Sjöström Audio DIY forum

Projects => Power supplies => Topic started by: raq on April 16, 2012, 03:07:49 PM

Title: JSR03 @ 43Vout
Post by: raq on April 16, 2012, 03:07:49 PM
Hi P-A,
Hi everybody,
I am trying to put together some calcs for 43V out w/ the JSR03 and parts I have available in my bin.
- op amp Vref - prob. (2) LT1034 Vrefs @ 2.5V, to make Vref 5V;
- I am not sure about Dz1, which should be half Vout; some low noise Vref? Any ideas?
- R8 ~ 7.6 kohm if R9 is 1kohm; R7 = R8IIR9 etc.;
- T power - D44H11 is fine.
Any thoughts?
Thank you,
Radu.
Title: Re: JSR03 @ 43Vout
Post by: raq on April 16, 2012, 06:45:18 PM
...this might also exceed the ratings of LM317T. Vin max is 37V. Any workout available, P-A?...
Thanks much.
Radu.
Title: Re: JSR03 @ 43Vout
Post by: peranders on April 16, 2012, 07:35:49 PM
LM317 = 35 V ACROSS Vin and Vout but if you use it you must have at least 5-7 Volts more than out.

Which opamp do you use?
Title: Re: JSR03 @ 43Vout
Post by: raq on April 16, 2012, 07:54:56 PM
I am thinking MAX410, though I obviously have to sort out something about its power supplying (which maxes out at 12V). Maybe try a voltage divider @ R12 considering its current draw - 3.3mA). Or maybe altogether replace R12 with an 7810. Having said that, I do need to think well how split supply vs. single supply works in this case. Any indications or thoughts?

I can also use LM49710.

Also - I am considering three 14V LEDs for "DZ1" (half Vout). I like LEDs for low noise (and they look pretty cool too!).

Thank you, Per-Anders.
Title: PS
Post by: raq on April 16, 2012, 11:52:13 PM
...I think I have 47V Vin.
Title: Re: JSR03 @ 43Vout
Post by: raq on April 17, 2012, 06:45:56 PM
...on the op amp:
- I like LME49710 (which worked for me well on my 7.1V application). MAX410 is not too likely to end up being my choice, due to very low voltage rating for Vs. I am unsure how that will interact with the relatively high Vout requirements of the application. 
- other possible choices: OPA227/228; LT1128; OPA27/37; and so on (134, 604, etc.). I lean towards trying OPA228 out, but OPA134 seems to be favored by you, so I'll give it some serious thought. I'm trying to roughly stick with the lowest noise, highest Vs rating option out there, as I will have to drop down from the 47V or so I have in.

Any specific suggestions or thoughts regarding this?

Thank you,
Radu.
Title: Re: JSR03 @ 43Vout
Post by: peranders on April 17, 2012, 08:44:49 PM
R13 = 33 k
R11 = Adjust for the double current of the opamp supply current
Add a 33 V Zener across C10
DZ1 = 24 V

Reference voltage preferable 21.5 V and gain = 2
Title: Re: JSR03 @ 43Vout
Post by: raq on April 17, 2012, 09:58:47 PM
Thank you for your input. Some more questions below.

> R13 = 33 k

I guess this is adjusted for safe LED operation @ roughly 43V.

> R11 = Adjust for the double current of the opamp supply current

I am not sure I understand this. I'd assume what I want is to drop Vs from the 43V to whatever's safe for the op amp by using a R. R calculated by Ohm to drop required V in relation to current drawn. Correct?

> Add a 33 V Zener across C10

Can you please explain? Would this not force the op amp to be fed with 33V? Which would limit choice of op amps considerably. 

> DZ1 = 24 V

I thought this needs to be half Vout, 21.5V

> Reference voltage preferable 21.5 V and gain = 2

Am I getting DZ1 and IC2 mixed up? On the schematic I'm relying on for this, the V reference for the op amp (LM431, for instance) is "IC2," while "DZ1" can also be an LED, per schematic.

Going back to Vref (IC2) - it could ultimately be almost anything, provided the gain is adjusted accordingly, correct? Is there a reason to keep gain as low as possible? Anything wrong with a gain of 17.2, for instance?

