[OpenTRV-dev] H-bridge transistor selection

[Mike Stirling]

I'll sketch something out later. The one you have in your head is probably right, though. You need to pull down the low-side and pull up the high-side, and remember that no current will flow through any of these resistors with the bridge turned off because there is no voltage across them.
 
----------------original message-----------------
From: "Damon Hart-Davis" EMAIL ADDRESS HIDDEN 
To: "Closed list for developer discussions" EMAIL ADDRESS HIDDEN 
Date: Thu, 16 May 2013 14:03:13 +0100
-------------------------------------------------
 
 
> I will try hard to avoid the last option!
> 
> Thanks
> 
> Damon
> 
> PS. Do you have a diagram to hand for a power sipping design? In my head I'm seeing 
> 'leaky' pull-ups, though they could at a pinch be pulled up via my intermittent 
> I/O rail for example...
> 
> 
> On 16 May 2013, at 13:40, Mike Stirling wrote:
> 
>> Yes. Drive each high-side transistor from the collector of the opposite 
>> low-side one. This also solves the problem of high-side drive when V+ is higher 
>> than logic level - not an issue in our case.
>> 
>> This topology leads to the following truth table (and no need for 
>> tri-state):
>> 
>> 00 - Idle (all off)
>> 01 - Motor run forward
>> 10 - Motor run reverse
>> 11 - Magic smoke (all on)
>> 
>> Mike
>> 
>> ----------------original message-----------------
>> From: "Damon Hart-Davis" EMAIL ADDRESS HIDDEN 
>> To: "Closed list for developer discussions" 
>> EMAIL ADDRESS HIDDEN 
>> Date: Thu, 16 May 2013 12:30:46 +0100
>> -------------------------------------------------
>> 
>> 
>>> Hi,
>>> 
>>> I *should* be confident with this, but I am not.
>>> 
>>> Is it possible to run a bipolar H-bridge from 2 tri-state outputs properly 
>>> 'class 
>>> C' ie avoiding ever having shoot-through by being able to let all four 
>>> transistors be off at once?
>>> 
>>> It's trivially possible with 4 tri-state outputs, or even 4 
>>> non-tri-state 
>>> outputs.
>>> 
>>> Rgds
>>> 
>>> Damon
>>> 
>>> 
>>> On 16 May 2013, at 11:38, Stuart Poulton wrote:
>>> 
>>>> On 16/05/13 11:32, Mike Stirling wrote:
>>>>> Interesting. The poor performing transistors do so because he is 
>>>>> running 
>>>>> them far too hard, although he does acknowledge this. He is also 
>>>>> running out 
>>>>> of gain at the lower voltages, and would get an improvement there by 
>>>>> reducing 
>>>>> the value of the 1k base resistor. This is evidenced by the 
>>>>> reduction in 
>>>>> voltage drop with increasing supply voltage. The important 
>>>>> figures in the 
>>>>> datasheet are Ic from the absolute max ratings, which needs to be 
>>>>> well above 
>>>>> the expected stall current of the motor, Vce-sat, which should be 
>>>>> small, and 
>>>>> the worst case hFE (large signal current gain), which should be 
>>>>> high enough 
>>>>> to allow for an acceptably low base current during motor stall 
>>>>> conditions. 
>>>>> Mike 
>>>> 
>>>> Mike,
>>>> 
>>>> Coming up with a nice H-Bridge for battery operation would be great, 
>>>> esp if it 
>>>> could be driven with a low pin count from a uP.
>>>> 
>>>> I did measure the stall current of my dissected TRV head,
>>>> 
>>>> 2 - Motor run current (no load) 13-15mA
>>>> 3 - Motor Stall current - 95-100mA
>>>> 
>>>> Cheers
>>>> 
>>>> Stuart
>>>> _______________________________________________
>>>> OpenTRV-dev mailing list
>>>> EMAIL ADDRESS HIDDEN 
>>>> http://lists.opentrv.org.uk/listinfo/opentrv-dev
>>>> 
>>> 
>>> _______________________________________________
>>> OpenTRV-dev mailing list
>>> EMAIL ADDRESS HIDDEN 
>>> http://lists.opentrv.org.uk/listinfo/opentrv-dev
>>> 
>> 
>> -- 
>> 
>> 
>> 
>> _______________________________________________
>> OpenTRV-dev mailing list
>> EMAIL ADDRESS HIDDEN 
>> http://lists.opentrv.org.uk/listinfo/opentrv-dev
>> 
> 
> _______________________________________________
> OpenTRV-dev mailing list
> EMAIL ADDRESS HIDDEN 
> http://lists.opentrv.org.uk/listinfo/opentrv-dev
> 

--