Basic problem with transistors

[Nigel Goodwin]

What? When did the circuit have 3 batteries? And why would it only get 1.2V?

Feeding the motor from the emitters of the transistors is the worst possible method, and will lose you the most voltage possible - using only a 3V supply to power an h-bridge is really far too low, you will lose more then the motor gets.

 

[giftiger_wunsch]

Feeding the motor from the emitters of the transistors is the worst possible method, and will lose you the most voltage possible

What would be a better method?

- using only a 3V supply to power an h-bridge is really far too low, you will lose more then the motor gets.

Perhaps I should increase the voltage of the power supply then; I was sticking with the D-Cells since they were already being used to power the robotic arm, but if the voltage drop due to the transistors is going to be a problem for the motors, perhaps I should instead use 3-4 AA NiMH cells (I have quite a large stockpile of these ). It's been mentioned that 3V motors can typically be used at much higher voltages, and making the supply 3.6 - 4.8V should account for some of the voltage losses.

But of course if there's a way to avoid some of the inefficiency rather than simplying throwing more power at it to overcome it, then I'm all for that idea

 

[marcbarker]

Has anyone actually Simmed this circuit yet? The Discussion/Reality ratio here is becoming not unlike a NewsGroup discussion! **broken link removed**

It was suggested that the motor will only get 1.2 V (40% speed probably OK?). Can some analysis (or testing if easier) be posted to prove it?

My understanding of this minimalist circuit has not gone very deep, as far as I'm aware it behaves like a car's PAS, and the base current is analogous to force feedback from the wheels and steering rack. I do hope this circuit works, if it doesn't then go with something like the IC solution posted earlier which seems to be more the 'done thing' for motor control.

 

[Hero999]

Here's how it should be done.
How much current does the motor need?

You may need to reduce the resistor values a bit.

As shown, it should be fine up to about 240mA.

 

[Nigel Goodwin]

What would be a better method?

Driving the motor from the collector, the voltage drop is a LOT lower, and you've also got far more base current available to drive the transistor ON properly.

 

[giftiger_wunsch]

Here's how it should be done.
How much current does the motor need?

You may need to reduce the resistor values a bit.

As shown, it should be fine up to about 240mA.

Motor requires ~200mA. Sorry if this is a silly question, but what are CW and ACW? From the schematic it seems they refer to PA0 and PA1 on my uC. Clockwise and Anti-Clockwise?

Driving the motor from the collector

Sorry, but I don't quite understand the difference. Could you provide a basic schematic or point me to a webpage which explains the difference? Meanwhile I'm going to try to look it up myself, the tutorial I was reading previously probably highlights the differences.

 

[Nigel Goodwin]

Sorry, but I don't quite understand the difference.

With the motor driven from the collector you lose probably 0.2V, with it in the emitter you probably lose 1V or more, particularly as you have little current to drive the base with, so the transistor isn't turned on very well.

 

[marcbarker]

Yes, there you go Giftiger, the circuit posted by H_999 is more 'the done thing'.
Although it's not as minimalist as the 2Fet 2BJT solution, this has the advantage that it uses cheap transistors, you know they're going to run saturated (that's good). As Nigel says, the 2 good points are it's driven from collectors and the base drive current is minimal.

Back in the very old days of electronics it was the 'done thing' to use the mininum number of transisitors, but that was because they originally cost more than valves, also came in individual cardboard boxes, that was how expensive they were. Nowdays you throw as much semiconductors as you can at the solution, because they are so cheap.

I still like the minimalist circuit though, it has a kind of 'organic' appeal to it! I would like to see it working, even if only a spice simulation.

 

[giftiger_wunsch]

But I thought the transistor only allows current to pass in one direction between the emitter and collector... *goes back to looking at the transistors tutorial*

 

[giftiger_wunsch]

the circuit posted by H_999 is more 'the done thing'.

I'm going to take a good look at it and make sure I understand how it works

 

[marcbarker]

I think you need to look in the tutorial the difference between "Emitter Follower (common-collector)" and "Common Emitter" configuration.

Having said that I don't think the upper transistors in the 2fet/2bjt circuit fall into either those 2 configurations, I suspect it fall somewhere between the 2.

