Samuel Groner said:If I understand correctly, the dropout voltage (for good regulation) is something like 25 to 30 V (in to out difference), so you can't expect more than ~360 V at the output.
Samuel
No; LM317HV guarantees good regulation for anything between 3 and 60V. But the preregulator (TIP50+associated resistors) introduces a voltage drop that's load-dependant (about 15V at 100mA). Since there is a minimum current of 12.5mA (due to the voltage-reference across the 100r resistor), the minimum drop in the prereg is about 3V. So, in theory, the maximum no-load voltage is about Vin-6V.Samuel Groner said:If I understand correctly, the dropout voltage (for good regulation) is something like 25 to 30 V (in to out difference), so you can't expect more than ~360 V at the output.
Samuel
That's true if the Zener conducts, but there are cases where the voltage between base and output is too low. If the output voltage is set at 300V, the Zener does not conduct but the regulator works perfectly.moamps said:The voltage on the 317 reg is defined with Vz-Vbe-47Ic ,
With the actual 100r resistor at the adjust pin, the no-load current is 12.5mA. No need for a bleeder resistor.The darlington transistor here should be used with much higher base resistor's value, or output should be all the time loaded with minimum 10mA (27kOhm, 5 or 7W resistor on output), IMO.
abbey road d enfer said:That's true if the Zener conducts, but there are cases where the voltage between base and output is too low. If the output voltage is set at 300V, the Zener does not conduct but the regulator works perfectly.
Exactly what I wrote...moamps said:abbey road d enfer said:That's true if the Zener conducts, but there are cases where the voltage between base and output is too low. If the output voltage is set at 300V, the Zener does not conduct but the regulator works perfectly.
The Zener doesn't conduct only if Vin-Vout is smaller than Vz .
abbey road d enfer said:Exactly what I wrote...moamps said:abbey road d enfer said:That's true if the Zener conducts, but there are cases where the voltage between base and output is too low. If the output voltage is set at 300V, the Zener does not conduct but the regulator works perfectly.
The Zener doesn't conduct only if Vin-Vout is smaller than Vz .
Why? The zener is there just to make sure that Vin-Vout stays within reasonable limits and the TIP50 takes its share of the dissipated heat, but I have simulated the circuit, it works even with with only 3V across the prereg. But the available current is nil. At 6V across the prereg, there is a total of 35mA available, so 22.5mA for the load.moamps said:Nope. In this circuit (contrary to the simple reg without pre-regulation, where protection Zener doesn't conduct in normal operation mode) Vin-Vout must be higher than Vz to ensure normal operation, and the Zener conducts all the time. If the Zener doesn't conduct, the circuit doesn't work, IMO.abbey road d enfer said:Exactly what I wrote...moamps said:abbey road d enfer said:That's true if the Zener conducts, but there are cases where the voltage between base and output is too low. If the output voltage is set at 300V, the Zener does not conduct but the regulator works perfectly.
The Zener doesn't conduct only if Vin-Vout is smaller than Vz .
abbey road d enfer said:Why? The zener is there just to make sure that Vin-Vout stays within reasonable limitsmoamps said:Nope. In this circuit (contrary to the simple reg without pre-regulation, where protection Zener doesn't conduct in normal operation mode) Vin-Vout must be higher than Vz to ensure normal operation, and the Zener conducts all the time. If the Zener doesn't conduct, the circuit doesn't work, IMO.
and the TIP50 takes its share of the dissipated heat, but I have simulated the circuit, it works even with with only 3V across the prereg. But the available current is nil. At 6V across the prereg, there is a total of 35mA available, so 22.5mA for the load.
Why? The regulator can work with anything between 3 and 60V. The choice of Zener voltage is dictated by the targetted max current and associated dissipation in the LM317. [/quote]moamps said:abbey road d enfer said:Why? The zener is there just to make sure that Vin-Vout stays within reasonable limitsmoamps said:Nope. In this circuit (contrary to the simple reg without pre-regulation, where protection Zener doesn't conduct in normal operation mode) Vin-Vout must be higher than Vz to ensure normal operation, and the Zener conducts all the time. If the Zener doesn't conduct, the circuit doesn't work, IMO.
The Zener is there to keep Vin(reg)-Vout constant, independent of Vin-Vout voltage. It is essential in this circuit.
and the TIP50 takes its share of the dissipated heat, but I have simulated the circuit, it works even with with only 3V across the prereg. But the available current is nil. At 6V across the prereg, there is a total of 35mA available, so 22.5mA for the load.
