Fair enough Jeff. My objection is to the insistence that there is only one universal 'true". A single answer does not apply to all questions. This thread is about direct coupled amplifier implementation - not whether direct coupled is better or worse than any other, and not "what is the best single-ended amp of all time". Clearly, many people have executed a variety of successful direct-coupled amps - to their satisfaction.
The poster that you responded to, specifically expressed a desire to build from a kit. I expressed my opinion (based on my own satisfying experience with a similar fully DIY build) that the direct coupled amp he referenced, looks to me like a nice execution both electronically and physically. It looks quite manageable for a builder without a lot of experience and comes with a well executed assembly manual. I personally enjoy building from scratch and adding my own tweaks. BTW, I used a power supply implementation based on a technique that you have advocated - but that doesn't mean that everyone should want to do the same.
I don't have a lot of experience directly comparing different transformer makes, so I didn't address that issue. But the Magnequest transformers appear to have a reputation as quality units - as do many others. I would hazard a guess that jebercovici would also be able to source equally high quality OPT's in the EU for a reasonable price.
Sheldon
The poster that you responded to, specifically expressed a desire to build from a kit. I expressed my opinion (based on my own satisfying experience with a similar fully DIY build) that the direct coupled amp he referenced, looks to me like a nice execution both electronically and physically. It looks quite manageable for a builder without a lot of experience and comes with a well executed assembly manual. I personally enjoy building from scratch and adding my own tweaks. BTW, I used a power supply implementation based on a technique that you have advocated - but that doesn't mean that everyone should want to do the same.
I don't have a lot of experience directly comparing different transformer makes, so I didn't address that issue. But the Magnequest transformers appear to have a reputation as quality units - as do many others. I would hazard a guess that jebercovici would also be able to source equally high quality OPT's in the EU for a reasonable price.
Sheldon
Thanks for you replies. My drivers are 95dB efficient, but it is for a quite small room Android with music I listent, 3W I think would be sufficient.
Il could build thé amp From scratch, Just thé time for the enclosure that I think I will miss.
Jeff, I'm curious about your advice, I'll write to you.
At the beginning, I intend to build à 45 DC amp, but those tubes are quite difficult to source and expansive.
Il could build thé amp From scratch, Just thé time for the enclosure that I think I will miss.
Jeff, I'm curious about your advice, I'll write to you.
At the beginning, I intend to build à 45 DC amp, but those tubes are quite difficult to source and expansive.
A well done 45 SET amplifier can be quite special, so I can easily recommend building one despite the cost of the tubes. Just make sure you get a good set. My personal preference is Sylvania (and I've tested over 90 tubes, multiple brands). As for transformers, I've been fortunate enough to have a large number of them at my disposal, including Magnequest, ElectraPrint, Hammond, James, Hashimoto and Transcendar. In my personal experience with these, the Hashimoto is the clear winner. If you make that choice (and it's an expensive one) also get their power transformer and filter choke. you won't be disappointed. I built a pair of 45 monoblocks (all Hashimoto) over 8 years ago and they have been flawless to date. The difference between a 2-watt and 3-watt amplifier is less then 2dB, so it's unlikely you'll notice any difference at normal listening levels.
Regards, KM
Regards, KM
The yen might be at a favourable exchange rate, but you will have to check on the price to be sure.
My favourite 2a3 OPT is Tango xe-20s but they are sadly no longer in production. I have heard that Hashimoto H-203s are a superb substitute. James are a bit more than half the price but I never tried them.
Make sure if you go for magnequest that they will fit...
Another thing you might look at is getting your own chassis and using one of TubeLab's SE boards. The asano amp is wired point-to-point, so it will be more work will have greater risk of a wiring mistake...
Look at George's page here Tubelab | Dedicated to advancing the state of the art in affordable high end audio.
This would be a more reasonable way to start out and George is very kind at offering advice.
Ian
The Hashimoto H-203S is their smaller 7-watt OPT, I have these and they are excellent. However I would reference the H-20-3.5U (which I also have) as a replacement for the Tango XE-20s.
As for globe 45s, I have 6 of them... they spec perfectly fine, but I find them to be quite sensitive to mechanical excitation, which is due to the lack of internal damping supports. Normal sound levels from speakers playing in the same room will set them off a bit, kinda like sound-on-sound ;-) The other problem with the globe versions is the higher hum levels in the same circuit (with AC filaments) vs the better ST-glass versions.
Overall, I prefer the 45 over the 2A3... I have a large quantity of these (2A3) as well in several brands. Once again, I prefer the Sylvania version over most others, especially with the coil-spring tensioners. Single Plate 2A3 tubes are nice, but require DC filament supplies to make them quiet and nowhere near as robust as the later dual-plate versions which replaced them. The plates are quite thin and will red-plate easily if driven a bit too hard. I can run dual-plate 2A3 tubes at 18 watts idle and they are perfectly fine.
