1/2 the bias current of the 12HG7. However, I'd probably have to give some of the advantage back by having to run with a somewhat higher Lp, as the 6EJ7 is likely to have a higher Rp in triode mode than the 12HG7. Also, the distortion will likely be higher. If I choose this route, the 6JC6A in triode mode might also be worth a look.
Normally, I'd be really antsy about running the screen grid of a sweep tube over ratings. In pentode mode, this would start to be a real problem as soon as the plate swings significantly below the screen (screen becomes surrogate plate - fizzzzzz...). With the plate tied to the screen, this is less of a problem, especially with a resistor between plate and screen to help prevent significant power from being dumped in the screen. It's worth a try anyway, especially as I'm experimenting with $2 sweep tubes - not exactly the high-priced spread. I also have a bunch of cheap 6DQ5s (Raytheon) that might see some time in the sockets. I might use a certain rusty pair of transformers in the breadboard, but will switch to the Fishers when I get the kinks worked out.
However this turns out, I no doubt will be messing with the interstage transformer for quite a while, unless I cave in and get someone else like Edcor or Electra-Print to wind it. First off, I've got to get on the horn and locate a box of lams before I go anywhere with this project. I also have my 6CW5 NLD project and my "New Year's Contentious Amp" to get finished (among several others).
However this turns out, I no doubt will be messing with the interstage transformer for quite a while, unless I cave in and get someone else like Edcor or Electra-Print to wind it. First off, I've got to get on the horn and locate a box of lams before I go anywhere with this project. I also have my 6CW5 NLD project and my "New Year's Contentious Amp" to get finished (among several others).
I've done some nosing around, and it looks as if the 12HG7 is the way to go for the input/driver tube. I can choose a pentode with lower bias current, but the transconductance is lower, and the plate resistance and distortion are higher. The 12HG7 appears to give the D3A a run for its money.
My knee-jerk assumptions about the interstage transformer were off in terms of basic concepts, but when I modeled the transformer in terms of DC and AC flux, the results came out pretty close to my original assumptions. I started over with an assumption that the flux resulting from the DC bias current in the input stage/driver (30 ma) would be 5kG. I also added to this the resulting flux from a worst case square pulse of 100V, 25ms (a horribly clipped and overdriven 20Hz 1/2 sine....). This I assumed to be 10kG, for a total DC+AC flux at horrible overdrive of 15kG. The bobbins I mentioned in a previous post will support a 1" thick stack of EI75 laminations. This adds up to ~5cm^2 cross-sectional area. At any rate, for a peak flux density of 15kG and a full-on assault at 20Hz, I get 5000 primary and secondary turns, a gap of 0.38mm, and a primary inductance of 41H. This is pretty damn close to my original numbers of 5k turns and 50H primary inductance. The next step is to download some tube models and check things out assuming ~ 1-2% leakage inductance and a few guesses as to intrawinding capacitance.
Interstage transformers are no walk in the park, I may look at a paraphase drive transformer. This will be easier to design, but will require a higher B+/more power blown in plate load for the same drive amplitude.
Interstage transformers are no walk in the park, I may look at a paraphase drive transformer. This will be easier to design, but will require a higher B+/more power blown in plate load for the same drive amplitude.
hey-Hey!!!,
I share your apraisal of 12HG7/12GN7 but in practice it seems to lack 'something'. The lower gm variants like 12HL7 and 12BY7 sound better FWIW, IME...blah-blah-blah.
cheers,
Douglas
I tried it in a Citation II( and got operating points comparable to 12BY7 ), so thisone goes to straight pentode with FB. I tried it in a triode/pentode cascode active load in a 12B4 linestage. Tried using it in both positions, and in either...so as straight pentode and triode open loop for that one. They never sounded as good as the 12BY7 or 12HL7 did, even though the specs of the GN7 fit the app better IMO.
I suspect if they could have made the valve with more gm they would have, and as a design running at the limits of what's possible I think *SOMETHING* got lost. Same sort of experience with 6H30 v. 6H6...I prefer the 6H6 in all cases.
cheers,
Douglas
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.