LT SPICE transformer

[Sceadwian]

Hearing loss is additive, so even if you aren't bothered or able to notice the sound it's level with your hearing aids off you can still be further damaging your ears, I'd strongly recommend earplugs, it should also be done in a closed room with no other living critters nearby.

 

[MrDEB]

CRITTER UPDATE!!
The dang deer were no where to be found . Maybe due to time of day etc. I had to turn off after 20 minutes as the TO-22 pkg transistors got really warm.
Thinking I should either just add heat sinks and/or change to a TO-3 package. Add some heat sinks.
Change the base resistors to be equal. Audio guru recommends 270. To be conservative I'm thinking 330 -400 ohm. Don't want to drive the PIC into oblivion.
For now just add some heat sinks to the to-220 's
Then go out looking for more deer. The cattle looked up, horses didn't seem to have any reaction.
using TIP 31 and TIP 42 transistors. Pretty sure I can bolt a strip of aluminum to connect all together as tabs are isolated?. Need to verify before.
I still can't get over how loud it is. Can hear slightly while driving.

 

[audioguru]

The metal tabs of TO-220 transistors connect to their collectors. The schematic shows that one NPN collector connects to one PNP collector.
Heatsinks have a high surface area since they have fins. Black anodizing causes better radiation of heat.

 

[MrDEB]

So I need to install seperate heat sinks for each transistor (except the PNP.
Thinking of just going with TO-3 package as I have lots of 2n3055 and 2n3054s. Just need to configure a heat sink. Thinking a piece of tin with fins or just an aluminium plate/ Have seen heat sinks fashoned from copper pipe as copper I think is the best? but alumiium is easier and cheaper.
Using the TO-3 package I will have a larger surface area..
Shouldn't affect performance by using 350-400 ohms on the PIC to the 2n2222"s. Will just get the two the same resistance.
Reason I changed from 150 to 300 was current draw on the PIC output. 30ma vers 16ma,(thinking of going down to 10ma draw(400ohm)

 

[audioguru]

The TO-220 output transistors should not get hot if they have enough base current because they are switches, not linear amplifiers.
They produce a square-wave, not a sine-wave.
When they turn on their current is high but the voltage across them is low so their heating is low.
When they turn off the voltage across them is high but their current is zero so they do not heat.

I corrected your schematic and added voltages when the PIC output current is high.

 

[MrDEB]

Stupid but just got to ask?
R9 has 4.5v - 1.6v. am assuming the 1.6v is the drop across the resistor?
same goes for R 2
math is not my best subject.
Dim light in Idaho

 

[audioguru]

R9 has 4.5v - 1.6v. am assuming the 1.6v is the drop across the resistor?

Yes.
One end of R9 at the output of the PIC is at about +4.5V and the other end is at the base of an NPN transistor at about +1.6V so the resistor has the difference across it.

same goes for R2

Yes again.
One end of R2 at the output of the PIC is at about +4.5V and the other end is at the base of an NPN transistor at about +0.8V so the resistor has the difference across it.
These resistors limit the current from the output of the PIC.

 

[MrDEB]

I tried to figure out HOW you compute voltage drop across a resistor??
very elementary question but curious as to calculation.

 

[audioguru]

The voltage across a resistor is simply to subtract the voltages at the ends.
I know that a PIC drops about 0.5V when it has a load on its output of 20mA to 25mA.
I also know that a power transistor uses 0.8V at its base and its driver transistor also uses 0.8V at its base and the two bases are in series and add to make 1.6V.

The inverter transistor has about 0.8V at its base.

 

[MrDEB]

Just for a learning experience I am going to simulate using LT SPICE as well as TINA
Not that I don't believe you but maybe , just maybe I might accidently learn something. I have done the same in the past as POW! I actually learned something .
Thanks again.
Tomorrow I plan on making the resistor changes and trying to find some deer to scare.

 

[MrDEB]

TINA simulation BUT??

current draw is not what I expected on T7 to T! = -2.38pa
Must be a simulation issue.
Input is 10khz
You were close on most everything else.
My plan is to place tweeters in a box to muffle the sound then measure voltages in real time.
NOTE the R4&,R7 (R%&6 in your posted schematic are 50 ohm instead of 33ohm)
This shouldn't be an issue as I have lots of 50 ohm resistors on hand buit couldn't find 33 ohm.

 

[audioguru]

current draw is not what I expected on T7 to T! = -2.38pa

It is correct. The inverter transistor T7 is turned on so it turns off the entire right side of the H-bridge. The entire left side of the H-bridge is turned on and draws current. When the input signal goes low then the inverter turns off, the entire right side of the H-bridge turns on and the entire left side turns off.

Your simulation shows +3.48V at the base of T2 and +10.66V at the collector of T4 which are very wrong.
The base of T2 is two base-emitter voltage drops above 0V which is about +1.6V to +1.8V.
The collector of T4 should be about +11.5V since T2 and R4 turn it on very well for it to supply about 1.4A to the 8 ohm load.
The voltages and currents will be different when the low current piezo tweeters replace the 8 ohm load.

 

[MrDEB]

Thats what I kinda figured on the tweeter load.
But it is interesting none the less.
So no issue with the 33 ohm resistor (curious as to why 33 ohms) vers the 50 ohm I used?
Supposly by changing the R9 (your schematic) to a 150 ohm, changing R2 from 2.7k to 1.2kand eliminating R10 (150 ohm) the transistors shouldn't get warm?

 

[MrDEB]

updated schematic

To avoid confusion here is the updated schematic

 

[MrDEB]

oups R5 & R6 are supposed to be 50 ohm
my bad

 

[audioguru]

I used 33 ohms so that the output transistors have enough base current to drive a 3A motor. An 8 ohm speaker has a peak current of only 11V/8 ohms= 1.4A so the 33 ohm resistors can be as high as 68 ohms. 50 ohms is not a standard value but can be used. The 33 ohm resistors dissipate about 1.5W each so 2W resistors should be used. 50 ohm resistors dissipate 1W each so 2W resistors should also be used.

The transistors work fairly hard and there is a moment when both sides of the h-bridge are turned on so they short circuit the supply for the moment with a very high current.

The TIP31 and TIP42 transistors are fairly slow so that they ramp instead of switching quickly which creates some heat. The "ramping" also increases the amount of time that both sides of the h-bridge short-circuit the supply.

 

[MrDEB]

The TIP 31 and TIP42 are what I had as far as pairs.
Suggestion on faster switching transistor?
I need to purchase some larger wattage resistors

 

[audioguru]

You might not notice much difference in heating if you use faster output transistors.

 

[MrDEB]

Still contemplating using the TO-3 package transistor (2n3055 or 2n3054)n An MJ2955 is the complimentary PNP
The TO-3 pkg allows lower profile and perhaps a larger heat sink area.
Will keep you posted as I need to go purchase the 2watt resistors. I have lots of 300 ohm 2 watt resistors, may just put into a parallel configuration.

 

[MrDEB]

I replaced the 50 ohm 1/4 watt 50 ohm resistors with 2-100ohm 1watt resistors (All Rat Shack had in 1 watt.
The 1/4 watt resistors are slightly off color (read ting of black/gray color like it got hot))
Now to find some deer and install some heat sinks