power amp current

[razorrick1293]

hello,

im fairly new to tubes, built a few things (12v 50v and 300v designs all SE) but i wanted to move on to something bigger, i wanted to try and build an el34 push pull amp but design it from scratch. though how do i work out the current il need to supply the tubes with? ive got a 350vRMS transformer at 200mA and ive already designed a pre amp with 2 channels and effects loop. overall im using 5 12ax7s or 4 12ax7s and a 12at7 for a phase inverter. and with this current is it possible to maybe add 2 more el34's to bring the power up to 100W?

any help on how to design this power stage would be greatly appreciated or at least somewhere that provides good help

kind regards
Razorrick1293

[ChrisA]

Originally Posted by razorrick1293

hello,

im fairly new to tubes, built a few things (12v 50v and 300v designs all SE) but i wanted to move on to something bigger, i wanted to try and build an el34 push pull amp but design it from scratch. though how do i work out the current il need to supply the tubes with? ive got a 350vRMS transformer at 200mA and ive already designed a pre amp with 2 channels and effects loop. overall im using 5 12ax7s or 4 12ax7s and a 12at7 for a phase inverter. and with this current is it possible to maybe add 2 more el34's to bring the power up to 100W?

any help on how to design this power stage would be greatly appreciated or at least somewhere that provides good help

kind regards
Razorrick1293

You are not going to get 100W output. I think maybe 35W

I think you could build the amp described as "Class AB" in the EL34 data sheet
http://www.r-type.org/pdfs/el34-1.pdf
But you'd need more than 100ma per EL34 tube.

I think you need either a smaller output tube or transformer with more current.

[razorrick1293]

hmm, how much current would be enough?

[DF96]

The normal procedure is to design the amp first, then design the power supply. That way you know how much current you need before you start working out how to supply it. It is possible to work backwards, but you may end up with a weaker design. In either case the place to start is the EL34 data sheet, then look at what other people have done (including the Mullard 5-20 amp).

The sort of questions you are asking suggest that you have a lot of reading to do before you start any real design work.

[bayermar]

Hi Razorrick1293
From your initial post, I would guess you are talking about a guitar amp not a Hifi amp?
For the more experienced persons WRT guitar amps it might be of interest if you are planning to deliberately design the power supply underrated to have the B+ voltage drop at high volume peaks.
100W from a pair of EL34 still sounds wrong - I think Marshall runs their "100W" heads (4 EL34) with a pair in the "50W" setting. If this really means true 50 Watt output power I wonder...

Cheers,
Martin

[kevinkr]

A quartet of EL34 with fixed bias and plate voltage of just over 400V can easily provide more than 100Wrms of output. Here are some possible thoughts on how to fairly easily achieve your goals. (I'm not a guitar amp builder but have fixed more than a few diy builds for others.)

The following set up more or less interpreted from the Mullard data sheet will give you 100Wrms or more. This output stage would operate with fixed bias in pentode mode. Depending on configuration more than 110W can be achieved.

Transformer Primary Z: 1.75K plate to plate

Plate voltage:400V

Screen: 1.5K individually to 400V supply or all screens in parallel with 400 ohm power resistor to supply. (About 10% more power) I think the individual screen resistors are a safer option from a stability standpoint.

Bias: -35V (adjustment range -30 to - 40V)

Idle current would be on the order of 130mA, (32mA per tube) but max combined plate and screen currents at full power would be in the range of 500mA..

Idle dissipation per tube would be in the vicinity of 14W per tube.

I'd make the individual bias adjustable preferably per tube or worst case per pair driving each primary phase of the OPT. Regardless each tube would have a 10 ohm cathode resistor for monitoring quiescent current when setting the bias. Matched tubes should be used.

Your power transformer should probably be rated in the region of 300 - 350mA (say 70% of rated power) and should sag significantly at full load to get you some crunch.. Use a dual rectifier set up with either 5U4 or 5AR4. I'd probably go for something in the ~700VCT range (need to know winding dcr) which would result in significantly higher voltage under light loads, but should be designed to sag 20% or so at full power. Use Duncan amps PSUD to get this approximately right. Consider adding an option to switch between solid state and tube rectification with the flip of a switch on the back panel of the amp.

Note that some tweaking is possible, the transformer Z could be increased slightly as can the plate voltage, idle currents can be decreased slightly as well.

I'd learn to do simple loadline calculations and make sure you have an operating point that makes sense.

Make sure that with no speaker connected the amplifier is connected to a 100W resistive load to prevent accidental destruction of the OPT. (I'd make this occur automatically whenever speakers are not connected to the amp.)

I'd use a 100W Marshall repro or replacement OPT for this application if you can get your hands on one.

Hopefully Tubelab will chime in with his comments as he has a lot more experience with guitar amps than I do.

[wakibaki]

350vrms * 0.2A(rms) = 70VA.

Class B push-pull has a maximum theoretical efficiency of 78.5%

So a class B amp of 100W using a transformer with 350vrms output would require a transformer supplying (100 * (100/78.5))/350 = 0.364A ABSOLUTE MINIMUM current. In practise this would not be achieved. You can probably get ~50W from your existing xformer.

'Course this is for continuous sinewave output power. As noted, for a guitar amp you probably want a bit of sag.

Like DF96 says, you need to do a lot of reading, particularly of datasheets, and spend some time doing trial calculations such as the one above so that they become second nature before launching into a design.

w

[kevinkr]

Comment about supply sag should probably be 10 - 15% from idle to full power.. Rated current is DC load current just to clarify.. Too late to edit unfortunately..