Really simple low-volt valve amp

[markeebee]

Where would I find one of these transformers Mark? Can't seem to spot one unless i'm searching for the wrong thing.

It's one of these:

http://uk.farnell.com/jsp/displayProduct.jsp?sku=1778009&CMP=e-2072-00001000&gross_price=true

You can get them at Maplin as well. Mine was from eBay.

Use the 2W tap.

[Scruffie]

Cheers Mark! This looks like it should suffice... http://www.ebay.co.uk/itm/100V-LINE-TRANSFORMER-0-25-0-5-1-2-4W-PA-SYSTEM-VOLUME-CONTROL-/310330501667?pt=UK_Consumer_Professional_RL&hash=item484123aa23

Why the 2W tap? Can different ratios be achieved with the others?

[markeebee]

Yeah, that'll crack it.

TBH, the maths that dictates the choice of transformer tap is beyond me, especially as data on the ECC86 as an output valve is fairly scant.  I found this document http://cool386.tripod.com/6bm8/6BM8.html which gives info about the relative impedance of the taps on similar 100V line transformers.  The 2W 5kohm tap just felt right, so I tried it and it worked.

If anybody can enlighten with regard to the theory I'd be very grateful.

[slacker]

If you google "line transformer as output transformer" you should get a link to ozvalveamps which has loads of info and a link to another page with the maths for calculating the impedances for different transformers. I would post the link, but I still haven't figured how to do that on my phone.
I built something very  similar a while back and just tried different taps until I found the one that sounded best. I've also used a line transformer as the output tranny on a real valve amp, works great.

[markeebee]

Thanks Slacker, that's good stuff.  TBH, though, the txf wasn't the real problem; I was more stymied trying to ascertain the load impedance of the valve

[iccaros]

treat the load like a resistor, which it is, ohms law will do the rest.  you have a datasheet with max dissapation and you know your voltage, so you have current and voltage..

to give you ideals
http://www.diystompboxes.com/smfforum/index.php?topic=41776.0

and my 24v tube amp
http://www.diystompboxes.com/smfforum/index.php?topic=91606.0

[brett]

Hi
not sure why the comments about low output from a 12AU7. They are similar to the 12U7, which was the output tube for valve radios in cars (@12V). About half a watt??
If you want more, you could always run it through a little transistor amp (5 clean watts are easy to get from the 20V supply into 8 ohms).
cheers

[iccaros]

12U7 were preamp tubes in car radio's. 12K5 and such like it, were the output tubes.  Most ran to transistors of the day.
While the 12u7 and 12AU7 were simular, they are not the same. I have read the internet rumors that 12u7 were 12AU7's that did not pass all of the test, but they react differently, if not as much to NOS 12AU7 they do with newer tubes.

You could put a 12AU7 or 12U7 as an output tube.. but you would not be happy I believe..

[Ben N]

You could put a 12AU7 or 12U7 as an output tube.. but you would not be happy I believe..

Really? 
http://www.ax84.com/index.php/oldprojects.html?project_id=firefly

[iccaros]

Low voltage (12 - 24).. Not 230v like my firefly... There is a huge difference..  the 12au7 @ 12v and 3 -5 ma is not very loud.  A 12K5 at 50ma is not very loud but much louder than 12au7 and at more reasonable loads. sitting at 1K and not 22,000

But I will put one together, simple enough and see how much power and what it sounds like.. I will post

[slacker]

If you search the forum for rascal you'll find a little amp I built basically the same as this using 12au7s or 12at7s in the power section. That worked fine at about 20 volts although it was much better and louder at 30ish.

[markeebee]

^ Ah, yeah, of course.  I remember the Rascal now.  I have probably inadvertantly ripped you off, and many apologies for that.  Please send the invoice to the usual address.

I recall that was going to build a Rascal, and you kind of told me to go for one of Frequency Central's amps instead.  Oooops - he'll probably send me an invoice now, as well  .

[iccaros]

If you search the forum for rascal you'll find a little amp I built basically the same as this using 12au7s or 12at7s in the power section. That worked fine at about 20 volts although it was much better and louder at 30ish.

Good sound clips, Transistor source follower is a good add, It sounds quite, which is not bad, I am always ready to admit when I am wrong.. Will have to build one.  I wonder if a ECC99 at 50V would be better though..

[brett]

Hi
RE: 12U7 and 12AU7. Just sub one for the other - in the circuits I've built (9 to 32V supply) they sound and scope more-or-less the same.

If you want to avoid loading the 12AU7 (it is quite a low load at present) an option would be to stick a class A MOSFET output stage on the end. About 6 components and is perfect for a 20V supply. The most expensive part - an MTP3055 or similar - only costs about $2, which is cheaper than the line transformer. There have been several threads about simple class A MOSFET amps.
cheers

[allesz]

Oh boys I really hate you now: first comes markeebee and his amp, then comes the rascal.... Ok I will buy another 12au7 and start experimenting, it seems I have no alternatives right now.

[iccaros]

Oh boys I really hate you now: first comes markeebee and his amp, then comes the rascal.... Ok I will buy another 12au7 and start experimenting, it seems I have no alternatives right now.

I just bought 20 off ebay for $20.. some may be bad or weak, but at that price.. look around you can find really good deals. Preamp tubes almost never fail and in Hammond organs they would change them every year no matter what, so there are a lot of them sitting in a box in someones house and they sell them on ebay. like here
http://www.ebay.com/itm/RCA-12AU7-Clear-Top-Tubes-/120898639830?pt=Vintage_Electronics_R2&hash=item1c261ed7d6#ht_1801wt_1396

Most likely some of these are weak, and one may not work, but a "Weak" tube will be OK for this design as we are not pushing high voltage.

[PRR]

> enlighten with regard to the theory

Life is too short.

Use Rules-Of-Thumb.

Triode Power Amplifer: find the plate resistance. This varies with plate voltage and grid bias, and on a "tuner" tube will vary a LOT. You want the zero grid voltage line from zero to your nominal plate voltage.

ECC86 data:
http://www.mif.pg.gda.pl/homepages/frank/sheets/030/e/ECC86.pdf

Page 9: Vg=0V hits Va (plate voltage) of 20V at 8mA current.

20V/8mA= 2.5K ohms.

You have two sections in parallel, 1.2K.

Rule-Of-Thumb:

For good power and distortion, the load should be double that.

For less power and lower distortion, or for high supply voltage, go 3X to 5X.

To force maximum power from insufficient supply voltage, rough-match the plate resistance. With a "good audio triode" the distortion will be a bit obvious. With this tuner triode the distortion will be gross, but not sour (up to a point).

Since your voltage is awful low, you would compare 2X (2.5K) and 1X (1.2K) loading, to see how the power and distortion work out.

In each case you "must" re-optimize standing bias to center the tube on the available current/voltage limits for that load. Put DC voltmeter on cathode bias resistor. Note reading at idle and at FULL WAIL. If it idles 0.7V and wails 2V or 0.3V, that's way too much bias shift (unless it sounds kewl). Change bias resistor until the idle/wail shift is under 50%, under 20% for "clean". Or until you decide you like it not-centered (a valid musical choice).

[markeebee]

Thanks PRR.  That makes sense and is very helpful.

I found the plate resistance (like iccaros suggested above, thanks iccaros) but I wasn't aware of the rule of thumb about doubling it. Seems I fumbled my way there anyway.

I have to say that I quite like the sound of this little thing, but on my next build I'll try a lower impedance tap and monkey around with the bias as you suggest. A bit more output and a bit less distortion might be good things.  Thanks again.