advertisement


Introducing the PowerReg

is it possible to configure this power reg to output +-57vdc? current draw is around 2A. if thats possible how? whats the replacement for LM317? TIA
 
Hello. Thanks for your scheme.

I can't find D44H11 and D45H11. Maybe it can be replaced with some more wide-spreaded ones?
For example somthig from banzaieffects.com (I'm planning to buy some part from there).
 
Hi Everyone,

As Mike mentioned above we are discussing the possibility of a high current regulated power supply over at Diyaudio.

A couple of things we are after;
  • 30V +/-
  • Up to 10A of current per rail
  • Large buffer capacity.

Mike suggested the PowerReg as a possibility for this PS, sorry for not looking for more information as i am currently mobile, but i was wondering if anyone has any input about using the PowerReg for this task.
 
rhys h said:
Hi Everyone,

As Mike mentioned above we are discussing the possibility of a high current regulated power supply over at Diyaudio.

A couple of things we are after;
  • 30V +/-
  • Up to 10A of current per rail
  • Large buffer capacity.

Mike suggested the PowerReg as a possibility for this PS, sorry for not looking for more information as i am currently mobile, but i was wondering if anyone has any input about using the PowerReg for this task.
Click to expand...

Hi Rhys,

Just read through the thread and I see the powerreg is capable of 10A with sufficient heatsinking.

There were PCB's made up for a group buy earlier. Here is the image:

http://i26.photobucket.com/albums/c114/neiljadman/PRScan.jpg

This was the group buy thread, might be more info on the PCB design there:

http://www.pinkfishmedia.net/forum/showthread.php?t=47030&highlight=teddyreg
 
Trancera, Teddy, Puffin,
What about using TIP2955 (PNP 60V 15A 90W TO218 ) and TIP3055 (NPN 60V 15A 90W TO218 ) instead of D44H11 and D45H11 ?
 
I am using the PowerReg with D44/45H11 to power a bridged gainclone at +/- 27V. Even while driving a 4 ohm speaker it performs well (with good heatsinks), but required a lot of experiments to find the appropriate capacitor combination. I didn't find any other transistor with high hfe which will also be fast as the D44/45H11
 
The 2955 and 3055 are the Horseshoe Crabs of power transistors. Slow, and with low Hfe that drops quickly at high current, almost anything else here would be more appropriate.
 
I suggested MJE15030/MJE15031 a while back because they have similar low Re (emitter resistance) to D44Hxx/D45Hxx. The enquirer didn't like them. The MJEs have half the DC gain and FT of 30MHz vs 40MHz of D4 series. They are widely used as drivers in power amps however and might be more easily obtainable.

Oh, and only 8A Ic max.

David
 
If you want to play with high power regulated supplies, you need to think about the following:

1) Current limit strategy. If something goes wrong, destructive amounts of current can be drawn. At some point, the system has to go into limiting mode to protect itself. Will it just limit the current? Go into a shutdown mode?

2) Thermal stability and overload. The regulator potentially dissipates a lot of heat, and this has to be managed; when you take all the thermal resistances into account you need the junction temperature to stay within bounds. Remember that the allowable dissipation drops rapidly as the junction temperature goes up. You also need to have some method of shutting things down if the regulator heatsinks get too hot. A simple thermal switch at say 70C is better than nothing.

3) Secondary breakdown in the pass transistor. Transistors destroy themselves very fast (in milliseconds) when exposed to forbidden combinations of current and voltage. This could happen if a loud passage (high current drain) follows a quiet section where the raw supply has drifted up to a high value. Again, you need to design in how the system will cope.

4) You need to at least think through how the overall system will deal with catastrophic faults in the regulator; do you have fuses in the right places and have they been sized correctly (in both peak current and I^2T) to protect the other components if something bad happens.

High power design is not an area for bodging; you absolutely have to take a serious engineering approach. If you want others to build it, you need to be able to produce the calculations that show it is a safe design.
 
PigletsDad said:
If you want to play with high power regulated supplies, you need to think about the following:

1) Current limit strategy. If something goes wrong, destructive amounts of current can be drawn. At some point, the system has to go into limiting mode to protect itself. Will it just limit the current? Go into a shutdown mode?

2) Thermal stability and overload. The regulator potentially dissipates a lot of heat, and this has to be managed; when you take all the thermal resistances into account you need the junction temperature to stay within bounds. Remember that the allowable dissipation drops rapidly as the junction temperature goes up. You also need to have some method of shutting things down if the regulator heatsinks get too hot. A simple thermal switch at say 70C is better than nothing.

3) Secondary breakdown in the pass transistor. Transistors destroy themselves very fast (in milliseconds) when exposed to forbidden combinations of current and voltage. This could happen if a loud passage (high current drain) follows a quiet section where the raw supply has drifted up to a high value. Again, you need to design in how the system will cope.

4) You need to at least think through how the overall system will deal with catastrophic faults in the regulator; do you have fuses in the right places and have they been sized correctly (in both peak current and I^2T) to protect the other components if something bad happens.

High power design is not an area for bodging; you absolutely have to take a serious engineering approach. If you want others to build it, you need to be able to produce the calculations that show it is a safe design.
Click to expand...

This is obviously a bit of a problem with the PowerRegs, since they have no overload protection. Have you any suggestions PD?

Andy
 
A direction that might be worthwhile investigating is to achieve overload protection by biasing the jfet.

The jfet can limit the output transistor curent, with the TeddyReg it acts as overload protection (you can short-circuit the output and nothing happens, I tried it!). On the PowerReg however the hfe of the output transistors is much higher, but maybe a bias resistor between the jfet source and ground can help.
 
Teddy,

Rhys and I are interested in implementing the powerreg on a 6 x Lm3886 bridged 300w amp (BPA300). I think its looking very possible as it can provide upto 10A. What are your thoughts on it regarding any safety issues and required capacitance?
 
teddy_pardo said:
A direction that might be worthwhile investigating is to achieve overload protection by biasing the jfet.

The jfet can limit the output transistor curent, with the TeddyReg it acts as overload protection (you can short-circuit the output and nothing happens, I tried it!). On the PowerReg however the hfe of the output transistors is much higher, but maybe a bias resistor between the jfet source and ground can help.
Click to expand...

I don't think that will fail safe; the problem is that once you hit the current limit, and the output voltage falls, the voltage drop on the pass device goes up. This can easily take you into the secondary breakdown region, as you have 10s of volts on the device and maybe 10A or more of current through it.

The other problem is that if one rail goes into limiting, you probably want to shut the whole thing down, to avoid speaker destroying DC offsets.

Normally, high power regulators use some sort of crowbar circuit that shuts the supply right off, and you cycle the mains to reset. Look at the regulator for the NAP250 or 135 for an example.
 
You must log in or register to reply here.


advertisement


Back
Top