Lucaslucas
just a brand new second hand
I bought them from Cricklewood about two years ago, I cant see them in stock but perhaps they still have some. I think they were NOS: http://www.cricklewoodelectronics.com/Cricklewood/home.php?cat=111
Salas Shunt Regulator Build
- Thread starter Bemused
- Start date
Lucaslucas
just a brand new second hand
Many thanks for the link, I will break it down and work it out.
The placid bp will drop the Voltage if not shunting enough current but this seems odd because with a 90ohm dummy load connected the Salas IRF+- both measure as calculated, red leds come on, measures around 350ma current through r101 (2.5v/8r2).
The placid bp will drop the Voltage if not shunting enough current but this seems odd because with a 90ohm dummy load connected the Salas IRF+- both measure as calculated, red leds come on, measures around 350ma current through r101 (2.5v/8r2).
Lucaslucas
just a brand new second hand
Right you are, thanks.
90r makes only 166ma which would be too little to test it properly. Also, yes (2V5 / 8R2) = 304mA, it would help if I quoted the correct maths
wouldn't it. Well at least it seems to be doing that ok with that dummy (..and this one), which is a start. I used 90r as it was the closest I could get to with the resistors I had available, should have been 75r for 200ma (plus shunt).
I measure that my placid bp feeding legato settles at 250ma +pos and 260ma -neg for ~60ma to be shunted for each channel. I was working with that goal and the latest spread sheet BiB calculator.
Having said all that, the legato manual states:
"The Legato presents a nearly constant load of about~350mA "
so maybe it needs more current..
90r makes only 166ma which would be too little to test it properly. Also, yes (2V5 / 8R2) = 304mA, it would help if I quoted the correct maths
wouldn't it. Well at least it seems to be doing that ok with that dummy (..and this one), which is a start. I used 90r as it was the closest I could get to with the resistors I had available, should have been 75r for 200ma (plus shunt). I measure that my placid bp feeding legato settles at 250ma +pos and 260ma -neg for ~60ma to be shunted for each channel. I was working with that goal and the latest spread sheet BiB calculator.
Having said all that, the legato manual states:
"The Legato presents a nearly constant load of about~350mA "
so maybe it needs more current..
Lucaslucas
just a brand new second hand
Back on these after a couple of months off, think im getting nearer the problem but don't get what is wrong.
[As previous, built salas shunts v1.1, one BJT based for buffalo dac, works perfectly.
Two IRF shunts not working as desired, looking for circa 290ma including shunt @12-15v. Transformer is 2x15v 30va. Measured dc after shunt rectifier as ~20.5v]
Both IRF same problem - With 75R dummy load; V can be set to 12-15v no problem. With real load; legato 3.1 IV stage, no red leds lit, V drop to near 2v.
Across (8r2 /3w) R201/301 measures 2.5v.
So.
Measuring the negative shunt current out to dummy load (75r) I notice that it is too low to power the legato 144ma at 13v (hence no leds and V drop?) and also that it rises with increase in V to 169ma at 14.5v. Can I conclude anything from this?
With CCS separate,
shorted current out is 312ma
2r load 310ma
8r2 load 302ma
75r load 133ma
[As previous, built salas shunts v1.1, one BJT based for buffalo dac, works perfectly.
Two IRF shunts not working as desired, looking for circa 290ma including shunt @12-15v. Transformer is 2x15v 30va. Measured dc after shunt rectifier as ~20.5v]
Both IRF same problem - With 75R dummy load; V can be set to 12-15v no problem. With real load; legato 3.1 IV stage, no red leds lit, V drop to near 2v.
Across (8r2 /3w) R201/301 measures 2.5v.
So.
Measuring the negative shunt current out to dummy load (75r) I notice that it is too low to power the legato 144ma at 13v (hence no leds and V drop?) and also that it rises with increase in V to 169ma at 14.5v. Can I conclude anything from this?
With CCS separate,
shorted current out is 312ma
2r load 310ma
8r2 load 302ma
75r load 133ma
john.luckins
pfm Member
There's a couple or three things that I can think of for you to try.
