Faulty dual rail power supply

[drewc]

Hi I need some help to resurrect an old synthesizer that had stopped working possibly due to a failure in its power supply circuit. The parts are diffucilt to locate so I thought if I used equivilent parts I may be able to get it to work. It has 44v AC input and a 12v and 5v + and - DC output.
I will post a pic of the circuit. It has 44v AC input going through a bridge rectifier and then through 2 x 4700mf 25v caps. So the filtered supply to the front of the circuit is 25v + and - with a gound from the transformer.
This then is stepped down to 15v by two regulators. One I cannot find the same part number and is the one that is possibly faulty.It is marked as uPC 14315H. I have tried a 7815 in its place but did not have any output after firing it up.It did get hot so I switched off.
I have made another supply that used similar parts but I am not very good with electronic design so I made this section similar to the first part of the cicuit with an LM 317 for + and an LM337 for the - and a variable pot for both sides to get to 5volts.
When placed back into the syn I had the keyboard light up but no sound and when I checked the voltage I had no 5 v output and extremely hot IC's. I think my design was flawed because on further research the 5v section
has two transistors to boost the current level as opposed to my two 1 amp regulators.
Can anyone please help with an alternate design or subsitute part numbers for the original supply as I am an amateur in respect of electronics and would like to try to get this synth back up and running. Thanks Tom

 

[sheldonstv]

what you need to do first is isolate the rest of the circuit following both regulators -the 5v positive and 5v negative supplies will not b present then....check the supplies feeding yr regs and the supplies coming out of the regs....if you have fitted the correct regulators in circuit the correct way round you will have a positive and negative 15 volt supply relative to ground...the reg should be a 7815 for 15v positive supply and 7915 for the negative 15v supply and you need to check the pin outs are correct regarding board layout where they are fitted if you find yr regs are producing the correct plus and minus 15v supply reconnect the circuitry to give you your plus and minus 5v supplys make sure the supply is disconnected from the rest of the equipment when doing this-this will then give you an idea where the fault may be in the power supply if you have still got supplies missing you may find the fault is external to the powersupply in which case you need to do further checks to locate the fault

 

[bountyhunter]

Hi I need some help to resurrect an old synthesizer that had stopped working possibly due to a failure in its power supply circuit. The parts are diffucilt to locate so I thought if I used equivilent parts I may be able to get it to work. It has 44v AC input and a 12v and 5v + and - DC output.
I will post a pic of the circuit. It has 44v AC input going through a bridge rectifier and then through 2 x 4700mf 25v caps. So the filtered supply to the front of the circuit is 25v + and - with a gound from the transformer.
This then is stepped down to 15v by two regulators. One I cannot find the same part number and is the one that is possibly faulty.It is marked as uPC 14315H. I have tried a 7815 in its place but did not have any output after firing it up.It did get hot so I switched off.
I have made another supply that used similar parts but I am not very good with electronic design so I made this section similar to the first part of the cicuit with an LM 317 for + and an LM337 for the - and a variable pot for both sides to get to 5volts.
When placed back into the syn I had the keyboard light up but no sound and when I checked the voltage I had no 5 v output and extremely hot IC's. I think my design was flawed because on further research the 5v section
has two transistors to boost the current level as opposed to my two 1 amp regulators.
Can anyone please help with an alternate design or subsitute part numbers for the original supply as I am an amateur in respect of electronics and would like to try to get this synth back up and running. Thanks Tom

The +/- 5V supplies are discrete linear regulator designs using a single pass transistor and an op amp to run them (the transistors are not current boosters they are voltage regulators for +/-5V outputs). They down regulate from the +/- 15V lines. Discrete regulators have no current limiting. If something failed, it could blow out the parts like firecrackers. The IC regulators like a 7815 have internal protection so they should survive. If they are getting hot, the +/- 5V circuitry or downstream loads probably have a problem drawing excessive current.



EDIT TO ADD: I notice the design has some very precise 0.1% resistors used to set the voltage on the +/-5V outputs so IC regulators like a 7805 can't be used for the +/-5V outputs. It looks like they are making those output voltages very precise for some reason(?) Note that they have a trimpot on the LM336 reference used to precisely set the 5.00V.

The first thing to do is cut the load lines on these supplies and see if they still get hot without load. If the circuit failures are downstream of the supplies, the supplies will get hot, read wrong voltages and possibly fail if that circuitry is not fixed first.

 

[Mr RB]

The +/- 5v supplies are supposed to be matched very well and precise at 5v. The synths analog VCO/VCA/VCF etc are all voltage controlled and need an exact 5v supply.

This is probably a valuable collectable old synth being butchered. As for "all the IC's getting very hot" that's just scary.

Disconnect the PSU from all the rest of the synth, make sure the PSU is working ok when connected to some little light bulbs or some 5W resistors etc first before frying the entire synth. You can probably use 7805 and 7905 for the +/- rails, IF you pick 2 that are exactly 5.00v and both the same, AND both stay the same when under load. Some brands of 7805 have much better regulation.

