SE 300B Integrated Amp: Transformers

[xenithon]

Hi all. I am going to have a SE 300B integrated amplifier put together by a local amplifier builder. My technical understanding and knowledge is limited, but I have done a lot of reading to try and "catch up"

Thus far, the following has been decided on already:
- Amp will use an Aikido circuit design
- Tubes will be: 12AV7 signal tube, 12BH7A driver tube, 300B output tube
- Capacitors for the power supply section will be Angela polypropylene and oil caps
- Signal capacitors will be the Angela copper foil paper in oil signal caps
- Bypass caps will either be Angela or Black Gate
- Blue Velvet volume pot
- 4-ohm and 8-ohm speaker taps
- 1/4" headphone jack (precede with 120 ohm series resistor)

Now the decisions are about transformers (and chokes). I want to mention in advance, that we're opting not to use 2K5-3K5 as with most 300B amps, but more in the 4K-5K region - the reduced power is fine and we'd prefer the better distortion and damping factor.

We were initially intending to use the Hammond 1628SEA 5K. I'm a little hesitant though having read about possible issues with hum and frequency dips. We are considering, as an alternative at around the same price, the One Electron UBT-2 4.8K. Do you think this would be the better option?

In terms of power transformer, we initially planned on using the large Hammon 378CX 465mA. I have seen this used on a number of 300B amps such as this one. If the One Electron would be a better OPT choice, would it be better to use something like the BFT-1B power transformer from One Electron?

Finally, the amp builder is thinking of using one choke. Would one choke be ok to use or would two be needed for any reason? Also, we wanted to use the Hammond 193M Choke. Which would be the right one to use from One Electron (if I wanted to go for all One Electron components) and would I need one or two?

Thanks in advance for any assistance and suggestions!
X

[kmtang]

I would recommend to go for the Lundahl LL1620 SE OPT. You have the choice of standard or amorphous core. I have built two 45 tube SE amp with them. Surely the amorphous one is much better but the price also good indeed.

By the way, my personal opinion is that it is better off for the 300B to match with 3.0K OPT. With 5K OPT, the 300B sounds rather dull and tight.

I have two used Hammond 1627SE OPT for sale which is in good condition. They were used in my 45 tube amp which I had them replaced with LL1620. Do you want them??


Johnny

[xenithon]

Thanks for the offer, but I need to find something locally (too costly to import) - also a reason why I need to make do with what is available to the builder, which includes Hammond and One Electron, and a few other unknowns

Interesting to read what you say about the 300B with 5K OPT's . I read that with the higher K transformers yield less power but with better distortion and damping factor performance. I do have the option of using the One Electron UBT-3, which is 3K.

[Tweeker]

I would go with the UBT over the Hammond.

12AV7 would not be my first choice in linear amplification service, especially without feedback. It also might be short gain in an Aikido(~18).

As to 3k v 4.8k, operating point and speakers play a large role here. Some speakers may like more 2hd, and even net less distortion overall. Some listeners and/or sources may also like more 2hd for that matter.

Is this a cathode biased or fixed bias amplifier?

The One Electron power transformer has multiple primary and secondary taps if you want to play with the operating point. B+ might run rather high with the Hammond unless the power supply is using choke input, or cathode bias is used.

Assuming a capacitor input supply, consider splitting it with 2 193B chokes after input capacitor for quasi dual mono supply.

[xenithon]

Hi there and thanks for the reply.

My primary listening "speakers" aren't actually speakers - I do all my listening with the AKG K1000 earspeakers, which were made to be run of a proper "hi fi" amplifier via speaker taps. They are rated 120-ohm.

This will be a cathode biased amp and will use solid state rectification. I must admit my understanding of everything is limited so I am not really sure what to select. The majority of 300B SE amp using 2K5 to 3K5 OPT's, but I have seen a few using higher impedance OPT's and [url=http://www.tubelab.com/AssemblyManualTubelabSE/TubesAndTransformers_TSE.htm]this page here[/urk] which talks about matching tubes to OPT's mentions:

Most 300B amplifiers operate at 350 to 370 volts and use a 2.5 K, 3 K or 3.6 K ohm load. This is the sweet spot for most 300B tubes, and provides about 7 to 9 watts. It is also possible to use a 5K ohm load for reduced power with better distortion and damping factor. 6 or 7 watts will result. Most users go for a 3K ohm OPT and about 360 volts of B+. Some also use a 5K ohm OPT and a B+ nearing 400 volts, this is my preference.