But that's the thing about the design biz, when you are working among peers, you need to carry out design methodology which is mainstream and understandable. There are some really great and clever designs have been around which are economic, reliable and efficient, but unfortunately they fall down because they are difficult to understand (or classify), so these designs die out.

 

[Hero999]

Yes, ACW is anti-clockwise and CW is clockwise.

I missed the protection diodes off.

Another option is to use MOSFETs for the low side and BJTs of the high side.

I'd just go for the all BJT option though, although it uses more transistors, it'll be cheaper, more reliable and the components will be easier to get hold of - no search for exotic low threshold MOSFETs.

 

[indulis]

MOSFET's have body diodes, so you don't need any external.

 

[audioguru]

Hero's circuit is an H-Bridge. To run clockwise then TR1 turns on TR4 and TR5. To run counter-clockwise then TR2 turns on TR3 and TR6.

The analysis is:
1) If the output voltage of your micro is 2.9V then the base current of TR1 is (2.9V - 0.7V)/1k= 2.2mA.
2) The collector current of TR1 and base current of TR5 is (2.9V - 0.8V)/100 ohms= 21mA.
3) The collector current of TR5 is 21mA x 10= 210mA for the motor and for the base current of TR4.
4) The motor gets at least 210mA - 21mA= 189mA at 2.6V.
If your transistors have very good spec's then the motor will get more current.

 

[marcbarker]

To run clockwise ...

Will that be true if the motor was wired backwards

If your transistors have very good spec's then the motor will get more current.

Can you elaborate a bit more on what exactly is "transistors that have very good specs" ?

 

[giftiger_wunsch]

Thanks for the explanation audioguru. I'll probably redraw the schematic so that I am sure I understand each component of it, as well as expanding the positive and negative rails to show the battery and uC ground attached. I believe I should be able to use my original idea of using PA1 for PA0's ground and vice verca, simply by connecting the green section of my schematic to the 0V rail on this schematic.

The main difference really just seems to be that this schematic uses one pair of transistors to activate a further 2 pairs of transistors, whereas my schematic used the uC to directly activate 2 pairs of transistors. I assume this is mostly because of the current issue.

Looking at audioguru's analysis of the schematic, it seems to fulfils the requirements - the current through the uC is around 2.2mA, well within the limit, and the current through the motor seems adequate.

What is the purpose of the protection diodes? I know that a relay coil produces a high-voltage spike which requires a protection diode to protect ICs driving them, but didn't realise this also applies to BJTs?

 

[giftiger_wunsch]

Speaking of redrawing the schematic, can anyone suggest some good freeware schematic drawing tools? Using powerpoint is becoming tedious even though I've already created most of the common components now. It has problems with correctly aligning wires and I'm a bit of a perfectionist.

 

[audioguru]

Will that be true if the motor was wired backwards?

Isn't everything upside-down in Australia and in New Zealand and motors run backwards?
One end of a motor turns clockwize while the other end of the same motor turns counter-clockwize. Isn't that amazing?

Can you elaborate a bit more on what exactly is "transistors that have very good specs" ?

Transistors have a wide range of spec's. Some saturate much better than others even if they have the same part numbers and are made by the same manufacturer. Most Zetec transistors saturate very well.
Can you run as fast and as far as me?

 

[giftiger_wunsch]

Can marcbarker and audioguru please stop fighting >_> marcbarker, I thought 'good specs' was enough explanation on the transistors, not every point made has to be accompanied by firm figures. And I don't see how the comment about the motor being wired backwards was relevant, because the wiring is very clear from the schematic, and I felt audioguru's explanation was also very clear and useful. I don't know what the facetious over-analysis of virtually every suggestion made by audioguru has been about but unless you have a criticism which will add to the discussion or help with the schematic I'd appreciate it if you didn't bury the actual issues with posts of unrelated sniping at another poster's comments.

Thank you to both audioguru and marcbarker, you have both been very helpful in improving my schematic. Let's all just get along

 

[Chippie]

Speaking of redrawing the schematic, can anyone suggest some good freeware schematic drawing tools? Using powerpoint is becoming tedious even though I've already created most of the common components now. It has problems with correctly aligning wires and I'm a bit of a perfectionist.

download Eagle from Cadsoft...It will do pcb and schematic layouts...
The schematic editor is brill..