1/2 hp SCR-based (?) DC motor controller

[mark_in_manc]

I have a 30-40 yr old separately-excited 1/2hp DC motor on a drill in my shed, with an equally venerable speed controlled attached. I rescued it from a skip and it's been great since it allows speed control unlike a normal single-phase ac induction motor. I knew nothing whatsoever about it until tonight, when it stopped working - and I wondered if folks here have experience of these things.

So far I worked out that the field sits at a constant ~210v dc (not put a scope on it yet - this is just a DMM) and the armature volts change depending on whether the motor is running, or not. Unfortunately the speed control pot is acting like an on-off switch; this is the nature of the fault. (The pot is not broken - my first check!)

Having read around a bit I guess the armature is meant to be turned on by the scr for various durations of the half-cycle, giving rise to increasing current in the armature and hence increasing speed. But at the moment the controller succeeds only in either switching the thing off entirely, or, once it starts, running away up to full speed and staying there.

There are two big fat diodes in heat sinks on the board - one is I guess the SCR, and the other may be for half wave rectifying the field. But reading around suggests a 'free-wheeling' diode across the armature whose absence prevents the scr turning off - so maybe I should go looking for that next?

Taken off the motor, should I be able to see the switch-on square wave changing shape (mark-space) on a scope as I twiddle the speed control - and which terminal is which on an unknown SCR?

This is old and I've drawn a blank online looking for circuit diagrams. And - this may be stretching the board's remit a bit - in which case mods please delete.

[mark_in_manc]

oh - and one last thing. Can I run the armature from the rectified output of a variac, whilst I try to fix the controller?

[McMurdo]

I'd have thought the pot itself would be the problem. If the triac or thyristor has failed it would normally run continuously at full speed or not at all, unless that's what you're saying...is the pot switched?

The field will need excitation if you want to run the motor armature by some other means (dc of course). As a bodge you can join the field (excitation) to the armature and supply a variable DC supply. But the reason the field is normally fully powered and the armature varied is to optimise the torque and stabilise the speed under different loads. Is the dc controller branded? What's the scr type?

The freewheeling diode is unlikely to be faulty.

It won't be a square wave..that's PWM, rather exotic for an old drill. It'll be phase angle half-controlled with possibly a degree of armature voltage feedback (regulation). . Dont bother scoping it, it wont show anything meaningful without a load anyway and you usually get alot of noise and hash on a simple controller. Try some cold checks of the power devices.

[mark_in_manc]

That's great, thanks.

I was unclear - it runs at full speed if the control knob is anywhere except at 'fully slow', in which case it does not run at all. The pot is not switched.

I thought I might find a square wave (or at least a pulse train) at the gate of the scr (assuming I can find the right leg of the device!), to turn the thing on at the right point in phase each (say) positive going half cycle. If I've understood what it is doing correctly.

But OK, I found a page last night on checking an scr so I'll pull it out and do it. In the process I hope I'll learn what type number it is.

The device and motor are by 'Neco Controls' for what it's worth, who now seem to make garage door openers.

[mark_in_manc]

Here's a couple of pics. The black square box resting on the cover is normally a cover for the square circuit board which you can see sitting on top of the main board.

What I think is the SCR is sat on a big heatsink NW of the big electrolytic to the right of the board. Just below it is what looks like a power diode in another heatsink (pressed over the can), which might be to 1/2 wave rectify the field.

What I think is the scr is attached to the board (and its pcb track) by a big fat threaded stud - but the red and yellow wires attached to it (although solid core) don't half look weedy given it's meant to drive a 1/2 hp motor!

[Herald1360]

Half a horsepower is only 373W so even allowing for losses only a couple of amps. Those weedy wires are fine.

[TrevorG3VLF]

If you run with the field and armature connected together then, in theory, the speed will not change as the voltage changes.

A Variac does not have isolation so even though it might work, I would not do it

The field needs to be present before the armature is supplied or the current will go through the roof..

It probably has phase angle control which will be a combination of the speed potentiometer and a capacitor. I would suspect the capacitor, if this is open circuit, then the SCR will turn on as soon as control power is applied. There are some small chips there, presumably for feedback control.

Will the maker supply a circuit diagram?

[mark_in_manc]

OK, I've started on cold checks on power devices. SCR is a BT108.

It behaved oddly using a DVM, so I lashed it up to switch a 12v 21W bulb, touching the gate to the anode through a small current-limiting resistor. It switches on and latches OK. So... I guess I'm in the dark!

Oh - and I think the garage door people are a different comapny - Neco Electronics (Europe) comes up with status 'dissolved' on a web search, so no chance of a circuit diagram there.

[McMurdo]

Neco were quite common in their day and were nice controllers with full armature voltage feedback for good speed holding.
Have you checked/substituted the speed pot?

Those preset pots on the board are known for developing hairline cracks in the carbon.

Check for cracked tracks around the power resistors because continual expansion & contraction can stress the joints.

There will be a feedback resistor from the armature that ends up going back to the control board, quite a high value, that generates the negative feedback. If that goes open circuit it will give you the symptoms of all or nothing.

You can certainly run a shunt motor by connecting the field and armature together. Energy in = energy out. I do it as a rough test of shunt motors on the bench. But at slow speeds there'll be poor torque. Never run the motor without its field. The armature will try to take infinite current and try to run at infinite speed. Recipe for disaster!

PS The garage door company is the same company. Neco used to be called Normand Electrical, they made rather nice little motors a bit like Parvalux. They were bought by Henderson Doors, maybe they used to supply their motor equipment.

[mark_in_manc]

Well, it looks like the armature takes a feed from the SCR, and the other end is earthed via a 0.5 ohm resistor - which looks to my ignorant eyes like possible strategy for a current detector. The odd thing is it is a power resistor (the one just above the tag strip in my photo) but it doesn't get at all hot, unlike the two long ones at the top of the board. At the mo the controller less motor is on my electricals bench in the attic and I'm using a 60W bulb as a load on the armature, nothing on the field.

