Spice Simulation – BGW 750C

Spice – BGW 750C

For those who are into electronics, Spice is an invaluable tool. I remember vividly the early years when computer cpu’s were intel 386. I was running DOS5.1 if I remember correctly.  To facilitate designing power amplifiers, I bought B2 Spice because it was extremely time consuming breadboarding circuits and in the process burning up parts especially power transistors as it was a pretty costly affair then.

Now that I’m back to analyzing circuits, it’s a good time to re-visit Spice. Presently, I’m not using Spice to design amplifiers. Instead, I’m using it for a deeper understanding of the circuits in some well known models. This will show me the strengths and weaknesses and prepares me for what to expect should I decide to prototype the circuit. For the first simulation, I’ll use the BGW 750C. 

Fig 1 – Spice circuit of BGW 750C Amplifier

Fig 1 is the Spice circuit of the BGW 750C using ideal components. I’ve removed all superfluous components so that Spice is not burden by them. My objective is firstly to determine the DC operating points in the circuit. This is an approximation but it does give me an idea of what is happening. 

Note that the op-amp is driving a “load” resistor of 680Ω directly. No transistors are used. What this implies is that the op-amp needs to supply enough current when the amplifier is pushed hard. In the original BGW 750C, the op-amp is a LM318. I believe a suitable replacement like a NE5534 can be used instead.

After some tweaking, I managed to get the op-amp output at 13.57mV. Ideally, it should be zero but 13mV is excellent. BGW recommends +- 300mV which seems a bit high. From the simulation, the 680Ω resistor bias the first stage at 4.93mA.

After this, the signal is then injected into the base of the VAS transistor. In this second stage, the idling current is at 25.33mA. This is set by the 100Ω resistor at the transistor emitter. Now that the signal is amplified, it is fed to the Current stage.

Due to ten power transistors in the output, the designers decided to use a Pre and Main Driver. This will prevent the VAS from being “loaded” when the amplifier is pushed to the max.

The Pre-driver current is at 14.82mA, the Main driver at 22.5mA and finally, the power transistors are biased to 23.72mA. These idling currents are achieved by adjusting some resistor values. Had I followed the BGW schematic blindly, it would flag some ridiculous values.

Fig 2 – Spice Transient Simulation @ 1kHz into 8Ω

With the DC operating points established, I proceeded to do a Spice Transient Analysis (Fig 2). This will flag me if the circuit is amplifying correctly.

For this analysis, I altered the gain structure of the 750C to 32x. In the original schematic, it was 19x. What this indicates is a much higher input voltage is needed. 

With this revised gain, the output into 8Ω is 100V peak to peak with only an input of 1.5V.

Summary

Spice is capable of doing much more than what I did. When I bought B2 Spice in 1985, there was no such thing as free Spice. In fact, internet was not available then.

Today, one can download a free working copy of LTspice which is a lot more powerful than what I had. If LTspice is not to your liking, there are many other apps that’s free.

Spice is a fantastic learning tool. I encourage those who are interested in electronics to try it out.