LM723 adjustable bench PSU not working

[Dannyx]

Re: LM723 adjustable bench PSU not working

Based on the tiny heatsinks usually put inside these power supplies. You put a bigger heatsink on, it will do more current safely.

It is advisable to keep transistor case temperature under 60C. If it burns your finger, you need a bigger heatsink.

So basically I CAN go higher in current as long as I manage to cool the thing....makes sense. A fan should definitely be used too IMO...I need this info to hopefully build my own from scratch one day. LONG ago, I put together a very crude design of how I imagined it should look based on Dave's design I saw in one of his videos. That one used a simple plain ol' op-amp (well not really, he eventually resorted to using an LT3080, but that's friggin' expensive), but now I got another IC I can use, this LM723 guy which is purposely built for this and I've got a couple lying around. I thought it'd be easier to upgrade this one, because it's fairly stable. The only thing it lacks is current: can't go higher than 1.5a MAX, so upgrade the transformer, tweak the current limit and it will hopefully do what I imagine.

[eccerr0r]

Re: LM723 adjustable bench PSU not working

Also note that heatsinks will saturate with heat too, so you likely need a fan for small heatsinks. Larger heatsinks may be able to passively dissipate the heat.

[Th3_uN1Qu3]

Re: LM723 adjustable bench PSU not working

Based on the tiny heatsinks usually put inside these power supplies. You put a bigger heatsink on, it will do more current safely.

It is advisable to keep transistor case temperature under 60C. If it burns your finger, you need a bigger heatsink.

[Dannyx]

Re: LM723 adjustable bench PSU not working

A single 2N3055 will be good for 2A continuous, no more

Why is that exactly ? Because of the eventual heat buildup ? If I have a 20v input and I want a 1v output at 2A, according to your formula, that would give me 38w - is that correct ? If so, it seems way below what a 2n3055 can dissipate. I'm assuming you suggested such a low power to keep it in the "safe" zone with the minimal heatsinking.

[Th3_uN1Qu3]

Re: LM723 adjustable bench PSU not working

And since this is a linear PSU, the voltage drop on the transistor is going to be input voltage (as measured after the bridge rectifier) MINUS output voltage.

So for instance if you have 20V at your bridge rectifier output and you want 5V regulated output at 3A from the transistor, your transistor will dissipate (20-5)*3=45W. If you want 15V at 3A then it will be (20-15)*3=15W. As you can see the worst case conditions are with low output voltage and high current - not high output and high current as logic would lead you to believe.

A single 2N3055 will be good for 2A continuous, no more. Especially at the minimal heatsinking on cheap products like that. You can add more in parallel to increase current capability and reliability - even with the same heatsink, using more power devices at the same output current will make a more reliable supply as the load is spread between them.

When you parallel bipolar transistors like this, they are going to need low value resistors in series with their emitters, 0.22 ohms or thereabouts. These create local negative feedback which balances the current between the transistors. If you just wire them in parallel without any resistors, the hottest transistor will take the most current and it will keep getting hotter until it eventually fails. This is an inherent behavior of the bipolar transistor, the hotter it gets, the lower the voltage drop between C-E gets. The resistors are going to need to be rated for at least 3W, with 5W recommended. If you open an audio amplifier you are going to see the same. Anywhere bipolars are paralelled, the same rules apply.

Only bipolar transistors have this behavior. MOSFET type transistors increase in resistance with temperature rise, so when multiple devices are connected in parallel, they balance themselves out without requiring any resistors.

[budm]

Re: LM723 adjustable bench PSU not working

Because 19v*10a=190w (which exceeds the power dissipation of the 2n3055) ? Am I correct ? If so, I should limit the maximum current to around 5a so it's 19v*5a=95w ? I always suck at these, I admit...

Have one of those - found it in the dumpster. Not exactly huge, but definitely powerful....don't know how much it can whack out, but I've ran some automotive halogen lamps on it and it didn't even get warm

Voltage drops on the pass device x current = power dissipation.

[redwire]

Re: LM723 adjustable bench PSU not working

Best to have more than one 2N3055. At 20V their absolute max. is 6A as you see in the SOA curve. For long life, I usually half those curves.

Most 3A 0-20V lab PSU's use 2 or 3 2N3055's for the pass-transistor.

2N3055 datasheet

[Dannyx]

Re: LM723 adjustable bench PSU not working

If you have a 19V input and try to draw 1.5V out at 10 amps, you're going to well exceed what the 2n3055 can dissipate

Because 19v*10a=190w (which exceeds the power dissipation of the 2n3055) ? Am I correct ? If so, I should limit the maximum current to around 5a so it's 19v*5a=95w ? I always suck at these, I admit...

Other sources for massive transformers are the portable heavy old style car battery chargers if anyone is interested...

Have one of those - found it in the dumpster. Not exactly huge, but definitely powerful....don't know how much it can whack out, but I've ran some automotive halogen lamps on it and it didn't even get warm

[eccerr0r]

Re: LM723 adjustable bench PSU not working

Keep in mind that even though the 2n3055 may be able to support 15A, there are a lot of other limits that need to be honored else it just won't work as it should - main one being power dissipation (115W typically - if you have a good heat sink). For these series dissipative power supplies, if you have a 19V input and try to draw 1.5V out at 10 amps, you're going to well exceed what the 2n3055 can dissipate.

Other sources for massive transformers are the portable heavy old style car battery chargers if anyone is interested...

[Dannyx]

Re: LM723 adjustable bench PSU not working

Aha... yeah that's what the schematic looks like what it would do. I was looking for ideas in changing my CV/CC homemade PSU from using a LM317+TIP42 pass to one that uses a 2n3055 + LM723 on hopes for more current and also it seems this combo can go down below 1.2V without a negative supply - I'm thinking it can go to zero just fine.

Alas I want CV/CC so I'll have to modify this circuit a bit. The transformer I'm using also is using a 24V 2A CT transformer that I want it to automatically switch between 12V and 24V in hopes that it will both go well past the 12V limit that I have now (since I'm using the other winding as a negative supply) without dissipating double the power...

I might have to use a RRIO op amp to do current sense for current limiting (CC) but if I could get away without using an external op amp, that would be good...

Instead of just fixing this cheapy thing of mine and leaving it as it was before, I was thinking of trying to upgrade it a bit and increase its current rating. If you think about it, the 2n3055 has a 15A maximum current rating, so the series pass part is capable of being pushed some more. Assuming the only limitation was the small transformer originally in it, I can swap that for a big ol' UPS one and hope for the best. I'll also need to tweak the upper threshold for the current limiter so it no longer trips at 1.5A, but instead at, say, 10A....in the bottom right of my first picture, just above that pot, there's an orange trim pot which most likely sets the maximum current threshold...I'll try turning that to see if it changes anything.
Another problem might be the PNP transistor driving the 3055, since it might not be able to push enough current into the base of the 3055 to reach 10a...there's a formula to calculate this, Ic / hFe = Ib, where Ic is the collector current I want the 3055 to reach, hFe is the gain of the 3055 (duuh) and Ib is the base current required to reach that Ic...the numbers might be too high for that PNP.

[eccerr0r]

Re: LM723 adjustable bench PSU not working

Aha... yeah that's what the schematic looks like what it would do. I was looking for ideas in changing my CV/CC homemade PSU from using a LM317+TIP42 pass to one that uses a 2n3055 + LM723 on hopes for more current and also it seems this combo can go down below 1.2V without a negative supply - I'm thinking it can go to zero just fine.

Alas I want CV/CC so I'll have to modify this circuit a bit. The transformer I'm using also is using a 24V 2A CT transformer that I want it to automatically switch between 12V and 24V in hopes that it will both go well past the 12V limit that I have now (since I'm using the other winding as a negative supply) without dissipating double the power...

I might have to use a RRIO op amp to do current sense for current limiting (CC) but if I could get away without using an external op amp, that would be good...

[Dannyx]

Re: LM723 adjustable bench PSU not working

I was staring at the schematic here for this particular PSU, when the current limit is reached, does it shut off or does it fold (voltage goes down to ensure it doesn't go over the current limit...)

Just trying to understand this schematic...

My particular one turns off for like a second, makes a LOUD beeping sound (notice that piezo on the board), then turns back on. If the overcurrent condition is still present, it will do it again and again until it's removed, so it's not a proper constant current supply in that respect. It's a constant VOLTAGE supply....still haven't fixed it BTW I replaced the 2N3055 but haven't tried it out, mostly out of laziness....was gonna put an UPS transformer on it, since it doesn't matter as long as the input is roughly 20v input.

[eccerr0r]

Re: LM723 adjustable bench PSU not working

I was staring at the schematic here for this particular PSU, when the current limit is reached, does it shut off or does it fold (voltage goes down to ensure it doesn't go over the current limit...)

Just trying to understand this schematic...

[Dannyx]

Re: LM723 adjustable bench PSU not working

This looks similar model is PS-1502d under many names

Yes, doing a search on google, the overall look of the PSU is identical to the one I've got (except for the position of the plugs and knobs which varies here and there).

[Dannyx]

Re: LM723 adjustable bench PSU not working

Actually, this 2N3055 is being used in "forward active" mode not "saturated" switch mode, so it's not supposed to be "full on".

Unfortunately without more details of what was done with the unit, I can't give any guesses as to why it died with the laptop PSU, assuming it was hooked up correctly... Perhaps the input voltage was too low and the current limiting on the output failed to kick in, and too much base drive was applied? Don't know.

Yes, I believe it was hooked up correctly. I left it exactly as it was, except instead of the transformer which was dead, I used the laptop power brick and connected it between the AC legs of the main bridge rectifier. It worked, but when I started loading the output, for some reason the 2n3055 quit. The other PNP driving it is fine though :|

[eccerr0r]

Re: LM723 adjustable bench PSU not working

Actually, this 2N3055 is being used in "forward active" mode not "saturated" switch mode, so it's not supposed to be "full on".

Unfortunately without more details of what was done with the unit, I can't give any guesses as to why it died with the laptop PSU, assuming it was hooked up correctly... Perhaps the input voltage was too low and the current limiting on the output failed to kick in, and too much base drive was applied? Don't know.

[sam_sam_sam]

Re: LM723 adjustable bench PSU not working

Thanks for the info

[Dannyx]

Re: LM723 adjustable bench PSU not working

The 2n3055 is bad: I got a new one and I get a reading with my meter in diode mode across B-E, whereas with the faulty one I get nothing....still, it doesn't answer the question why it died. My only idea is that the laptop power brick I tried using didn't provide enough current to drive the 2n3055 all the way "on", so it got too hot and died.

[Dannyx]

Re: LM723 adjustable bench PSU not working

What's the voltage at the base of the 2n3055 too, check for bad wire?
Are you measuring with any load on the output? With these emitter follower designs, it gets confusing when the output (emitter) has no load...

No load attached. The base of the 2n3055 is around 21v, the same as the collector of the PNP one, so I can only assume the 2n3055 is dead like STJ suggested.
No diodes appear shorted.
Also, yes, the schematic does look very similar so thanks for that - saves me the headaches

[sam_sam_sam]

Re: LM723 adjustable bench PSU not working

Check the diode at the output I had one short out on one power supply like one you have and if I remember right once you got to a certain voltage the current would go to full current but would not show anything on the display