It looks like you are sure that the Legato negative rail only needs 260mA (and this might be worth checking using a measured split rail supply which isn't current limited, both rails connected), and are sure the CCS of both Salas Regs supplies 60 or so mA more than this, then the fault should lie in the voltage reg bit of the negative Salas reg. The Legato does however present a capacitive load to the Salas which it doesn't like. I'd suggest powering it through something like 2.2 ohm resistors to eliminate this as a cause of instability once you've got things working on dummy load.
If the available current to the load is restricted by the voltage regulator bit of the Salas it can only mean that it is shunting more current than the voltage setting requires. No LED's lit means either the LED/voltage setting resistor string is disconnected, or one of the active devices there is faulty or wrongly connected. It is best to swap out Q203, Q204, Q205 and Q206 while checking that they are the right way round and confirming that their Idss is good (3-5 mA). I've made the mistake of putting the wrong BC BJT tranny in before now and it is easy to get a jFet 180 degrees out. I'd check/swap just the BJT BC device first to see if that gets the red LED's lit and a b-e voltage across it when under load. Then progress to the others. You should be using an IRF530 for Q206 on the negative reg, yes?
If the negative Salas doesn't power the Legato properly then it will have an effect of how the positive one operates when they are both connected. So unless the negative reg supplies the current the 75 ohm dummy load needs it won't work on the Legato. (Yeah, I know, seems obvious when you say it but I made that error too).
Hope that helps.
John
It looks like you are sure that the Legato negative rail only needs 260mA (and this might be worth checking using a measured split rail supply which isn't current limited, both rails connected), and are sure the CCS of both Salas Regs supplies 60 or so mA more than this, then the fault should lie in the voltage reg bit of the negative Salas reg. The Legato does however present a capacitive load to the Salas which it doesn't like. I'd suggest powering it through something like 2.2 ohm resistors to eliminate this as a cause of instability once you've got things working on dummy load.
If the available current to the load is restricted by the voltage regulator bit of the Salas it can only mean that it is shunting more current than the voltage setting requires. No LED's lit means either the LED/voltage setting resistor string is disconnected, or one of the active devices there is faulty or wrongly connected. It is best to swap out Q203, Q204, Q205 and Q206 while checking that they are the right way round and confirming that their Idss is good (3-5 mA). I've made the mistake of putting the wrong BC BJT tranny in before now and it is easy to get a jFet 180 degrees out. I'd check/swap just the BJT BC device first to see if that gets the red LED's lit and a b-e voltage across it when under load. Then progress to the others. You should be using an IRF530 for Q206 on the negative reg, yes?
If the negative Salas doesn't power the Legato properly then it will have an effect of how the positive one operates when they are both connected. So unless the negative reg supplies the current the 75 ohm dummy load needs it won't work on the Legato. (Yeah, I know, seems obvious when you say it but I made that error too).
Hope that helps.
John
Lucaslucas
just a brand new second hand
Thanks John,
Based on using a placid HD BP, my legato takes around 200-210ma for each, actually a bit more on the negative. The salas reg is set up via the calculator to put out ~300 so there should be plenty of shunt in theory from the setup but on measuring Im not actually getting that much out. Ive tried a load of different configurations, it seems that the input cannot deliver enough to provide correct V and current in to a load higher than 50r.
I have the correct components in definitely, Ive at least triple checked (like looking for something lost in the house), also measured them. Ive been discussing this at diyaudio too.
Im not sure but it looks like the transformer might not be enough for this reg.
Might be worth me getting a 2x18v 40-50va
Has anyone else used the twisted pear 2x15v 30va with a salas reg to power a legato 3.1 IV?
Cheers,
Lucas
Based on using a placid HD BP, my legato takes around 200-210ma for each, actually a bit more on the negative. The salas reg is set up via the calculator to put out ~300 so there should be plenty of shunt in theory from the setup but on measuring Im not actually getting that much out. Ive tried a load of different configurations, it seems that the input cannot deliver enough to provide correct V and current in to a load higher than 50r.
I have the correct components in definitely, Ive at least triple checked (like looking for something lost in the house), also measured them. Ive been discussing this at diyaudio too.
Im not sure but it looks like the transformer might not be enough for this reg.
Might be worth me getting a 2x18v 40-50va
Has anyone else used the twisted pear 2x15v 30va with a salas reg to power a legato 3.1 IV?
