EZTube Power Amp

[ruffrecords]

The last time I built a tube power amp was back in the 60s. For quite some time I have been meaning to build a small tube power amp for listening to background music in my workshop and listening to my mic pre and EQ designs. I came up with a simple single ended design using the EL84 and added a first stage using an ECC88 with dc NFB similar to the EZTube  Eurochannel design.

I had an old guitar amp chassis to build it on and acquired a couple of suitable output transformers. For the supply I had an suitable mains transformer left over from some early experiments. I used the HT250 board fitted with 500V capacitors and 5W resistors to cope with the expected total current. First I just built the output stage, powered it up and verified the current draw from each tube was OK. I had a pair of Tannoy book shelf speakers connected to it and there was a fair bit of hum audible. A lot of this is undoubtedly induced by the mains transformer straight into the output transformers. I have ordered a small toroid replacement which I hope will fix this.

Being impatient, I decided to hook a 10K:10K transformer direct to the EL84 grid to see if I could get any sound. Wired the transformer input to a TRS socket and plugged it into the output of my Focusrite USB thingy. To my surprise there was plenty of output level. So I quickly wired up the other channel and soon was listening in stereo.I will change the mains transformer when the toroid comes but right now I am not sure if I will bother to include the ECC88. Thta must be the simplest stereo power amp ever, just 2 x EL84. When I have got rid of the hum I will definitely make some distortion, frequency response and noise measurements and then decide if I need to include the ECC88 stage. I'll post some pics later.

Cheers

Ian

 

[ruffrecords]

I have conducted some tests with an 8 ohm load.

Distortion is 5.5% at +4dBu and 3.4% at 0dBu. frequency response is 3dB down at 100Hz and there is a 15dB peak at 35KHz. This results in the response being +2.8dB at 20KHz.  +4dBu theoretically produces 188mW into an 8ohm load. It is feeding a Tannoy  Mercury F1 bookshelf speaker and sounds quite loud so I an surprised the actual power is so low.

Anyway, it's pretty poor performance.  I am beginning to see why some NFB is needed. I am definitely going to add the ECC88 stage.

Cheers

Ian

 

[abbey road d enfer]

I have conducted some tests with an 8 ohm load.

Distortion is 5.5% at +4dBu and 3.4% at 0dBu.

That seems excessive; that's the values you would typically get at 2-3W, not 0.1-0.2. At what current are you running them?

frequency response is 3dB down at 100Hz and there is a 15dB peak at 35KHz.

Looks like your OT's are somewhat lacking...

 

[PRR]

No negative feedback = no damping factor.

Try EL130. 4Vrms in gives 10 Watts SE. As most "+4" outputs will deliver 10V if cranked, you can rig 6dB NFB and find some damping, better than none. But 100mA cathode current!

 

[ruffrecords]

I have conducted some tests with an 8 ohm load.

Distortion is 5.5% at +4dBu and 3.4% at 0dBu.

That seems excessive; that's the values you would typically get at 2-3W, not 0.1-0.2. At what current are you running them?

frequency response is 3dB down at 100Hz and there is a 15dB peak at 35KHz.

Looks like your OT's are somewhat lacking...

I agree, the distortion figures are puzzling. However, the loudness seems to me to be a lot more than a couple of hundred milliwatts. The speakers I am using are Tannoy Mercury F1 which have a sensitivity of 87dB. THis definitely needs further investigation. At present I have a 220R cathode resistor across which I measure 8.4V which makes the cathode current  about 38mA

I measured the primary inductance of the OP transformers  before using them - it is about 5H which is about 3K at 100Hz. The transformer is supposed to have a 5K reflected impedance at the secondary which sort of ties in with the 3dB drop.

The secondary is multi-tapped and the documentation rather ambiguous. I think I have it wired correctly for 8 ohm speakers but it is possible I have made a mistake.

I initially had a lot of trouble with hum coupled to the OP transformers from the mains transformer. I have no fitted a toroid mains transformer and cured that problem. I need to repeat the measurements.

Cheers

Ian

 

[ruffrecords]

No negative feedback = no damping factor.

Try EL130. 4Vrms in gives 10 Watts SE. As most "+4" outputs will deliver 10V if cranked, you can rig 6dB NFB and find some damping, better than none. But 100mA cathode current!

I am a complete noob when it comes to tube power stages. However, a pentode plate resistance is very high so the normal simplification for small signals is gain = gm.Rl where presumably Rl is the load impedance reflected to the primary of the transformer. In that case I am not sure how there can be any damping at all - or is that what you are saying?

Cheers

Ian

 

[PRR]

> I am not sure how there can be any damping at all

Exactly.

Damping Factor of naked no-NFB power pentode will be around 0.1, maybe 0.2 counting OT core losses.

Old-old-old-skool speaker design assumed DF of 2 or more (triode drive).

By 1959 we liked DF numbers over 10. Even before that a true Williamson would do DF over 40.

By 1969 DF numbers over 1,000 were loudly touted, no matter that this was polishing a turd-table.

There are speakers adapted for low damping. Good amp damping can extend bass response/efficiency for a given box size, but sometimes box is unlimited.

In non-hi-fi work:

Open-back Fender guitar amps often ran DF near 1 or 0.5. The bass bump cancelled some of the bass-slump from open-back, gave almost double the sound power in the room and better throw. Marshall copied but with quite different speakers in closed boxes, found they liked DF more like 2 or 4 to control the different response. At an extreme, one model of Ampeg VT-40 with four cheap Tens in an open box ran no NFB and no damping at all, following a long tradition of lesser guitar-amps and even console radios, and was a kick-butt sound.

Playing at fairly low acoustic levels with short-bass transformers, all of this may be moot. It sounds like you don't have much below 100Hz, and if you did have a little then Fletcher-Munsen suggest you would not hear it.

 

[ruffrecords]

Thanks PRR. Fascinating info. The specified ra of the EL84 at 40 something mA/250V is 40K. So, even with 20dB of NFB this only drops to 4K giving a DF of about 1 (assuming for the moment the XFRMR is not included in the NFB loop).

Cheers

Ian

 

[abbey road d enfer]

Thanks PRR. Fascinating info. The specified ra of the EL84 at 40 something mA/250V is 40K. So, even with 20dB of NFB this only drops to 4K giving a DF of about 1 (assuming for the moment the XFRMR is not included in the NFB loop).

Cheers

Ian

that has to be seen in conjunction with the speaker's impedance.
In primitive systems (typical gramophone), damping is needed to tame the large impedance hump at resonance. Since the impedance there is about 10x nominal, the effective damping factor is about 10:1. Phew!