300b SE powered by bench supply for Owston

[Paul Barker]

The four abreast holes forward of the valve bases, I’ll Araldite matt black mesh underneath so power resistorse can vent through them by convection. Mount the gold type resistors on blank din rails beneath the vents.



I had a test of the EL38 amp and if turned up too much as Nick predicts about se amps “it turns to mush”. The Apple of the EL38 hasn’t fallen far from the trea of the EL37. El37 is predicted by the manufacturer to suit push pull driver for push pupp output stage. So the horrendous harmonics are cancelled, probably EL 38 PP would drive SE 300b, like Stephie drove the 845 with balanced amp in front. But thats a bit complex. So I was shopping for a less distorted driver. The 6b4g has more credibility, but the curves of the PL36 as introduced at
ETF 2006 seems suitable, though by my own measurements I get 4% mostly second h. I dont think the 6b4g is much better either. Kind of simpler to have indirect heating. Only one pair of valves to go dc with. Also the build needs to be simple for my mind arms hands fingers.

I may achieve this build. Big relief being able to use the existing bench supply which can provide all the volts required up to 550mA.

We’ll see how this sounds at Owston, I have enough 300b’s to drive the 300b with a 300b, certainly I’m a little embarrassed the driver has more distortion. The driver is supposed to make the sound. So after Owston might build as above. But before Owston it needs to be an achievable goal……… which I think will show the the likelihood of a good 300b amp if its built in my own time and place, not for a deadline.

I may never build a high voltage transmitting valve for 40 watts agan. Doesnt mean I’ll sell the Tributes…. yet!

[Paul Barker]

But I’m open to offers for VT4C’s 833a’s V1505’s/4212’s/A300’s and some of my GM70’s, of which only four are copper which would be last to go. A pair of GM100’s.

[Paul Barker]

Although the 300b distortion is 2.73% the EL36 driver is 4.2% 2nd, however the 6b4g isn’t better. You have to get the 801@ out the drawer and dust it off but then you have other challenges which means you end up considering not worth it except for driving a powerful efficient and lower distortion GM 70, see the GM70 thread.

So the 300b is going to get driven with the EL36 for now.



No compromises, no Owston attendance.

Anyway the interstages just fit on top so the low distortion 801@ all be it down to 50hz only. The same 300b setting for the bench supply fits the 801@.

The e88cc srpp pair Va as usual.

Opts inside

[Paul Barker]

Maybe its better to drive 300b with 300b, with which the inductive reactance of the AE interstages are better matched, if both levels of 300b have same operating point the distortions should cancel. 2.63 2nd h drives 2.63 2nd harmonic out of phase. The ecc88 shouldn’t add much distortion.

Also found a better way to keep the Interstage transformers further away from 300b’s.

[Paul Barker]

Unfortunately the AE interstages are limited to 50mA dc and the 300b at 50mA is not any more linear than the EL36 triode connected which is 4.2%.



So the next plan, since the pl36’s only cost £15 each and I have sufficient. A pl36 srpp driver is back on the table. As the cathode of the top bottle would have lower output impedance it would drive the 300b better. The distortion may be better because of the greater load impedance than the AE transformers. With the AE’s out of the way no issue adding two bottles.

[Paul Barker]

The EL36 SRPP worst case scenario 2nd h 3.2% drive impedance 150 ohm.

[Nick]

If you wanted to use dht interstage dht 2a3 would be an obvious choice for those interstages.

[Paul Barker]

[simon]

Unfortunately the AE interstages are limited to 50mA dc and the 300b at 50mA is not any more linear than the EL36 triode connected which is 4.2%.

Is that a hard DC limit, or would it just reduce inductance if current was more than 50mA?

[Paul Barker]

It would reduce inductance which is barely enough at 50mA. But to get the same distortion as the output valve it would require 100mA. So inductive is a non starter due to the poor distortion of the driver were it a 300b on a inter trans or a LCL.

Here is the 6b4g option. The distortion is considerably better than the EL36, so imagine 4 6b4g’s in place of 4 EL36’s.



The grid resistance limit for 6b4g is 500k so coupling cap change to 0.47.

I think the srpp driver is better than inductive load output resistance of driver is 190 ohm.

The 6b4 g’s are running at 60mA and distortion negligible ( so probably wont mitigate the output tube distortion of 2.6 ) but if there is small level of distortion that will mitigate to a small factor, compared to the EL36 which had more distortion than the output valve. So I have found a better overall package and I dont count the cost of 6b4g’s as I don’t have to buy any. When I bought them they were probably £30 each. They are USA made.

[Paul Barker]

I cant be exact about the distortion of any srpp as both valves are the load for each other and that is not known in my brainbox, but its a lot bigger load than 3k. So Its a question of try it and see how it sounds.

[Paul Barker]

[Paul Barker]

At last I have mounted which valve bases that I have. Had to order 8 different octal bases that I hope will fit the holes punched originally for Jumbo 4 Pin. Always loads of patience required for things to arive. But a mental victory that I began.



The next thought though is now check the distortion if the PL36 is arranged SRPP. Which is guess work / estimate anyway. Less dc supplies required. Especially considering SRPP stereo requares 4 independent dc supplies.

I like doing the calcs anyway.

[Paul Barker]

Yes Nick I looked at that with the 6b4g but the distortion was considerably more than the output stage.

So that, and the roll off in bass with inductance limitation not favoured, I looked at the srpp, then decided to make the srpp out of el36’s for heater simplicity; hoping the distortion is tamed to the same level as the output stage. So they cancel.

Anyway, when the bases arrive I’ll hook up the last circuit diagram, and audition it. I tried to calc the distortion but too difficult.

[Paul Barker]

OK I have found a 2a3 (but substituded with 6b4g made in USA) an operating point pushing the dc current past 50mA to 54mA (the higher the current the lower the distortion) with the almost the same 2hd 3rd 4th and thd as the 300b output stage.

Paying attention to Nick’s point and Simon’s idea of slightly exceeding dc current ideal.



and just two 6b4 g’s two 300b’s on one row switched to the front should clear the interstages sufficiently.