Attempting a Western Electric Loftin White SE tube amplifier

[Hari Iyer]

I have been dreaming of my own DIY amplifier now for over 1+ year and that dream has not yet seen the light of the day. Three months ago i was searching the web for some published schematic and came across the Western Electric Loftin White amplifier using the 205D tubes. The schematic is as below,



Now issue with this schematic is, the 205D tube is not available and though available are replica models from Psvane and TJ Full music which themselves are 500$+ for a pair. Thats lot of money for a DIY project. Not to say about other stuff like output transformers, power transformers and chokes. More over in a class A configuration the 205D gives around 800mW of audio output which means you will need a very high sensitivity speaker to use with this amplifier. Another alternative suggested in the same article was a 45 tube with some modification for the B+1 and B+2 voltage along with the filament voltages. Now again 45 tube has also become very hard and rare to find. The 45 tube gives around 2W of audio output with the same schematic. The main feature of this schematic is the directly coupled triode without the use of any interstage transformer or a coupling capacitor which fancied me. But the availability of both the 205d and 45 tube meant that my dream amplifier project need to wait for a while.

I again started working on this schematic and was wondering if other power tubes like 2A3, PX4 or 300B tubes can be used instead of the 205D or 45 tubes and was working what modifications will be required to this schematic to make that work. As i was already having a spare 300B tubes i modified the voltages to suit the 300B. I will be using a 250V plate voltage, 30mA plate current and a grid bias of -50v. I had in haste ordered a pair of amorphous c core output transformer last month with 5K primary and 0-4-8 ohm secondary and this limited my design to use the above combination for the 300B. A 2.5K primary would have been a better option imo. I am trying to cancel the transformer order but not yet succeeded in that. As per the 300B data sheet, this combination gives around 3.2w of audio output which should be sufficient for my 93dB SPL FC speakers.

Now everything is just in paper and other than the output transformer ( i already have the 300B pairs with me) nothing has been ordered. I am not planning to order anything for the next 4 to 5 months due to supply chain perils due to the lockdown.

Any ideas / suggestions on my thought process or replacing the original tube with 300B or probably a 2A3 or PX4 are welcomed.

Thanks for looking.

 

[Hiten]

Haribhai,
from what I have read few years back about loftin white amplifier has safety issues regarding powerlines etc. I am not technically knowledgeable but while handling and playing extra precautions needs to be taken.
Regards

 

[Hari Iyer]

Haribhai,
from what I have read few years back about loftin white amplifier has safety issues regarding powerlines etc. I am not technically knowledgeable but while handling and playing extra precautions needs to be taken.
Regards

Thanks Hiten for bringing it out. I do agree on the safety issue of the above schematic too. If by any chance the driver tube B+2 voltage fails, then the output tube will melt as the grid voltage becomes 0. Or if by any chance the 10 ohm resistor or the voltage regulator fails then the full ca+ voltage is on the grid and the output tube will melt. All parameters and connections should be 100% secure and there should be no chance of any loose connections. The voltage regulator tube has a maximum limit of 50ma without any chance for any headroom. Hence to minimise risk, i am using a lower bias current of 30mA. Also the driver tube should not fail for any reason as this will make the grid voltage again 0 causing meltdown. For sure, this is a risky amplifier to built but again this is what diy is all about. I do have experience working on high voltages and will take all precaution.

Also the powersupply needs to be well tested outside the amplifier with simulated load to ensure stability. For my recommended plate voltage and current the load on the powersupply worksout to be around 8333 ohms. I am planning to use a 15 watt incandesent lamp as load and test the powersupply for 4+ hour of continous run so that i am sure that the powersupply supply that current to the plate continously without drifting. Similar test willbe made to the plate filament of the output valve so that the filament supply does not fail.

