Switch efficeincy

[Dragon Tamer]

As some of you may know, I have been working on my shoe box radio. I have been building the power supply for it and I decided to include some TTL switches as part of the power supply. The problem is that the radio sucks up an enormous amount of power (not too bad though) during high power base frequencies. It sucks up so much power that the radio will "stutter" because the built in filter capacitor can't handle the load. I've added some external filter capacitors to help it cope, but the problem persists.

In the old design the batteries power went directly into the voltage supply for the radio with a standard SPST switch. In the new design however, it has to power the TTL switches, and go through a PNP transistor (configured as a collector follower) before getting to the actual radio. I am worried because I think that the PNP is the actual cause of the problem by not enabling enough current to go to the radio. If I can't fix the problem then I'm going to have to get rid of the transistor and replace it with a relay which is not what I want to do (I don't want replace an electromechanical device that is going to wear out if 5 months).

Does anyone have any ideas on how I can fix this problem? (I will post a schematic later once I finish it)

 

[bychon]

Sweat thee not about relay survival. In air conditioners, we use relays that provide 100 amp starting power and they live a quarter of a million cycles.

 

[tvtech]

Sweat thee not about relay survival. In air conditioners, we use relays that provide 100 amp starting power and they live a quarter of a million cycles.

We all have our specialised fields here. Share brudda, share.

Cheers.

 

[bychon]

Can I get a "Halleluia"?

 

[Dragon Tamer]

The radio gets turned on and off a lot. My relays are rated for 100,000 cycles (probably about half that now) they were exposed to a situation where they wer forced to osccillate at a high speed and an unknow number of cycles occured (by the way is a cycle a full on and off or just on or off?)

But the TTL switches consume about 50mA from a 2500mAh battery, (doesn't seem like that much) however the unit could be unpluged for several months at a time and that would drain that battery in a heart beat.

 

[Reloadron]

OK, so your initial design used a switch correct?

In the old design the batteries power went directly into the voltage supply for the radio with a standard SPST switch.

Why did you depart from using a simple switch which apparently worked just fine?

Next, I agree with bychon in that the use of a mechanical relay is not a bad thing. I am surrounded by relays that have beek cooking around here for over 30 years of regular use. If you wanted you could use a relay in a latching configuration for on / off using momentary buttons.

However, why worry about incorporating electronics to turn something on and off when a simple switch will work?

Ron

 

[Dragon Tamer]

I was trying to make a better system becuase with the old switches, the switch would get cought on things and flip on but there was no way to know if it was on or off unless you looked directly at it.

For this design I wanted to get rid of all of the old switches and replace them with push buttons that can share an input with a microcontroller. The old design would not allow a microcontroller to bypass the switch. (not very easily)

 

[Reloadron]

OK, now I understand a little more. When you get it together how about just posting a schematic or drawing of your current on / off portion.

Ron

 

[audioguru]

What is a "TTL switch"? Power-hungry TTL logic ICs have not been used in new designs for about 40 years.

Since you keep making new threads without any links to your old threads then I might spend some time searching for your other threads to see what you are talking about.

 

[Dragon Tamer]

I have a PNP transistor and a 1N400X diode before the switches actually get the voltage from the battery (the PNP is connected to a battery monitor for low voltage cut off) There is also a diode from the 12V regulator that operates off of the AC wall voltage (to keep the battery from powering everything else in the circuit). I have also put a push button switch on the positive end of the battery so I can shut off all of the power to the circuit.

In the schematics I have posted the connectors that say optoCOMP# are all connected to the same photo-comparator that shuts off the LEDs if it gets too dark (additional battery saver), the connectors that say uC are where the micro-controller's input will go, each connector that says VU, Amp, FM, and Charger go to the respective portion of the circuit.

The first circuit is of the circuit before the TTL switches, the second circuit is all 4 circuits connected together, and the third circuit is just a single TTL switch.

 

[audioguru]

Your schematics do not show the voltage of the battery so I guess it is 15V. Then the 470 ohm base resistor for the MJ2955 transistor produces a base current of 29mA. Then the MJ2955 will switch only 290mA well.

I do not see any "TTL switches" anywhere on your schematics.

You show an N-channel IRF510 Mosfet but the circuit needs a P-channel Mosfet.
You show many 2N3906 PNP transistor that are supposed to be 2N3904 NPN transistors. Then the pairs of LEDs can be connected in series and driven from a single 2N3904 transistor.

Why use current regulators for the LEDs instead of a simple resistor?

 

[Thunderchild]

ahem: ever heard of a mosfet ?

 

[Dragon Tamer]

I have been using MOSFETs in part of the design, but a friend recommended that I try an analog switch. Would that make much of a difference? The ones for audio applications seem to have a low resistance state.

 

[audioguru]

A CD4066 has a fairly low on resistance but it should have a load of at least 10k ohms.
It is called a Cmos bilateral switch not a TTL switch. The CD4051, CD4052 and CD4053 are multiple switches.

 

[Thunderchild]

you could get something like and IRF9540 that has a on resistance of 0.117 ohms and will handle up to 28 amp (33 if you get the "N" model)

 

[Dragon Tamer]

For the radio, the switches have to have an on state resistance of less than 10 ohms othrwise it will affect the performance of the radio. I have been trying to track down some switches that are also bidirectional (I have some other things that need to have switches replaced). I have been looking on Maxim's web site (because I can get free samples ) and have yet to find anything that fits my needs.

The TTL part of the circuit (which is actually CMOS, sorry) is the JK flip-flops that enable the dilfferent outputs. I have thought of a few ways to improve circuit efficiency though, the biggest way is that I'm going to replace the all of the 555 timers with 1 TLC555, and 2 TLC552 timers.

There is the PNP transistor that is powered by the ICL7665 that I would also like to replace with an analog switch, but it needs to be able to handle at least 1 Amp of current (to allow adequit room for the current to vary during opperation).

 

[Thunderchild]

I think an analogue switch is essentially a mosfet as it will pass power both ways

 

[Dragon Tamer]

Not all of the analog switches I was looking at would pass current in both directions. Most of them would only pass current in one direction. I also thought that a MOSFET would also only allow current to pass in one direction, or am I wrong about that?

 

[audioguru]

A Mosfet switches DC. An analog switch switches AC or DC. A CD4066 has an on resistance of 90 ohms and a CD4051, CD4052 and CD4053 have an on resistance of 120 ohms but it varies with signal level which causes distortion. If the load is 10k ohms and higher then the distortion is very low.

 

[mneary]

I also thought that a MOSFET would also only allow current to pass in one direction, or am I wrong about that?

A MOSFET will pass current in both directions. It is like a low value resistor in both directions when it is fully turned on. In one direction there is an intrinsic diode in parallel, but if the MOSFET is turned on, you will not see it.