Finding an old stereo tube console at the dump or 'Free' on CList would be a good start at a simple amp. What are good tube consoles to look for year brand etc?
I don't think you will fit a 2A3 amp into a '$300 max for everything' budget....but you might manage it. I TOTALLY AGREE WITH YOU!
I don't think you will fit a 2A3 amp into a '$300 max for everything' budget....but you might manage it. I TOTALLY AGREE WITH YOU!
I still recommend a 6L6CG UL Amp (Possibly switchable to Triode mode). Edcor GSXE15-8-5K transformers are about $41.70 each plus shipping.
I have used about 8 pairs of Edcor transformers and have found the GXSE15 series to be the best bang for the buck. The larger core of the 15W transformer gives lower distortion at low frequencies than the 10W version and are much better than the XSE series. Don't skimp on the transformers.
Antek AS-2T300 - 200VA 300V is $38.50 plus shipping.
Tubes are relatively inexpensive, a new pair of matched JJs will run less than $40. Russian tubes are less.
That is your major costs and a little over half your budget.
I have used about 8 pairs of Edcor transformers and have found the GXSE15 series to be the best bang for the buck. The larger core of the 15W transformer gives lower distortion at low frequencies than the 10W version and are much better than the XSE series. Don't skimp on the transformers.
Antek AS-2T300 - 200VA 300V is $38.50 plus shipping.
Tubes are relatively inexpensive, a new pair of matched JJs will run less than $40. Russian tubes are less.
That is your major costs and a little over half your budget.
ECL86 / 6GW8 is the final evolution of the Noval triode/pentode output tube for radio/TV. The pentode has more power and the triode more gain relative to the predecessor ECL82 / 6BM8. This is needed on my application because the source will be the headphone output of a iPhone, and the loudspeaker is a small Klipsch R15m. By the way, PCL86 / 14GW8 is the same tube but with different heater and is available at much lower price.
I am currently building a small amplifier based on ECL86; the schematic is adapted from the Philips technical library book "Hi-Fi circuits" by E.Rodenhuis. Output transformers have been hand-built by a local supplier at low price, the weight is about 600g each; the power transformer is recovered from a older amplifier. I'm still optimizing it and the result is already surprisingly good; it's so good that I like this amplifier better than another EL34 china-sourced amplifier I bought previously for twice the price. I choosed a very small chassis (about half the size of the Jeff one)
and I feared hum coupling from the power transformer, but with careful placement of the ground connections there is no background noise at all. This small size also enabled a orderly and easy point-to point wiring with only three 5-lug terminal strips, one for each tube socket. The only glitch until now has been a ultrasonic oscillation on the right channel triode preamp. I cured it with the 4k3 grid stoppers, a ferrite bead in the cathode, and by rerouting the output transformer primary wires as far as possible from the triode grid circuit. I apologize for the pencil schematic but I'm still perfecting it.
Comparing this schematic with Eli Duttman's one I see some differences. Since the amplifier is still on the workbench, I'm ready to explore modifications, should this improve the result.
1) The resistor between the G2 and the ultraliner tap is 1.8K. On most UL designs with ECL86 and EL84 I've found much lower values, so I put a 120 ohms there. Is there any advantage by using the 1.8K value? Why 5W?
2) I've put a 1M input resistor on the triode grid (maximum value according to the datasheet), because according to my calculations this should be the right value to get a flat response at high frequencies with a 68nF capacitor.
3) The capacitor on the global feedback network is connected in a different way. What is the advantage of the connection directly to the ground instead to the 100 ohm triode cathode resistor?
The B+ capacitors on my build are rated 450V because this is what I found in my drawer and also because I planned to exchange the power transformer in the future with another one to get the almost 300V B+ specified on the original schematic to get the full power (cathode bias on the pentode sections will need to be adjusted). But since the amplifier is already OK for my needs, I probably will leave it as is.
I have built previously another 3-tube cheap amplifier with two EL84 and one ECC83 at the same overall expense: the output transformers were identical, and I fitted cheap guitar amp tubes. Total output power with EL84 is slightly higher but I don't hear any substantial difference compared to the ECL86 one. The ECC83 does have a faint hum from the filament
supply, while the ECL86 doesn't, but it may simply be a bad tube. At this low price level and with basic circuit schematic the overall performance of the amplifier is dominated by the output transformer limitations, I guess.
I am currently building a small amplifier based on ECL86; the schematic is adapted from the Philips technical library book "Hi-Fi circuits" by E.Rodenhuis. Output transformers have been hand-built by a local supplier at low price, the weight is about 600g each; the power transformer is recovered from a older amplifier. I'm still optimizing it and the result is already surprisingly good; it's so good that I like this amplifier better than another EL34 china-sourced amplifier I bought previously for twice the price. I choosed a very small chassis (about half the size of the Jeff one)
and I feared hum coupling from the power transformer, but with careful placement of the ground connections there is no background noise at all. This small size also enabled a orderly and easy point-to point wiring with only three 5-lug terminal strips, one for each tube socket. The only glitch until now has been a ultrasonic oscillation on the right channel triode preamp. I cured it with the 4k3 grid stoppers, a ferrite bead in the cathode, and by rerouting the output transformer primary wires as far as possible from the triode grid circuit. I apologize for the pencil schematic but I'm still perfecting it.
Comparing this schematic with Eli Duttman's one I see some differences. Since the amplifier is still on the workbench, I'm ready to explore modifications, should this improve the result.
1) The resistor between the G2 and the ultraliner tap is 1.8K. On most UL designs with ECL86 and EL84 I've found much lower values, so I put a 120 ohms there. Is there any advantage by using the 1.8K value? Why 5W?
2) I've put a 1M input resistor on the triode grid (maximum value according to the datasheet), because according to my calculations this should be the right value to get a flat response at high frequencies with a 68nF capacitor.
3) The capacitor on the global feedback network is connected in a different way. What is the advantage of the connection directly to the ground instead to the 100 ohm triode cathode resistor?
The B+ capacitors on my build are rated 450V because this is what I found in my drawer and also because I planned to exchange the power transformer in the future with another one to get the almost 300V B+ specified on the original schematic to get the full power (cathode bias on the pentode sections will need to be adjusted). But since the amplifier is already OK for my needs, I probably will leave it as is.
I have built previously another 3-tube cheap amplifier with two EL84 and one ECC83 at the same overall expense: the output transformers were identical, and I fitted cheap guitar amp tubes. Total output power with EL84 is slightly higher but I don't hear any substantial difference compared to the ECL86 one. The ECC83 does have a faint hum from the filament
supply, while the ECL86 doesn't, but it may simply be a bad tube. At this low price level and with basic circuit schematic the overall performance of the amplifier is dominated by the output transformer limitations, I guess.
Attachments
How would it sound hooked up to my Cornwalls? I am pretty set on using a S.E.T. amp though.
Thanks Dan
I'm never very 'picky' if I can get my hands on a stereo tube console, at all.
I think in the US, Magnavox and RCA were pretty common. Different in Canada where Eatons and Simpsons (think Montgomery Ward and Sears) branded consoles were popular, along with Electrohome and RCA.
The stereo tube console era didn't last very long, as SS replaced tubes not very long after stereo became common. Mono consoles can provide power transformers, (and tubes) but getting a 'matched pair' of output transformers for stereo can be difficult.
Considering the low cost of decent iron from Edcor and Triad, if you want something later this winter, it would be best to buy parts rather than depend on the dump. And not many consoles have iron that's better than Edcor's 'basic' line. But if you see an old console at the curb, it's worth taking a look...
Having a bunch of old (perhaps sub-par but functional) tubes is very useful when debugging/troubleshooting/firing up for the first time.
Dan-
Do you mean that a 'triode connected' power tube won't do the trick - i.e. only a true triode?
To pcan: #3 - the cap to ground will roll off response at a few hundred KHz. Given the low cost iron, this offers a reasonable way to prevent unwanted oscillation, should any occur.
To Danconway327 - the SE84 Knockoff was indeed aimed at the Decware SE84, but with the utmost respect. Some of us on the Decware Board had asked Eli for help in designing some basic amp topologies.
I think the trioded 12ab5 will sound a little more like the SET on your speakers. At the same time, the 6BM8 is still single ended and has some 2nd order harmonic sound as well. If I had to pick, I think you'd like the 12ab5 amp. I now use a E180CC/7062 for the 12AV7 without circuit mods. This design has good sound with definite in-the-room midrange presence. The 6BM8 design is a little more laid back. I can concentrate on work with it in the background for hours st a time. The 12AB5 makes me want to pay attention to the music. I use that amp for active listening. So the 6BM8 is blue, and the 12AB5 is red. 😉
To Danconway327 - the SE84 Knockoff was indeed aimed at the Decware SE84, but with the utmost respect. Some of us on the Decware Board had asked Eli for help in designing some basic amp topologies.
I think the trioded 12ab5 will sound a little more like the SET on your speakers. At the same time, the 6BM8 is still single ended and has some 2nd order harmonic sound as well. If I had to pick, I think you'd like the 12ab5 amp. I now use a E180CC/7062 for the 12AV7 without circuit mods. This design has good sound with definite in-the-room midrange presence. The 6BM8 design is a little more laid back. I can concentrate on work with it in the background for hours st a time. The 12AB5 makes me want to pay attention to the music. I use that amp for active listening. So the 6BM8 is blue, and the 12AB5 is red. 😉
It will sound just fine. The little bit of extra damping factor, courtesy of the GNFB loop, may be just what the woofers of the "Cornwalls" need. Klipsch speakers are sensitive, but voice coil control matters, along with adequate power. The 6BM8 set up IS single ended. Remember, UL mode has been described as partial triode.
If more money was budgeted, other options would be possible. Maybe you can do the DECWARE SE84 "knockoff" within budget, but it will be touch and go. Without the linearizing help of GNFB, I'd like true full range O/P "iron" to be used. The suitable, Edcor CXSE25-5K base cost is $91.02 each. 🙁 Triode wired 6V6 family tubes, including the 12AB5, don't need the 25 W. power handling capability, but that's the only 5 Kohm primary model shown on the Edcor website. Perhaps you could get Edcor to make something that's full range and a 5 Kohm I/P, with less power handling capability, at lower cost.
TANSTAAFL will be with us, forever.
For less than 10W SE, the GXSE15-8-5K is more than adequate. I have used them in the past and am building two more amps with them now. There is no reason to use 25W iron for a less than 10 W amp.
Here is the build thread on the amps I'm building:
http://www.diyaudio.com/forums/tubes-valves/190797-another-se-amp-design-6.html
Here is the build thread on the amps I'm building:
http://www.diyaudio.com/forums/tubes-valves/190797-another-se-amp-design-6.html
Last edited:
If "Gimpy" thinks the $41.70 Edcor GXSE15-5K is adequate, go for it. That "iron" makes the triode wired 6V6 family amp a realistic $300 possibility.
Please keep in mind that the 6V6 family make very nice triodes. It has been said by people other than me that in this power range you need to go to COSTLY #45 DHTs to get better sound.
Please keep in mind that the 6V6 family make very nice triodes. It has been said by people other than me that in this power range you need to go to COSTLY #45 DHTs to get better sound.
Meaning I don't want a push pull amp. From what I have read and what I am trying to accomplish a S.E.T. amp is what I am looking for.
Eli, I think the 15w Edcor is a great transformer which allows for reasonable bass response while still giving good transient and microdynamic response. The core is large enough to support the bass without incurring the losses which effect small signal level signals.
My concern with the 25W core is loss of transient response on small level signals as the larger core imposes more losses.
A couple of points on building an SE amplifier.
1. Power supply- You need to have a very low ripple power supply for the driver stage and also fairly low ripple for the output stage as there is no cancellation (rejection) in an SE amp like there is in a push pull amp.
Your speakers are very efficient and will produce a lot of hum in the background if the PS is not very quiet.
