Amplifier based on 2N3055

Hi Madaudio,

Nice find you have over there.This circuit dates back to 1982, Elektor Summer Circuits.I built this one with minor tweaks and it's a real beauty.
Actually, this is what I'm listening to right now.I have a few extra PCB's for
this amp.If you are feeling like having a go, drop me a PM at selim_ardali at yahoo dot com and I'l be glad to supply you the PCB's.

Cheers
Selim ARDALI
 
Hi Madaudio,

Nice find you have over there.This circuit dates back to 1982, Elektor Summer Circuits.I built this one with minor tweaks and it's a real beauty.
Actually, this is what I'm listening to right now.I have a few extra PCB's for
this amp.If you are feeling like having a go, drop me a PM at selim_ardali at yahoo dot com and I'l be glad to supply you the PCB's.

Cheers
Selim ARDALI

I wouldn't mind having a couple of those pcbs myself. I have most of the parts to build this (old stock) and it would be a nice little diy project for fun.

Can we have a look at those pcbs please? How much would you ask for them?
 
With +/-35V the oputput devices are running right on their maximum Vce (at peaks and troughs in the signal) I would change for higher voltage types that 2N3055s. Assuming you are running off smoothed but unregulated supply (ie transformer-rectifier-caps or similar) you should allow for mains +10% or so and take that 10% off your nominal supplies so in this case take off at least 3.5V (=31.5V absolute max nominal supply) . You *may* get away with 35V peak across these devices but as it's your speakers at peril, a few cents/pennys extra for higher Vce devices does make sense!

N.B the modern datasheets show the Vce as 70V max with 100R resistors across the base although this criteria is satisfied in the above circuit, I baulk at the idea of risking speakers with such a shortfall - 100V Vce transistors are not exactly difficult to come by anymore...
 
Last edited:
With +/-35V the oputput devices are running right on their maximum Vce (at peaks and troughs in the signal) I would change for higher voltage types that 2N3055s. Assuming you are running off smoothed but unregulated supply (ie transformer-rectifier-caps or similar) you should allow for mains +10% or so and take that 10% off your nominal supplies so in this case take off at least 3.5V (=31.5V absolute max nominal supply) . You *may* get away with 35V peak across these devices but as it's your speakers at peril, a few cents/pennys extra for higher Vce devices does make sense!

N.B the modern datasheets show the Vce as 70V max with 100R resistors across the base although this criteria is satisfied in the above circuit, I baulk at the idea of risking speakers with such a shortfall - 100V Vce transistors are not exactly difficult to come by anymore...

My two cents:

I think it would be a cool thing to make a 3055 amp in a grounded bridge topology. Keeping the rails at reasonable levels for the 3055, adding pairs for better handling of lower load impedance.

The grounded bridge was really aimed at getting more power from lover voltages and lower Vce0 transistors, without cascoding them.

I wouldn't mind working on such a little project, I have a bunch of old 3055 waiting to get to work. It would be cool to see how much power we can get out of those babies...
 
I thought spookkydd meant 'single-supply but bridged' which is quite workable. You won't need output decoupling caps in this scenario as both speaker terminals are connected to an amplifier output and will thus be at 1/2 supply.
You inherently double the voltage to the load bridged without increasing supply volts at all due to the balanced nature of the output signal. This means the Vce of the transistors is no longer such an issue (for a given power), but you DO need to be able to supply double the current (or more, especially thinking about 4 ohm bass units used often today).
 
Suggestion:
55V DC nom single supply unbridged amplifier can give approx 19Vrms output signal (losses ignored for now)
Bridged amplifier, same supply the output voltage doubles to 38V rms
this will give 180W into 8 ohms 3.4A peak or 2.4A rms current
or 360W into 4 ohms, 6.8A peak or 4.8A rms current

Allow supply caps sufficient ripple current rating to deliver the 4 ohm peaks (not normally difficult, these days!) At 55V nom you should be able to safely use 63V types.

