Buck converter for high impedance source

[Marcus4audio]

Hello guys. I'm a long time lurker here and at first i want to thank U for great ideas. Here is my first question. i try to "design" current converter from unregulated r/c voltage source (5780ohm/200uf) . At the end of r/c voltage is 35V @ 40ma and I would like to convert it to 7v @ 130ma. At first it looks like job for simple buck converters but I have a problem because the load current is approx 250ma in cold condition (heater wire). I experimented with current limit circuits and LM2597 but it worked partially. In some situations Vin voltage drops under the buck safe vin and chip stuck untill next restart. I would be happy even with non regulated converter 5:1 ratio (transformer for dc) just to keep input current at 40ma even the Vin voltage changes like mains voltage changes any ideas where to start or how to improove LM2597?

 

[Les Jones]

R/C type power supplies are potentially dangerous as the output is is always live at mains potential. Please give more details of your proposed application and a schematic of the existing power supply and your modifications. Also what is the mains voltage where it will be used. I do not think you will be able to make this work as an r/c supply is more like a current source with a shunt regulator on the output.

Les.

 

[Marcus4audio]

**broken link removed** Hello again. No fancy things to draw but here is the schemo. My problem is that R/C has to be as it is. No mods are possible

 

[Nigel Goodwin]

You need to give FAR more information, what you've offered so far is pretty meaningless - you need to be EXACT and detailed, and WHY would help as well.

 

[Marcus4audio]

Hello Nigel. Thanks for chime in. I supposed that asking simple question will be better, but OK. I try to modify old valve piece of gear with modern valve but with minimal mod's around existing layout. This is my desperate attempt to put in order this old guy. I don't know why somebody will help, but I will be happy if someone can. What else I can add to this? I have no clue.
Oh, yes, this "thing" works partially but it's not stable as expected. So, how to make it more stable?

 

[Les Jones]

Hi Marcus,
This seems a very complicated way to get 7 volts (Or is it 6.3 volts ?) A complete schematic of the piece of equipment would help with the old valve type and the one that you want to use to replace it ( A B7G or B9A device ?) Is the 105 volt point the original HT supply ? or are all of these components ones you have added ? If the valve is a triode then it may be possible to use a power mosfet and then you would not need a heater supply.

Les.

 

[Blueteeth]

With 36V @ 40mA, that's ~1.44W in. For 7v @ 200mA, thats 1.54W out. So you'll need more current in, and the simple switcher converters often only go above 80% efficiency in their mid-current range (>1A). I believe the HV version can run up to 60V input, so just reducing R7 would allow say, 50V @ 40mA (2W) in. Of course the current on the input of the regulator depends on the output voltage and current.. and input voltage..

Say you need 7V @ 250mA out max, that's 1.75W. With an estimated conservative efficiency of 75%, input = ~1.75/0.75 = 2.3W. Clamping the input voltage to 50V, that's 46mA. Also at higher input voltages I suspect the quiescent current of switchers (with internal switch) increases, so add 15-20mA. But as its an R/C source, the more current it draws the lower the input voltage, allowing it to settling somewhere lower than 50V.

I would just try adding resistors in parallel with R7 OR raising your zener diodes voltage (making sure you don't disconnect the diode when running..) until you get a stable output of 7V @ your max current. Then you can measure the voltage across R7 to see the current draw, and perhaps optimize by choosing a different input voltage.

 

[Nigel Goodwin]

Hello Nigel. Thanks for chime in. I supposed that asking simple question will be better, but OK. I try to modify old valve piece of gear with modern valve but with minimal mod's around existing layout. This is my desperate attempt to put in order this old guy. I don't know why somebody will help, but I will be happy if someone can. What else I can add to this? I have no clue.
Oh, yes, this "thing" works partially but it's not stable as expected. So, how to make it more stable?

Sorry, but you're still telling us nothing useful - how about posting a circuit of the original design?.

 

[ronv]

Marcus4audio
If you can't make R7 smaller the only hope I see for it is to make the time constant of the heater shorter than that of the input voltage decay. So in order I think.
Make R7 smaller.
Make the zener voltage higher.
Make the input supply cap much larger and the turn on delay longer so it can fully charge.
Maybe a thermistor on the output for current limit.

 

[chemelec]

What is ZD40 on your schematic?
A 40 Volt Zener?
If so, it consumes Current.

36 Volts in at 40 mA is 1.44 Watts
7 volts at 200 mA is 1.4 watts

Also No room left for Efficiency of conversion.

Why not use a more suitable transformer, With an LM317 as a constant current source?

