Help with Regulator/Rectifier circuit for my motorcycle

After my cheap chinese reg/rec decided to up and melt itself I've been looking into alternatives. The difference between a market $100 and $180 reg/rec is simply how many cooling fins are used. I've found this schematic which is a shunt type reg/rec



with components as follow:
Q1 - pnp TIP42
Z1 - 12v Zener diode IN4742
D1-D5 - si diode IN4148
R1 - 47 Ohms, 1/4W
R2 - 75 Ohms, 1/2W (can be varied from 50-100 Ohms to vary reg output)
R3-R8 - 680 Ohms, 1/2W
C1 - 10uF/50v

Issue is S1-S3 are 25A, 800v scr's (NTE5460) which are supposedly obsolete (or at least difficult to find for my time frame). I was wondering if there were any substitutes I could use for this, ie NTE5554 which is 25A, 400v as the voltage is rpm variable and 400v does sound reasonable (service manual rates performance of 14-15v, 18.5A @ 5000rpm)?

Also wanted to use a pre-made rectifier to save space and couple better with the cooling fins. Will it matter how much greater the current/voltage rating on the rec is? I've been told to use double the current as a guideline but there are rec's with cooling fins that have much higher current ratings.

And Finally for this application will a homemade PCB handle the currents or should I solder thicker gauge wires between components on a blank board?

So long as your alty doesnt produce more than 400v the alternative you suggested looks ok, the data sheet specifies it has the same trigger current and gate dissipation so it'll probably work, but you just dont know without trying.
While the battery is connected the alty wont produce much more than 12v, its when the battery is disconnected with the engine running that a very high voltage is produced.
A rectifier is only as good as its heatsink, you need to look at the datasheet and work out the dissipation and then work out what size sink you'll need, allthough you can go off the original and use something similar, just **** it and see.
Dont forget to put the thyristors on the sink too, they will dissipate soem energy esp at high revs.

You can make the reg up on veroboard or a pcb, you'd be better putting the thyristors and rectifier on a heatsink and make direct connections, or use some tagstrip to support the legs of the devices from vibration, and build up the smaller components on the veroboard/pcb with a piece of ribbon cable between, commercial units tend to be potted to keep them safe from impact/vibration and cack from the road/track.

My thoughts:

1.

I assume that's the field coil of the alternator. Would it be carrying 2 amps?

I recall that 4 amps is a typical current to energize ordinary automobile field coils. That's for alternators putting out 40, 60, 80 amps.

Maybe a motorcycle alternator puts out half that? That makes 2/3 amp load through each of 3 wires. The table says 22 awg minimum is needed for .7 amp. This is just theory.

2.

The scr's may have to endure hi voltage spikes which can occur when switching off those coils. Perhaps that's why the spec is for 800V.

You may point out that scr's are not insulated gate devices. And you may point out that the scr's automatically stop conducting when current drops to zero, which means that they are not being used to switch off coils strictly speaking.

It's just that you never can be sure. If you want to use 400V scr's then it would be wise to check with a scope as to whether spikes are getting at them.

This will be hard to do under operating conditions. You'd have to spin the alternator with a load attached. Attaching a load makes the alternator hard to spin.

3.

As to the A rating of the rectifier, you may get away with using the same value amps as you measure in everyday operation.

However someday the battery may have been depleted while the engine is off. A low battery draws higher than normal charging current.

Or you may use your motorcycle to jumpstart another vehicle, requiring all the current output your alternator can give.

Or someday one diode may fail, putting the burden on the remaining diodes.

Etc.

So we hear traditional wisdom saying to use components with double the rating of what they might be exposed to in all circumstances.

Sounds words from Brad, only one thing, cheap chinese bike gens tend to be permanent magnet, the coils on the drawing are the output not the field windings, the regulator dumps excess energy when the battery is charged, the gen is designed to withstand having its output shorted for a period.
This is a very ineeficient way of doing it, I spose its cost effective, 25 amps at 12v though is over 1/3hp and for a small engined racing pitbike that might mean losing or winning, however for a runabout its not a major concern.

the coils on the drawing are the output not the field windings

Click to expand...

In that case the scr's, coils and rectifiers are carrying the entire 18.5 A of charging current.

Scratch what I said about using 22 gauge wire.

I see the table recommends 10 gauge or thicker to carry 18 amps.

Divided by three, then each third can use 12 gauge.

==============

One more thing, test resistance going into ground. Particularly in the left-hand circuit.

You always want as close to zero as you can get. Clean bright contact surfaces are a must. You probably know this already.