Cheers,
Lucas
avr300
pfm Member
Non exactly, but I think I have to share my experience. Running the Placid BP into the Legato 3.1 I found that trimming the PSU to match the current demand steady state will not be sufficient to start up the Legato.
My find is that the Legato draws more current "cold" upon start-up than it does steady state (once powered and operational).
Problem is that the increased current (to make the Legato operational) is producing more heat.
Kind of a dilemma... ;-)
My find is that the Legato draws more current "cold" upon start-up than it does steady state (once powered and operational).
Problem is that the increased current (to make the Legato operational) is producing more heat.
Kind of a dilemma... ;-)
john.luckins
pfm Member
Lucas,
What's the voltage across the CCS when it is sourcing 260mA. The FETs in the CCS need around 5 volts across them to deliver that and more when they become hot. Add that to the volts dropped across R201 and take the total away from the volts available from the rectified 30VA transformer and that'll be the very maximum voltage the regulator is capable of. To check if this is the fault try setting the regulator to give a low voltage output, say less than 10 volts and see if the red LED's light with a smaller load current. If they do then the voltage headroom from the supply might be the cause and you will have proven that the voltage regulation part of the Salas works.
I would have thought a non centre tapped 4 diode bridge from a 15-0-15v 30Va would be enough but it might have quite severe regulation of its own, i.e. greater then 30%.
John
What's the voltage across the CCS when it is sourcing 260mA. The FETs in the CCS need around 5 volts across them to deliver that and more when they become hot. Add that to the volts dropped across R201 and take the total away from the volts available from the rectified 30VA transformer and that'll be the very maximum voltage the regulator is capable of. To check if this is the fault try setting the regulator to give a low voltage output, say less than 10 volts and see if the red LED's light with a smaller load current. If they do then the voltage headroom from the supply might be the cause and you will have proven that the voltage regulation part of the Salas works.
I would have thought a non centre tapped 4 diode bridge from a 15-0-15v 30Va would be enough but it might have quite severe regulation of its own, i.e. greater then 30%.
John
themystical
pfm Member
Hi Guys
I have built a +-15V Salas Regulator. I have used R101, the current setting resistor of 10 ohma giving me a current of 300mA and the test load is 140 ohms so 300ma total current with 100ma going to the load and the regulator sinking 200ma. The heat sinks that I have are around 10C per watt but they are getting quite hot. I can only touch them for 2-3S before i have to remove my hand. Is this too hot? Any ideas as to what heat sink I should be using and how hot it should be getting?
Salas Regulator by mystictraveller, on Flickr
I have built a +-15V Salas Regulator. I have used R101, the current setting resistor of 10 ohma giving me a current of 300mA and the test load is 140 ohms so 300ma total current with 100ma going to the load and the regulator sinking 200ma. The heat sinks that I have are around 10C per watt but they are getting quite hot. I can only touch them for 2-3S before i have to remove my hand. Is this too hot? Any ideas as to what heat sink I should be using and how hot it should be getting?
Salas Regulator by mystictraveller, on FlickrYes i'd say you need more heatsinking. You could use a section of L shaped aluminium to bolt both transistors. Alternatively you can mount the transistors from the bottom of the PCB with the legs facing up and bolt the transistors to the bottom of the case (if you are using a case with a metal bottom)
martin clark
pinko bodger
15v*300mA is 4.5watts, so with only one 10degC/W heatsink you'd expect things to run about 45degC above ambient - say 60-65degC. If split evenly between the two heatsinks things will stiill run around 40- 45degC which certainly corresponds with your subjective 'touch for 2-3s'
Yes, that will be too hot to touch comfortably; it's not a problem for the transistors though, either way.
Yes, that will be too hot to touch comfortably; it's not a problem for the transistors though, either way.
themystical
pfm Member
Thanks for the input guys. It looks like I can run it as is for testing/tweaking purposes until it gets mounted in a case when I can sink the transistors to the case. I might see if I can get a thermometer and measure the temperatures of the sinks. What temperatures can the transistors/sinks run at Martin?
martin clark
pinko bodger
For long-term reliability that's about as hot as I would take them.
But silicon junctions will work well past 100degC (175 is often quoted, though life wil be very short)
But silicon junctions will work well past 100degC (175 is often quoted, though life wil be very short)