For that you probably don't need to regulate it. You could just use the big filter caps.

I think what I currently have is only rated for 60V which would be pushing it on an unregulated supply and would any noise from an unregulated supply be fed into the LC circuit?

 

I think what I currently have is only rated for 60V which would be pushing it on an unregulated supply and would any noise from an unregulated supply be fed into the LC circuit?

If yours is like most of them I have seen you will already have almost 60 volts across the FETs due to the LC circuit, and close to 12 amps. So you need to be careful you don't way overshoot. If you want to post a schematic I will simulate it for you.

 

If yours is like most of them I have seen you will already have almost 60 volts across the FETs due to the LC circuit, and close to 12 amps. So you need to be careful you don't way overshoot. If you want to post a schematic I will simulate it for you.

It's this old classic with a few differences. You're absolutely right I got it the wrong way round it is rated for much high voltages it was the current that was my problem in that it draws about 30A at 60V under load. So I was planning on setting my upper Vout to about 48V and a 20-25A load

 

Here is a sim at 48 volts without any regulator.
It's hard to say with certainty what the current is because of the work piece, but this will give some ideas.
Most people have trouble with the main filter caps because the ripple current is so high. So make sure to check that.
The other trouble makers are the inductors. If they saturate bad things happen. What I have seen is more than one core for the inductor.

 

Here is a sim at 48 volts without any regulator.
It's hard to say with certainty what the current is because of the work piece, but this will give some ideas.
Most people have trouble with the main filter caps because the ripple current is so high. So make sure to check that.
The other trouble makers are the inductors. If they saturate bad things happen. What I have seen is more than one core for the inductor.

Thanks for taking the time to run that for me So I would probably ok without the regulator but I think I may still have a go at it as the transformer and smoothing cap alone are about 80% of the price of the project.

 

So I think I will attempt the regulator as it's not much more expense than the transformer and capacitor. But I'll get the cap and transformer first and try it . This is what I've got and I think I'm ready unless anyone see a problem (please tell me if you do ).



Changes are

• 240/40v 1000VA for T1 to give me 56.56 Vp (nice and close to 51V average to give 48V out)
• LM317T has replaced the LM317HVT as a 60V differential is no longer desirable, 40V is more appropriate.
• MJ15003G/4G 250W NPNs and PNP respectively
• R1 set to 20Ohm to draw 50mA across it to switch on PNP for a total of 100mA across the regulator.
• R3, VR1 & VR2 have been increased in value to lower the power dissipation across them
• VR1+2 =22000ohm for range 30V-48V (8556Ohm-17578Ohm)
• R4 was only described as 30mA minimum load in the datasheet so I've just put a 30mA LED in and an appropriate resistor for the range.
• D3 will most likely be a 100A stud fit diode.

Heat sinks

• LM317 won't be dissipating much at any range so I intend to mount it internally with a small heat sink and forced air
• 3xNPN and 1XPNP will be mounted on the heat sink below, externally with natural convection in order NPN-NPN-PNP-NPN

So if that's all good I think I will start getting bits together

 

So I think I will attempt the regulator as it's not much more expense than the transformer and capacitor. But I'll get the cap and transformer first and try it . This is what I've got and I think I'm ready unless anyone see a problem (please tell me if you do ).

View attachment 99593

Changes are

• 240/40v 1000VA for T1 to give me 56.56 Vp (nice and close to 51V average to give 48V out)
• LM317T has replaced the LM317HVT as a 60V differential is no longer desirable, 40V is more appropriate.
• MJ15003G/4G 250W NPNs and PNP respectively
• R1 set to 20Ohm to draw 50mA across it to switch on PNP for a total of 100mA across the regulator.
• R3, VR1 & VR2 have been increased in value to lower the power dissipation across them
• VR1+2 =22000ohm for range 30V-48V (8556Ohm-17578Ohm)
• R4 was only described as 30mA minimum load in the datasheet so I've just put a 30mA LED in and an appropriate resistor for the range.
• D3 will most likely be a 100A stud fit diode.

