PSU Repair- Transformer outputting wrong voltage. Help much appreciated =)

[jamman1125]

I am trying to fix a power supply to an Alesis Multimix 16 audio mixer that gave out. I have purchased a new transformer to replace the original that was in the supply and made sure the voltages match etc. The problem i have is that the new transformer is not giving me the correct voltages that i need. I require 18v on rail 1, 0v on rail 2 and 18v on rail 3. However when i test the AC output voltage i am getting 0v from rail 1 and 2, but 21v from rail 3. Here are some pictures to assist explanation.

Here is the link to the transformer: http://uk.rs-online.com/web/p/chassis-mounting-transformers/0504640/

And here are the original supply specs

Input:230-240V-50hz
Output 1:18v-700mA
Output 2:18v-700mA
Total output power: 25.2VA

Any help on solving this would be much appreciated, im no expert in this field.

Many thanks

James

 

[(*steve*)]

You're trying to paste links from your email inbox. Unless you want to give us your username and password (hint: don't!) we can't view them.

Upload them somewhere we can all access, or upload them here as attachments.

 

[jamman1125]

Thanks, all sorted!

 

[73's de Edd]

.


Sir jamman . . . . . .




I also am having to rely on your "micro pictures" but I CAN see that you are measuring AC voltage at a connector coming from the T-former.
Thereby you are measuring STATIC voltage . . . .you need to be measuring a DYNAMIC voltage, (under load).
To measure DYNAMIC voltage you need to optimally get two 25 ohm 20 watt resistors for loading down the two 18V AC supplies and then read the loaded voltage produced.

The transformer is constructed with the AC line primary wound on first and then one of the two 18 V secondaries wound on next and the final secondary is wound above the first 18 V secondary.

With you measuring an UNLOADED unit, the first 18V secondary is the closest to the core of the transformer and will receive a specific magnetic flux to induce its output voltage .
The second winding will be spaced out further from the core and will additionally have the mass of the first winding being in between.
That closest winding of the two will be having the greatest voltage reading, with the outer winding having a lesser voltage . . .if reading unloaded windings and getting their STATIC voltages...

The manufacturer has designed the first winding to yield your 18VAC at a 700 ma load and THEN tweaked the second winding by ADDING on corrective extra turns such that the same 18VAC at a 700 ma load loaded voltage is produced by that second winding.
That, taking into consideration, that the final winding is spaced out further in the cores flux path density, along with the quasi-shielding of the first winding being in between.

Either bench test the unit as specified or install the unit in your equipment for the produced loading.

Thasssit . . . . .

73'd de Edd


.

 

[Martaine2005]

.you need to be measuring a DYNAMIC voltage, (under load).

Hey Edd, under those conditions, the voltage would normally be higher.
Loading it would give the proper specs. But without load as the OP is doing, I would expect 21+ volts.
It also sounds like the OP is measuring this as a centre tap transformer. It is not. It has two 18v outputs.
0 - 18v and another 0 - 18v. 4 terminals in total..

Martin

 

[cjdelphi]

That's all well and good...

But why did you buy a replacement transformer? What was wrong (if anything) with the old one?

If your old one died, it would be due to excessive current caused by what ever's attached...

(Connecting it won't fix it, only blow it again, if indeed it's dead)

 

[jamman1125]

Thanks for your replies, so from my understanding I need to give the transformer the same load that the actual mixer would to test the dynamic output voltage?

Cjdelphi: The transformer on the original supply gave out, and it is a known common problem for that mixer. This is why I thought I'd build my own rather than buying a new original one, just for it to die again.

 

[GPG]

as a centre tap transformer. It is not. It has two 18v outputs.
0 - 18v and another 0 - 18v. 4 terminals in total..

You need to connect 2 terminals together and repeat your measurements. If you get ~ 0V between the unconnected terminals reverse the connections of one winding

 

[jamman1125]

You need to connect 2 terminals together and repeat your measurements. If you get ~ 0V between the unconnected terminals reverse the connections of one winding

Which 2 terminals are you referring to?
Also if I reversed the windings I wouldn't get 18v on the correct terminal

 

[73's de Edd]

.


Sir jamman1125 . . .
. . . .

.My prior info was more related to why a person would be confused by imbalanced reading on a transformer with duplicate windings.
With a given, that unloaded transformers will potentially read higher voltages, than when being loaded down.
I am used to seeing a "Pwemium Bwewed" Amerikanski brands, like Thordarson / Merit / Triad transformers, only pull down ~ 10%
on a full spec load down.
Some of the "Cheapies(/cut corner units) will drop down upwards of up to ~20 % and run hot.
THEN we have some Mil Spec units are so good that they only load down in the order of 5%.
(Toroidal wound units can also fall in that last category. )

Lets refer back to the transformer manufacturers sheet that you gave us a link to.
Hey . . . . . it looks like you had added on a cable to it, in your voltage measuring .
Sooooooooo in just looking at the " naked" transformers, lets assign RED . . . .A-B-C-D
identifiers to their secondary winding terminals.
Ideally and probably being expected will be that if you put a wire jumper across terminals B-C that will establish connectin
between the two separate windings and thereby get yourself a common center tap between the units.
If that is correctly establishing a "series aiding " interconnect between the two windings, then in using your AC meter to
measure between a shared BC connection over to A terminal should read 18VAC, or additionally , BC connection
over to D terminal should read 18VAC. Then a final reading from A-D should reveal a totaled 36VAC.
Therefore those two windings were series connected in the correct phasing to add up to 36 volts.

IF you don't get this, then the windings would have been in opposed phasing wiring and you won't end up getting that 36 VAC additive voltage.
THEN, if finding that situation, to correct, go back and reinstall the jumper from terminal B-to-D and let that be the newly acquired
center tap. In testing with the new B-D center tap, it should read 18 VAC to either A or C and then, a final confirmation is made with
there being a full 36 VAC between terminals A and C.

Now hook it up to your system in that way.

A REAL TIME ANALOGY . . . . .

You are all of 4 years old and are exploring Uncle Bills pick up truck and are looking around in the glove compartment and find this flashlight.
You turn it on and off and on and off and play with it all that you can and than notice that the bottom end will turn.
You turn and turn and all of a sudden it pops apart and two long round thingys drop to the floor.
PANIC . . . . you find those two things and then rapidly stuff them back in the flashlight tube and with some musing, find that if you press hard,
the part you had come off on you, will go back on.
Of course, that's with some pressing and some early on self acquired / OTJ education of "Righty tighty and lefty loosey."
Finally that cap is tight and you immediately turn it on to play with it some more . . . . .but ALAS . . . . it no workee !
What a sinking feeling in the stomach . . . . .certainly your britches are in for a dusting !
You take it apart another time and this time . . . TOTALLY by chance . . . . a battery button end to a flat battery end is contacting,
and upon reassembly . . .
Behold ! . . .there was light.
Unknowingly you experienced "polarity" and "series voltage adding".
Definitely time to put that thing back where you got it and go in the back yard and play with the cat that walks up to greet you and
puts its paw up to shake your hand . . . .but that's another story . . . . . .

Now, with your transformer wiring just mentioned , there is a bit of both that polarity, and voltage adding aspect also being involved , but with the
additional complexity of it being an AC voltage.

Thasssit . . . . . . .

HOW-TO-DO-IT-TO-IT :

73's de Edd

 

[jamman1125]

.


Sir jamman1125 . . .
. . . .
Thasssit . . . . . . .

HOW-TO-DO-IT-TO-IT :


73's de Edd

Thanks allot for your very helpful reply, i will try your suggested method and report back. The only query I have about the first centre tap method, is why the A (0v) terminal produce 18v's? Why would the centre tap terminals (BC) produce 0v when they are 0v and 18v?

Much appreciated

 

[(*steve*)]

Pretend it's DC for a moment.

you want -18, 0, 18 you have 0 - 18 and 0 - 18

another way of looking at -18, - 0, 18 is 0, 18, 36 In the former case we measure relative to the center rail, in the latter the end rail (al voltage measurements are relative).

From the latter, you can see that if we join 0 - 18 to another 0 - 18, we will get 0, 18, 18+18. that is 0, 18, 36 and we already know this is equivalent to -18, 0, 18.

