I would claim that he is actually using the choke in differential mode. Common mode chokes have current compensation (both dots at the same side). In the example from NEM the currents are adding up, this is effectively a choke input supply with the choke in both legs of the supply.
yup , as usual , ZM's brainfart
what I meant is OK , what I posted (sketch) is OK , while name isn't - it is differential choke , polarized for flux canceling
ZM,
I put "common mode" in quotes since it is not really a common mode connection. I know I added to the confusion.
I know how to wire it, I am concerned about the current handling capacity.
More I have thought about it I figure I would use it with coils paralleled to get maximum current capacity.
I am more concerned that I have no idea (well, more of a vague idea) of what is needed to keep it from overheating. I had always planned on using one for each rail.
Anyway, these things offer much more inductance than the HAMMONDs and the 1.7R, at least using DUNCAN PS tool, this resistance does no harm to the voltage output. I am getting actually a little more voltage than with the resistor. I got the best looking result with a first cap of 6300 uF and a second cap of 10000 uF.
Will a choke with a 5 amps rating survive here?
I, too had my doubts about the ANTEKs but when I see that my SIT 1s are using them I figure they cannot be too inferior.
I do not doubt the TALEMA superior but the one I would need is not stocked in the US. The amplifiers are used between 100 and 500 Hz. I figure the 300 VA ANTEKs will be sufficiently good in this application.
I can say the ANTEKs use larger wire than any other toroid I have used and they are shielded - though I am not sure I want to use that feature. With as much junk as the toroidal by its nature allows into the circuit I worry the shielding would allow even more junk to float around the case..
If I was to make a true financial sacrifice in the quest of a good toroid I would get them from TOROIDY, I think they are whom I have heard many good things. But a 120 volts primary would be a special order if they offer them at all. Their prices are good, about 50% more than the ANTEKs but I have no idea what it would cost to get them here. I bet the freight would be as much as the transformers themselves if not more.
With PLITRON I would have to pay $250.00 for one transformer and then the freight.
I put "common mode" in quotes since it is not really a common mode connection. I know I added to the confusion.
I know how to wire it, I am concerned about the current handling capacity.
More I have thought about it I figure I would use it with coils paralleled to get maximum current capacity.
I am more concerned that I have no idea (well, more of a vague idea) of what is needed to keep it from overheating. I had always planned on using one for each rail.
Anyway, these things offer much more inductance than the HAMMONDs and the 1.7R, at least using DUNCAN PS tool, this resistance does no harm to the voltage output. I am getting actually a little more voltage than with the resistor. I got the best looking result with a first cap of 6300 uF and a second cap of 10000 uF.
Will a choke with a 5 amps rating survive here?
I, too had my doubts about the ANTEKs but when I see that my SIT 1s are using them I figure they cannot be too inferior.
I do not doubt the TALEMA superior but the one I would need is not stocked in the US. The amplifiers are used between 100 and 500 Hz. I figure the 300 VA ANTEKs will be sufficiently good in this application.
I can say the ANTEKs use larger wire than any other toroid I have used and they are shielded - though I am not sure I want to use that feature. With as much junk as the toroidal by its nature allows into the circuit I worry the shielding would allow even more junk to float around the case..
If I was to make a true financial sacrifice in the quest of a good toroid I would get them from TOROIDY, I think they are whom I have heard many good things. But a 120 volts primary would be a special order if they offer them at all. Their prices are good, about 50% more than the ANTEKs but I have no idea what it would cost to get them here. I bet the freight would be as much as the transformers themselves if not more.
With PLITRON I would have to pay $250.00 for one transformer and then the freight.
Rick - LL2733 will work as I sketched , without problems , no need for one per rail ....
considering stock Iq of your amp , current capacity of windings is right on spot , so one LL2733 in differential mode per channel PSU
regarding Donuts - what's good for Pa , it's good for you (us) too
I had opportunity to be disappointed in quality of too many established names , so if you are going to over-think things about smartest purchase , there is greater chance to get a expensive dude (according to Murphy)
ever heard buzzing Lundahl ?
