Old circuit needs to be re-born Smaller

[Tanker135E]

Hi As this is my first post I might not get the attachment loaded so be easy on me. Little background on me and the circuit. I am a Model Railroader and use different circuits that others have made and I employ. My electronics knowledge is minimal. That is I understand somethings but mostly don't know why they work. I am a smart guy but needs to be told more than once. The circuit I am trying to recreate (make smaller) is what I have come to learn is a amp/booster, they who build it called it an H-bridge.(1994) What it does for me is, it fits between a throttle/decoder and protects it from thermal overload/over amperage. I am working with a fairly old command control system that is mostly an analog.(Keller Onboard 1979) Just in case I can't load the image I'll tell you about the circuit. It receives power (14v) thru the BR and then to 4 transistors 2 NPN,2 PNP there are 2 1K 1/4w resistors. the throttle sends its signal via Brn Org wires then this circuit sends the signal to the motor (engine) this little booster takes all the heat there by saving the no longer made throttle. I just thing it can be made smaller and easier since we are 15 years further advanced? Ok when I try to upload it says I am missing a security token? steve Try this link and you can view the schematic
https://i44.tinypic.com/2iux5lc.jpg

 

[Ubergeek63]

At what current?

https://www.electro-tech-online.com/custompdfs/2008/12/3959-1.pdf with only a few control inputs and a pot for speed control will control your engine up to 3A according to the sheet ... if that is actually what you are doing.

Dan

 

[Tanker135E]

The engines draw less then .6 amp and the throttle/decoder is rated at 1 amp. The problems that occurs is from high current heating of the throttle This little booster called a H bridge amplifier make the throttle a preamp and makes the throttle run cold at a constant 1/10 amp and the booster takes all the current/heat. It still works today as it did in 1995 but I am trying to see if there is a better way such a an IC that does this in a nice little package. steve

 

[Tanker135E]

Hey all please see the first post as i was able to get a link to the schematic

 

[Ubergeek63]

Well it would not act the same, but that IC would work. It takes a linear voltage in to control the motor current and a direction signal that can be tied to a switch.

It looks like your controller just controls the voltage to the engine, the IC would give you low end torque so it would no longer seem as sluggish starting up. (speed is voltage after losses)

 

[arhi]

Check out the L293 it can take 1A easy. It is 4x 1/2 H-bridge (so 2x what you need)
L293

 

[Roff]

Check out the L293 it can take 1A easy. It is 4x 1/2 H-bridge (so 2x what you need)
L293

L293 requires digital inputs. Tanker's input is a pot. The circuit he posted is just two push-pull emitter followers. It's all analog. PWM should work, but would require a bigger redesign.

 

[arhi]

yes, my fingers were faster then my brain

 

[bigkim100]

reinventing the wheel

Ok..as per usual, we are trying to reinvent the wheel....and then reinvent that invention. I could make the original circuit on a 1ich by 1inch piece of perfboard just using the original (or replacement) transistors and bridge. Does it REALLY have to be smaller than that?..usually cars have a TON of space inside to throw in another circuit board.
As far as easier...it really doesnt get any easier than 4 transistors and a bridge.

 

[Roff]

Ok..as per usual, we are trying to reinvent the wheel....and then reinvent that invention. I could make the original circuit on a 1ich by 1inch piece of perfboard just using the original (or replacement) transistors and bridge. Does it REALLY have to be smaller than that?..usually cars have a TON of space inside to throw in another circuit board.
As far as easier...it really doesnt get any easier than 4 transistors and a bridge.

The transistors probably need heat sinks.

 

[Tanker135E]

Well I was waiting for this type of question. Yes it is small but with the rest of the components it rarely fits in the engine where it belongs so the engine can act like the prototype. If you place it in a car its always married to that car. I have to do this in many cases as it is already with the sound chips and speakers. Most of the time I have to run wires to the 2 sets of transistors that are located 5 inches apart from all the other components. I running HO Scale. The throttle is .375 x 1.375 x .700 those transistors get hot they melt plastic in minutes! Its not the end of the world here but I was thinking 15 years after I first started using this it could have been done smaller, that's all. steve

 

[Tanker135E]

I found this in one of my searches and thought this works for analog circuits. Can anyone confirm this. Or if it might be my answer?
https://www.electro-tech-online.com/custompdfs/2008/12/mc33886.pdf

 

[Roff]

I found this in one of my searches and thought this works for analog circuits. Can anyone confirm this. Or if it might be my answer?
https://www.electro-tech-online.com/custompdfs/2008/12/mc33886-1.pdf

It won't work for analog. It requires pulse-width-modulated digital inputs.

 

[Ubergeek63]

Ok..as per usual, we are trying to reinvent the wheel....and then reinvent that invention. I could make the original circuit on a 1ich by 1inch piece of perfboard just using the original (or replacement) transistors and bridge. Does it REALLY have to be smaller than that?..usually cars have a TON of space inside to throw in another circuit board.
As far as easier...it really doesnt get any easier than 4 transistors and a bridge.

