Insufficient Current Output from 300V DC Adjustable Power Supply

[spica]

Hi

I built a 0 to 300V (DC) adjustable power supply circuit based on the design I found after googling and stumbling upon this link. { http://www.oocities.org/tjacodesign/300vsup/300vsup.html }

I managed to get the circuit working but I have a problem with controlling the current as the load effect of this power supply is not very well regulated. And the max current I can get from this is only 100mA.

My question is how to increase the output current so that I can have a stable regulated output at different loading. (Preferably, up to 500mA or even up to 1A at 300V). Can I use any power amp to increase the current? If so, any part number recommendation? I posted this in one of the TI Forums but no one seemed to know the solution so far. Also, it is ok if you have an entirely different circuit design or idea as long as the requirement of 0 to 300 V (DC) output with at least close to 500mA is met.

Thank you very much.

Regards

 

[IanP]

Hi

I built a 0 to 300V (DC) adjustable power supply circuit based on the design I found after googling and stumbling upon this link. { http://www.oocities.org/tjacodesign/300vsup/300vsup.html }

I managed to get the circuit working but I have a problem with controlling the current as the load effect of this power supply is not very well regulated. And the max current I can get from this is only 100mA.

My question is how to increase the output current so that I can have a stable regulated output at different loading. (Preferably, up to 500mA or even up to 1A at 300V). Can I use any power amp to increase the current? If so, any part number recommendation? I posted this in one of the TI Forums but no one seemed to know the solution so far. Also, it is ok if you have an entirely different circuit design or idea as long as the requirement of 0 to 300 V (DC) output with at least close to 500mA is met.

Thank you very much.

Regards

Look, in this circuit ( http://www.oocities.org/tjacodesign/300vsup/300vsup.html ) current is limited by T2, and in particular by the value of R2.

The rough formula for current limit in this circuit is:
I[limit] = U(be) / R2
Where:
Ube = Ube(T2) = 0.65V
R2 = (here) 3R3 = 3.3ohm

Just remember one thing: the power MOSFET has to be mounted on a decent heatsink, and another hint – you may try two, three (or even more) MOSFETS connected in parallel, as they behave like “controlled” resistors.

And another thing, with a load the unregulated voltage across R3 will drop and that means that the output voltage will also drop, so the regulation is not very good.
Here you may try to add, say, 50kohm resistor on top of the 500kohm pot and add 250V Zener diode in parallel with 500kohm pot. That will stabilize the reference voltage that is then divided and delivered to the gate(s) of the power MOSFET(s).
:wink:
IanP

 

[spica]

Look, in this circuit ( http://www.oocities.org/tjacodesign/300vsup/300vsup.html ) current is limited by T2, and in particular by the value of R2.

The rough formula for current limit in this circuit is:
I[limit] = U(be) / R2
Where:
Ube = Ube(T2) = 0.65V
R2 = (here) 3R3 = 3.3ohm

Just remember one thing: the power MOSFET has to be mounted on a decent heatsink, and another hint – you may try two, three (or even more) MOSFETS connected in parallel, as they behave like “controlled” resistors.

And another thing, with a load the unregulated voltage across R3 will drop and that means that the output voltage will also drop, so the regulation is not very good.
Here you may try to add, say, 50kohm resistor on top of the 500kohm pot and add 250V Zener diode in parallel with 500kohm pot. That will stabilize the reference voltage that is then divided and delivered to the gate(s) of the power MOSFET(s).
:wink:
IanP

Thank you very much, Ian. Appreciated much. :smile:

So, in this case, assuming Vbe is 0.65, it was designed to give only 196.97 mA. Can I then change the value of the R2 (3R3) to a lower value instead so that I can get a higher current output? Or what if I removed the current limiter portion (i.e. T2 and R2) altogerther?

Yes, I used a relatively big heatsink on power MOSFET. BTW, could you explain more about your idea as I am not very experienced in it?

Regards
:smile:

 

[IanP]

Thank you very much, Ian. Appreciated much. :smile:

So, in this case, assuming Vbe is 0.65, it was designed to give only 196.97 mA. Can I then change the value of the R2 (3R3) to a lower value instead so that I can get a higher current output? Or what if I removed the current limiter portion (i.e. T2 and R2) altogerther?

Yes, I used a relatively big heatsink on power MOSFET. BTW, could you explain more about your idea as I am not very experienced in it?

