advertisement


weird SKA behaviour

A

AussiePete

pfm Member
Hi,
I have recently built another couple of SKA GB150 amp modules and have a weird issue with one.
I need to say that this board was working perfectly until the power ground wire released itself from the power supply (GB300S), which caused it to blow a test resistor wired into the F2 fuse holder.
I replaced that and everything was good; it biased up to the 1.2V I usually run it at, and let this burn in for a couple of days and connected a speaker, only to have the test resistors blow.
I changed these again and they blew on switch on.
It will boot up without issues without a speaker attached but, with speaker attached it will only boot up if I turn the bias right off and then turn it back up to the normal 1.2V without issues or any burning.
It plays sweet music (although just fired up with my Iphone and cheap TV speakers), runs cool and apparently perfect.
I noticed today that the resistor across F1 (+ rail) was at 1.14V and the one across F2 (- rail) had 1.31V across it.
The other side module was even on both sides, and they both run cool to the touch and even temperature.
I believe the schematic is available, if not, I'll open a flicker account seeing as I've locked myself out of photobucket:mad:.
Can anyone shed any light on what might be causing this?
Thanks,
Pete
 
AussiePete said:
Hi,
I have recently built another couple of SKA GB150 amp modules and have a weird issue with one.
I need to say that this board was working perfectly until the power ground wire released itself from the power supply (GB300S), which caused it to blow a test resistor wired into the F2 fuse holder.
I replaced that and everything was good; it biased up to the 1.2V I usually run it at, and let this burn in for a couple of days and connected a speaker, only to have the test resistors blow.
I changed these again and they blew on switch on.
It will boot up without issues without a speaker attached but, with speaker attached it will only boot up if I turn the bias right off and then turn it back up to the normal 1.2V without issues or any burning.
It plays sweet music (although just fired up with my Iphone and cheap TV speakers), runs cool and apparently perfect.
I noticed today that the resistor across F1 (+ rail) was at 1.14V and the one across F2 (- rail) had 1.31V across it.
The other side module was even on both sides, and they both run cool to the touch and even temperature.
I believe the schematic is available, if not, I'll open a flicker account seeing as I've locked myself out of photobucket:mad:.
Can anyone shed any light on what might be causing this?
Thanks,
Pete
Click to expand...


Why not ask the designer himself? He was extremely helpful when I was building my GB150D a few years back :)

http://www.ska-audio.com/Forum/YaBB.pl?custompage=products

John.
 
PigletsDad said:
Put a scope on it, and see if it is oscillating.
Click to expand...

Thanks PD, the only problem is that I don't have one and the only person I know with one is overseas.
Wouldn't the module run hotter than the other one if it was oscillating?
BTW, I forgot to mention that DC offset is also on par with the other channel.
I did note last night that the transformer on that channel is ever so slightly warm whereas the other one is totally cool.
Pete
 
If oscillating, it would probably run a bit warmer.

OK, without scope, it would start using a DVM to compare voltages between good and bad channels, when unloaded.

The other thing you can try, is putting a signal in, and look at AC and DC voltage at the output, comparing good and bad channels. You expect small DC -compared to AC; if something is badly wrong, the waveform may distort assymetrically, which can give a clue to where the trouble lies.
 
Can you measure the voltages across these resistors: R15, R16, R17 and R18?

I would expect VR15 close to VR17 and VR16 close to VR18.

Can you also measure voltages on D3 and D4?
 
PigletsDad said:
Can you measure the voltages across these resistors: R15, R16, R17 and R18?

I would expect VR15 1.45v close to VR17 1.68v and VR16 1.46v close to VR18 1.68v.

Can you also measure voltages on D3 and D4?Both 51.3v closest to Mosfet and 47.8v on outer end
Click to expand...

Thanks PD, voltages in red.
Pete
 
Both 51.3v closest to Mosfet and 47.8v on outer end ???? I assume this is placement on the board!

Surely the other way round in the schematic? The outer ends in the schematic are at the voltage rails.

None of those voltages sounds silly, so lets check the output stage bias.

Can you tell me what values R28 and R29 are, and can you measure the voltages across them VR28 and VR29?
 
PigletsDad said:
Both 51.3v closest to Mosfet and 47.8v on outer end ???? I assume this is placement on the board! It is placement.

Surely the other way round in the schematic? The outer ends in the schematic are at the voltage rails. You are probably right

None of those voltages sounds silly, so lets check the output stage bias.

Can you tell me what values R28 and R29 are these are OR15/2w mox, and can you measure the voltages across them VR28 and VR29? The voltages across these resistors is 17mV but referenced to 0v they are rail voltage
Click to expand...


Hi PD, here are the measurements you asked for.
Regards,
Pete
 
OK, I have now changed Q9, Q10 and when nothing changed I then changed Q1, Q2 for still no change.:confused:
Another interesting thing is that not one component on the -ve rail, which is .5V higher, is any warmer than any other.
Unfortunately I don't have BC556C's here so will need to get some from Greg seeing as they are as rare as rockinghorse crap unless you buy in bulk from Mouser.
When I get them I will change out Q3, Q4.
Regards,
Pete
 
I'm still fighting this thing.
All of the troubleshooting tests have been done a couple of times and everything checks out.
I have replaced Q1 to Q10, removed and checked C4,5,6,7,10,11 (C9 linked), soldered a .1uF cap across R14, split power rail and added OR15 to each Mosfet instead of one resistor for each pair, swapped power supplies around, changed VR1 trimpot, changed D5, D6 to perfectly matched pair and it is still giving 16V across F1 resistor on turn on and burning up the resistors.
On the couple of occasions that I've been able to get through bootup with normal bias, it takes about 20 seconds to drop down to preset bias voltage.
Is it worth fitting C9 again?
Any ideas would be appreciated?
Pete
 
You must log in or register to reply here.


advertisement


Back
Top