24V Reg. shutdown Question

After about 4-5 minutes my 24V rail turns off while my 48V rail stays on. Here are the details. This is with all the LEDs and Relays on, 4 Phantom Supplies on.

I calculate 482 mA for Everything

PS is SSL tech with +/-24V( 7824 7924) and Phantom (LM317)
-Heatsinks on all 3
2-990C
2-2520
12 Relays (41.7 mA each)
16 LEDs

30VA 25V Toroid:
http://www.avellindberg.com/transformers/y23_range_specs.htm


Do I need a bigger toroid?

I can post pics soon if that would help.

[quote author="guavatone"]After about 4-5 minutes my 24V rail turns off while my 48V rail stays on. Here are the details. This is with all the LEDs and Relays on, 4 Phantom Supplies on.[/quote]
Your transformer should be OK, as long as its not getting too hot. When you say the +24V rail turns off, do you mean it shuts down to zero volts, or it droops to some unstable lower voltage?

From the figures you gave, it looks as though a substantial fraction of your regulated current is going into the relays. I personally would not run the relays from the same voltage regulator as the electronics. Relays do not need accurate regulated voltages, and the current pulse during energising can kick back into the op amps. It could be that the unregulated d.c. from the rectifier is within the tolerance range of the relays. If you are using mechanical switches for switching the current through the relay coils, you could try distributing them between the +24 and -24 (raw) rails to reduce the maximum imbalance in current draw.

7824 will shut down if it gets too hot. try with heatsink.

Thanks, I'll check it out Boswell, but it powers down enough to kell the power to the relays and LEDs.

Maybe I'll find a chuck of aluminum to add more sink, I am using small-ish ones now.

Hey GuavaT

I've found that an old front panel cut into strips works really well for a heat sink.

I have a bunch of them cut.. I'll send them with your knobs

Cannakin, do they just hnag off the regs?

yup -- a little grease and your good to go

A little cuting, some berilium. On for over 1 hour without therman shutdown, but the case is damn hot. I have gear that gets hot but, what the ____? even drilled about 30 small holes above the regs for venting.




I was a little impatient david...

12 realays @ 41mA = half an amp (plus the rest of the circuitry), even tho the regs are rated @ 1 or 1,5 amps, they will get hot.

I'm sure you'll see that the -24v reg will be much cooler than the +24v reg.

For an 8 channel Green (about 800mA draw), I bolt the regs to the case & the case is fairly warm in that area.

Peter

I don't think it's related to current draw alone, I think it's related to the reg getting fed around 35vdc. you are dropping 10v AND/or you are riding damn close to the 36v limit that most regs will shutdown at. Couple this with the current draw and that reg will be hot!

Would it be better to go with the 22-0-22 xfmr? or Higher current toroid of 50VA?

P = V * I

In this case (no pun intended), V is the drop across the regulator. If you're feeding it 35V on its input, then V is 35-24=11V

11V * 0.5A = 5.5W

That is a fair bit of heat to dump, so a good sized sink is in order.

Now what's this about Berylium? You realize that stuff is highly toxic, right?

You don't have to worry about the 37V limit here--that's the spec for in to out differential. With a large enough heatsink, you could have 24+37 = 61V on the input.

<edit> To reduce power dissipation, reduce the input voltage. be sure to leave about a 3V margin in to out (i.e. no less than 27V on the input) for reliable regulation.

A P

Heatsinks are commonly spec'ed by a parameter named k(w), which is the temperature rise in degrees Celcius for each watt dissipated in it.

Low k(w)'s are high efficiency, high k(w)'s are low effiency.

Knowing the k(w) of your heatsink and the power dissipated, you can precisely predict heatsink temperature.

Jakob E.

[quote author="AnalogPackrat"]

Now what's this about Berylium? You realize that stuff is highly toxic, right?
[/quote]
Thanks-
It's that white thermal compound. Sh-t! my tube doen't say berylium but I thought it was. -and I am concerned about amalgam -sheez.

Let me take a moment to wake up and digest this stuff (not berylium)

I can't seem to find my formula for the correct xfmr (eg how we came up with 35V from 25) -the 35V value makes me think I should just use a lower Voltage xfmr so the PS is more efficient.

I probably wouldn't worry about beryllium in your thermal compound. The main ingredients are usually zinc oxide and silicone grease.

Metallic beryllium in fine powder form is quite nasty if inhaled, hence dangerous to machine, but BeO which you might encounter (and could I suppose even be in some thermal grease) is quite inert, and also an excellent dielectric with decent thermal conductivity.

I recall Lew Soloff used to have (maybe still does have) a mostly-beryllium trumpet. It was remarkably light. I don't know who braved the hazards of working with the material to make it.

Formula for figuring DC voltage from a transformer:

In a full-wave bridge rectifier, if Vt is the voltage on the transformer secondary, then the DC voltage will be Vt times the square root of 2, or about 1.414, minus the diode drops, typically 2.5V. So a 25V secondary will yield 35.35V minus 2.5V, or 32.85V. You're actually getting 35V which indicates that your transformer is putting out more than the rated voltage (or your wall power is above normal).

For a 24V regulator, ideally your unreg. voltage should be 30V; that allows for a 10% drop to 27V during summer brownouts without going out of regulation. If the power is +10% at some point, your unreg. voltage will then be 33V and your regulator dissipation will be 9V x 0.5A = 4.5W. So you need a good chunk o' heatsink. Or rig separate regulators for relays and audio circuits.

Peace,
Paul

Thanks Paul ( enjoyed the Impedance article)

OK I think I finally got it and hopefully retain this info.

The toroid is 25V from 115V and my mains are putting out 121V so I figured I take the ratio of 121/115 to figure what I am getting for Power Transformers right? eg. the ratio of 121/115=1.052 so I multiply that by 25V(transformer 115v to 25V) in order to figure my real world value which I figured now to be 34.7V Rectified DC and Power Disipation is 5.35.

In my studio my AC is 123-124(from Equitek balanced power) and rarely goes below 120. After a bunch of calc buttons, I found 22V xfmr to have 3-3.5W and 18V under a Watt but then we are at risk in low power situations.

Oh, your .5A number is from my Current Calculation right?

For some reason I thought that higher Current Power xfmr's could keep down heat. I assume that is a different scenario where the Load exceeds the transformer rating.

Oh, so the best way to deal with this is probably the 22 V xfmr from avenell. In the short term, for heat-sinking I was thinking about adding these small 2W "fins". At this point I'm a bit scunched for space.

These are the sinks I've got

BTW all this is with no mics connected and the 4 DOA preamps not "working" on amplifing a signal. -I'm not sure how much different it will be when it is used to mic a drum kit.

[quote author="guavatone"]
30VA 25V Toroid:
http://www.avellindberg.com/transformers/y23_range_specs.htm


Do I need a bigger toroid?
[/quote]

deleted post
suggested avel Y236105 before noticing the 2nd page of this thread. sorry.

Sh--t I went with the 22V prw xfmr which is rated 115 to 22V adn 680mA. Everything was cool now(literally), until I found that I left it on last night after testing. Uncool(figuretively I think), my one amp fuse blew and my srudio stunk. The regs are most likely fried - something near the reg sparked when I turned it on, but they down't seem shorted.

I am not sure if the XFMR is opperating in spec anymore. My mains are 121VAC, Should I be getting 26.7 VAC on secondary between center tap and one side?

I wanted to get this going for a session tomorrow and I thought I was in the clear. Anyway, from my previous post, I was thinking that the ratio was about 1.05 and this ratio ofr the 26.7/22=1.214, quite a difference.