I want to build kinda "universal amplifier PSU", adjustable from 25 to 40 volts and giving enough current to drive GC, MyRev, Symasym and others. I got a pair of 30 V / 240 VA dual secondary transformers and no clue. What regulator is recommended? All I found is the LM317HV, but it doesn´t give enough current ...
Don´t want to spend so much for transformers anymore.
Thanks.
Don´t want to spend so much for transformers anymore.
Thanks.
Look on the LM317 datasheet for how to add pass transistors to boost current.
Given the cost and heatsinking required for this project, I would just build a simple fixed unregulated PSU from a 100VA 2x25V transformer. There's rarely any need to have any more voltage than this just for testing.
Given the cost and heatsinking required for this project, I would just build a simple fixed unregulated PSU from a 100VA 2x25V transformer. There's rarely any need to have any more voltage than this just for testing.
No,
VDiff for LM338 is 35V, If properly designed, fits OK for your needs.
It must be a problem in case of short-circuits, but, since you don´t need less than 25V, you can pre-regulate or drop the voltage to achieve what you want.
But, for linear power supplies, is not a easy task achieve well regulated 40V from more or less 43 volts. The transformer must be very good. For this level of voltage I sugest you work in non regulated mode. A switch to bypass the regulator could work.
Best regards,
VDiff for LM338 is 35V, If properly designed, fits OK for your needs.
It must be a problem in case of short-circuits, but, since you don´t need less than 25V, you can pre-regulate or drop the voltage to achieve what you want.
But, for linear power supplies, is not a easy task achieve well regulated 40V from more or less 43 volts. The transformer must be very good. For this level of voltage I sugest you work in non regulated mode. A switch to bypass the regulator could work.
Best regards,
SoP
It is clear that you wish to use your existing transformers and that you would like a variable supply rail voltage, so some of the earlier replies have been less than helpful. Here are a couple of suggestions.
One option is to build a conventional unregulated supply using the transformers you have and to feed the transformers from a Variac. This will enable you to alter the primary voltage thus giving you variable supply rails with little loss of efficiency.
If you prefer a regulated supply there will be reduced efficiency, particularly at the lower rail voltages, due to the power dissipated in the pass device. An LM338K could be used, with care, because it is a floating regulator and the voltage limitation is on the input to output differential voltage rather than an absolute voltage to ground.
However, this is not an approach I would recommend as the LM338K has a lower maximum junction temperature and a higher junction-case thermal resistance than a good power transistor so the heatsinks will need to be larger and you still run the risk of the regulator tripping out due to its thermal limiting circuitry when you have a high input to output differential voltage (ie when you have it set for +/-25V rails).
My preference would be for the current boosted LM3x7 circuit (there is a schematic on my website) using robust pass transistors (ie ones with a 200degC maximum junction temperature and a junction-case thermal resistance of 0.7degC/w, such as the MJ15003/4), or perhaps even two in parallel if your amps are going to drive a low impedance load. You will still need fairly large heatsinks but it should be workable.
It is clear that you wish to use your existing transformers and that you would like a variable supply rail voltage, so some of the earlier replies have been less than helpful. Here are a couple of suggestions.
One option is to build a conventional unregulated supply using the transformers you have and to feed the transformers from a Variac. This will enable you to alter the primary voltage thus giving you variable supply rails with little loss of efficiency.
If you prefer a regulated supply there will be reduced efficiency, particularly at the lower rail voltages, due to the power dissipated in the pass device. An LM338K could be used, with care, because it is a floating regulator and the voltage limitation is on the input to output differential voltage rather than an absolute voltage to ground.
However, this is not an approach I would recommend as the LM338K has a lower maximum junction temperature and a higher junction-case thermal resistance than a good power transistor so the heatsinks will need to be larger and you still run the risk of the regulator tripping out due to its thermal limiting circuitry when you have a high input to output differential voltage (ie when you have it set for +/-25V rails).
My preference would be for the current boosted LM3x7 circuit (there is a schematic on my website) using robust pass transistors (ie ones with a 200degC maximum junction temperature and a junction-case thermal resistance of 0.7degC/w, such as the MJ15003/4), or perhaps even two in parallel if your amps are going to drive a low impedance load. You will still need fairly large heatsinks but it should be workable.
I think Eva´s suggestion is the best for low losses.
However, since the maximum power is 240VA/transformer, one or two LM338K (paralleled) correctly designed will be fine with no adittional necessities. Just take care of the heatsinks and maximum vdiff over the CI. The losses are (much) higher, though.
best regards,
However, since the maximum power is 240VA/transformer, one or two LM338K (paralleled) correctly designed will be fine with no adittional necessities. Just take care of the heatsinks and maximum vdiff over the CI. The losses are (much) higher, though.
best regards,
Good choice. There are some nice deals on variacs on German ebay if you have 3-4 weeks patience. I searched for "(stelltrafo, stelltransformator, regeltrafo, regeltransformator, variac) -märklin -marklin -piko -pico -trix -roco -minitrix -fleischmann -primex" to ignore the rail model crowd ;-)
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.