The resistor measures OK, and I've re-flowed te solder joints onto the board. No joy.

The not-earth end of this resistor returns a signal to the control board via one or two of those preset pots you mentioned - although I don't know which tag is the wiper on them, none of them measure open cct across two tags. They're proving impossible to get off the board at the mo.

[McMurdo]

you don't mean earth of course.

The current shunt is indeed to measure the current. There will be closed loop current limiting so you don't overload anything. It won't get hot.
The long green resistors will be a voltage divider to provide the power supply for the board. They do get stinking hot which is why I said check the tracks/joints.

Go over the 1/2w resistors on the board with the multimeter, see if any are high.

I can't see if the pots are marked but they're usually min speed, max speed, current limit and either acceleration or possibly IR Compensation. Min speed will return the bottom end of the panel pot to 0V, max speed usually feeds the top end of the pot with 10V or thereabouts (or sometimes trims the armature feedback voltage), IR Compensation feeds back a small amount of armature current into the speed demand and current limit of course divides down the voltage across the current shunt.

[Alistair D]

The wiper on those preset pots is the pin that is at the centre of the diameter of the body. the track end pins are closer together than the distance between them and the wiper.

The large electrolytic on the small pcb is probably the smoothing for the low voltage supply for the chips. It would be worth checking it out as excess ripple on the LT could cause all sorts of random operation. By excess ripple I mean volts.

Al

[mark_in_manc]

Folks - time for an embarrassing question. I blew a fuse earlier by trying to use the scope to measure the waveform across a component, where (I now assume) one side of the scope input is earthed, and one side of that component...wasn't.

So - since I have something which will serve as an isolating transformer - if I plug the speed control into this, can I then assume I can measure the waveform across whatever I need to?

since I have something which will serve as an isolating transformer

Yes you can, the old (and not recommended) way was to lift the 'scope ground. We used to have an IEC (kettle) lead with the ground disconnected, to indicate this the unattached earth wire was stuck out of the plug by about a foot (a plug with a hole to inspect the connexion). We used this for big kit the size of a room or two where an isolating transformer would be rather expensive and hard to move. And we never did it alone.

[mark_in_manc]

Originally Posted by McMurdo

Go over the 1/2w resistors on the board with the multimeter, see if any are high...

...I cant see if the pots are marked but they're usually min speed, max speed, current limit and either acceleration or possibly IR Compensation.

Thanks Kevin. I'm a bit slow on the uptake - I've read goodness knows how many people on various threads here saying 'have a look at high value resistors, they sometimes go up and up' but it took you saying it to me directly to get me to remember! So a 680k was o/c, and for now I've cobbled a 500k-ish out of 3 random used ones from the bottom of the drawer (imagine the poverty and squalor of my stock arrangements) - and it works! On a light bulb, anyway. I've also worked out which presets do max and min speed (middle 2) but not worked out the others yet.

Now I know how to use the scope across the armature supply without blowing fuses (thanks MerlinM.) I'll have a look at waveforms and what the SCR does to them.

The only downside of this great forum is not being able to buy you a drink when, in person, it would be just the right conclusion.

cheers
Mark

[mark_in_manc]

And it works on the motor too

Better, it seems the torque (leftmost preset pot, it turns out) was turned right down for its previous incarnation in a piece of undergrad apparatus, presumably to stop students breaking their fingers. Now I've turned it up it'll run dead slow but is very reluctant to stall. I'm tempted to take it off the drill and put it on my lathe.

While I'm feeling positive, I have a smaller DC Parvalux (mentioned up thread) motor with a worm gearbox on it and a homebrew speed controller which came from the same skip, and which does not work. The circuit is much more simple - I'll try to draw it out.

Kevin, you have a PM

Result! If you are ever in my vicinity a beer or three could be exchanged. My local (which does very good food as well as beer) is 300 yards up the hill (slope) and half way back down there is a (very good) curry house.

Back on topic, the voltage waveforms you see will look very sharp and 'orrible. It's the current that matters for motors, you could put an ohm or two in series with the winding and look at the voltage across that, the current will be voltage/resistance e.g. a 1 ohm resistance would give 1 volt per amp. This is very interesting as you will then see the inductance of the winding smoothing out all those spiky volts.

[Radio Wrangler]

DC motors with chopper controllers on lathes can do wonderful things. Take a shufti on youtube for "Monarch 10EE" to see a real beast in operation (I want one! but rare on this side of the pond). I have a 1 horse 3 phase motor and a panasonic inverter that is going to go on my myford when I get round to it. The control of the induction motor is superb, and DC injection braking is quite a trick It's almost as good as a DC machine and no brushes.

David

[mark_in_manc]

Merlin - I had a quick look at the volts across my light bulb earlier (not much L there...) and it wasn't too bad - a train of positive half waves which turn-on sharply later and later through the half-wave as you turn the speed down. There was quite a ring on the top of the turn-on step, but apart from that it looked sensible. I'm rarely in the South, so for now that curry will have to remain notional!

And David - you have PM since I don't want to make mods cross with machine tool speak

[McMurdo]

The pot that 'improves' the torque at the bottom end will be the IR Compensation. Essentially it compensates for speed-to-volts droop at the low end due to the resistance of the armature (hence losses) coming into play with lower drive voltages. It's exactly there to prevent stalling or speed variation under load, though of course you should be on a different gear/pulley ratio for slow speed work.
The trade off is that at higher speeds, too much IR compensation can make the motor hunt.
Modern machine tools are all AC motor/inverter driven and many have hydraulically shifted gearboxes into the bargain.