Controller efficiency

We use a couple of Electric "mopeds" here in the local island society. 24 V dc series (shunt?) motors, 1,5 to 2,5 kW, controlled by Sevcon Micromos units and similar. I wonder about the efficiency of the controller; how much of the controlled power is becoming heat in the controller itself?

Seems they need quite substantial heat sinks to avoid down-regulation by the integrated thermo-protector.

Anyone, CDK?

 

Hi baeckmo, It depends, which controller you are using GEN4? Normally the controller itself takes approx 60 mA at 24 Volt, but where the loss and heat comes from is from the MOSfets in the controller. If you have 24 Volt and 2.4 KW you draw 100 Ampere. Take the resistance between Drain and Source of 10 milliOhm per leg, you have a loss off 100 x 100 x 0.01 = 100 watt loss per leg. Do they only use 1 MOSfet, 1 leg, 2 MOSfets or 2 legs or 4 legs?. If they use 2 MOSfets parallel for each leg, you have 100 x 100 x 5 mOhm = 50 watt per leg.
We need more details to advice properly. Do you have all the details? Thanks
Bert

 

Thanx for that Bert. I have no idea as for how many MOSfets or legs are involved. Unfortunately the unit label is bleached and unreadable. I took a pair of photos, maybe they could tell you something. That controller is one of the older, I'd say from 1995. There is a marking on the top card saying 630 Rev 3, which ought to refer to the 630 series, but other than that ???

Anyway, with heat losses in the 100 W + range, the malfunctions we see are completely logical. The controls are mounted to a 1mm alu plate (together with a fwd/reverse contactor), which in turn is bolted to the bottom of a steel box without any ventilation. Paint and corrosion is destroying any thermal contact between the heat sink and the environment after short time in this climate. I guess there is a thermal protector that will cut off or reduce working current, leading to reduced performance or stop.

Again thanks, now off to the alu plating bin to create some cool (!) heat sinks.

 

That looks like a an SCR (silicon controlled rectifier) based unit. The typical heat loss of an SCR is about 1.5watt per ampere conducted per SCR , and it looks like you have 2 of them in that unit. This heat is transferred to the heat sink. You can work out your amps via the motor output and voltage / ~0.8 for motor efficiency factor. This gives you an idea of the controller efficiency, ballpark anyway...

Th SCRs are the puck type by Looking at it, they are closest the motor terminals. Probably a simple PWM type controller, where th output current is chopped up by the SCRs to control the output power.

 

Enjoy it. The front ones are diodes from Sevcon type 840-44245. They are ordinary automotive high current diodes as used in a alternator bridge. Probably 60 Ampere each stud diode. I cannot figure out what else they are using. Have you tried to get a technical manual from Sevcon? I tried, got the telephone number from a call girl. That certainly does not help very much. Bert

 

Hi Bodo,

An efficiency of 85% is considered to be a good trade-off between size, cost and performance. The controller is a device that rapidly switches on and off. Using higher frequencies allows the controller to be smaller and cheaper, but the penalty is more heat loss because the relatively slow MOSfets dissipate a lot during the transition from off to on and back.

 

Power black-out today, so I just saw the pictures. This is an ancient device with diodes in the end stage, so the efficiency is a bit lower. They each drop a little over 1 volt, at 50 Amps the dissipation from these two is another 50 Watts (only one conducts for half a cycle).

 

My apology Beckmo, CDK did you get those globes/bulbs which carry on burning when a power outing take place? Bert

 

No I didn't, they cost a fortune here. I'll mail you tomorrow.

 

Hi CDK, They cost here Rand 130 each, about Euro 9,- each.

Baeckmo, have you been able to get the technical manual?
Bert

 

No, it seems your girl was still on duty....... not too electrifying she was either..

I'll wait and check the effects of the improved cooling on this ancient micromos; work is in progress.

If there still are issues, I guess I have to come back to you guys for advice on substitutes.

 

Baeckmo, please note, should the stud diodes be faulty, bear in mind you have cathode to stud and anode to stud, get the correct ones.
bert