High School power supply update.

[geologie]

Hi.  I'm a sciences teacher and i want to upgrade our power supply units.  What we have for years are linear 5A with a beefy transformer that are ''regulated'' with a SCR,   That was ok 20 years ago for light bulb and ceramics resistors experiments but programs changed and we now have experiments that needs regulated DC.

We don't have budget to buy new power supply (about 25 units).  So i bought some AC-DC step down on eBay and replace the module that was in the source with these LM317 or LM338 kit.  The problem is that the transformer output is about 36V and these kit don't last long or even don't work at all with voltage more than about 28V.  I experiment a newly rebuilt unit with a LM317 kit with  a variable AC input at about 100V insted of 120V and the power supply like that because the AC output drop to about 28V.

My question is how, without a variable isolation transformer, can i lower the 36v output of the transformer (or the 120V input) so i can put 24V to 28V AC insted of 36V AC in the  LM module ?


These power supply are still sold today and they still have the same bad design for about 550$ CAN! :

https://www.fishersci.com/shop/products/variable-ac-dc-power-supply-ac-dc-power-supply-built-in-panel-ammeter-voltmeter/s441856

The case, voltmeter,  Ammeter, rear breaker, terminal post and transformer are heavy duty; they are built like  tanks that's why i want ot keep them and upgrade.  The only problem is the way it is regulated...

Thanks for helping me and my teen students!

[lmester]

Hello,

The LM317 and LM338 are rated for a maximum differential voltage of 40VDC. Your 36VAC transformer will have a no load DC output of about 51VDC.

As you've already found, this is far too high for the LM317/LM338. Off hand, I can't think of any linear voltage regulators that can handle 51VDC. Also, a quick Ebay search didn't turn up any pre-assembled modules that can handle this voltage.

A simple solution would be to use resistors to make a voltage divider to drop the transformer voltage.  This solution has several problems. This worst is that it would generate a lot of heat. Your power supply would now be an electric heater. Also, power resistors are fairly expensive.

Unfortunately, the only practical solution that I can think of is to just replace the transformer. An 18VAC 40VA transformer should work. This would give you about 20VDC output at the 1.5A maximum current that your LM317 modules can provide. Transformers are fairly expensive. A cheaper solution would be to use an external 24VDC power supply to provide power to your power supply.

A transformer is about $20-$25. A 24V 3A power brick type external supply is about $10-$15.

Some Ebay search links:

Replacement transformer: (If you can find one that will fit inside your power supply case)
https://www.ebay.com/sch/i.html?_odkw=18v++transformer&_sop=15&_osacat=0&_from=R40&_trksid=m570.l1313&_nkw=18v+40va+transformer&_sacat=0

Power brick:
https://www.ebay.com/itm/ABI-24V-3A-AC-Adapter-Power-Supply-Driver-for-24V-LED-Strip-Light/361306136292?epid=1258644087&hash=item541f861ee4:g:laYAAOSwXMBakX1W

Unfortunately you're now getting to the place where the only parts left of your power supplies are the case and meters.

I don't know what your budget is for this. I also work for a school system so I should say lack of budget!

Add up the cost of the LM317 module, a replacement transformer or power brick and your time to do the modifications.

If you ignore the insanely over-priced education lab equipment suppliers, you can get a brand new cheap Chinese generic  bench power supply for about $60. Do some research on this message board if you decide to buy a generic bench power supply.  Some are good and others are of very poor quality.

Here is what I found with a quick Ebay search:

New PS:
https://www.ebay.com/itm/30V-3A-110V-220V-Adjustable-DC-Power-Supply-Variable-Digital-Lab-Grade-Bench/202172867223?epid=2025936273&hash=item2f12712297:g:1hcAAOSwCcZaK08n

[BradC]

Have you considered whether replacing the 317 with a 317hv ( rated to an input of 60V ) might work?

[David Chamberlain]

Like Imester I would recommend purchasing a dedicated DC bench supply to compliment the current AC/DC units you have.

My argument though is more around limiting your personal liability should something go wrong and one of your students gets electrocuted. In that circumstance the total cost could be a human life, your livelihood and potentially your liberty.

I don't mean to come of as negative or overly dramatic but if your hell bent on doing it yourself it would be nice to get some assurances you've crossed this with the relevant school authority, and plan on getting the modifications independently safety tested.

[radar_macgyver]

How about using a DP50V5A or similar with your transformer, a bridge rectifier and caps?

[lmester]

Like Imester I would recommend purchasing a dedicated DC bench supply to compliment the current AC/DC units you have.

My argument though is more around limiting your personal liability should something go wrong and one of your students gets electrocuted. In that circumstance the total cost could be a human life, your livelihood and potentially your liberty.