2. SE amps require lower impedance power supplies than PP class A amps. A PP amp will have a pretty much constant current demand. A SE amp will have transients imposed on the power supply to a much greater extent. This can be handled by using a low DC resistance inductor in the PS and using a CLC (Capacitor, Inductor, Capacitor) or CLCRC (Capacitor, Inductor, Capacitor, Resistor. Capacitor) filter sequence.
My concern with the 25W core is loss of transient response on small level signals as the larger core imposes more losses.
A couple of points on building an SE amplifier.
1. Power supply- You need to have a very low ripple power supply for the driver stage and also fairly low ripple for the output stage as there is no cancellation (rejection) in an SE amp like there is in a push pull amp.
Your speakers are very efficient and will produce a lot of hum in the background if the PS is not very quiet.
2. SE amps require lower impedance power supplies than PP class A amps. A PP amp will have a pretty much constant current demand. A SE amp will have transients imposed on the power supply to a much greater extent. This can be handled by using a low DC resistance inductor in the PS and using a CLC (Capacitor, Inductor, Capacitor) or CLCRC (Capacitor, Inductor, Capacitor, Resistor. Capacitor) filter sequence.
A Triad N-68X isolation trafo feeding a SS diode "full wave" doubler meet those requirements, at modest expense. The 820 μF. caps. in the doubler stack crush the ripple fundamental. The "hash" filter and LC reservoir section finish things off. Low low noise and low impedance await your listening pleasure. 🙂 Over 100 mA. of high quality B+ is available and that's plenty.
What Eli said, plus Use fast diodes like the UF4007 or HER207. Don't use standard 1N4007 diodes as the turn off transients play havoc with your system.
I'm not the do-all know-it-all for SE amps. In fact I probably have a lot less experience than many members here.
I've built a number (ok, 5) SE amplifiers. Plus I' have bread-boarded an additional amps to see how they behaved. I build them, test them, and tear them down.
I've played around with transformers to see how the quality of the transformer translates to the sound of the system.
From what I've learned, there is a basic requirement of a certain level of performance for each component of the system.
The GXSE15 series seems to be a minimum threshold for high quality sound.
I built two SE amplifiers using the XSE series and felt that the sound was not as good as the two amps I built with the GXSE series transformers.
I've purchased several Hammond transformers but have felt that the performance was not up to the same level as the Edcor equivalents.
In a 6W SE amplifier I would tell the difference between the GXSE10 series and the GXSE15 series. The 15 always sounded better whether I was driving it with a 6P1P (12W) or a 6P41S(20W+) tube.
That said I have not tried going beyond the 15W series although I have a set of UBT1-3 Transformers which I will experiment with next for a 45 / 2A3 amp.
I've built a number (ok, 5) SE amplifiers. Plus I' have bread-boarded an additional amps to see how they behaved. I build them, test them, and tear them down.
I've played around with transformers to see how the quality of the transformer translates to the sound of the system.
From what I've learned, there is a basic requirement of a certain level of performance for each component of the system.
The GXSE15 series seems to be a minimum threshold for high quality sound.
I built two SE amplifiers using the XSE series and felt that the sound was not as good as the two amps I built with the GXSE series transformers.
I've purchased several Hammond transformers but have felt that the performance was not up to the same level as the Edcor equivalents.
In a 6W SE amplifier I would tell the difference between the GXSE10 series and the GXSE15 series. The 15 always sounded better whether I was driving it with a 6P1P (12W) or a 6P41S(20W+) tube.
That said I have not tried going beyond the 15W series although I have a set of UBT1-3 Transformers which I will experiment with next for a 45 / 2A3 amp.
Attachments
The schematic Jeff Yourison was kind enough to upload shows UF4007s. Between the "hash" filter and the ferrite beads in several spots, there really is no need to buy something more costly, like Cree high PIV Schottkys. The little bit of switching noise UF4007s generate gets killed off, before it can cause trouble.
Gimp thanks for your replies, I am pretty new to this, you and Eli posted are easy to read and mostly understood. Do you know of a schematic etc, that would work for your suggestions. I totally understand about taking care to get the proper transformers for this build, I had been looking at some other amps that used the CXSE25-10K but they are 91 each.
I really appreciate your help.
- Status
- Not open for further replies.