 

[Marcus4audio]

Thanks guys. here is the original schematics. **broken link removed** The original tube use 36V for heating and I try to substitute it with some small available tubes with up to 125ma @ 6.3V. the 0,7V is used for current limit sense. Anyways, tube goes up to 200ma when start to heat but as soon it heat up it goes down to 125ma. Unfortunately I cant change the R7 and the surrounding components. Of course, everything around the buck or the buck can be replaced. BTW Mosfet as triode is not bad idea at all but I have no idea where to start

 

[ronsimpson]

I think you can get parts with a 6.3V filament voltage and can not get a 36V filament.
It looks to me that:the 105V supply, and R4, R3 and the filament resistance makes a voltage divider to get 1.1V on (K).
So cut the filament out and change R4 or add a another resistor to get the same voltage divider.
Now with the filament not connected you can make the 6.3V any way you want.
I would probably use a "wall wort" or any small 6V power supplies.

 

[Les Jones]

Hi Marcus,
I assume this is a hi input impedance microphone pre amp and the component KK47 is some form of capacitor microphone. Can yo conform this ?
Some suggestions.
1 Use one of the DF series valves that were used in battery radios in the 1950s (A DF96 heater takes 25 mA at 1.4 volts.)
2 As a wild guess (As I do not have a crystal ball and you do not seem inclined to give much helpful information.) I suspect you may have chosen to use an EF86 as it is a low noise device. (But very microphonic.) If this is the case then as it is indirectly heated the heater supply does not effect the biasing. Use a very basic series regulator from the top of C4 consisting of an high voltage NPN power transistor of suitable voltage, current and power rating connected as an emitter follower. The collector to the top of C4, the base to the + end of the 40 volts zener diode and the emitter (Output) to the switching regulator. If you are building the switching regulator yourself (Rather than buying a ready made one of ebay.) then make sure that you use low ESR capacitors. This will make the supply to the switching regulator much stiffer so it should work correctly.
3 Redesign the unit using a fet input op-amp followed by a transistor stage to give a high enough voltage swing at the output transformer.

Have you considered the problem you may have with electrical noise from switching regulator getting into the high impedance input of the preamp ?

Les.

 

[Marcus4audio]

Hi Les. My idea was to use 6cw4 nuvistor or similar. By measuring I found that filament is approx 125-130ma. Again, my idea was not to modify psu or internal layout/values but to substitute the tube with buck/tube because it will make possible to exchange this mod easilly with original tube (if I find it one day). Also, since nuvistor is in metal pack it's more immune to switching noise. Btw, I tried that with sucess withouth any noise in audio range because buck was used jut for heating. Tube bias was taken from high non regulated voltage.

 

[Blueteeth]

Unfortunately I cant change the R7 and the surrounding components. Of course, everything around the buck or the buck can be replaced.

Thats going to be quite tough, if the 1780 Ohm cannot move - but I did suggest adding another resistor in parallel to reduce its resistance

Also, if you disconnect the heater pin from the 29 ohm resistor, and connect it to ground via a 10R, you can connect the output of your buck directly to the heater+, and use the heater- resistor as the feedback to your buck. This will provide a constant current of 1.25 / 10 = 0.125 = 125mA. The price you pay is that, if the 'heated resistance' really is 50 ohms, you'll have a drop of (50 + 10)*0.125A = 7.5V.

That is pushing almost a watt of output power, and the 36V line *will* sag when drawing more than 40mA, so whilst the zener is there for protection and current shouldn't flow through it when the buck is operating, you're still looking at lower than 36V as your input power, further narrowing the gap between input and output power.

Give Roman Black a pm, he made a nice simple two-transistor buck that might not have world class efficiency, but at higher input voltages its surprisingly good!

 

[Marcus4audio]

Thanks Blueteeth. I like the idea of simple 2 tr buck and I wonder is it possible to modify it to fixed ratio (pwm) buck. I like the "modern" regulated buck's chips for some other situations but here i tink that primitive "transformer like" design will be much better and more flexible. I really don't need the regulation, just the current conversion.

 

[Les Jones]

I do not see a solution to your problem as you are trying to make a plug in replacement for the VF14M without making any change to the original circuit. My mosfet idea is no use as the gate to source capacitance would be too high.

 

[chemelec]

I Think Ron Simpsons idea is probably the best way to go.

Just add an 845 Ohm resistor in series to ground with R4 to maintain your 40 mA.

 

[Marcus4audio]

I Think Ron Simpsons idea is probably the best way to go.

Just add an 845 Ohm resistor in series to ground with R4 to maintain your 40 mA.

I agree, but only problem is that I have no mains or empty pin on cable to bring AC inside the mic body. Still thinking about that damned current conversion from 36V/40ma to 7v/125ma :/

 

[chemelec]

What is the AC Output Voltage on the Transformer TR1?