Home |
Search |
Today's Posts |
#1
Posted to rec.audio.tubes
|
|||
|
|||
Voltage Swing of D3a??
My friend just sent me an email for his project and he is building
some 2 stages big triode SE amp (845 or GM70 I guess) with D3a. He says the D3a wire in triode is able to swing 240V, and its operating volatge is around 165V with red LED for cathode bias. I'm wondering is this particular pendode wire in triode can swing that much voltage??? Any thoughts?? |
#2
Posted to rec.audio.tubes
|
|||
|
|||
Voltage Swing of D3a??
On Sun, 19 Feb 2012 01:59:48 -0800 (PST), King Kong
wrote: My friend just sent me an email for his project and he is building some 2 stages big triode SE amp (845 or GM70 I guess) with D3a. He says the D3a wire in triode is able to swing 240V, and its operating volatge is around 165V with red LED for cathode bias. I'm wondering is this particular pendode wire in triode can swing that much voltage??? Any thoughts?? Probably, if it's choke or transformer loaded. The 'two tube' triode SE is a popular design and the D3a is supposed to be 'super duper' in triode mode. Here's one example http://www.audiodesignguide.com/Claudio845/GM70amp.html |
#3
Posted to rec.audio.tubes
|
|||
|
|||
Voltage Swing of D3a??
On Feb 19, 8:59*pm, King Kong wrote:
My friend just sent me an email for his project and he is building some 2 stages big triode SE amp (845 or GM70 I guess) with D3a. He says the D3a wire in triode is able to swing 240V, and its operating volatge is around 165V with red LED for cathode bias. I'm wondering is this particular pendode wire in triode can swing that much voltage??? Any thoughts?? The data for pentode D3a is here.... http://www.audiodesignguide.com/Phono/D3a.pdf Its not what I would ever use to drive an 845 because that requires up to 122Vrms of more. For what I thnk about driving 845, see http://www.turneraudio.com.au/monobloc845se55.html I never try to use just two stages to avoid the use of a third tube. With 3 stages, and medium gain in each, you should find you'll get better sound than if you have just one input / driver tube which has high gain, along with complications of an IST, which I still don't yet want to embrace. I quite like choke feed to a driver stage with a build out resistance between choke and anode. This L+R dc carrying load gives an extremely high ohm value for the tube to drive so it becomes a "negligible load" above 50Hz. The added R isolates the triode anode from the phase shift and distortion generated by the choke because of irn properties and the shunt L and C at F extremes. What you want for good sound is the most intimate connections possible between input - driver - output tubes, and without muck included from the damn iron. RC coupling is excellent IMHO. When you have a DC carrying L&R network, the following grid bias R may be brought down from silly high values above 100k to less than 47k, and "reverse grid current" at idle rarely becomes a problem. Tubes develop positive grids as they age. Seems like a positive charge delvelops over time at the grid, and if Rg was say 680k like in Quad-II, you can find there is sometimes 4Vdc across the 680k, ie, grid is 4 volts more positive than the bias voltage. 6550 go the same way, hence makers suggest low Rg values, which many ppl ignore because they don't like loading the driver stage, and don't like using a larger value coupling C. Yes, you'll need a 1uF cap for pole of 3.4Hz with Rg = 47k. Its OK, try it, you'll like it. Everyone including the local barber and butcher in town knows where an EL84 or EL34 may be obtained. These sound rather well if not fantabulastic compared to some high gm iddy biddy German pentodes which the barber and butcher don't know exist, so nobody knows where ta get 'em. The other Fantabulistic tube to drive OP stages woud be the 45, OK, DH, but sound, mate, jus' sompin else. I've seen 2A3 and 300B used as drivers. MingDa used an LTP with 300B to drive 845, only 7mA if Ia per tube, and a 3 stage amp, and after re-engineering to correct Chinese Foo Pahs about basic tube use, but retaining their panash for sprectacular tube choice I'd not have thought of, WOW, plenty of dynamics and punch and detail and low noise. Things like 6C45pi belong in phono input stages. But, D3a probably also would make a very good phono stage because gm is up to 60mA/V. Its like having 50 halves of 12AX7 in parallel, so noise should be 1/7 of the noise in 12AX7 grid input. I say should. Predictions about noise are never to be taken seriously because noise performance and microphony of tubes varies, especially with anything made 50 years ago, despite the best engineering efforts of extremely keen German tube maker masters - and they WERE GOOD. Then there is the Telefunken E8020, another German marvel....... I could go on and on. Probably 50 yo NOS Gerry toobes are going to be predictably better performing than say what somebody has bought from some E-bay sale where occasionally you see 6SN7 and 6SL7 advertised cheap and NOS In Boxes. Both tube types offer marvellous music. Trouble is, many such sales do give you NOS tubes, but you find its all 1941 ex military spares stocks, and there's barely one tube in 10 that is usable