• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

DIY Hybride amplifiers

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
re-thinking Hybrid designs

Every time I see a "Hybrid" amp design, it seems to just be a SS class aB output stage stuck onto a tube gain stage. Now, this is understandable, since the output xfmr and tubes are costly and SS has the advantage of large bandwidth if you want to use global NFB. But, I believe the reason many people don't like SS amps is the output stage crossover distortions. So one should at least use a class A output stage or it is time for some innovative thinking. Here are a couple ideas.

Mosfets have near square law transfer functions, which means their transconductance (derivative of V to I transfer function) is linearly increasing with current. This makes them good for class A output stages since the two opposing Mosfets then sum to a constant transconductance (one transconductance increasing at the same rate as the other is decreasing), giving low distortion without crossover problems. Unfortunately, class A waists a lot of power and heat, so is only reasonable for low output power.

Two get more power output, one would like to resort to class aB, like used in P-P tube amplifiers for more power. Tube amps seem to do better at avoiding crossover distortion in class aB than SS amps. The reason being the more gradual turnoff of the 3/2 power law of tubes versus the exponential law for bipolar transistors or the square law for Mosfets. Gradual turnoff of the output devices prevents a sudden kink in the output transfer function that produces the unpleasant higher harmonics characteristic of typical SS designs. So, why not make the output stage itself hybrid.

The simplest way is to make a Sziklai / CFP (complementary feedback pair) using a tube for the first active element and a Mosfet or bipolar for the output element. This allows the tube to set the gain rule, or 3/2 power transfer function. Since we are limited to N type tubes, we must use an output xfmr and go with a P-P design. So, only the power output tubes are eliminated this way.

Finally, let me suggest a couple of NEW approachs. Lets modify the SS devices to have 3/2 power transfer functions, then we can still use the usual SS totem pole output approach.
For bipolar transistors, put a damper tube rectifier in series with each of the base connections, with appropriate polarity for conduction. The SS biasing network will have to be modified to handle more like 10 V bias instead of 1.2V bias due to the voltage drop on the rectifiers to get idle current setting. Now the transistor will have a 3/2 power law for voltage input to current output just like a tube.
The preceding tube voltage gain stage should have a lowish output impedance so as to look like a voltage source drive.

For Mosfets, one can build a hybrid current mirror for each output device. This means putting a small thermionic diode across the Source to Gate connection in appropriate polarity for conduction with drive signal. (6JU8, 6AL5 ...) Probably will want to use NON lateral Mosfets to get sufficient voltage drop across the gate to source for class aB idle current setting. Now the diode will perform 2/3 power current to voltage conversion, and the Mosfet will perform square law voltage to current conversion. The result being a 4/3 power current to current power law. This might possibly give even better results than tubes for class aB crossover due to its even slower turnoff. Keep in mind that this topology is for current input, so should be driven from a high impedance preceding tube voltage gain stage. The idle drive current level setting will set the idle current in the output stage.

Don B.
 
Eugene Karpov Hybrid

I went to Next Tube and viewed the two PDF's in Russian.
Then I e-mailed Mr. Eugene Karpov and asked if an Enlish translation existed. His responce was "unfortunately English version of article still is not present. I have the plan to translate article, but I have no time for this work yet." Does anyone on this site who reads Russian want to take a stab at it for those of us who don't speak Russian?
 
help on hybrid, please

Hello,

Could somebody help me please ?

I intend to build this hybrid amplifier (joined). It is a scaled version of this amplifier :
http://ptsoundlab.com/sechybrid/transitube35wca/transitube35wca.htm

It will deliver 7Watt between 100-20000Hz. (20 to 20000 with 0.5% atenuation at 20Hz)

I have all the tubes but IRF 140 mosfets, not IRF 150.

The diferrence between them (from Farnell catalogue) :

IRF 150 : TRANSISTOR MOSFET; Conforme RoHS:NON; Transistor type:MOSFET; Polarité transistor:Canal N; Tension Vds maxi:100V; Courant, Id cont.:40A; Courant, Idm impul.:160A; Power, Pd:150W; Résistance, Rds on:0.055Ohm;

IRF 140 : TRANSISTOR MOSFET; Conforme RoHS:NON; Transistor type:MOSFET; Polarité transistor:Canal N; Tension Vds maxi:100V; Courant, Id cont.:27A; Courant, Idm impul.:108A; Power, Pd:125W; Résistance, Rds on:0.085Ohm;

My questions are :

1. can I use IRF 140 without any other modification ? (i never worked with mosfets)
2. What can I do to have a better bandwith ?

Many thanks,
André
 

Attachments

  • ampli.gif
    ampli.gif
    8.8 KB · Views: 982
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.