I don't mean to come of as negative or overly dramatic but if your hell bent on doing it yourself it would be nice to get some assurances you've crossed this with the relevant school authority, and plan on getting the modifications independently safety tested.

This is a very important reply! Even though I work for a public school system, I was only thinking of this as an electronics upgrade problem. Upgrade cost versus replacement cost.

The possibility of electrocuting a student,  the heartbreak that you would feel if a student died, your legal liability and possible time in prison are just not worth the risk.

I'll now suggest only the last option in my reply. Buy new power supplies.

Also, I have dealt with this type of situation in my school district.

Does anyone remember the  legionnaires disease outbreak?

One cause was bacteria in fluid cooling tower sumps. Fluid cooling towers are used in building HVAC cooling systems to remove heat. There are several fluid cooling towers in my school district. To keep the bacteria from building up, you only need to put a trickle of chlorine bleach into the sump water.

We have a contract with an outside company to pump bleach into the sumps and test the water to verify that there is no bacteria. It's the only thing that this company does. We now have no legal responsibility for this. Also, since sanitizing cooling towers is the only thing that this company does I would expect that they would never forget to refill the chlorine bleach drum. Much less likely than a forgotten monthly task for our overworked maintenance staff.

[Gyro]

Your best option is probably to construct some add-on post-regulators rather than modifying the existing PSUs (assuming that the existing PSUs are still considered electrically safe).

This approach would isolate the students from the safety risk of modified PSUs and give you the headroom benefit of a maximum of 25V input. This option would allow you the option of well regulated fixed outputs, eg. 5V using simple 7805 regulators for your sensitive stuff, while keeping your 5A capability for the 'dirty' experiments. Include a bridge rectifier to protect it in case someone connects it to the AC rather than DC output.

It needn't be too expensive to construct 25 units, especially if you can get a discount on Aluminium boxes which could double as heatsinks. Even better if you could 'verbally enforce' a maximum input voltage, of 8-10V to limit dissipation. Worst case you have on-chip thermal shutdown.

I think you need to work our what 'clean' voltages and maximum currents you actually need for your experiments and which ones can be fixed. Variable would be a complication for an add-on unit due to the second voltage adjustment.

[nuno]

The PSU description says

- Offers continuously variable outputs of 0-20VDC and 0-25VAC with a maximum load of 5A

[David Chamberlain]

The PSU description says

- Offers continuously variable outputs of 0-20VDC and 0-25VAC with a maximum load of 5A

If I were to guess why you mentioned this I would say OP is asking because the output is a halfwave sine (via the SCR) smoothed with some capacitor but that the resultant ripple does not provide a narrow enough tolerance for what? We have to guess again. It would be good to know what sort of experiments they are trying to do, I know my microcontroller would not love that sort of waveform to be considered vcc

[Gyro]

The PSU description says

- Offers continuously variable outputs of 0-20VDC and 0-25VAC with a maximum load of 5A

I know it does!  - but what does he want to use a quiet supply for ?  If it's for simple analogue electronics, logic etc. at lower current then a fixed output may be exactly what he needs. Only the OP can answer that!

[nuno]

The Op asks

"(...) can i lower the 36v output of the transformer (or the 120V input) so i can put 24V to 28V AC insted of 36V AC in the  LM module?"

But there are lower voltage outputs already, according to the PSU's link he provided.

[Someone]

The PSU description says

- Offers continuously variable outputs of 0-20VDC and 0-25VAC with a maximum load of 5A

I know it does!  - but what does he want to use a quiet supply for ?  If it's for simple analogue electronics, logic etc. at lower current then a fixed output may be exactly what he needs. Only the OP can answer that!

Exactly, the solution might be as simple as a RC filter added inside the supply to a new set of "quiet" terminals or as an external filter box, or some robust fixed regulators, or using something already better suited such as batteries.

[Someone]

The Op asks

"(...) can i lower the 36v output of the transformer (or the 120V input) so i can put 24V to 28V AC insted of 36V AC in the  LM module?"

But there are lower voltage outputs already, according to the PSU's link he provided.

When unloaded the raw output can rise above the nominal voltage rating, and the lower output voltages rely on setting the knob correctly. Give students control over the voltage and one or more will try it at all the settings no matter how much you try to explain it will damage things, equipment for teaching needs to be incredibly robust and the experiments constrained to avoid destruction (some students just destroy things for the sake of it and some without paying attention).

[John Heath]

For educational purposes clarity of thought of concepts bring presented is a high priority. You do not want details of a lab experiment getting in the way of a concept for a better understanding. Getting a message from A to B is the beginning and end of it. With this in mind I would recommend this power supply. The prices are quoted in Canadian dollars.

[geologie]

We bought this one x 15, thanks! we are very happy

[David Chamberlain]

This one has a happy ending.