I know the idea is to maintain a reasonable voltage across the LM317, and the TIP50 to take its share, which is the case when the output voltage is more than 20V below unreg B+, but it does not mean that the circuit does not regulate when Vin-Vout is lower than 20V. In that case, dissipation in the LM317 is limited and the TIP50 does not need to take a share.
This circuit is not intended to work in Vin-Vout≤Vz regime.
Why? The regulator can work with anything between 3 and 60V. The choice of Zener voltage is dictated by the targetted max current and associated dissipation in the LM317.The Zener is there to keep Vin(reg)-Vout constant, independent of Vin-Vout voltage. It is essential in this circuit.
No. With 3V across the prereg @ 12.5mA, the transistor has about 800mV Vce; admittedly, it's close to saturation, but not completely.Have you simulate the circuit with transistor's hfe=30 or so? Your math stands only if the transistor is in saturation.
In fact some combinations of Vin-Vout and current put the transistor in saturation; but it does not prevent the regulator to do its job.
I know the idea is to maintain a reasonable voltage across the LM317, and the TIP50 to take its share, which is the case when the output voltage is more than 20V below unreg B+, but it does not mean that the circuit does not regulate when Vin-Vout is lower than 20V...This circuit is not intended to work in Vin-Vout≤Vz regime.
No, I did not. Ther is no need for this voltage to be constant. It just needs to be clamped at a reasonable voltage.moamps said:Why? The regulator can work with anything between 3 and 60V. The choice of Zener voltage is dictated by the targetted max current and associated dissipation in the LM317.The Zener is there to keep Vin(reg)-Vout constant, independent of Vin-Vout voltage. It is essential in this circuit.
You answered your own question.
What you don't understand is that the transistor, being essentially in voltage-follower mode, is self-regulating. Whenever the current demand increases, the voltage across it increases too and the base current increases in such a way that the transistor never runs out of juice.No. With 3V across the prereg @ 12.5mA, the transistor has about 800mV Vce; admittedly, it's close to saturation, but not completely.Have you simulate the circuit with transistor's hfe=30 or so? Your math stands only if the transistor is in saturation.
In fact some combinations of Vin-Vout and current put the transistor in saturation; but it does not prevent the regulator to do its job.
If Vin-Vin(reg)=3V and I=12,5mA then:
Vre=47I=0,6V, Vrc=100I=1,25V
Vrb=Vin-Vin(reg)-Vbe-Vre=3-0,7-0,6=1,7V
Irb=Vrb/Rb=1,7/10000=0,17mA
for hfe=50, Ic=Ie=Irb x hfe=0,17x50=8,5mA !!!
So you haven't enough base current to drive transistor to get minimum needed current, without any load.
I know the idea is to maintain a reasonable voltage across the LM317, and the TIP50 to take its share, which is the case when the output voltage is more than 20V below unreg B+, but it does not mean that the circuit does not regulate when Vin-Vout is lower than 20V...This circuit is not intended to work in Vin-Vout≤Vz regime.
Practically, you don't know that for sure.
For example:
Vin-Vin(reg)=10V, I =35mA
Vre=1,65V, Vrb=Vin-Vin(reg)-Vbe-Vre=10-0,6-1,65=7,75V
Irb=Vrb/Rb=7,75/10000=0,775mA
to get Ic=35mA at Vce=5V, hfe must be 50 what for power transistors at so small currents isn't so usual.
If the loading changes, you can be partially in unregulated regime.
So I proposed earlier use of darlington transistor for prereg transistor, where higher value of Rb can be chosen (the bleeding current to output via Zener is small) without problems with driving the prereg transistor.
I have used this topology few times, and have nothing more to add here.
What you don't understand is that the transistor, being essentially in voltage-follower mode, is self-regulating. Whenever the current demand increases, the voltage across it increases too and the base current increases in such a way that the transistor never runs out of juice.
Also you based your comments on the fact thet hfe is 50; the datasheet for TIP50 shows that it can vary between 20 and 150. I agree that in the extree case of hfe=30, operation with 3V across will not be sustainable, but for hfe=50, operation at about 4.5 V is stable. What is 1.5V difference compared to 390?
This is just silly argumentation.
This is turning into a perfectly useless dick contest. You are trying to prove that something does not work at all past some limits, and I'm just saying that operation in degraded mode is not a total failure.moamps said:The silly argumentation here is advocating something what's working only if you are lucky when you pick the transistor from the bin.
Enter your email address to join: