Bench Power Supply

[AnalogKid]

BTW, that diode arrangement is part of a ring modulator, used for frequency-division multiplexing of multiple voice channels onto one wideband telephone line. As a spinoff, it was used in early voice scramblers.

https://en.wikipedia.org/wiki/Ring_modulator

ak

 

[davenn]

Post #41, Figure 26 - look at the bridge.

ak

haha I was just reading this old thread and was looking at the diode bridge config in that fig 26
then continued scrolling till the end of the thread and noticed you beat me to the comment

Hi AK, I think I see what you mean in post #43 - the input of the negative regulator needs to be fed from the positive side of the bridge rectifier, while the output of the reg. is held to ground, this would leave the bottom rail to 'go more negative' as the regulator adjust to keep its Vref stable.

If I got it correct, I will draw up a new schematic and post it.

yes... but it's worse that that, the bridge is drawn totally incorrectly, it cannot be just rotated
Man, thats is a huge error from a professional electronics manufacturer !!!!

here a drawing of mine with the diodes drawn correctly ......




cheers
Dave

 

[chopnhack]

True, I didn't pickup on that! I too sometimes forget - two heads, two butts is how I try to recall it.

Any thoughts on the previously asked questions?

 

[davenn]

True, I didn't pickup on that! I too sometimes forget - two heads, two butts is how I try to recall it.

Any thoughts on the previously asked questions?

specifically, which one was that ?

 

[chopnhack]

specifically, which one was that ?

Back to the LM350/LM338!

1. Would it be wise to get one of each regulator from the same manufacturer? My thought was that if there were any small deviations, if the chips were from the same company they would be closer together in their variances.
2. I see that the datasheet states built in short circuit protection - would it be prudent to add foldback current circuitry to prevent the chance of the IC being overheated?

- omit- I would like to have current control and voltage control per channel, is that something I could achieve using this IC? -I think so as this is just another linear reg-

1. a. The other option I am weighing is a pcb and parts for about $10, it wouldn't be a negative rail, but for $20 I could have two channels of pos. voltage. The transformer I was looking at is below as well.

PCB
Transformer

Your thoughts as always!
Thanks in advance

I know that its a bit apples to oranges as one would not yield a neg/pos supply, but in case I am over my head in trying to design one, I left you all an easy way out ;-)

 

[davenn]

Back to the LM350/LM338!

1. Would it be wise to get one of each regulator from the same manufacturer? My thought was that if there were any small deviations, if the chips were from the same company they would be closer together in their variances.

if it's going to be a variable PSU, this it wont be an issue

I see that the datasheet states built in short circuit protection - would it be prudent to add foldback current circuitry to prevent the chance of the IC being overheated?

if you follow through with the variable current aka current limiting that that also shouldn't be an issue

omit- I would like to have current control and voltage control per channel, is that something I could achieve using this IC? -I think so as this is just another linear reg-

variable voltage is easy see the circuit I posted a short time ago
variable current is a little more involved

a. The other option I am weighing is a pcb and parts for about $10, it wouldn't be a negative rail, but for $20 I could have two channels of pos. voltage. The transformer I was looking at is below as well.

looks pretty cool for the price ... wish there was a schematic to confirm what they are doing


Dave

 

[chopnhack]

if you follow through with the variable current aka current limiting that that also shouldn't be an issue

I will continue down this road then. Thanks for the input. I will post a schematic later today.

looks pretty cool for the price ... wish there was a schematic to confirm what they are doing

Yeah, I hate to black box that whole assembly as I am trying to learn, but if I just couldn't get it, I figured this might be a viable option if it even works!! LOL

 

[chopnhack]

After much research today, availability of a matched set has been narrowed to LM317+/LM337- regulators. These would have a cap. of 1.5A of output current. Which is not bad in retrospect. My wants exceed my capabilities as always .

I did find a nifty little board already made up which has some of what I want out of a power supply, I was thinking for the ~$8 they want for it, it might not be a bad board to learn on and modify. Here is the schematic below. The transformer is not included, but if the output caps are only 25v, I doubt I could use a higher value for the xformer. Luckily, this kit is not assembled so I could swap out the caps that are rated too low...

For academic purposes, can someone explain how this circuit allows more current to flow to Vout while maintaining Vreg?
As always, thank you for your patience and support!

 

[hevans1944]

The LM317 maintains Vreg (or Vo) by varying its conduction from Vi to Vo. When the load at Vo draws sufficient current from the LM317 it will bias the 2N2905 PNP transistor on through the voltage drop across the 22 Ω resistor. As PNP collector current rises, it turns on the power NPN by forward biasing it from current through the 500 Ω resistor. The threshold current where this occurs depends on the 22 Ω resistor. Higher values mean the 2N2905 will turn on at lower load currents drawn from the LM317. The PNP and NPN transistors become part of the LM317 feedback loop that controls the output voltage, so stability of this loop may be a concern.

This circuit is often used with multiple LM195 integrated circuits in parallel. These devices emulate a "bulletproof" power NPN transistor with current limiting and thermal shutdown protection. Seems like a good idea to do that. I've also seen a similar circuit using a single PNP power transistor to share the increase in LM317 load current instead of driving one or more paralleled power NPNs. Power PNP transistors may be difficult to find.

I have not used this circuit, so I can't verify it works without problems, but it seems to be popular.