because of noise and or microphony despite never having been used. I tested a batch recently, and not all were very good. I also tested a batch of 1941 6SH7 - a potentially marvellous pentode, good when triode strapped, but of 15 of those very microphonic, even though a few had less than 1uV of equivalent grid input noise. I could tell the milspec spare tubes hadn't ever been out of their boxes, because the boxes are made so that once inserted into a box, tube removal tears bits of the box to pieces. Some boxes had a rivet through the cardboard folds. Rivets had 70 years of corrosion. Some 6SH7 had rust on the bottom of the metal ring around the base or on the case, depending on whether it was a metal envelope or glass envelope with metal ringed base. Musta been stored where condensation gradually had its effects. Probably microphony is what causes signature voicing that occurs in some tube gear. Some don't mind the euphonics, ie the slight rattling of inner works by sound in the air, but I do mind, so I'm careful about what tube I use at inputs of amps. I like sound NATURAL, without artifacts added. An EL34 in triode with Ia = 25mA give max gain = 8.6 approx, Ra maybe 1k5, low enough, and with Ea at 350Vdc, and load of 47k, you can get a swing of over 200pk volts at very low THD, and this is good for driving almost anything, incuding the most difficult, 845, then others slightly easier such as GM70. 211and some others will maybe need grid current drive which means that for highest PO the grid drive becomes low impedance for a large part of the driving grid voltage. So plain CR coupling from an anode is not much good , so one has to consider a cathode follower buffer and / or maybe an IST, so that the sec provides stable biasing though a coil to the grids. There are many more ways to do class A2 or AB2 than I have time to mention here, and so far I've never bothered to persue A2 amps, or AB2 PP amps. EL86 is another tube like EL84, but triode Ra is 1k4, µ = 11, and it likes lower Ea than EL84, so its a fabulous input/driver/whatever tube IMHO. Just one makes a splendid preamp, no NFB needed, no output buffer, very simple indeed. But its no longer made. Probably far more EL86 exist and are usable than D3A. I've used EL84 in triode for headphone amps and driver stages and its quite superb. Plenty still being made. Patrick Turner. |
#4
Posted to rec.audio.tubes
|
|||
|
|||
Voltage Swing of D3a??
Patrick:
An EL34 in triode with Ia = 25mA give max gain = 8.6 approx, Ra maybe 1k5, low enough, and with Ea at 350Vdc, and load of 47k, you can get a swing of over 200pk volts at very low THD, and this is good for driving almost anything, incuding the most difficult, 845 EL34 in triode mode (Rg2=2K2, Rg3=1K) with Eb = 600V, Rl = 20K, Eg = -26V, (Ia = 18mA) can swing 300V p-p at reasonably low THD into 845, using RC coupling C=0.47uF 1000V, Rg=220K to -150V fixed bias, as I have done. Glenn. |
#5
Posted to rec.audio.tubes
|
|||
|
|||
Voltage Swing of D3a??
On Feb 22, 4:46*pm, glenbadd wrote:
Patrick: An EL34 in triode with Ia = 25mA give max gain = 8.6 approx, Ra maybe 1k5, low enough, and with Ea at 350Vdc, and load of 47k, you can get a swing of over 200pk volts at very low THD, and this is good for driving almost anything, incuding the most difficult, 845 EL34 in triode mode (Rg2=2K2, Rg3=1K) with Eb = 600V, Rl = 20K, Eg = -26V, (Ia = 18mA) can swing 300V p-p at reasonably low THD into 845, using RC coupling C=0.47uF 1000V, Rg=220K to -150V fixed bias, as I have done. Sure, I've done similar to what you suggest.Say you have 18mA in RLa = 20k, then VdcRL = 360V, and if cathode is at 0V then Ea = 240V with B+ at +600V. For EL34, this Ea is a bit low, and +350V is better, but then to get a decent Ia level in EL34, the B+ with 20k would need to be much higher. The solution is to either use a solid state CCS instead of RLa, one that will withstand the full B+ across it, or use a choke of about 70H or more with a resistance in series of at least 5k0 and with Ea at +380V and Ek at +30V, Ia can be 25mA, Vdc across 5k0 = 125V, so B+ may be 505V. The choke increases the RLa value and becomes a negligible load, and this maximises the V swing possible at low THD. My load line analysis shows that if the total RL for EL34 was 40k, then you can get 305Vpk, or 215Vrms at the anode. This is about 1.7 times what is needed to drive an 845 with grid bias of -175V. The voltage ability seems excessive, but you'd find the THD is as low as anyone could get with a single EL34 in triode with a load of 26 times Ra. A driver with a pair of EL34 in an LTP with PP 845 could also have the same operating V and I and produce amazingly low THD because most 2H dissapears, and you just get a tiny amount of 3H. Above, where I said RLa = 47k, that is the C coupled grid Rg to following 845, and there would be a choke feed to EL34 plus "build out" anode resistor. This R is importatnt because it allows the EL34 to still have good gain at extremely low F when the choke XL goes very low, and also at extreme HF where choke self capacitance also shunts Vo and causes phase shift. Patrick Turner. Glenn. |
#6
Posted to rec.audio.tubes
|
|||
|
|||
Voltage Swing of D3a??