(edit) YAY! This is my 555th post on this forum! "Posts: 555" Is that good "timing" or what?? hehehehe

 

[sheldonstv]

u old timer poster u..............

 

[marcbarker]

Some tips I have... ICs in older equipment often run hotter than you'd expect.

This is a controversial one: If there's a short circuit on a 5 V power rail and you can't find it, connect a bench power supply of 5V to the shorted rail and turn up the current and leave it for a while. The chances are if it's an IC, it's dead anyway, so look for hot spots appearing, or even better smoke leading you right to the offending part. If still can't find it, despite feeding 5 A into the PCB, get a can of Freezer spray and cover the PCB in frost, then you'll see some tracks/ICs/capacitors thaw out first, you can literally see the current path as though you're using a thermal imaging camera!

 

[bountyhunter]

You can probably use 7805 and 7905 for the +/- rails, IF you pick 2 that are exactly 5.00v and both the same, AND both stay the same when under load. Some brands of 7805 have much better regulation.

Absolutely not. Even if you found one that was 5.000V at room temp, they have both a tempco and load regulation effect. The envelope for those would be 3% or more on that type of regulator. Those are what we built in the last place I worked...... I won't mention their name out of fear that if I write National Semiconductor, it might influence somebody to buy product from one of the crappiest semiconductor houses on earth.

 

[bountyhunter]

Some tips I have... ICs in older equipment often run hotter than you'd expect.

This is a controversial one: If there's a short circuit on a 5 V power rail and you can't find it, connect a bench power supply of 5V to the shorted rail and turn up the current and leave it for a while. The chances are if it's an IC, it's dead anyway, so look for hot spots appearing, or even better smoke leading you right to the offending part. If still can't find it, despite feeding 5 A into the PCB, get a can of Freezer spray and cover the PCB in frost, then you'll see some tracks/ICs/capacitors thaw out first, you can literally see the current path as though you're using a thermal imaging camera!

There used to be a film type material we bought in sheets that was temp sensitive. If you laid it over a PCB and powered it, the warm one showed up quickly.

 

[bountyhunter]

The +/- 5v supplies are supposed to be matched very well and precise at 5v. The synths analog VCO/VCA/VCF etc are all voltage controlled and need an exact 5v supply.

Sure looks like it considering how much money they threw at the 5.000V rail. The 0.1% resistors shown are temperature compensated (since they have the little braces around their schematic symbols). Those suckers are not cheap. And the 2.49V reference is trimpot adjusted for zero tempco and to dial in the 5.000V rail voltage. hand adjusted = cost $$$.

 

[drewc]

Thanks for the response.

This is probably a valuable collectable old synth being butchered.

This is exacty what I don't want to do. The PSU I am using as a subsitute does not use any of the original parts as I have kept that pretty much intact.
When first tested with the psu still connected I had no 15v + or + and - 5v output but I did have - 15v .I checked the voltage at the output of the 14315H but had not output at all.
This led me to think this part may be faulty.I cannot find this part number anywhere (uPC 14315H). Can I use a 7815 in its place ?
Now that I have the PSU out of the synth I will put power back through it and check the voltages again with no load. I will put the 14315 back in and if still no voltage I will try the 7815.
Is there other parts I can use as substitutes for the originals to maintain the design as I cannot find them here in Australia (original parts).
I will replace the caps with new equivilents and will put the original 15v reg back in. ( uPC 14315H ) and post when done.
Thanks for the advice

 

[bountyhunter]

Thanks for the response.



This is exacty what I don't want to do. The PSU I am using as a subsitute does not use any of the original parts as I have kept that pretty much intact.
When first tested with the psu still connected I had no 15v + or + and - 5v output but I did have - 15v .I checked the voltage at the output of the 14315H but had not output at all.
This led me to think this part may be faulty.I cannot find this part number anywhere (uPC 14315H). Can I use a 7815 in its place ?
Now that I have the PSU out of the synth I will put power back through it and check the voltages again with no load. I will put the 14315 back in and if still no voltage I will try the 7815.
Is there other parts I can use as substitutes for the originals to maintain the design as I cannot find them here in Australia (original parts).
I will replace the caps with new equivilents and will put the original 15v reg back in. ( uPC 14315H ) and post when done.
Thanks for the advice

I don't see how changing the +15 regulator to a 7815 could be a problem unless the load current exceeds it's limits which is about 1 - 1.5A.

I have never heard of "uPC 14315H" either. It could be a custom number.

If it was mine, I would cut the load lines to the external circuitry and use some lab supplies to power the unit to see if it works and measure the load current. No sense throwing money at the power supply if the downstream circuitry is fried.

