The datasheet on LM317 states a maximum voltage between Vin and Vout at 40 volts and while this is mentioned a few times on here people seem to suggest not using them with higher supply voltages.
If I understand the datasheet correctly one could regulate say 35 Volts from a 50-60 Volt supply using a 317 or am I missing something?
On the TL431 its stated 36V relative to the anode/ground and max 100mA so theoretically I should be able to regulate say 35 volts and 50 mA using these but it does not work, they break (newer ones almost all but a batch of 20 year old ones coped much better)
So what would work for the TL431, am I pulling to much current through it or is 35 volts to close to its maximum?
If I understand the datasheet correctly one could regulate say 35 Volts from a 50-60 Volt supply using a 317 or am I missing something?
On the TL431 its stated 36V relative to the anode/ground and max 100mA so theoretically I should be able to regulate say 35 volts and 50 mA using these but it does not work, they break (newer ones almost all but a batch of 20 year old ones coped much better)
So what would work for the TL431, am I pulling to much current through it or is 35 volts to close to its maximum?
TO-92 can take typically 0.625W
So, at 35V you can only take like 18 mA
0.625/35 = 0.0178
So, at 35V you can only take like 18 mA
0.625/35 = 0.0178
Yes I found the TL783 when looking for options. I need a +- supply though so did not look into it further but maybe it would work stacking them like one would do with batteries.
Anyone know if they are better or worse wrt output impedance compared to the LM317?
That is correct. You can use LM317 at 60V in, and take 35V at the output.
As long as the difference is max 40V.
But you should care about the max power across LM317. And possiblly fit LM317 to a suitable heatsink.
Power would be: (60-35) x the current. I would say max current is possibly 250mA at 25V across LM317.
But I have not found the max power LM317 can take.
Do you have LM317T .. the TO-220 package?
Last edited:
A lm317 is a so called floating regulator.
You can use it at 300V as long as you keep input - output voltage low enough.
Fyi
TL783 is on TI chopping block so are many of the original TI branded regulators such as lm237kc, lm317kc, they have kept the more expensive and older National line up of lm317T etc. I’d have to look to see the status of lm317hv
TL783 is on TI chopping block so are many of the original TI branded regulators such as lm237kc, lm317kc, they have kept the more expensive and older National line up of lm317T etc. I’d have to look to see the status of lm317hv
The LM317HV is still available but not the 337HV.
I should be ok with the 317/337 though, I'll feed them 42 volts and maybe take out 35-37.
I should be ok with the 317/337 though, I'll feed them 42 volts and maybe take out 35-37.
Here is how you can use TL431 to put out 35V 125mA.
Using some power transistor.
To set the voltage you can adjust R5.
Using some power transistor.
To set the voltage you can adjust R5.
Thanks Lineup, subjectively I have always preferred a shunted supply based on the TL431. I have a few PNP 2N2905 since before that might work boosting the TL431 as per the data sheet.
Not sure I will retain the sonic signature of the TL431 though.
The idea now would be to use the 317/337 as pre regulators and then use the TL431 to decouple these and do the last few volts of regulation.
I'll do some listening and measurement tests with the 317/337 over the weekend and we'll see if I think I need to add a shunt regulator.
Not sure I will retain the sonic signature of the TL431 though.
The idea now would be to use the 317/337 as pre regulators and then use the TL431 to decouple these and do the last few volts of regulation.
I'll do some listening and measurement tests with the 317/337 over the weekend and we'll see if I think I need to add a shunt regulator.
Last edited:
I've had good success connecting a 30 volt, 5 watt zener diode between LM317 input and output pins. This guarantees that the LM317's absolute maximum rating for Vin-to-Vout is always met, even during startup, even during shutdown.
Want to build an LM317 regulator with Vout=80V, using Vin=90V? This will do it just fine. Naturally it won't work in situations where your Vin is more that 30V greater than your regulated Vout. In those cases you'll need a second power transistor installed upstream of the LM317, whose job is to drop (and pre regulate) the voltage applied to the Vin of your LM317.
Want to build an LM317 regulator with Vout=80V, using Vin=90V? This will do it just fine. Naturally it won't work in situations where your Vin is more that 30V greater than your regulated Vout. In those cases you'll need a second power transistor installed upstream of the LM317, whose job is to drop (and pre regulate) the voltage applied to the Vin of your LM317.
Yes this is a solid strategy that can be extended a surprisingly long way. I've one that provides 200Vdc for an ancient valve tuner, and has done for 15yrs now.
The abs imperative note as you push potential voltage input is (1) be wary of the energy stored in the Vadj and Vout capcitors - keep these low values; and (2) the reverse bias diodes from Vadj to Vout and Vout back to raw supply are absolutely imperative (even as low as 20vdc raw-in on an LM337). (3) - remmeber the TO220 tab will also be at high voltage; review your insulation strategy carefully.
Mark's point about 5W rated Zener is ideal - this diode limits both inrush deltaV, and provides the switch-off protection. I usually prefer to limit the Vdiff to 15-20v, but as Vin uncertainty rises, with higher input raw DC voltages, and mains fluctuation, that can get a bit fine. 30V is def safe.
If you are going this route - it does a lot of good to make the raw dc supply for this longsuffering 3-pin reg to be C-R-C and deliberately 'drop' several volts there for your desired output current. It provides better HF filtering, and can usefully slow rate of voltage rise into the hot mess at switch-on.
Last edited: