Author |
PSU issue 3 40V output |
br>joe p |
br>Hi! First off, praise to Tony for the beautiful pcbs and extremely thorough documentation!
Now, on to my troubles... I've just built an issue 3 PSU powered by a Yamaha PA-20, thats outputting almost 40 volts across the +/- 15v outputs, with adjustment trimmer fully cranked. Both show half this to 0v outputs. I've checked values and polarity, and it all seems right. Any hints? br> br> |
|
br>Synthbuilder |
br>Some things to check:
The diodes D1 and D3 could be around the wrong way.
The ADJ trimmer is 10K.
R6 should be 12K and R7 should be 10K.
Q1 should be TIP35 and Q2 should be TIP145 and not the other way around.
Tony br> br> |
|
br>joe p |
br> Synthbuilder wrote: | Some things to check:
The diodes D1 and D3 could be around the wrong way.
The ADJ trimmer is 10K.
R6 should be 12K and R7 should be 10K.
Q1 should be TIP35 and Q2 should be TIP145 and not the other way around.
Tony |
Hi Tony, thanks for the reply.
Those are correct, except I've got a TIP147 in Q2. Could that be an issue? br> br> |
|
br>Synthbuilder |
br>The TIP147 is fine - so the problem must lie elsewhere.
If both the + and - supplies are the same value, ie. both 20V, this would suggest a problem of the circuitry around the LM723. The negative supply tries to match that on the positive output so if they are the same magnitude then the problem is probably within the positive supply.
So if all the parts are correct there is possibility of a short somewhere. Check the voltage on pins 4 and 5 - both should be the same and be around 7V if the LM723 is working properly.
Tony br> br> |
|
br>joe p |
br> Synthbuilder wrote: | The TIP147 is fine - so the problem must lie elsewhere.
If both the + and - supplies are the same value, ie. both 20V, this would suggest a problem of the circuitry around the LM723. The negative supply tries to match that on the positive output so if they are the same magnitude then the problem is probably within the positive supply.
So if all the parts are correct there is possibility of a short somewhere. Check the voltage on pins 4 and 5 - both should be the same and be around 7V if the LM723 is working properly.
Tony |
4 and 5 on LM723 both show around 12V br> br> |
|
br>Synthbuilder |
br> joe p wrote: | 4 and 5 on LM723 both show around 12V |
I'm thinking that the LM723 is not working. Just check pin 6 and measure from the top and on the chip's leg directly. This should be roughly 7V with respect to 0V. If it isn't then the LM723's internal reference has been damaged and the chip will need replacing.
If it is 7V and pin 5 is 12V then check that any of the LM723's leads haven't been squashed underneath the chip's housing.
Tony br> br> |
|
br>joe p |
br> Synthbuilder wrote: | joe p wrote: | 4 and 5 on LM723 both show around 12V |
I'm thinking that the LM723 is not working. Just check pin 6 and measure from the top and on the chip's leg directly. This should be roughly 7V with respect to 0V. If it isn't then the LM723's internal reference has been damaged and the chip will need replacing.
If it is 7V and pin 5 is 12V then check that any of the LM723's leads haven't been squashed underneath the chip's housing.
Tony |
Pin 6 was also 12V. Tried another LM723 - same result... I got them from Mouser so they're not some shady ebay-chips. Is the LM723 very sensitive? br> br> |
|
br>Synthbuilder |
br> joe p wrote: | Pin 6 was also 12V. Tried another LM723 - same result... |
Not good. Either the reference voltage is bad or something is pulling it high which is also not good.
Quote: | I got them from Mouser so they're not some shady ebay-chips. Is the LM723 very sensitive? |
The LM723 is very rugged - I wouldn't expect one to die easily. And two would be really unlucky.
What's the voltage at pin 7 on the LM723? It should be 0V. If it's not then pin 7 may not be connected to the board. Either there's a board fault (which shouldn't happen since the boards are supposedly electrically checked for continuity and shorts), the soldering hasn't wetted the socket pins or the socket is bad.
Tony br> br> |
|
br>joe p |
br> Synthbuilder wrote: | joe p wrote: | Pin 6 was also 12V. Tried another LM723 - same result... |
Not good. Either the reference voltage is bad or something is pulling it high which is also not good.
Quote: | I got them from Mouser so they're not some shady ebay-chips. Is the LM723 very sensitive? |
The LM723 is very rugged - I wouldn't expect one to die easily. And two would be really unlucky.
What's the voltage at pin 7 on the LM723? It should be 0V. If it's not then pin 7 may not be connected to the board. Either there's a board fault (which shouldn't happen since the boards are supposedly electrically checked for continuity and shorts), the soldering hasn't wetted the socket pins or the socket is bad.
Tony |
Pin 7 is at 0V.
However, I measured the resistance of R7 and I only get 8,9k... could that be far enough off spec to have this effect? br> br> |
|
br>Synthbuilder |
br> joe p wrote: | I measured the resistance of R7 and I only get 8,9k... could that be far enough off spec to have this effect? |
Did you measure it while the resistor is still in the board? If so, the resistance value will always be lower than it should be as there are other paths within the whole circuit the current will flow. However, 8.9K is close enough to tell me that the 10K resistor is probably fine.
The real problem is that your reference voltage is 12V when it should be 7V. That would normally suggest that the LM723 is bad.
But, one more thought comes to mind; is your voltmeter correct? I once had a voltmeter that would always display much higher voltages when the battery was running down. Do you have a way of checking your meter or perhaps using another meter?
Tony br> br> |
|
br>joe p |
br> Synthbuilder wrote: | joe p wrote: | I measured the resistance of R7 and I only get 8,9k... could that be far enough off spec to have this effect? |
Did you measure it while the resistor is still in the board? If so, the resistance value will always be lower than it should be as there are other paths within the whole circuit the current will flow. However, 8.9K is close enough to tell me that the 10K resistor is probably fine.
The real problem is that your reference voltage is 12V when it should be 7V. That would normally suggest that the LM723 is bad.
But, one more thought comes to mind; is your voltmeter correct? I once had a voltmeter that would always display much higher voltages when the battery was running down. Do you have a way of checking your meter or perhaps using another meter?
Tony |
Actually, the battery is low... I'll swap it and measure again... and hopefully feel a bit dumb [/b] br> br> |
|
br>joe p |
br>So yeah... It was the battery in the multimeter, because of course it was. I have 30V across, +/-15V, LEDs are behaving like they should. I get 610 mA on the + side and 750 on the negative, so a bit high on - and a bit low on + compared to the documentation. Is this fine? br> br> |
|
br>Synthbuilder |
br>Yeah, been there, done that. I have a better meter these days - it comes up with a little battery symbol when the battery needs replacing.
joe p wrote: | I get 610 mA on the + side and 750 on the negative, so a bit high on - and a bit low on + compared to the documentation. Is this fine? |
Yes, that's fine. The actual value varies with temperature, resistor accuracy and meter characteristics.
Tony br> br> |
|
br>joe p |
br> Synthbuilder wrote: | Yeah, been there, done that. I have a better meter these days - it comes up with a little battery symbol when the battery needs replacing.
joe p wrote: | I get 610 mA on the + side and 750 on the negative, so a bit high on - and a bit low on + compared to the documentation. Is this fine? |
Yes, that's fine. The actual value varies with temperature, resistor accuracy and meter characteristics.
Tony |
Great! Many many thanks for the help! br> br> |
|