This is not DC, it is AC -- but a similar situation exists, that of phase.

If you look at the two diagrams above, the first will likely measure (relative to the far left) 0VAC, 18VAC, and 36VAC,

The second will likely measure 0VAC, 0VAC, 18VAC (the last is the "centre"). In this case the phase relationship is such that the second winding subtracts from the first.

Depending on how your rectifiers are set up, the first WILL work, the second may not.

 

[jamman1125]

Pretend it's DC for a moment.

you want -18, 0, 18 you have 0 - 18 and 0 - 18

another way of looking at -18, - 0, 18 is 0, 18, 36 In the former case we measure relative to the center rail, in the latter the end rail (al voltage measurements are relative).

From the latter, you can see that if we join 0 - 18 to another 0 - 18, we will get 0, 18, 18+18. that is 0, 18, 36 and we already know this is equivalent to -18, 0, 18.

This is not DC, it is AC -- but a similar situation exists, that of phase.

If you look at the two diagrams above, the first will likely measure (relative to the far left) 0VAC, 18VAC, and 36VAC,

The second will likely measure 0VAC, 0VAC, 18VAC (the last is the "centre"). In this case the phase relationship is such that the second winding subtracts from the first.

Depending on how your rectifiers are set up, the first WILL work, the second may not.

Okay so I tried BOTH of the suggestions from Edds post and for the first BC centre tap, i got 21v across BC-A and 21v across BC-D and a reading of 42v across A-D. As this clearly didn't work I tried the second suggestion and still gave 21v but 0v across A-C.

Is there something wrong with the transformer? Am I making a mistake, or would this voltage change under load?

 

[(*steve*)]

The voltage will decrease under load.



It's also possible that if the primary is rated for 230V but your mains is higher that the voltage won't drop as far as 18V. This should not be a problem

 

[jamman1125]

The voltage will decrease under load.



It's also possible that if the primary is rated for 230V but your mains is higher that the voltage won't drop as far as 18V. This should not be a problem

Okay great, so the specifications of the original PSU would be rated under load, and not just static. That seems like a bad idea to me

 

[(*steve*)]

That seems like a bad idea to me

Why? It would be like rating a sprinter by how shiny his shoes are rather than on how fast he runs.

 

[elebish]

Judging the transformers as marked, do this: Before starting, ohm the 2 zero points. Should be open. Now ohm the two 18 volt points. They should also be open. If so, connect the 2 zeros together. Measure for ac from there to each 18 marked point. Should get a few volts above 18 volts ac at each one. Now measure between the 2 points marked 18 and should get 0 volts. That tells you that the two 18 volt windings are in phase and you have 2 separate 18 volt sources, isolated from one another. Now connect one zero point and one 18 point together and measure the 2 unconnected points. You should now have 36 volts ac since both windings are in series. This looks like a doorbell transformer? Ed.

 

[jamman1125]

Why? It would be like rating a sprinter by how shiny his shoes are rather than on how fast he runs.

Because then how would you know what the static voltage rating should be? And dont most appliances say they require a power supply that GIVES a certian voltage, not the rating that the supply comes down to after being connected.

Also the transformer is rated to give 2, 18v rails anyway, so how comes it would give me somthing different to whats specified?

Thanks for helping me out with this

 

[HellasTechn]

Okay so I tried BOTH of the suggestions from Edds post and for the first BC centre tap, i got 21v across BC-A and 21v across BC-D and a reading of 42v across A-D. As this clearly didn't work I tried the second suggestion and still gave 21v but 0v across A-C.

Is there something wrong with the transformer? Am I making a mistake, or would this voltage change under load?

Edds sugestion sounds correct. The 21 Volt you get at each winding pair if you connect them and make it "center tapped" will come down to 18 or 19 Volt under the devices load and your device will probably work. Come to think about it if the device is rated to work on 18 Volt it should also be ok with 21.

If i where you i would wire the transformer just like edd suggests and i would fuse each voltage rail to the rated 700ma and then connect it to the device and power on. But again that is what i would consider doing !

 

[HellasTechn]

It is normal for larger current demanding appliances to use unregulated power supplies (transformers) so i think that even the original transformer without load it would produce 20-21 Volt output !