No you couldn't.

The circuit is acting as a linear regulator, you neglected reading the OPs posts: the engines draw 0.6A at 14V at full speed. That means the circuit is dissipating up to 8W and what you propose is to dissipate 2-4W on 1/4W transistors.

There is a difference between "reinventing the wheel" and making the wheel safe to touch.

 

[Tanker135E]

Arhi and Roff and others I think I need to provide more information. I just not sure what it might be. Arhi said maybe a L293 Roff says it requires a digital input, won't work? Now I'll try this again this unit as is works. I wanted it smaller. I was looking as the throttle schematic and it has a L293 on it? I was also thinking I have found many IC that take Analog and convert it to digital. Why would that not work for me If I am sending out analog and I convert it to digital the Dc motor will still work probally better.
? steve

 

[arhi]

Arhi said maybe a L293 Roff says it requires a digital input, won't work?

Yup, mine fingers were faster then my brain. The L293 will do similar thing that you need but with your circuit you can control the output current by changing the input current so your controls are analog. With L293 the output is "off" or "fully on" so the way you control the speed of the motor with L293 is you do PWM (in short, you do a fast on/off with different lenghts of the on and off signal). In order to convert your analog input to generate PWM to drive L293 will probably make the circuit even bigger, and L293 also dissipates some heat...

In short, I think it would not make your circuit better nor smaller.

Now I'll try this again this unit as is works. I wanted it smaller. I was looking as the throttle schematic and it has a L293 on it? I was also thinking I have found many IC that take Analog and convert it to digital. Why would that not work for me If I am sending out analog and I convert it to digital the Dc motor will still work probally better.
https://groups.yahoo.com/group/onboard-keller/files/Throttle/
see if you guys can look at this page. It may enlighten you to something i haven't said? steve

That link ask for me to sign in so I ignored it, but in general, there are a lot of digital train controllers around, but going digital would mean that you have to change bunch of stuff ?! Not sure how is that feasible...

Looking at your schematic, you get the power from the rails (track) and you control the output from brown/orange that comes from T1 or T2 (whatever T1 and T2 are). Now, if you can show us how the "thingy" that outputs T1/T2 looks like, and if that "thingy" can be changed, we might be able to help you replace both the "thingy" that controls the motor speed and this "power pax"...

What I fail to understand is, motor is in the "locomotive", the device you want to "reduce in size" (power pax) is also in locomotive and it is passing trough the current from the tracks to the motor. Now, this part is clear, what is unclear is, where the hack T1/T2 comes from, how do you control this, I do not believe you have some wires sticking out from locomotive going into some joystick ...

If you had only one locomotive on the track you would be able to remove the "power pax" out of the locomotive and power the motor directly from the tracks (as cheap nrc models do) but if you want to have more then one locomotive on the track, the "power pax" has to be in the locomotive, and it has to be controlled by "thingy" that either receives "commands" via RF/IR or trough the modulated signal inside the tracks. Now, as I already mentioned, if you can show us that "thingy" maybe (just maybe) we can make that part digital ..

Before all that, you need to check what "device" can be small enough, control and pass 1A without producing too much heat. The L293 get fairly hot after few hours running ~.5A (without heat sync), you can add heat sync but do you really have room for them / that heat still need to be vented out somehow... If locomotive is made of metal it would be able to use the whole car as a heatsync

 

[Ubergeek63]

Yup, mine fingers were faster then my brain. The L293 will do similar thing that you need but with your circuit you can control the output current by changing the input current so your controls are analog. With L293 the output is "off" or "fully on" so the way you control the speed of the motor with L293 is you do PWM (in short, you do a fast on/off with different lenghts of the on and off signal). In order to convert your analog input to generate PWM to drive L293 will probably make the circuit even bigger, and L293 also dissipates some heat...

In short, I think it would not make your circuit better nor smaller.

And the https://www.electro-tech-online.com/custompdfs/2008/12/3959-2.pdf that I recommended does the PWM and synchronous rectification for you. All you have to do is supply and analog current and direction. How is the raw bridge of an L293 better than that?

 

[arhi]

Ubergeek63, sorry for my ignorance but this A3959 also has digital inputs (or I'm reading the ds wrong way around)?! how will that replace the existing circuit? (I do not thing l293 is right way to do things here but do not have a working idea neither)

 

[Roff]

Arhi asks an interesting question about how the throttle wires get connected to the locomotive. Tanker, could you answer that, please?

 

[Ubergeek63]

Ubergeek63, sorry for my ignorance but this A3959 also has digital inputs (or I'm reading the ds wrong way around)?! how will that replace the existing circuit? (I do not thing l293 is right way to do things here but do not have a working idea neither)

it has a digital direction input that can be brought out to a switch, digital mode inputs that can be tied off as desired, an analog current sense, and an analog current input that can be used to regulate the speed.

If you prefer to use a voltage drive you can divide the output voltage down and regulate it instead of the current.

Dan