Design a particular push pull 600W output stage

Status
Not open for further replies.
Jump to Latest
#41
luixssj3 said:
In this moment the load will be resistive, I have no idea of what is in the chief mind, the only thing I know in that this is a first experiment to see if tubes is more resistent than mosfet when peak overloaded so I think that the output stage will be driven with a squarewave and also the load will be lowered to create stress.
Click to expand...

It is not about tubes versus solid states, it is about how the topology reacts on overloads! Some solid state designs are intrinsicly immune to overload, while tubes in others can overheat and run off with red anodes.
 
#42
it costs money to hide the real requirements from the engineers

yes splitting the amp output xmfr by frequency bands, allowing switching is more practical, cheaper

combining 2 output xfmr wtih a XO needs a better load definition, and a frequency response flatness tolerance


the Real product requirements are really, really helpful for design recommendations
 
Last edited:
#44
"You design for me a product, but I will keep from you details". It does not work such a way."

Oroville dam is proof!
 

Attachments

  • oroville-spillway.jpg
    oroville-spillway.jpg
    103.3 KB · Views: 85
Last edited:
#45
It is well known that OTL output level is very much load dependent.

This is sim result of my 6 pairs 6c33c.

RL(Ohm)/Pout(W) @1Khz
2/650
4/550
8/820
16/500
32/275
64/150
100/100

How to get Pout 600W into 96 ohms?
 
Last edited:
#49
You need pentodes to get real power out. 6C33 (triode) is too constrained on max B+ and minimum voltage drop. They will just make a lot of heat.

300 KHz, full power, on a steel core will make for smoking hot laminations. Nano-crystalline, amorphous core or ferrite needed. Ferrite won't work reasonably for the LF end however (too low Bsat).

To get 20 Hz to 300 KHz bandwidth will require a toroid big enough for the primary winding to be single layer progressive wound. (and Circlotron to reduce # turns by 1/2)
 
Last edited:
#50
The efficiency of tube OTL is also very much load dependent, the higher the load the more efficient it gets, maybe < 10% @2 ohms 11%@8 ohms to 20%@100 ohms or more with specific design and use.

So if matching transformer is 2/100 ohms for max power 600W, how you account for efficiency? If a particular tube like pentode you may get 600W at 100 ohms, how about low load 2-8 ohms, how many watts you can get? I believe the final product is not only for 96 ohms, it would have limit use. So I suppose if pentode can do 100/(2-100) matching to cater for 2-100 ohms load that is more realistic, IMO. Is there such a pentode otl out there? For 6 pairs of power pentode mean 1.2k each rather than 100R or less for triode.

Edit: the sch I post earlier post no.4 class ab amp efficiency of 80%, this OTL output 200W more than that, but same operating points and supply. What do think the efficiency is?
 
Last edited:
#59
Tubelab_com said:
An OTL configuration has been suggested, and this may be possible, but you state that you want to use "current production tubes." Most OTL's bump up against, or violate the output tube's peak cathode current rating. This may be possible with music since the peaks are seldom seen, but the amp will fail if operated contiguously at full power into the minimum specified load impedance. The tubes that I use with an OPT are rated for 1.4 amps of peak cathode current. You will need a lot more current handling capability to drive an unspecified load impedance without an OPT. I am not aware of any current production tubes with a peak current capability even near 1 amp.
Click to expand...

Thank you very much for your suggestions and congratulation for your power amp! It looks great!

So, I'm thinking about an OTL mixed transistor driver/output tubes, I need +-300V B+ power supply (or 350 to stay safe) because power pentode or tetrode usually have a minimum of 50V Vak when Vg=0V.
The only tube that I have in mind for this kind of amp is the EL509, rated for 1.4A max but dissipation is about 40W, calculating a 500W dissipation I need at least 12 tubes (six couples).

I think it is also possible to use two or more transformer with different frequency bands.
I found the Shuguang 805 tubes, with 1.5kV and four or six of this tubes I can obtain about 600W, anyone know how reliable are they?

This is only an experiment to see if tubes can be useful today, I know that I can design the same with MOSFET (I don't know if SiC can be used for linear applications) but we want to see if tubes have some advantages, I know that transistor is a better choice but the chief want to experiment with tubes, for me is an occasion to get fun.
Maybe after experimentation phase all the stuff will be throw in the rubbish.
 
#60
"The only tube that I have in mind for this kind of amp is the EL509, rated for 1.4A max but dissipation is about 40W, calculating a 500W dissipation I need at least 12 tubes (six couples)."

If psu is +-300v, 1.4A each for 12 tubes this is 16.8A peak, total power in=300x16.8=5KW, minus 600W output still 4.4KW tot tubes dissipation, not 500W.
 
Last edited:
Status
Not open for further replies.
Top Bottom