My new PL36 Push-Pull amp

[hidnplayr]

I created this amp from scratch, using only parts I had laying around.
It's the first (decent) amp I designed so I'd like some feedback from the tube-guru's to what I did good/wrong

It's not very sensitive, but it's designed to work with my preamp.
B+ = 300V
B++ = 268V
-Ubias = -42V

datasheets here:
http://www.tubes.mynetcologne.de/roe..._triode_v2.pdf
http://www.mif.pg.gda.pl/homepages/f...30/p/PCC85.pdf

[Yvesm]

I like it !

Wich B+ and B++ values ?
Just for chatting, you may increase (double or triple) the values of R15 / R16 and lower C1 / C2 by the same proportion.

Congratulations !

Yves.

[radiostar]

Nice, PL36 could sound fine, I never tried it in AF, just in HF, serial heating?

[hidnplayr]

Originally Posted by Yvesm

I like it !

Wich B+ and B++ values ?
Just for chatting, you may increase (double or triple) the values of R15 / R16 and lower C1 / C2 by the same proportion.

Congratulations !

Yves.

Thanks, the values of B+ and B++ are in the first post, around 300V and 265V
I did not really calculate the values for the coupling capacitors, I just happen to have a large bag of phillips mustard capacitors i bought really cheap.

I lend my book with all the formulas (Morgan Jones, valve amplifiers) and dont know the formullas of the head. I just used values wich made the amp sound good.

Later today, I replaced the UCC85 with an ECC88, wich is far more lineair in the used region, but has even less gain. (and higher price)

Radiostar:
It's still a protype, so I used a pair of bench supply's to power the heaters.
A benefit of using a UCC85 is it can be placed in parallel with both PL36 heaters so only a 25V heater supply is needed.

I used a cheap 230V isolation transformer for the B+, together with a 15H choke I got together with the Jukebox Iron (from an old seeburg, sounds actually very good!)

If anyone is interessed, I might draw the power supply schematic tomorrow.

[ilimzn]

IRFP450 have a VERY high input capacitance, which is also somewhat non-linear - although this is helped out by the fact you normally have quite a bit of voltage between D and G. However, this nonlinearity will become apparent when you drive the output hard enough to have less than about 6-8V between G and D of either MOSFET. These are high current parts (14A), which you really don't need in your application. IRF820 (on a heatsink, of course) should do much better here - it's a 2.5A part which also has proportionally lower capacitances compared to the IRFP450. There are even better choices in the 2SK*** camp. Also, be sure to put protection zeners across G-S of each MOSFET - the way it's connected now, driving your amp to clipping will kill the MOSFETs due to overvoltage between G and S - MOSFETs are VERY sensitive to this.

[hidnplayr]

Thanks for the info ilimzn, I just picked some mosfets I had laying around.
I figured the capacitance didnt matter much in this circuit?
I will add the zeners, it's not because they where free, I want to destroy them

Are there other important parameters of the mosfets besides the input capacitance and Max Drain-Source voltage I should care about? how about power?

[hidnplayr]

After reading a bit about screen driving (Synola SE 509, also read the PDF article) I decided to give screen driving the EL36/PL36/6CM5/6P31S... a try.
I found the curves for screen driven EL36 on Tom Schlangen's Homepage

I have attached my newest brainchild here (B+ is 320V)
The potentiometer in the CSS is used to set up the BIAS of the whole amplifier. The other potentiometer is used to balance the output tubes.
I used -24V on Grid1 of the output tubes because I had it available in the prototype

It doesnt really sound as good as expected, tips to improve that are very welcome

PS: sorry for the crappy CSS placing on my schematic.

[MRupp]

It doesnt really sound as good as expected, tips to improve that are very welcome

From the Tom Schlangen datasheet: "CF driver for screen must be able to source at least 15-20mA at Eg2 = 200V". Maybe I am misreading your schematic but your source follower does not have enough current (3-4mA ???).

[MRupp]

1. You could replace R9 / R10 by current sources which might further improve the source followers.

2. Allen Wright would tell you to convert the output stage into a truly differential amplifier by biasing it from a current source. The tricky bit is then to get the current through both halves equal. One possibility would be to use some kind of autobias, which could potentially be done via G1.

My 2 cents.

[hidnplayr]

I am currently happy with the means of biasing and balancing the valves, it is also not the cause of the awful distortion at higher volume I believe.
At low volume, the amp sounds pretty sweet.

My power supply however is one of the limiting factors right now (I'm using a tube rectifier, EZ81)
When I crank up the amp, B+ is only 290V and B++ 250V.

You must be right about the source followers however, I mistakenly figured the current in G2 would flow from B+ throug the mosfet, through the tube, to GND.
Since it's obviously flowing the other direction, using P-mosfets could solve that problem I guess.
The IRFU9310PBF looks good on paper to me (and also the price hehe)

Too bad i dont have any laying around to try right now..

EDIT: or am i now mistaken about the current again?