So some guys at work needed to modify a piece of equipment we are trying to field by adding an external speaker http://www.radioshack.com/product/index.jsp?productId=2104031&tab=features.

Come to find out it only works on a little less than half of the units we have.

The ones that don't work start to chirp and eventually (sometimes very quickly) alarm.

My task is to find a component in the circuit that can be replaced by the same component with different tolerances. I suspected hFE of Q3, so I tested four transistors from working units and four from the bad ones but they were all similar.

Should I have my solder shop remove Q4 so I can test them?

Is there a simple Filter solution I can stick in the base of the speaker without modifying the equipment any further (to much bureaucracy and time for that)?

Hopefully the schematic is descriptive enough to help you help me if not ask and I'll elaborate
(The red is the mod)

p.s.the speaker itself is modifiable easy enough, its the further modifying of the equipment that my boss wants to avoid.

p.p.s. If its uncool to ask about work stuff I'm sorry
thanks for your time
joe

 

How are you getting 18 volts pk to pk with a 9.5 volt supply?

The RS horn is rated at 50 watts / 8ohms
I can't see a 2N4403 providing this, nor a 9.5 volt battery?

I guess your problem is output impedance as you say. Adding a second 8 ohm speaker in parallel will reduce the load unacceptably.

Please give some more detail

 

I also note the original speaker is fed through a 10 ohm resistor (R2), whereas your addition is not?

 

How are you getting 18 volts pk to pk with a 9.5 volt supply?

The RS horn is rated at 50 watts / 8ohms
I can't see a 2N4403 providing this, nor a 9.5 volt battery?

I guess your problem is output impedance as you say. Adding a second 8 ohm speaker in parallel will reduce the load unacceptably.

Please give some more detail

the 18 volts is one reason I came here, I should have specified that that was taken off the junction of the the 2 ten ohm resistors. I guess I was thinking that the Q3,Q4 configuration was doing something that I couldn't see, because with those 2 transistors 18 Volts isn't unbelievable

The 50 watts is a maximum rating, if that helps, but it is loud enough (connected)for what we need it for. The problem, which I should have specified, only appears when the volume pot is near or at max. The nine volts is from a wall wart.

The second 8Ω speaker does dramatically decrease the Z but why would it work with approximately 40% of our units?

 

Some 2N4403 transistors have more current gain than others. The speakers work only when the transistors have a high current gain because the circuit was not designed for "worst case".

 

I also note the original speaker is fed through a 10 ohm resistor (R2), whereas your addition is not?

Could it be that simple?

I might of been over thinking this. Yeah I noticed the 10 Ω apparently my head has been up my own αƩƩ.

I'll tinker with it some more tomorrow

 

Some 2N4403 transistors have more current gain than others. The speakers work only when the transistors have a high current gain because the circuit was not designed for "worst case".

Well I tested 8 of the Q3s and they seemed to be a mixed bag between units that worked and that did not, that is both good and bad units had high and low gain transistors.

Studiot pointed out a concern I was having but failed to mention in my original post which is the 18Vpp @full volume and a 9.5V Vcc.

Should I be testing Q4

 

Well I still don't understand how you are fitting 18 or 20 volts peak to peak between 9.5 volts and earth?

Is the RadioShack horn the equipment you are trying to modify or the extension speaker?

What do you mean by

a piece of equipment we are trying to field

?

 

Well I still don't understand how you are fitting 18 or 10 volts peak to peak between 9.5 volts and earth?

Is the RadioShack horn the equipment you are trying to modify or the extension speaker?

What do you mean by ?

the entire circuit posted is the audio portion of an instrument, the red wire and RS speaker is what we added (plus BNC connectors).

we can mod the speaker if need be; the equipment we can only swap parts for the same part.

I was hoping someone could point out a characteristic of a particular component that would explain why some work and others do not, like gain.

I also wanted to see if anyone else could explain the 18 Vpp thing

 

But isn't the RS speaker self powered?

 

But isn't the RS speaker self powered?

If the one I posted is powered then i put the wrong one up.

The ones that we have had, for apparently some time now, came with two bare wires that we added BNC connectors to.

Powered would be ideal if battery life were not such a concern.

I have already demonstrated that a simple high impedance amp circuit solves the chirping/alarming issues however extra wall warts and/or batteries have been deemed "undesirable" from the upper echelon that purchased these instruments.

...thanks for the replies by the way.

 

Well it looks like a PA bullhorn.

 

It basically is, don't get confused by its name though, when we were looking for powered speakers on their site we came up with this same model, we of course already had several.

If you look under "Product Summary" you'll see the section "You'll also need", where it lists amplifier.

All my boss could find as far as powered speakers had some kind of Ipod affiliation.

 

Best I can tell is... If your new speaker is added in parallel like shown on your drawing, you went from an output load of 28Ω to about 15Ω. You're running this off of a wall wort? Probably pulling the voltage down to something low enough to make the circuit fail.

Management wants to get a high-powered output without providing enough power from the supply?? When they get that plan to work, let me know. I'll want to buy some stock in their company. Till then, either increase the power in, put a big capacitor across the supply to provide for the short peaks of high power requirement (might work), or arrange the outputs to not have apply more than the previous 28Ω load (preferred).

Why do some of them work? You just got lucky. When you're teetering on the edge, some don't fall off. Doesn't mean that those that did are broken, it just means that all of them need to be further from the edge.

 

Excellent assessment and summary with the information to hand Don.

What do you make of the 18 volts peak to peak fitted between 0-9.5 volt rails?