It also mentions that the 300B tubes should be biased at 60 to 75 mA. The One Electron spec sheets for their OPT's include measured operating points with different tubes. With the 300B, the UBT-2 (4K8) says 400V plate voltage, 65mA plate current; the UBT-3 (3K) says 325V plate voltage, 75mA plate current. As to how to differentiate / pick between them though - I am a little lost

Also regarding inductance, I see the Hammond 1628SEA (5K) is rated 48H but the One Electron 4K8 OPT is rated only 29H - is this anything to worry about?

For the choke, we were thinking of using a single Hammond 193M, 10H 300mA choke.

With regards,
X

[xenithon]

Also just to add - the 12BH7A would be the driver, the 12AV7 the input tube. What would you suggest as an alternative to the 12AV7?

[Tweeker]

Cathode bias allows 300Bs to be biased up to 100ma. Distortion (especially 3hd) tends to drop with increasing current, though bear in mind life will be shortened at high dissipation. Datasheet has some examples of operating points and general performance trends.

The lower inductance of the UBT-2 should not be a problem, LF cutoff should still be plenty low. This is the inductance at 110ma, it will be higher at your operating point.

Id suggest 5751 for input. Im assuming the 12BH7 is operating as Aikido cathode follower.

The 193M would work fine.

SS full wave rectification and Hammond 378CX with capacitor input will put you in the neighborhood 450V (550V B+) across the 300B, rather high.

Choke input with 372JX and SS bridge rectification could work as one alternative to the BFT-1.

The amplifier you linked in first post is unusual in that it is operating between choke and capacitor input by using a very small tuned input capacitor. This gives a B+ in between the two methods (approx VRMS *.9 for choke input, VRMS*1.4 for capacitor).

The Eyuda 80uF motor runs are a great bargain for psu caps, even with international shipping (assuming you dont get reamed on customs or by UPS broker etc) Am quite happy with mine.

[Tweeker]

It wont be a problem with phones, but even the 5751 (mu=70, 12AV7 mu=40) is marginal on gain here in Aikido (gain=.5mu-CF losses). You may need to resort to 12AX7s (mu=100) depending on source.

[Sheldon]

Originally posted by xenithon
My primary listening "speakers" aren't actually speakers - I do all my listening with the AKG K1000 earspeakers, which were made to be run of a proper "hi fi" amplifier via speaker taps. They are rated 120-ohm.

[i]Most 300B amplifiers operate at 350 to 370 volts and use a 2.5 K, 3 K or 3.6 K ohm load. This is the sweet spot for most 300B tubes, and provides about 7 to 9 watts. It is also possible to use a 5K ohm load for reduced power with better distortion and damping factor. 6 or 7 watts will result

Keep in mind that you are not listening to the input tube, output tube, output transformer, or even the amplifier. You are listening to the entire system. Therefore, we need to take general statements about components, etc., with a large grain of salt. For instance, I assume your amp is using grounded cathode input section. This inverts the phase, and to some extent, cancels 2nd harmonics of the output section. So, saying that one tube has low distortion, doesn't mean the final figures are lower, in fact they might be higher. And how that sounds to you is a matter of your taste.

If you are driving 120 Ohm headphones, the loadline seen by the amp will be much higher than the OPT specification. If the OPT is spec'd at 3k, that means the tube sees a reflected load of 3k (square of the turns ratio, times the speaker load) with an 8 Ohm speaker load. With a 120 Ohm load, the reflected load will be 45k (it's not actually that high, due to transformer imperfections, but you see the point). With headphones, you can run various values of power resistor in parallel with your phones to experiment with the load. If you are building for the first time, look for OPT's with multiple taps. Running an 8 Ohm load on 4 Ohm taps, effectively doubles the load seen by the tube.

Sheldon

edit: forgot you were using an Aikido input stage, so the comments on the input stage might be a little confusing. But the Aikido amp is inverting, so the point remains.

[xenithon]

I'd suggest 5751 for input. Im assuming the 12BH7 is operating as Aikido cathode follower.

Yes; the 12AV7 would have been the signal/input tube; the 12BH7 the driver tube for the 300B I took a look at this page which has a great matrix for all various tubes. I also see he has a useful way to calculate final gain.

With the 12AV7/12BH7 I get +/- 18 in total. I have no idea what number I should be aiming for though . I see that with the 5751/12BH7 I get ~28; which I assume is better for an integrated amp?

For the power transformer we would like to use the TBFT-1 from One Electron, to keep continuity with the UBT-2 OPT's. Would that be a better option to the 378CX?

For chokes, I believe I can either use a single 193M - simplest and most cost effective. Alternatively - more expensive but perhaps better, I would use something like two OE PRC-2's which are 12H, 150mA (one per channel).