[stackofhay]

Quote:

Originally Posted by dgardner Actually, I mispoke here. I ordered this from Plitron: 037035201 TOROIDAL XFMR 2 X 38VAC @ 80VA

I completely agree that the 160VA I mentioned earlier is waaaay overkill for a simple stereo unit, and even more overkill for a dual mono setup.

I bought it because it was the smallest model available at 35+35. I would strongly prefer something with about 1/2 the VA rating at 33-35VRMS per winding.

I saw the Plitron unit, and figured I'd look for something local...

Please let us all know how things turn out with the Plitron!

 

[bg4533]

Quote:

Originally Posted by kevin gilmore with the .2 ohm current limit resistors, the maximum current is about 3 amps. Certainly enough to run a dynomite, but not enough to run speakers. But if you are going to run a dynomite, you really want 2 power supplys for dual mono...

I am still trying to figure out this PSU... I just looked at the PSU schematic, but still don't fully understand it. Both OPA541s seem to be linked together. What is accomplished by this? Each OPA541 is capable of putting out 5A of current. Is the max current of the PSU 3A per DC out or total? How much is each OPA541 limited to here? If a PSU is only used to power one channel is the PSU still capable of the same max output or only half?

 

[stackofhay]

Take a close look at how they are hooked up.

The 5V voltage reference is multiplied by +6 by the first op amp, thus generating 30V for V+.

The second op amp uses the +V rail as an input. The second's gain is set to -1 so its output is -30V. Each rail is driven by a separate op amp.

This way, they both track as complements as components warm up etc... and theoretically, the servo won't have to do much work maintaining DC offset.

Kevin Gilmore mentioned in an earlier post that matched, low temp coefficient resistors for "the second op amp" are important.




Quote:

Originally Posted by bg4533 I am still trying to figure out this PSU... I just looked at the PSU schematic, but still don't fully understand it. Both OPA541s seem to be linked together. What is accomplished by this? Each OPA541 is capable of putting out 5A of current. Is the max current of the PSU 3A per DC out or total? How much is each OPA541 limited to here? If a PSU is only used to power one channel is the PSU still capable of the same max output or only half?

 

[ble0t]

The first OPA541 is setup to have a gain of 6 with an precise input of 5V from the REF02, thus producing the +30V rail. The second OPA541 has negative unity gain and is setup to follow the first OPA541, but produce a negative -30V rail. The 0.2R resistors limit the current to 3A in each case. The supply voltage for both OPA541's is supplied by the LM338's.

Hopefully, I'm looking at it correctly myself

EDIT: that stack of hay beat me to it

 

[stackofhay]

Quote:

Originally Posted by ble0t EDIT: that stack of hay beat me to it

 

[ble0t]

I've sent quotation requests to 4 different companies (Victoria Magnetics, Avel Lindberg, Amveco and Toroid Corp) for 10, 25 and 50 piece orders of a transformer with the following specs:

80VA
115V/230V dual primary
35V - 35V dual secondary @ greater than 1A

I've kept everything bare minimum on the quotes thus far (standard plate mounting, no shielding, etc.) to get an idea of the base cost. Hopefully I'll get some replies back by early next week.

 

[dip16amp]

Here's a couple of transformers that I tested on the power supply. Both are used with the output wiring in series so two transformers are needed per power supply. I use four transformers for my balanced dual mono dynahi. I used a couple of 5k bourns pots in place of the 3.5k resistors to adjust the LM output.
Measurements are: transformer output vac rms, diode output, LM output, OPA output.

Radio Shack 273-1512 25.2 VAC 2 amp 50 VA $10.49 each
no load: 27.9 vac, 37 vdc, 33 vdc, 30 vdc
with load: 26.9 vac, 33 vdc, 31 vdc, 28 vdc


Digikey PL56-28 28 VAC 2 amp 56 VA $18.14 each
no load: 34 vac, 45 vdc, 35 vdc, 30 vdc
with load: 33 vac, 39 vdc, 35 vdc, 30 vdc

The 25.2 volt transformer clearly does not have enough voltage under load and shows what the minimum volts that the LM (two volts) and OPA (three volts) need.

