Driving a transistor problem

[BrendanP]

I've canned the MAX713 battery charger and am now going with a linear design.

I have a 08M pwming a TIP35C NPN transistor, by varying the pwm rate I can controll the I going into the battery. I have a DS18B20 temp sensor encased in the battery pack. (btw battery pack heat shrink is a special product and works a lot better than common heatshrink). With this set up I can terminate fast charge on temp rise and time and delay fast charge if battery pack is under/over temp. I have it bread boarded and it all works well.

Problem is I need to be able to run the 08M via a V regulator off the same DC 13.8V supply that is charging the battery. I'm switching the low side of the battery supply with the NPN. If I try and run the picaxe of the same supply that is charging the battery then the batterys low side is always grounded by the picaxes ground.

Ive had a good look around on google for ideas of how I can get around this. I cant find anything probabaly because I dont even know what such a circuit is called. I thought there might be a commonly used soloution that someone on the forum would know of.

I learnt about H bridges and potential dividers by asking questions here in the past.

 

[Dippy]

Without a schematic in my face, I can't exactly follow what you are saying, but have you considered a PNP or P chan MOSFET?
I dod't know if there are any circuit issues that preclude either method.

 

[moxhamj]

I always get confused driving the low side. The earth isn't an earth or it is sometimes but not others. And since there is always an equivalent high side circuit, I'd look at high side switching. Use a PNP on the high side, and pull its base down through a NPN small signal transistor like a BC547 - via ?1k. Emitter of the 547 to ground.

Then drive the base of the 547 via 2k7 from the picaxe.

It is one of those generic circuits that just works.

 

[BrendanP]

Once again a big thank you Dr. I'll bread board it now.

 

[westaust55]

I presume that you are endeavouring to charge the battery while it is in circuit and thus the battery negative is still conencted to the circuit ground. ?

 

[BrendanP]

Yep Westaust that is precisely the problem. (I lived in Geraldton for a few years, use to go fishing at Kalbarri and at the Abrollouhos Islands, massive fish there.)

The battery pack powers a picaxe/gsm module in a small housing.

I want to be able to charge the device off a 13.8V auto supply. The battery is fixed in the housing.

Im not going to have time to get a switch mode working so the picaxe/ linear/DS18B20 is good stop gap.

 

[hippy]

High-side switching always seems simpler, more logical and more understandable to me although it may require slightly more complicated analogue control ( and I'm well out of my depth there anyway ).

 

[BrendanP]

Well, I have the good Doctors circuit breadboarded beside me and ITS ALIVE!
Battery is being charged and picaxe doing pwming and monitoring temp slope is being supplied (via V reg) off the same supply.

Im getting corrupted data from the DS18B20, I presume this is because of the switching noise on the V supply now. I'll try a few caps to smooth things tomorrow morning.

I presume a potential divider across the battery would be the way to monitor its V level?

 

[westaust55]

okay, then the circuit by Dr Acula's circuit is best based on PNP switching transistor.

What sort of battery are you charging?

Ideally and particularly of lead acid batteries, you should wait a period (30 min) for the "surface" charge to settle before using a voltage measurement to ascertain the state of charge.

 

[Dippy]

This is an appendix to Brendans other thread using the MAX charger where he was charging a couple of MiMhs in't it??????

I'm sure if he duplicates the MAX's charging schemes as closely as he can then he'll be just fine. No panics needed - unless he's already sent the gerbers off

 

[Andrew Cowan]

Measuring the voltage alone can't tell when if it is charged - an unloaded 9.6V Ni-Mh battey fresh of the chrger can measure 12V+. Are you thinking about measuring the voltage drop when the battery is charged? (Delta v?)

Andrew

 

[westaust55]

Measuring the voltage alone can't tell when if it is charged - an unloaded 9.6V Ni-Mh battey fresh of the chrger can measure 12V+. Are you thinking about measuring the voltage drop when the battery is charged? (Delta v?)

Andrew

True and exactly what I was saying.
You need to let a battery sit (unloaded) for a period at least 30 mins and preferrabley 2 hous before trying a voltage measurement

 

[BrendanP]

Yep I'm measuring a 3X NiMH in series.

I may just stick with temp rise and time to cease fast charge. I have noticed the battery pack voltage drops back to its "real" voltage quite quickyl after the cessation of charging.

It is also possible for me to put a load on the battery while checking its V to get a more accurate idea on what is going on.

V measurement would be good to have. Im not sure or even if its possible how I could go about connecting the battery pack so it its in circuit being charged and then when external power is removed it is then powering the circuit and then its V can be read with the method below.

I have been using a LED to measure V. Andrew Hornblow of Bright Sparks in NZ taught me this trick. It works well.


high 1 'Make pin 2 go high
readadc 1,b3
readadc 1,b3
readadc 1,b3 'Measure the forward voltage on the LED which
'indicates the battery voltage inversely
low 1 'Turn off the LED and clamp pin 2 low again