For the driver tube the plate voltage required with my configuration is around 95volts and the current is around 4mA. I will be testing that too. The voltage drop on the choke coil is calculated as 15 volts and the B+2 is around 110 volts in my case. I have to ensure this 95 volt is available at the driver plate which is directly coupled to the output grid. With the 150v voltage regulator gives me a grid voltage of -50volts. Agreed this amplifier cannot be built over-night even if you have all the components. It needs to be built and tested stage by stage before you can call it complete. I am not sure if any FMs have ventured in this space before of modifying a schematic to fit our own needs. If yes, then please pitch in with your suggestions and ideas.

 

[Hari Iyer]

Below is my modified schematic with 300B tubes.
I have deleted the schematics as there are some correction in the voltages. I shall upload tomorrow.

 

[Hari Iyer]

Uploading the modified schematic,

 

[Beast_of_burden]

Uploading the modified schematic,

This is a Direct coupled design which is unusual for a tube design I think. I frequently hear about interstage coupling being superior in tube designs, but in SS design direct coupling is often highlighted, John Curl is a big proponent. What are the benefits of direct coupling in a tube design? Is this the reason you mention earlier that this design is risky?

 

[Hari Iyer]

This is a Direct coupled design which is unusual for a tube design I think. I frequently hear about interstage coupling being superior in tube designs, but in SS design direct coupling is often highlighted, John Curl is a big proponent. What are the benefits of direct coupling in a tube design? Is this the reason you mention earlier that this design is risky?

Direct coupled design is not very unusual. Most vintage tube amplifiers were directly coupled. The benefits of less is more applies to direct coupling though i have never heard a DC design before, hence the curiosity to built one. The direct couple per se don't cause any risk imo, but how you have configured them can. For instance if you consider the original design with 205d tube, the B+1 is +500V and the plate voltage is +350v. If something goes wrong with the voltage regulator the entire +500v is presented to the plate. For the 205d the maximum permitted plate voltage is +400v. This will cause permanent damage to the tube. Also if something goes wrong with the driver tube, then the grid bias becomes -150v which can also cause the tube to fail due to high plate current and there is where the risk lies.

300B tube has much higher limits if its designed within the limits. In my modified schematic, even if the voltage regulator fails, the plate voltage becomes +400v instead of +250v which is within the maximum limit for 300B which is +450v. In case of failure of the driver tube the grid bias will be -150v here too and can cause the 300B to fail. But 300B permits higher bias current of 100mA and a higher grid bias voltage of around -105v which can still save the tube compared to a 205d or a 45 tube. So you need to weigh the risk vis-a-vis benefits and adjust the voltages accordingly. The inductor in the driver ckt also works like a constant current source for the driver tube and i am planning to use around 300H to reduce any possible power supply induced hum. This can cause the ca+ voltage some delay for the ca+ voltage to develop. I am not sure how much the delay will be and am expecting it to be around 1sec.

 

[Beast_of_burden]

Direct coupled design is not very unusual. Most vintage tube amplifiers were directly coupled. The benefits of less is more applies to direct coupling though i have never heard a DC design before, hence the curiosity to built one. The direct couple per se don't cause any risk imo, but how you have configured them can.

Interesting, my understanding is that coupling caps are used to prevent DC on the output. Newbie question, Please explain why that might not be an issue with tube design if direct coupling is used.

A 300B is not very cost effective in terms of initial investment, but apparently they can last a really long time. Sir, if you can source all the parts with the exception of tubes locally in India that will be huge.

 

[Hari Iyer]

Interesting, my understanding is that coupling caps are used to prevent DC on the output. Newbie question, Please explain why that might not be an issue with tube design if direct coupling is used.

A 300B is not very cost effective in terms of initial investment, but apparently they can last a really long time. Sir, if you can source all the parts with the exception of tubes locally in India that will be huge.

If you notice the grid bias is -55v though the CA+ is +95v. The -55 v grid bias is w.r.t. cathode voltage which is +150v. We measure -55v between grid to cathode i.e.+95v -150v = -55v. The audio signal is superimposed on this but the DC don't reach the plate as the grid is -ve biased. When the electrons flow from cathode to anode through the grid the audio signal gets modulated with the electrons and reaches the plate imo.

I have a pair of 300Bs already with me with a quad 12au7 which I shall use for testing. Unfortunately nobody in India makes audio output transformer else could have atleast used them for testing. Other parts like VR, chokes, rectifier are not very expensive and we can still afford them.