I did .....
considering stock Iq of your amp , current capacity of windings is right on spot , so one LL2733 in differential mode per channel PSU
regarding Donuts - what's good for Pa , it's good for you (us) too
I had opportunity to be disappointed in quality of too many established names , so if you are going to over-think things about smartest purchase , there is greater chance to get a expensive dude (according to Murphy)
ever heard buzzing Lundahl ?
I did .....
Last edited:
ZM,
What would DIY audio be without obsessive overthinking of details?
I am satisfied with the ANTEK product.
Maybe the day before I die I will understand the interrelationships of power supplies and not a moment before.
Of course, this obsession should be directed towards the SIT 1s since they are used in the range most prone to noise. If this works out well I will probably use the same choke in them.
I have not used enough LUNDAHL stuff to have had any problems but they do seem to be a step above the ordinary. In addition the fellow who distributes them in the US is a good fellow and I prefer sending money to people I like.
I will give them a try.
One more thing to torture you with: there seems to be a good effect from using a resistor (0.05R) after the rectifier. I would have to use a slightly higher voltage (20) which would cost a couple of dollars more. What do you think of doing that? The danger of dilettantes using the DUNCAN AMPS tool!
What would DIY audio be without obsessive overthinking of details?
I am satisfied with the ANTEK product.
Maybe the day before I die I will understand the interrelationships of power supplies and not a moment before.
Of course, this obsession should be directed towards the SIT 1s since they are used in the range most prone to noise. If this works out well I will probably use the same choke in them.
I have not used enough LUNDAHL stuff to have had any problems but they do seem to be a step above the ordinary. In addition the fellow who distributes them in the US is a good fellow and I prefer sending money to people I like.
I will give them a try.
One more thing to torture you with: there seems to be a good effect from using a resistor (0.05R) after the rectifier. I would have to use a slightly higher voltage (20) which would cost a couple of dollars more. What do you think of doing that? The danger of dilettantes using the DUNCAN AMPS tool!
Last edited:
Actually I need to correct myself in that the coils are in series, so the current does not add up, the inductances "add" up, so 1 x I and 4 x L (as in the lundahl data sheet).
Edit: Ignore this , it is a bipolar supply so currents do add up
Last edited:
I also often think like this, but not just about power supplies, but everything to do with Audio amplification.
Pailings on a picket fence at 1000mph.
hardly
speaking of Iq per channel , it can be anything between 1A5 and 2A , most likely something around 1A7-1A8
frankly , I never did measured voltage sag across that resistor , and there is possibility that you're getting wrong value , influenced with peak-ing current for second cap itself
what is better, to measure voltage sag across source resistors , say one "bellow" Q9 (lower SSouth, that one connected to neg rail)
speaking of Iq per channel , it can be anything between 1A5 and 2A , most likely something around 1A7-1A8
frankly , I never did measured voltage sag across that resistor , and there is possibility that you're getting wrong value , influenced with peak-ing current for second cap itself
what is better, to measure voltage sag across source resistors , say one "bellow" Q9 (lower SSouth, that one connected to neg rail)
Dear ZM,
You should know better than to ask ME what is the best way to measure something!
I made my order yesterday to Kevin Carter at K&K AUDIO the US distributor and he asked to me to measure the voltage across that resistor.
The amps were left on about five minutes and I saw no change in the reading.
I made the measurement three times. After finding out it was 3 amps I figured I HAD measured it wrong but I did not and I did measure both the + and - rail's resistor bank.
AS far as what is happening in the circuit I would completely defer to you.
I do know one thing I have "discovered" from playing around with the DUNCAN AMPS tool - the output voltage I get when configuring for even 1.5 amps output is greatly below what is actually there - in my case 22.35 volts. If I used a resistor load for a 3 amps draw the voltage would sag greatly. I think I remember 18 volts or less. Needless to say if I use the constant current mode it is next to nothing. OBVIOUSLY, I have little idea how to use this tool. So i tend to agree with you that 3 amps seems unlikely but I would hate to burn the thing up!
