Soundcraft CPS 650 PSU regulator replacement?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

syn

Well-known member
Joined
Oct 11, 2005
Messages
653
Hi

I'm having trouble with my CPS650. It is fully recaped and works flawlessly up to ~ +0.7A/~ -0.7A (12 ch + 3grps+ master section). If I load the PSU with more than ~0.7A per rail (add another channel ) , +/-17VDC drops to like +/-0.5VDC and mixer goes into crazy, noise, random LEDs going on, meters go full up etc. Taking just one modul out puts everything back to perfect order, voltage goes up to +/-17 etc. It can work for hours like this.
All channelss. groups. and master are tested OK, one by one, and all are recaped and in perfect condition. I can plug in any combination of Venue 2 moduls, but, I can not load the PSU with more than ~0.7A per a rail.

I disconnected the sensing circuit (BTW sensing circuit transistors tested OK), and tested the PSU with a dummy load. The regulators will give up holding steady voltage around 0.7A per a rail,at which point the voltage soares to almost 0V.
Think that regulators are broken. Why they work up to 0.7A or so flawlessly and then give up, I have no idea.

Question : LT1038 is obsolete(10A)! Can I replace it with LM338K - it is 5A and the same pinout. I don't think this console draws more than 2.5-3A per a rail (24ch-8grps-Master).

Also, why are those transistors in the sensing circuit so "esoteric", couldn't the same function be achieved with ordinary BJTs like BCs and MPSAs? I must be missing something.

Schematic attached.

Thank you

Uploaded CPS 650b U.G. with schematics to Soundcraft tech page (as well as Venue 2 service manual).
 

Attachments

  • CPS650_UG-13.pdf
    347.6 KB · Views: 37
Last edited:
I have seen both transistors and voltage regulators in the T0-3 metal can package behave like this when the screws that hold them into the board/heat-sink are even a little bit loose. Try tightening those screws as they are 1/3rd of the connections in the circuit.
 
I have seen both transistors and voltage regulators in the T0-3 metal can package behave like this when the screws that hold them into the board/heat-sink are even a little bit loose. Try tightening those screws as they are 1/3rd of the connections in the circuit.
Thanks, I wish it was that... The screws are tight, checked and rechecked... I will rebuild the PSU tomorrow checking each and every component again.
Any opinions on replacing LT1038 with LM338K?
 
I know that the later SC supplies used the LM338K. If I dig long and deep enough I might even find a copy of the service bulletin for that change. I'm pretty sure you're ok doing that, but hey, this is the internet...

Edit: I had a look through my Soundcraft docs last night and I couldn't find the TSB that I feel like I've seen at some point. :-(
 
Last edited:
I would bet replacing the regulators will take care of the problem, it seems like it's going into current limiting. I think the small signal transistors are there in case one rail goes out both regulators shut down. You could probably get away with any small signal transistor but these are 120v and low noise for whatever reason. You could replace the regulators and see if you still have the problem and then replace the small signal with generic xsistor to test (MPS -A06/56). Also, when replacing the regulators use new insulators (if they have them). I've had problems where a damaged insulator would fail when the current draw increased.

according to the schematic it should be able to deliver 4.2a - Soundcraft: CPS650 UG : Free Download, Borrow, and Streaming : Internet Archive
 
Not familiar with that PSU, but the CPS 275 is infamous for the bridge rectifiers going bad.

By any chance do you hear actual hum from the power supply with higher currents?
 
I know that the later SC supplies used the LM338K. If I dig long and deep enough I might even find a copy of the service bulletin for that change. I'm pretty sure you're ok doing that, but hey, this is the internet...

Great to hear that, I seem to have a memory of even seeing one long time ago, but i might be wrong.

I would bet replacing the regulators will take care of the problem, it seems like it's going into current limiting. I think the small signal transistors are there in case one rail goes out both regulators shut down. You could probably get away with any small signal transistor but these are 120v and low noise for whatever reason. You could replace the regulators and see if you still have the problem and then replace the small signal with generic xsistor to test (MPS -A06/56). Also, when replacing the regulators use new insulators (if they have them). I've had problems where a damaged insulator would fail when the current draw increased.

according to the schematic it should be able to deliver 4.2a - Soundcraft: CPS650 UG : Free Download, Borrow, and Streaming : Internet Archive

I was sure up to now that regulators either work or don't work at all, never seen this "living dead" type of behavior.

From CPS650B user guide :

"Under normal operating conditions TR5 and TR6 are both inoperative..." hence I don.t see any need for low noise devices, but I might be wrong on that.

I think they used those because the mixer's "peak" circuit is full of them and they had a lot of those in the stock room at the time.

Not familiar with that PSU, but the CPS 275 is infamous for the bridge rectifiers going bad.

By any chance do you hear actual hum from the power supply with higher currents?

New rectifier bridge installed 2 x 400V x 35A.
No hum, dead silent when it is working properly or burst of noise, distortion and all LEDs on when misbehaves.
 