Patrick,
On Friday, 24 February 2012 11:29:47 UTC+11, Patrick Turner wrote: ... Sure, I've done similar to what you suggest.Say you have 18mA in RLa = 20k, then VdcRL = 360V, and if cathode is at 0V then Ea = 240V with B+ at +600V. For EL34, this Ea is a bit low, and +350V is better, but then to get a decent Ia level in EL34, the B+ with 20k would need to be much higher. The solution is to either use a solid state CCS instead of RLa, one that will withstand the full B+ across it, or use a choke of about 70H or more with a resistance in series of at least 5k0 and with Ea at +380V and Ek at +30V, Ia can be 25mA, Vdc across 5k0 = 125V, so B+ may be 505V. The choke increases the RLa value and becomes a negligible load, and this maximises the V swing possible at low THD. My design goal was not to use CCS or choke or LTP. I also chose to use fixed bias rather than cathode bias. For the values I gave, there is a sweet spot for Eg of about -26 to -27 V which maximises the output swing for best 2H and 3H distortion performance, given the limited mu of triode mode. There is also a sweet spot for Rg2-A of about 2K2. Rg3-A has minimal effect on operating point and HD performance. There are other g3 connection possibilities in triode mode but I haven't had time to investigate the effect of these. The 801A DHT is almost a drop-in for this 845 driver scenario. I have a few and will be trying that next. Glenn. .... Patrick Turner. |
#7
Posted to rec.audio.tubes
|
|||
|
|||
Voltage Swing of D3a??
On Feb 27, 11:14*am, glenbadd wrote:
Patrick, On Friday, 24 February 2012 11:29:47 UTC+11, Patrick Turner *wrote: ... Sure, I've done similar to what you suggest.Say you have 18mA in RLa = 20k, then VdcRL = 360V, and if cathode is at 0V then Ea = 240V with B+ at +600V. For EL34, this Ea is a bit low, and +350V is better, but then to get a decent Ia level in EL34, the B+ with 20k would need to be much higher. The solution is to either use a solid state CCS instead of RLa, one that will withstand the full B+ across it, or use a choke of about 70H or more with a resistance in series of at least 5k0 and with Ea at +380V *and Ek at +30V, Ia can be 25mA, Vdc across 5k0 = 125V, so B+ may be 505V. The choke increases the RLa value and becomes a negligible load, and this maximises the V swing possible at low THD. My design goal was not to use CCS or choke or LTP. I also chose to use fixed bias rather than cathode bias. For the values I gave, there is a sweet spot for Eg of about -26 to -27 V which maximises the output swing for best 2H and 3H distortion performance, given the limited mu of triode mode. There is also a sweet spot for Rg2-A of about 2K2. *Rg3-A has minimal effect on operating point and HD performance. There are other g3 connection possibilities in triode mode but I haven't had time to investigate the effect of these. I don't know if EL34 in triode is better with g3 taken to cathode, or to anode. I doubt it makes any difference. In EL84 which sound fantastical in triode, g3 is internally connected to k, so there is no choice. For EL34, EL84, I use Rg2 of 220 ohms, and I like Ia at 25mA, Ea = 350V, I should get plenty to drive 13E1 with 33% CFB so up to 130Vrms is needed. Between anode and B+ of +500V, I used 4k8 plus choke which gives 63H at 47Hz. Following 47k Rg is bootstrapped to CFB winding making it a load of about 270k effective and EL34 gain = 8.79, close to its µ, and minimum THD. I'll know if I done good when I have a listen, but methinks it'll be fine. I also bypassed the cathode Rk with 1,000uF, so that Ra is now effectively *much* lower than before with only R load and no bypassing on Rk = 1k8. I think raisin' Ia and bypassin Rk and raisin RLa value must surely translate to better dynamics. I put SS on paralleled anodes of V1 input 6SL7 and raised Ia total to 4.2mA, and gain went from 44 to 64, with following Rg = 220k. It all helps IMHO. I once forgot to connect pin one ( g3) to pin 8 ( k ) in Quad-II amps in which I'd fitted EL34, Boy, sure got heaps of THD, much more than normal, had me wondrin' fo awhile, until I seen what I 'addent dunn. I guess the tubes were tryna work in some kinda wacky tetrode mode. Patrick Turner. The 801A DHT is almost a drop-in for this 845 driver scenario. I have a few and will be trying that next. Glenn. ... Patrick Turner.- Hide quoted text - - Show quoted text - |
Reply |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Forum | |||
Music ... Original CD Swing Out Sister | Marketplace | |||
Voltage swing | Vacuum Tubes | |||
HUGE LOT 450 tapes-Ebay-$10-C&W-EZLISTENING-BIGBAND-SWING | Vacuum Tubes | |||
HUGE LOT 450 tapes-Ebay-$10-C&W-EZLISTENING-BIGBAND-SWING | Tech | |||
Swing Live - DVD-A vs. SACD micing | High End Audio |