Regards, KM
As for globe 45s, I have 6 of them... they spec perfectly fine, but I find them to be quite sensitive to mechanical excitation, which is due to the lack of internal damping supports. Normal sound levels from speakers playing in the same room will set them off a bit, kinda like sound-on-sound ;-) The other problem with the globe versions is the higher hum levels in the same circuit (with AC filaments) vs the better ST-glass versions.
Overall, I prefer the 45 over the 2A3... I have a large quantity of these (2A3) as well in several brands. Once again, I prefer the Sylvania version over most others, especially with the coil-spring tensioners. Single Plate 2A3 tubes are nice, but require DC filament supplies to make them quiet and nowhere near as robust as the later dual-plate versions which replaced them. The plates are quite thin and will red-plate easily if driven a bit too hard. I can run dual-plate 2A3 tubes at 18 watts idle and they are perfectly fine.
Regards, KM
Thanks, KM - It must have been the H-20-3.5U that I was thinking about. I know Thomas Mayer has used them as a replacement for the Tangos.
I don't find any need to use DC filaments on 2a3's (or 45's for that matter) but I use a high quality separate heater transformer that has a center tap. Also from Thomas
The ST (shoulder top) 45's sound great in my opinion, and are far easier to obtain. Its important to note that the brand name is usually irrelevant though. Many of the manufacturer's helped fill competitor orders in those days. RCA, Raytheon, Tungsol, philco, national union etc. would all be more than adequate.
I'm running some lovely old philco's with spring filaments, but to be honest they probably don't sound any different than any of the other ST 45's I have from the 1930's through 1960's.
I think if the original poster wants to build direct coupled, he will also need to be well aware of how much current his 45's actually pass. Old, worn out globe 45's should not be considered for a direct coupled 45 amp. Again, I would re-iterate the possibility of using George's board too. Its relatively easy to do, and there's less likelyhood of a wiring mistake. George's amp is not direct coupled but I have build this kind of amp before and its pretty darn good.
I would advise looking for the ugliest looking 45's that test NEW or better.
I remember my mom had an old RCA console from the 50's. It was monophonic and used two 45's in push-pull. The design was probably from the late 1930's. It took something like 10 min to warm up and I'll never forget the beautiful sound it had.
I don't find any need to use DC filaments on 2a3's (or 45's for that matter) but I use a high quality separate heater transformer that has a center tap. Also from Thomas
The ST (shoulder top) 45's sound great in my opinion, and are far easier to obtain. Its important to note that the brand name is usually irrelevant though. Many of the manufacturer's helped fill competitor orders in those days. RCA, Raytheon, Tungsol, philco, national union etc. would all be more than adequate.
I'm running some lovely old philco's with spring filaments, but to be honest they probably don't sound any different than any of the other ST 45's I have from the 1930's through 1960's.
I think if the original poster wants to build direct coupled, he will also need to be well aware of how much current his 45's actually pass. Old, worn out globe 45's should not be considered for a direct coupled 45 amp. Again, I would re-iterate the possibility of using George's board too. Its relatively easy to do, and there's less likelyhood of a wiring mistake. George's amp is not direct coupled but I have build this kind of amp before and its pretty darn good.
I would advise looking for the ugliest looking 45's that test NEW or better.
I remember my mom had an old RCA console from the 50's. It was monophonic and used two 45's in push-pull. The design was probably from the late 1930's. It took something like 10 min to warm up and I'll never forget the beautiful sound it had.
Last edited:
Hi SM,
I've also found numerous differences in internal structure with ST 45s. Brands I have include Ken-Rad, Philco, Zenith, RCA, Sylvania, Tung-Sol and GE. There was always a lot of swapping between manufacturers as well. I have a pair of RCA 45s new in boxes with identical date codes. One is RCA made and the other an obvious Sylvania made 45. Zenith also sourced their tubes mostly from Sylvania (as they were competitors with RCA for many products). In any case, more robust Sylvania 45s have riveted supports to the micas and are noticeably better.
I also use AC on the 45 only.... and my better pairs obtain ~ 80dB S/N referenced to 1-watt output ( a couple pairs approach 90dB). I prefer AC on the dual-plate 2A3, but it simply won't work with original single-plate 2A3 tubes due to the filament structure. With the Hashimoto power transformer, I use the pair of 5.0V/6.3V windings with the 1.3V sections in series to get a center-tap. The extra 100mv or so is not a problem and the balance between the sections is very good. I use fixed DC balance (un-bypassed resistor from center-tap to ground) and an adjustable AC balance to null the hum. I use a 100-ohm potentiometer across the filament with the bypass cap from the wiper to ground. I also pad the 100-ohm pot with resistors; one from each end to the wiper to reduce the series resistance with the bypass cap and widen the notch adjustment a bit.
Regards, KM
I've also found numerous differences in internal structure with ST 45s. Brands I have include Ken-Rad, Philco, Zenith, RCA, Sylvania, Tung-Sol and GE. There was always a lot of swapping between manufacturers as well. I have a pair of RCA 45s new in boxes with identical date codes. One is RCA made and the other an obvious Sylvania made 45. Zenith also sourced their tubes mostly from Sylvania (as they were competitors with RCA for many products). In any case, more robust Sylvania 45s have riveted supports to the micas and are noticeably better.
I also use AC on the 45 only.... and my better pairs obtain ~ 80dB S/N referenced to 1-watt output ( a couple pairs approach 90dB). I prefer AC on the dual-plate 2A3, but it simply won't work with original single-plate 2A3 tubes due to the filament structure. With the Hashimoto power transformer, I use the pair of 5.0V/6.3V windings with the 1.3V sections in series to get a center-tap. The extra 100mv or so is not a problem and the balance between the sections is very good. I use fixed DC balance (un-bypassed resistor from center-tap to ground) and an adjustable AC balance to null the hum. I use a 100-ohm potentiometer across the filament with the bypass cap from the wiper to ground. I also pad the 100-ohm pot with resistors; one from each end to the wiper to reduce the series resistance with the bypass cap and widen the notch adjustment a bit.
Regards, KM
Well, much of this gets pretty subjective pretty quickly. The H-20-3.5U is a 20-watt OPT with double C core construction, while the smaller H203S is a 7-watt OPT with traditional core construction. With the larger core, expect the 20-watt OPT to have better low frequency extension and lower distortion at lower frequencies. You may possibly notice other improvements across the bandwidth as well. I also recommend using the 3.5K load for the 2A3. Run the voltages a bit higher and keep the cathode current in spec and performance is better, i.e., lower distortion, better linearity and slightly higher power, around 4.5-watts.
As for differences between the 45 and 2A3, again, it gets subjective quickly. My view is simple; the 45 is a single true triode and very linear, so one might say the "purity" of a single ended amplifier is realized. With the 2A3 (as most are double triodes in a single envelope) you have two sections that are in parallel and they will never be perfectly matched. Still, you have others who firmly believe that the 2A3 is the way to go.
In reality, the measurable differences aren't that great, assuming proper design, etc., but the 45 will generally yield slightly lower distortion and better S/N ratio with an AC filament supply. And finally, the 2A3 will yield an additional 3dB of power, or headroom if you prefer ;-)
Regards, KM
Many thanks for all those expert advices. Think I will go with H-20-3.5U as not so expensive actually with yen quite low, if it is slightly better and allow possibility to be used with 300B.
Do you think I could directly feed 2a3 SET DC amp from a 2V out DAC with a 250 or 500 k pot or step attenuaror ?
Do you think I could directly feed 2a3 SET DC amp from a 2V out DAC with a 250 or 500 k pot or step attenuaror ?
I can highly recommend the Hashimoto H-20-3.5U, excellent OPT. Driving the 2A3 to full output will require a driver stage capable of delivering 150 volts peak-to-peak. The 2A3 won't require quite this much drive (closer to 100 volts peak-to-peak), but you want the driver stage to be linear for it's required range. Also note that the Hashimoto OPTs are inverted output phase from most common SE OPTs, meaning the output from the OPT is the same phase as the input to the 2A3 grid. This is fine as you can use a two-stage input/driver which will yield non-inverting gain. My current input/driver is based on a 5814a that is DC coupled between stages and RC coupled to the 45 or 2A3. Voltage gain is about 150 volts-per-volt, or about 43.5dB. I would recommend a 100K input (pot or stepped attenuator).
Regards, KM
Another possibility for OPT would be Noguchi FM-6WS that cost much less than H-20-3.5U as it is only 6W, but is C-core with Finemet Nanocrystal.
Bandwith is a bit limited (-3dB at 20Hz, -0,5dB at 30kHz), but I listen with a 7" back loaded horn full range driver, so I guess it would be OK for bass extens.
Did some of you experiment this kind of OPT ? Especially benefit of Finemet ?
Julien
Bandwith is a bit limited (-3dB at 20Hz, -0,5dB at 30kHz), but I listen with a 7" back loaded horn full range driver, so I guess it would be OK for bass extens.
Did some of you experiment this kind of OPT ? Especially benefit of Finemet ?
Julien
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.