Heat sinks

• LM317 won't be dissipating much at any range so I intend to mount it internally with a small heat sink and forced air
• 3xNPN and 1XPNP will be mounted on the heat sink below, externally with natural convection in order NPN-NPN-PNP-NPN

View attachment 99594 View attachment 99595
So if that's all good I think I will start getting bits together

I have kind of a rule of thumb calculator for heat sinks. Teperature rise in C/Watt = 50/sq. root of the surface area in cm. So I think that will make yours 1 degree C per watt. You power is about 18 volts at 20 amps - so 360 watts.
You can see the problem.
Having said that, about 1 amp goes thru the 20 ohm resistor, so he needs to be bigger.
Let us know what caps you are looking at for the big filter.

 

Thanks Sorry my bad there I should have put in the thermal resistance for that, it's 1.5C/W. Little bit worse I am not expecting to be able to draw max current at a low voltage, I think this is probably just something I would inherently have to live with unless I use an enormous heat sink. Can you get bigger heat sinks that can dissipate that much power, I could have a look at bigger ones but what kind of size do you think I'd need to dissipate that much power?
I am a little confused about R1 now though because it is setting the maximum current for the LM317 at 50mA to create a voltage drop (R=1/0.05A) for the PNP to switch on and then another 50mA (limited by R2) is now flowing from the base of the PNP and across the regulator (100mA total regulator current). I may be wrong here of course it is my first implementation of such a circuit. So from that I just assumed Pd=20*(0.05^2) = 0.05W.
The filter cap is the right value at 33000uF and that should give my just under 10% ripple at 25A (just for headroom) C = 25amps/(1.414 x 5.656 x 100) = 0.0312F and that would need to be about 80V minimum.

 

Another slight change. I think I'm going to have to remove the base current limiting resistor (R2). I can't reconcile that voltage drop across it with the input need for the regulator, it would create a negative input voltage. It should be ok I suppose, it will only need 50mA max base current when maximum current is flowing through the NPNs.

 

So I finally got around to finishing this but I have a bit of a problem. If anyone can help it would be greatly appreciated please . It doesn't seem to want to regulate voltage at all and just outputs the input. All connections have been checked and match the schematic below. BUT I am using a LM317T instead of a LM317HVT. I assumed that the LM317T would be ok with an input of 58V and I would only have the differential output of 40V but is it the case that the LM317 won't work with voltages over 40V?

 

What happens when you short the ref pin to ground, ?

 

What happenes when you short the ref pin to ground, ?

Lol I bridged ADJ to GND (adj is ref I didn't have the right symbol in circuit wizard) and R3 caught fire and went pop and the output remained the same at 58V (60V in)

 

Just noticed this quote in the LM317HVT datasheet that is absent from the LM317 datasheet

"Since the regulator is floating and sees only the input to output differential voltage, supplies of several hundred volts can be regulated as long as the maximum input to output differential is not exceeded, or in other words, do not short the output to ground."

 

Sounds like Q1,4 maybe wrong or shorted out, Disconnect the transistors and try the regulator on its own.

 

Sounds like Q1,4 maybe wrong or shorted out, Disconnect the transistors and try the regulator on its own.

I am only measuring the output of the regulator at the moment. The transistors are still attached at their inputs but the output off the emitter resistors have been removed while I nail down the regulator problem. However they do appear to be conducting the full 60V across them independent of the regulator output at the moment. When their outputs are connected to the regulator output it decreases to match the drop across the regulator (58V).

EDIT: Maybe this is a good time to mention I got these regulators of eBay and the markings indicated they were 7 years old.

 

If the regulator is wired wrong it will give out full supply, try it on a lower input voltage first say 25v, then crank it up.

 

I am almost certain the regulator is wired correctly (assuming my diagram is correct). I am going to pull it out and thoroughly check again though. I'll photograph it while it's out

 

Right so checked it all again, started doing my continuity checks and turns out I have continuity between my input and output pins, no jumpers connecting them of course and I removed D2 to see if the regulator had problems. After removing D2 I still had continuity between Input and output. Is this a bad regulator?

 

Just checked by backup LM317T and there is no continuity between input and output and as there is no evident thermal damage so I suppose I assume this is a bad component. The other one I have is a much newer one but would you recommend I use the LM317HVTs (I have some) as I have to replace it anyway? OR do you recommend I wait, buy some off Farnell and use a piece I can rely on?

 

Ok so it's not the regulator as it works fine when I remove the connection to the PNP Base. In the application diagram there was originally a 5K base current limiting resistor but I was unsure as to if it would be necessary or work. I assume now that too much current is flowing from the PNP base and across the regulator.