Thank you very much for your continued support!
Title: Re: JSR03 @ 43Vout
Post by: peranders on April 17, 2012, 10:29:16 PM
The led should have approx. 1-2 mA

Check the datasheet of the opamp for supply current and take this times two. 43-33 = 10 volts across R11

DZ1 is a level shifter for the opamp output.
Title: Re: JSR03 @ 43Vout
Post by: raq on April 17, 2012, 11:28:55 PM
Quote from: peranders on April 17, 2012, 10:29:16 PM
The led should have approx. 1-2 mA

Check the datasheet of the opamp for supply current and take this times two.

I'm simply not clear as to where does that "times two" come from. For a moment, I thought it has to do with split vs/ single supply, but that doesn't make much sense

43-33 = 10 volts across R11

DZ1 is a level shifter for the opamp output.

Thank you.
Title: Re: JSR03 @ 43Vout
Post by: raq on April 18, 2012, 03:26:41 AM
P-A,
Also... if you can help clear out something else for me - I am not sure which op amp is best suited for this application between the following four. OPA227 or OPA228? OPA27 or OPA37?
They seem to have some some specific differences which I am unsure how thwy relate with the JSR03.
Thanks!
Radu.
Title: Re: JSR03 @ 43Vout
Post by: peranders on April 18, 2012, 06:18:29 AM
All will work but you must have a gain greater and 5 for OPA228 and OP37.
Title: Re: JSR03 @ 43Vout
Post by: raq on April 18, 2012, 01:53:30 PM
Quote from: peranders on April 17, 2012, 10:29:16 PM
Check the datasheet of the opamp for supply current and take this times two.

This questions still stands. Why would I calculate the power supplying circuits for twice the consumption of the op amp?
Thank you.
Title: Re: JSR03 @ 43Vout
Post by: peranders on April 18, 2012, 06:52:15 PM
Because of the zener you must add. You must have some current through the zener. Notice also that it can be complicated to get this thing working properly so I can only give you hints. I have no time to really verify this change.
Title: Re: JSR03 @ 43Vout
Post by: raq on April 18, 2012, 06:57:59 PM
Per-Anders,
Understood and thank you. I will test this and tweak these values if necessary, certainly. As before, chances are that this will add more proven setups to the library, and this would be pretty far off of ballpark, which is a good thing.
I appreciate you taking time with this.
Radu.
Title: Re: JSR03 @ 43Vout
Post by: raq on April 20, 2012, 04:48:00 PM
P-A,

One more thing. Do you see any problems with using a MAX410 with a Vref of 2.5V? I know it has to be within the common mode input voltage range (and it seems to be), but the data sheet is a bit confusing.

The other potential issue I see with this op amp is the very low Vs by specifications and how it limits the output V range, if at all relevant for this application, having in mind the high Vout required. This is more of a gut, undigested thought, so maybe it's not relevant but I'd appreciate your thoughts. 

Thank you.
Title: Re: JSR03 @ 43Vout
Post by: peranders on April 20, 2012, 08:54:07 PM
It's a disadvantage to have a very low reference voltage since you must have so high gain resulting in less feedback and slower performance.

If the opamp can't handle the output voltage you must insert a zener regulator for the opamp. You can check the schematics for the SSR01.
Title: Re: JSR03 @ 43Vout
Post by: raq on April 23, 2012, 06:12:40 PM
I understand your first point and thanks.

Regarding the second issue there - I was thinking more in terms of output voltage swing, though I'm unsure how much it needs to be in this application to maintain regulation. With such low Vs (I will probably supply this with 9V), voltage swing may also be limited compared to Vout and was wondering if this impacts the way this runs.

Thank you.
Title: Re: JSR03 @ 43Vout
Post by: peranders on April 27, 2012, 06:56:14 AM
The needed voltage swing is only a few volts but you must see to that the opamp not will get saturated. The DZ1 is a level shifter for the opamp.
Title: Re: JSR03 @ 43Vout
Post by: raq on May 01, 2012, 09:11:26 PM
Thank you, P-A.