Last edited:
I had a same problem with my CPS950 and there was thermal problem. I see that people mentioned that already but double check it because I repaired it after cleaning, adding new paste and changing the screws ;)
 
I would bet replacing the regulators will take care of the problem, it seems like it's going into current limiting. I think the small signal transistors are there in case one rail goes out both regulators shut down. You could probably get away with any small signal transistor but these are 120v and low noise for whatever reason. You could replace the regulators and see if you still have the problem and then replace the small signal with generic xsistor to test (MPS -A06/56). Also, when replacing the regulators use new insulators (if they have them). I've had problems where a damaged insulator would fail when the current draw increased.

according to the schematic it should be able to deliver 4.2a - Soundcraft: CPS650 UG : Free Download, Borrow, and Streaming : Internet Archive
Yes, I have heard of this issue with regulators going "soft" as well. Vague recollections of having one through the shop around 2013-14 and doing just that.
 
The 2SC22240BL and 2SA970GR are in no way "esoteric" unless you are considering the ease of availability. They have been the in-house standard low-power / small signal / switching / general purpose transistors at Soundcraft since before 1982 (which is when I joined the company).
As well as the 2SC2240BL and 2SA970GR, the MPSA06, MPSA56, BC212, BC184, BC214L, BC184L are all types that I have used for the mutual shutdown circuit.... it is not too fussy about the transistors. Ensure that the voltage parameters of the transistors used are suitable for the PSU into which they are being installed.

Many of the CPS range of power supplies suffered from poor contact performance of the sockets that the TO3 regulators plug into - heat is the enemy here, which is a bit ironic for a power supply! The idea of a quick and easy solderless replacement of failed regulators came back to haunt Soundcraft. In normal use over time when the PSU heats up, the TO3 socket contacts heat-up, they lose their spring pressure, go high-resistance, and that's when the trouble starts.

Trusting this helps.
 
I had a same problem with my CPS950 and there was thermal problem. I see that people mentioned that already but double check it because I repaired it after cleaning, adding new paste and changing the screws ;)
Thanks
Yes, I have heard of this issue with regulators going "soft" as well. Vague recollections of having one through the shop around 2013-14 and doing just that.
Good to know that , thanks.
The 2SC22240BL and 2SA970GR are in no way "esoteric" unless you are considering the ease of availability. They have been the in-house standard low-power / small signal / switching / general purpose transistors at Soundcraft since before 1982 (which is when I joined the company).
As well as the 2SC2240BL and 2SA970GR, the MPSA06, MPSA56, BC212, BC184, BC214L, BC184L are all types that I have used for the mutual shutdown circuit.... it is not too fussy about the transistors. Ensure that the voltage parameters of the transistors used are suitable for the PSU into which they are being installed.
Yes by "esoteric" I meant the ease of availability, probably not the best word chosen. Thank you so much for the conformation that other more readily available transistors could be used.
Many of the CPS range of power supplies suffered from poor contact performance of the sockets that the TO3 regulators plug into - heat is the enemy here, which is a bit ironic for a power supply! The idea of a quick and easy solderless replacement of failed regulators came back to haunt Soundcraft. In normal use over time when the PSU heats up, the TO3 socket contacts heat-up, they lose their spring pressure, go high-resistance, and that's when the trouble starts.
One LT1038 regulator is dead the other one survived ( I tested both on a another 2A tube heater supply),
I will loose TO3 sockets and solder the regulators directly to the PCB, ensuring good heat sinking and dissipation.
Would you consider a pair of LM338K as a good replacement for the LT1038s?

Thank you
 
For what it is worth the sockets used for LM338K (and TO3 power transistors) on many AMEK supplies constitute a 'standard' fault. The lugs on the 'screw inserts are tinned brass and only 'swaged' to the nut insert so by the time you have applied a big screwdriver to tighten the regulator/transistor down the chance of tearing the 'lug' from the nut insert is quite high. Then you have the issue of cracking the bakelite if you get too keen. After that you get the issue of the 'socket' contacts for the adjust and input (?) pin which are a bit 'cheap and cheerful' they are not really suitable for handling more than a couple of amps on a supply that does have a decent size heatsink so runs 'at a sensible temperature.
regarding the choice of the low power transistors, it is common for companies to buy 'shedloads' of only a few different types to get advantages of price and reduced effort in stocking. BC413 and BC415 were popular at Audix many years ago as well as BC301 and BC303 for 'higher power' requirements.
Matt
 
As Gareth commented, heat is the downfall of many power supplies and there is a good reason why the industrial supply makers (power One linears for example) only run 2N3055 transistors at between 2 and 3 Amps max per device. Similarly Coutant and Kingshill.
 
Would you consider a pair of LM338K as a good replacement for the LT1038s?
Yes, should work fine.

I was sure up to now that regulators either work or don't work at all, never seen this "living dead" type of behavior.
I have had regulators go flaky occasionally, I recall a similar problem like the one stated - might even have been a Soundcraft supply but it was long ago and my memory ain't what it used to be. :rolleyes:
 
Babbling about those little transistors again, the only time they do any 'work' is when one rail has started to fall out of specification by about a volt or so (can't be bothered to work out how far) so they represent the least of your problems at that point. Low noise specification usually refers to wideband 'hiss' for which there are so many capacitors on the audio rails that would be completely eliminated.
Thus their main attributes should be that they are not dead!
 

Latest posts

Back
Top