Transformer rating close to the RMS current into 4 ohms should be more than satisfactory (unless you really have 1 ohm nom. speakers) so 38-40V @ 6.5A or so will be more than fine.

Note there's nothing really complicated and the power supply as specified is generously rated compared to even some esoteric models out there!
 
Last edited:
Voltage and current wise, the grounded bridge is the same as cascoding them. Drive requirements are a little easier on the low (grounded) side, but you've still got that low beta to deal with.

Cascoding is more complex because of the extra number of parts, so the pcbs are larger and more parts cost more.

I think a simple design in a grounded bridge topo would be cool to attempt and that could make for a rather powerful amp with such old trannies as the 3055.

With a simple basic design, good enough for sufficient quality, considering what those 3055s can do, the pcbs can be rather small and we can keep the number of parts low.

On the other hand, the power supply, although much simpler, requires capacitors with double the voltage handling, but since we're only talking about 3055 with Vce0 around 60V, the symetric power supply's total voltage isn't so high as not to be able to find such capacitors.
 
I thought spookkydd meant 'single-supply but bridged' which is quite workable. You won't need output decoupling caps in this scenario as both speaker terminals are connected to an amplifier output and will thus be at 1/2 supply.
You inherently double the voltage to the load bridged without increasing supply volts at all due to the balanced nature of the output signal. This means the Vce of the transistors is no longer such an issue (for a given power), but you DO need to be able to supply double the current (or more, especially thinking about 4 ohm bass units used often today).

I'm not referring to the classic half bridge topo, but rather the grounded one with symetric supplies without a fixed ground, and that ground floats because of the low side of the amp.
 
Suggestion:
60V DC nom single supply unbridged amplifier can give approx 20Vrms output signal (losses ignored for now)
Bridged amplifier, same supply the output voltage doubles to 40V rms
this will give 200W into 8 ohms 5A peak or 3.53A rms current
or 400W into 4 ohms, 10A peak or 7.1A rms current

Can be done, in the grounded bridge topo.

In fact, we could think of the power supply in the grounded bridge arrangement as a single supply, not symetric, since it's made like a single supply without a fixed ground.

Note there's nothing really complicated nad the power supply as specified is generously rated compared to even some esoteric models out there!

I think this can be done. I wouldn't mind working on pcb layout.

That elektor schematic posted a few posts above looks rather nice and usage in this way I think. Anyone?
 
It's me I know, I'm still not quite awake but I'm not quite understanding what you're referring to as grounded bridge? :confused:

The grounded bridge was invented by crown/amcron. It's basically like an ordinary bridge configuration, but the difference is the speaker isn't connected only between the two amp sides, one amp side (low side) has its output tied to the ground directly and the power supply is not designed as the ordinary symetric one, but rather without a central ground, like a single supply, with twice the voltage (or let's say the same voltage as would be between +V and -V). The main difference is that the ground is left floating, and it's the low side amp that forces the ground to float by having its output directly connected to it.

I saw a description document of this somewhere some time ago. I will try to locate it and point to it here with a link.
 
Not sure about it at all looks at least like a minefield for hobbyists but certainly don't let me put you off...

It's true there is a need for a certain amount of attention, especially with the ground and power supply, since the power supply floats and there is one amp with its output tied to ground.

Without proper care, this can turn out to a disaster.

But still, it's an interesting arrangement, and with proper care, can work nicely.

I think since each amp (made of 2 amps) would always be meant to work together only, everything should be put on a single board. And to prevent too many possible errors, avoid as much wiring as possible. I think an all-in-one pcb is in order, with the least amount of wires. Done properly, with enough scrutiny, I think this can very well be a diy project.

Some diyers have built quite complex amps, some very advanced, I've seen those with a micro-controller added, with temperature sensors and all that.

With due diligence, a good design can be made that can be reproduced and built easily enough by ordinary diyers.