The 28 transformer under load shows a two volt margin on the LM regulator and a two volt margin on the OPA541. The diode output of 39 vdc is also under the maximum limit of the LM338 (40 vdc) that it will see on power up. The 35 vdc is under the 36 vdc operating range if the REF102 10 volt reference is used. I will change the 50k resistor to 20k and try the REF102CP next week.

Digikey and Mouser also sell the VPP28-2000 from Triad Magnetics that has the same specs (28 vdc, 2 amp, 56 VA) which should work. Mounting brackets or standoffs are needed for mounting on both of the 28 volt transformers.
I was also considering the digikey TE62083 toroidal with series output 30 vac 1.66 amps, 50 VA but I don't really need the higher voltage.

 

[sft]

What's the real value of AC-voltage(wall wallet) in California?Thanks in advance.

 

[amb]

Quote:

Originally Posted by sft What's the real value of AC-voltage(wall wallet) in California?Thanks in advance.

The US standard is 117V AC 60Hz.

 

[dip16amp]

My wall outlet currently measures 119 vac.

 

[sft]

Thank you everybody.

 

[Woody]

Quote:

Originally Posted by ble0t I've sent quotation requests .... a transformer with the following specs: 80VA 115V/230V dual primary 35V - 35V dual secondary @ greater than 1A .

Just noticed that this is a standard part for Plitron, P/N 037018201.

 

[kevin gilmore]

quote
The US standard is 117V AC 60Hz.

That may be the standard, but it is completely useless especially
around large cities with significant manufacturing areas.
In the summer i have seen my electricity as low as 105 and in the
winter as high as 122.

 

[Woody]

There is an amount of conflicting info in this thread, I'd dearly love to clarify some of it as it could help with sourcing a transformer. The current favourite seems to be a 2x35v 80VA transformer, which of course is not readily available in the UK !

Firstly, dip16amp has calculated minimum required voltages at each point in the PSU circuit, based on data sheets allowing for 30Vdc out @ 1A,

26.2 Vrms before diodes/ Output of transformer (Too low ?)
35Vdc before LM338
33Vdc after LM338 (And into OPA541)

Now here are some real world measurements of the Dynahi amp:

KG: 300mA per amp board
dgardner: 450mA per amp board

Now here are some real world measurements of the PSU:

Dip16amp , using 2x 28Vrms 100VA sees 33Vrms before diodes, 39Vdc before LM338 and 35Vdc into OPA541

KG, using an Avel ?????, sees 33.6Vrms before diodes, 40Vdc before LM338 and 36Vdc into OPA541.

MattN using a 30Vrms 160VA transformer, sees 39Vdc before LM338.

Observations:

1) KGs boards use 50% less power than dgardners.

2) Dip16amps 28Vrms (Digikey) transformer puts out 34rms with no load. So it either should be sold as a 33Vrms transformer or its regulation is appalling ? [edit] Or the line voltage was high ?

3) All 3 transformers end up with the desired 39/40 Vdc before the LM338, although this doesnt necessarily make sense. Now in both Dip16amps and KGs measurements there is perhaps a large voltage drop across the rectifier diodes, where as MattNs drop is smaller (closer to expected ?) The other possibility is that all three measurements were taken under different conditions !

This is all suggesting to me that a 30Vrms >100VA transformer may well be the way to go and if there is any problem with too low a voltage into the LM338, then adjust its output as required to maintain the necessary headroom (>2v), so long as >=35Vdc is maintained before LM338 and >=33VDC is maintained after it.

The good news is that 30VA 160VA transformers are readily available.

Can anyone else chip in with some more real world measurements ? Perhaps measuring with a fixed 1A load on both 30Vdc rails ?

 

[kevin gilmore]

The amount of power the amp boards pull are directly related to
the led voltages. I designed it such that with the bag of led's i have
(about 500 left) i get each board to pull 300ma which is 600ma total.
dgardner's boards are pulling more than that. Nothing wrong with that
either as long as you have the required amount of heatsinking.
The transformers i'm using are avel-lindberg. Because that is what
i have a number of laying around. I have ordered custom transformers
which should finally be here next week. My measurements were with
a true rms voltmeter. I did not measure my line voltage at the time.