 

[drewc]

That would be an ideal situation to try but unfortunatley I don't have access to such equipment. The synth stopped working 2 years ago and my son asked could I get it going. I have limited electronic experience and knowledge. My temp psu did power all the LEDS when hooked up but no sound was made so I disconnected it to avoid any further problems.
I will later today connect the original PSU back up with no load and check the voltages so see if it does indeed put out the 15 and 5v respectively. If it does then yes I believe the problem is further inside the synth and I am not sure what to do as I don't have the required test equipment or knowledge to go any further. The synth weighs too much to send away to get repaired, so rather than butcher it I may just put it back into storage and let it gather dust. Thanks again for your help.
Tom

 

[bountyhunter]

That would be an ideal situation to try but unfortunatley I don't have access to such equipment. The synth stopped working 2 years ago and my son asked could I get it going. I have limited electronic experience and knowledge. My temp psu did power all the LEDS when hooked up but no sound was made so I disconnected it to avoid any further problems.
I will later today connect the original PSU back up with no load and check the voltages so see if it does indeed put out the 15 and 5v respectively. If it does then yes I believe the problem is further inside the synth and I am not sure what to do as I don't have the required test equipment or knowledge to go any further. The synth weighs too much to send away to get repaired, so rather than butcher it I may just put it back into storage and let it gather dust. Thanks again for your help.
Tom

BTW: both the 5V and -5V rails have fuses which could be blown. Don't know if you have checked those yet.

 

[drewc]

Yes that was the first thing I looked at. Very much overlooked by most people and should always be the 2nd thing to check after making sure the power switches have been turned on.
Thanks Tom

 

[drewc]

Well the good news is after replacing the caps and checking all the connections and disconnecting all the feeds going away I was able to get + and - 15v and 5v and also able to adjust from 4.95 to 5.05volts. This was great news until I wired it all back up.
No lights, no sound , no voltage output and the IC's got hot. So I switched it off and placed all the screws back in and have placed in mothballs till the kids inherit it.
Thanks to all those that offered suggestions and advice. It was very much appreciated and at least I know that the PSU is not the cause.
Tom

 

[sheldonstv]

well thats a start what audio ics are fitted???give the type numbers and someone might be able to source them for you.......

 

[drewc]

Just another question for those knowledgable people, How do I tell which circuit or part is causing the synth to not function? (Korg Trident made in 1983)
When I connected the PSU I made rather than the original the 5v rails were connected just after the rectifer so that if they drew more current they would not be restricted by the 7815 and 7815 1 amp output. This also made the supplies seperate from each other. When I connected this up the fuses on the 5v cicuit did not blow and the LEDs on the synth lit up. The 7815 and 7915 ICs did get very warm. So I then swicthed it off.
When I checked the original supply was functioning and then connected it back up I had no lights and also the 5v fuses were intact.
Could I then assume the fault would be in the 15v circuit as the original design uses the 15v to power the 4558 ICs which in conjunction with the pot outputs a very stable + and - 5v ?
Is my reasoning correct? Can I disconnect all the 15v wires (14 in total) check that I have 5v + and - and if so am I best to connect one set at a time until the 5v stops outputing power and then check the components this powers? By doing this am I damaging something by disconnecting all the 15v power leads or is it ok as no power is no damage?
Any suggestions would be most appreciated Tom

 

[Mr RB]

Hmmm, a Korg Trident has a LOT of guts, it's regulators will get pretty toasty. I've never worked on one but the photo of a TridentII shows a big PSU heatsink on the back, that alone is rare for a synth. Does your schematic have any markings for PSU testpoints and voltages indicated etc?

You can solder some 0.1 ohm resistors across the fuesholders, and remove the fuses. Then power it up and measure the voltage across the 0.1 ohm resistors, it will tell you how much current each part of the downstream circuit draws, so you can see if there is an overcurrent fault on any supply line. Old school techs have blown fuses with resistors soldered onto them , that they use for tests like that.

Korgs normally use sockets for all the ICs. I would draw a layout diagram, then unsocket all the IC's and place them carefully on the diagram in their respective place. Check IC's for miscoloured legs that might indicate it has suffered an over current fault.

Then power it up again and check for overcurrent on the 0.1 ohm resistors to see if you have removed the fault. There may be other regulators like zeners etc in the circuit, zeners tend to fail as a short circuit and drag PSU rails down.

With the 4558's I would be temped just to buy a bag and replace them all they would be under $1 each, with the more specialised ICs like the VCFs you can plug them back in and see if the overcurrent fault returns, which will indicate if an IC is bad. A shorted IC that is causing overheating of your voltage regulators will probably show signs of heat damage on it or around its socket etc.

I don't see any problem with disconnecting the 15v supply to different parts of the circuit to see if one part is faulty, juts be carefula nd write good notes of what you do and where everything goes. That is a complex synth packed full of PCBs full of ICs... It's likely to take a while.

 

[drewc]

Korgs normally use sockets for all the ICs.

I wish they did but they are all hardwired. Thanks for the suggestions Tom