Regards
:smile:

You can remove the current limiting section of the circuit (R2 and T2) but you have to be aware that any short-circuit may end up with fireworks.
I’d use R2 = 1ohm, just in case …

As far as heatsink goes, the device has to be able to radiate heat being equivalent to
(Vin-Vout) * I[load]

So, for example, if you connect a load that will draw 500mA @ 30V, the equation looks like this:
Vin = 300Vdc
Vout = 30Vdc

(300-30)V * 0.5A = 135W

One BUZ326 MOSFET will not do it, you need at least 1 MOSFET per 30-50W, so 3-5 of them, and the heatsink size will be huge – for natural convection, or slightly smaller with a fan.
Calculations of sizes of heatsinks you can readily find on the internet, just to give you a rough idea, a 300mm heatsink of 0.7deg/W will increase its temperature in the example given above by 95deg, so it is much, much to small.

:wink:
IanP

 

[spica]

You can remove the current limiting section of the circuit (R2 and T2) but you have to be aware that any short-circuit may end up with fireworks.
I’d use R2 = 1ohm, just in case …

As far as heatsink goes, the device has to be able to radiate heat being equivalent to
(Vin-Vout) * I[load]

So, for example, if you connect a load that will draw 500mA @ 30V, the equation looks like this:
Vin = 300Vdc
Vout = 30Vdc

(300-30)V * 0.5A = 135W

One BUZ326 MOSFET will not do it, you need at least 1 MOSFET per 30-50W, so 3-5 of them, and the heatsink size will be huge – for natural convection, or slightly smaller with a fan.
Calculations of sizes of heatsinks you can readily find on the internet, just to give you a rough idea, a 300mm heatsink of 0.7deg/W will increase its temperature in the example given above by 95deg, so it is much, much to small.

:wink:
IanP

Hi Ian

Thank you for enlightening me. I am using this heat sink { https://sg.element14.com/jsp/search/productdetail.jsp?SKU=4621311 } on my IRF740A. :grin:

Yeah, I was thinking of 1Ω resistor too.

Also, if possible, could you please elaborate more on "50kohm resistor on top of the 500kohm pot and add 250V Zener diode in parallel with 500kohm pot."?

Regards
:grin:

 

[IanP]

Hi Ian

Thank you for enlightening me. I am using this heat sink { https://sg.element14.com/jsp/search/productdetail.jsp?SKU=4621311 } on my IRF740A. :grin:

Yeah, I was thinking of 1Ω resistor too.

Also, if possible, could you please elaborate more on "50kohm resistor on top of the 500kohm pot and add 250V Zener diode in parallel with 500kohm pot."?

Regards
:grin:

For some initial tests with small currents you can use that heatsink, but for real operation in the whole range of voltages it is way to small.

I have attached a drawing that should explain the modification around the reference voltage of 250V. It will improve stability of the output voltage by making it independent of the input voltage drops related to connected loads.
The 250V (doesn’t need to be exact, 240-270) can be achieved by connecting several Zener diodes in series, for example, 8 x ZD33V.

:wink:
IanP

 

[spica]

For some initial tests with small currents you can use that heatsink, but for real operation in the whole range of voltages it is way to small.

I have attached a drawing that should explain the modification around the reference voltage of 250V. It will improve stability of the output voltage by making it independent of the input voltage drops related to connected loads.
The 250V (doesn’t need to be exact, 240-270) can be achieved by connecting several Zener diodes in series, for example, 8 x ZD33V.

:wink:
IanP

Wow. That heat sink is still too small! :shock: Ok, thanks a lot for your help. I've learnt a lot.:-D

I will experiment with that R2 of 1Ω and your modification.

Regards
:grin:

 

[spica]

For some initial tests with small currents you can use that heatsink, but for real operation in the whole range of voltages it is way to small.

I have attached a drawing that should explain the modification around the reference voltage of 250V. It will improve stability of the output voltage by making it independent of the input voltage drops related to connected loads.
The 250V (doesn’t need to be exact, 240-270) can be achieved by connecting several Zener diodes in series, for example, 8 x ZD33V.

:wink:
IanP

Hi Ian

I have tried the circuit with 50kΩ and 270V zener diode together with R2=1Ω but now the circuit shows max output of 134~136 V (not stable) only.

Can you pls help?

Regards