 

[adas]

If you notice the grid bias is -55v though the CA+ is +95v. The -55 v grid bias is w.r.t. cathode voltage which is +150v. We measure -55v between grid to cathode i.e.+95v -150v = -55v. The audio signal is superimposed on this but the DC don't reach the plate as the grid is -ve biased. When the electrons flow from cathode to anode through the grid the audio signal gets modulated with the electrons and reaches the plate imo.

I have a pair of 300Bs already with me with a quad 12au7 which I shall use for testing. Unfortunately nobody in India makes audio output transformer else could have atleast used them for testing. Other parts like VR, chokes, rectifier are not very expensive and we can still afford them.

There are of companies making audio transformer in India.

 

[Hari Iyer]

There are of companies making audio transformer in India.

I know of Delta and Ahuja only. You can post links of others here.

 

[var1681]

Sir as you are into tube amps ... I am looking fr Tube pre to hook up with my Pass labs First watt 7 amp and AudionNote ANJ speakers. As Audionote lower end preamp output impediance doesn’t match with input impediance of First watt and the other models are very costly... I want to discuss the options which one can explore . Would be grateful if u drop in ur number on my mobile 9815082181.. Reagrds Varinder Singh ( Amritsar )

 

[drlowmu]

5-2-2022 ............................... Educational UPDATE..........................

I am now aware of four different hi fi people owning SE DC amps, where they lost the filament power of the Driver tube.

In all four cases, there was ZERO damage done to the Finals tube !!!!!! ................. Zero.

I am one of those four. Hari Iyer, in 2021 with his 2020 KT88 SE DC amp, is another of the four. ( Finals tubes were : 2A3, 45, KT88 and 6005 )

I am only guessing that the ultra high Ohmic value, of the Final's tube Rk, protects the Finals tube if it is at zero bias voltage.

No need to guess, or apply theory, we have four actual different actual examples above, including me personally.

Jeff Medwin

BTW, IMHO, the Western Electric Loftin White circuit shown early in this thread, is not very good. It should not be duplicated. A wasted theory-only effort, not practical as a build at all !!

 

[vkalia]

(A bit late to the original post, but since the topic has come up and in case anyone else is reading)

I had built a Loftin White 2A3 SE tube amp, with Magnequest custom transformers in the 1990s.

It was absolutely magnificent and resulted in my selling off my high-end solid state system and becoming addicted to tubes - that came much, much closer to the sound of live instruments than anything else i had heard before.

I did use a bigger, beefier power supply circuit than the original schematic, FYI.

 

[drlowmu]

(A bit late to the original post, but since the topic has come up and in case anyone else is reading)

I had built a Loftin White 2A3 SE tube amp, with Magnequest custom transformers in the 1990s.

It was absolutely magnificent and resulted in my selling off my high-end solid state system and becoming addicted to tubes - that came much, much closer to the sound of live instruments than anything else i had heard before.

I did use a bigger, beefier power supply circuit than the original schematic, FYI.

Nice report vkalia !! Yes, inquisitive people will read this.

In prior years, a large percentage of DIYers in Japan have built, and use, DC two stage amps - such as of the topic here.

I have been building DIY tube amps for over four decades.

For the last sixteen years, I am totally convinced that a two stage DC tube amplifier is absolutely the best amplifier topology possible. No other topologies I have heard can inherently better it. IMHO, it is impossible to couple two tube stages better than with a length of silver wire, VS: using any capacitor or an interstage transformer.

I think amplifiers ( not speakers ) need the most improvement, and that they have been, and are, the weakest link in our entire audio chain.

I believe we absolutely should build our entire audio system - around a truly superb Directly Coupled two stage tube amp .

If you start off with really good 101 dB speakers, you give yourself the possibility to achieve the highest possible level of in-home audio performance.

Have fun !! If easy to do, PM me a schematic of your 1990's DC 2A3 amplifier. Mine used DS-025s from Magnequest. Thanks.

Jeff