As time has gone by I do wonder about the coil in the ground. Some folks seem to think this is really bad - though it seems to me the resistance in the positive line is still resistance in the ground but I have no real understanding of how this works. But I would like to get the benefit of the series coils - 400 mH produces less ripple than 100 mH.
You should know better than to ask ME what is the best way to measure something!
I made my order yesterday to Kevin Carter at K&K AUDIO the US distributor and he asked to me to measure the voltage across that resistor.
The amps were left on about five minutes and I saw no change in the reading.
I made the measurement three times. After finding out it was 3 amps I figured I HAD measured it wrong but I did not and I did measure both the + and - rail's resistor bank.
AS far as what is happening in the circuit I would completely defer to you.
I do know one thing I have "discovered" from playing around with the DUNCAN AMPS tool - the output voltage I get when configuring for even 1.5 amps output is greatly below what is actually there - in my case 22.35 volts. If I used a resistor load for a 3 amps draw the voltage would sag greatly. I think I remember 18 volts or less. Needless to say if I use the constant current mode it is next to nothing. OBVIOUSLY, I have little idea how to use this tool. So i tend to agree with you that 3 amps seems unlikely but I would hate to burn the thing up!
As time has gone by I do wonder about the coil in the ground. Some folks seem to think this is really bad - though it seems to me the resistance in the positive line is still resistance in the ground but I have no real understanding of how this works. But I would like to get the benefit of the series coils - 400 mH produces less ripple than 100 mH.
I still prefer one dual inductor connected differentially than two solo inductors (one per rail)
-no need for gapped cores , so factor of chunkiness is smaller
-greater amount of inductance possible , or smaller Rdc for same inductance
from experience - proper transformer , regular Graetz bridge , common FW filter and common FW Iq - from 18Vac secondary you'll get something as 22Vdc rail
no brainer , shown in practice numerous times
what you are probably missing , playing with PSUD , is proper data for xformer
right click on it , go for edit , input full data
-no need for gapped cores , so factor of chunkiness is smaller
-greater amount of inductance possible , or smaller Rdc for same inductance
from experience - proper transformer , regular Graetz bridge , common FW filter and common FW Iq - from 18Vac secondary you'll get something as 22Vdc rail
no brainer , shown in practice numerous times
what you are probably missing , playing with PSUD , is proper data for xformer
right click on it , go for edit , input full data
ccs as load , what else
for example :
input 18V ,90mR for xformer
bridge - BRIDGE25
then C 47mF/10mR
then RC , made of 100mR and same C as previous
simulate for 50mS after reporting delay of 5S
edit - for one channel (remember also that you're simulating just one rail , but that's sufficient) use CCS load in value of one channel's Iq
for two channel PSU , just double value for CCS load
for example :
input 18V ,90mR for xformer
bridge - BRIDGE25
then C 47mF/10mR
then RC , made of 100mR and same C as previous
simulate for 50mS after reporting delay of 5S
edit - for one channel (remember also that you're simulating just one rail , but that's sufficient) use CCS load in value of one channel's Iq
for two channel PSU , just double value for CCS load
ZM,
How much current is specified for you CCS?
I have found it interesting to see what happens at 0 seconds, too. Supposed to be a measure of the supplies "quickness" and an easy way to look for overshoot.
So far I have found the best looking results with 6mF on the first cap and 18mF on the second with the coils in series. With them in parallel, almost as good but the values still work as good as any other I tried and I tried a great many.
Also, do you think one can get too low with ESR? There are these very nice 3000mF caps from KEMET that I would use as a pair for the first cap but they are already low for ESR and wonder if they could get too low in parallel.
Any opinion or cautions?
How much current is specified for you CCS?
I have found it interesting to see what happens at 0 seconds, too. Supposed to be a measure of the supplies "quickness" and an easy way to look for overshoot.
So far I have found the best looking results with 6mF on the first cap and 18mF on the second with the coils in series. With them in parallel, almost as good but the values still work as good as any other I tried and I tried a great many.
Also, do you think one can get too low with ESR? There are these very nice 3000mF caps from KEMET that I would use as a pair for the first cap but they are already low for ESR and wonder if they could get too low in parallel.
Any opinion or cautions?