Phase Linear 400 - DC offset problem
- Thread starter audiogymp
- Start date
Apologies, I missed that.
Q10 is suspect faulty.
Q10collect will be about rails voltage +70V
Q10base will be about 1.8-1.9V (from D9/Q8collector measurement, 1.28V)
Q10emitter is +912mV with respect to speaker line but will be +65-70V with respect to ground, this voltage with respect to ground is clearly
wrong. You could try diode testing Q10, in circuit, it may indicate failure(c-e short) however it may need a few volts to fail. Alternatively you
can measure the Q10c-e voltage drop. It might be 2-3 volts, not the 70-80V from fig III-5.
This still might be some obscure grounding issue but very unlikely. Replace Q10 and go from there.
62vauxhall
Addicted Member
I went through with a 700B what you're going through with that 400. I literally spent months with 125 forum posts trying to track down the cause of a bias problem on one channel. I finally bit the bullet and it got a White Oak treatment.
You are right about the original boards being hard to work with. They do not stand up well to manipulation and re-working.
A few months ago I picked up an early 400 as an opportunity buy. It actually worked OK but I did the White Oak thing to it anyway. The performance specs went up as did sound quality and I think it will be of more interest upon re-sale.
You are right about the original boards being hard to work with. They do not stand up well to manipulation and re-working.
A few months ago I picked up an early 400 as an opportunity buy. It actually worked OK but I did the White Oak thing to it anyway. The performance specs went up as did sound quality and I think it will be of more interest upon re-sale.
Hi mbz
There have been some crossing posts. I have to remind myself to refresh the thread because time difference usually means a bigger cycle between posts.
Q10 appears to be a problem. On diode tester it measured 483mV c-e, although resistor measurement didn't show conductivity. When I measured c-e voltage under load, I got only -4.0V (black-c; red-e), a long way from the almost full rail voltage we should see.
Any thoughts on a readily available replacement for the RCA 40327? If I use a substitute, will I also have to replace Q7 with a complimentary substitute? I expect that I will probably have to order a replacement from OS but is there anything I could easily source from here (e.g. Jaycar)?
Vauxhall - I suspect a new board may be a wise choice but I'm not quite ready to throw in the towel yet. I've already shelled out for a full set of output Xsistors and I'd like to get it going in original form before a going for a new fully complimentary circuit. Of course, if I hit expensive replacement components, it may change my thinking.
Thanks, Jon
There have been some crossing posts. I have to remind myself to refresh the thread because time difference usually means a bigger cycle between posts.
Q10 appears to be a problem. On diode tester it measured 483mV c-e, although resistor measurement didn't show conductivity. When I measured c-e voltage under load, I got only -4.0V (black-c; red-e), a long way from the almost full rail voltage we should see.
Any thoughts on a readily available replacement for the RCA 40327? If I use a substitute, will I also have to replace Q7 with a complimentary substitute? I expect that I will probably have to order a replacement from OS but is there anything I could easily source from here (e.g. Jaycar)?
Vauxhall - I suspect a new board may be a wise choice but I'm not quite ready to throw in the towel yet. I've already shelled out for a full set of output Xsistors and I'd like to get it going in original form before a going for a new fully complimentary circuit. Of course, if I hit expensive replacement components, it may change my thinking.
Thanks, Jon
Little busy at the moment so only have a few minutes. More detail later.
pos/neg speaker posts on problematic side. Compare with working side. Just to make sure there is no existing near short.
http://au.rs-online.com
As you know, should be non conductive in both directions.
MM generally don't push enough current to properly test transistors in resistance mode.
Clearly not good, needs replacement. Why has it failed. maybe a speaker short? With unit powered off/unplugged, measure resistance between the
pos/neg speaker posts on problematic side. Compare with working side. Just to make sure there is no existing near short.
Transistor matching is not my strong suit. Wait for others to chime in.
Not sure this is really required, depends how close a match for Q10.
I would avoid the car place for everything except hardware. Maybe RS however they are a little expensive and very limited range.
http://au.rs-online.com
Resistances across speaker terminals (power off)
Problem side:130K (black to black; red to red), 30K (polarity reversed)
Good side: 155K (like to like), 50K (reversed)
There is a fair bit of capacitance going on, so resistances tended to stabilise. The resistances generally started in the lower KOhm range. Please let me know if initial resistance is what you want to know.
Problem side:130K (black to black; red to red), 30K (polarity reversed)
Good side: 155K (like to like), 50K (reversed)
There is a fair bit of capacitance going on, so resistances tended to stabilise. The resistances generally started in the lower KOhm range. Please let me know if initial resistance is what you want to know.
Thanks for all the help, mbz. It's been great.
I thought you might know or be able to readily work out a Xsister substitute. I'll see if anyone chimes in and do some searches myself - I don't need to take up more of your time if you'd be doing searches I can do.
There will be a bit of a hiatus while I get parts. There's another amp that needs attention and a recap I've been considering, so I'll put together a shopping list for a $60+ Mouser order.
Can you think of any further tests before I order or any parts I should get with the order just in case?
I'll drop in a post once I have the parts - 10-14 days. I don't expect you'll have the time/inclination to continue after that but it would be great if you did.
Cheers, Jon
I thought you might know or be able to readily work out a Xsister substitute. I'll see if anyone chimes in and do some searches myself - I don't need to take up more of your time if you'd be doing searches I can do.
There will be a bit of a hiatus while I get parts. There's another amp that needs attention and a recap I've been considering, so I'll put together a shopping list for a $60+ Mouser order.
Can you think of any further tests before I order or any parts I should get with the order just in case?
I'll drop in a post once I have the parts - 10-14 days. I don't expect you'll have the time/inclination to continue after that but it would be great if you did.
Cheers, Jon
A basic speaker protection circuit covering power ON/OFF muting and also dc presence is mandator. 60Vdc on the speaker line would generally mean toast.
I've recently picked up a chinese one off the bay, AUD16? for assessment. Was going to build my own then saw this. I believe in full recapping if you want the
most out of your vintage gear, so support a recap of the PL400. The White Oak modes look interesting but a little costly, will need to sleep on that. Certainly
order more than one transistor. I buy quite a lot of stuff from mouse, shipped by UPS. Order monday pm, delivered thursday afternoon. Makes a mess of
Aust Post.
I think the 2N3439 is the way to go but would be nice to get confirmation.
First time I've looked at the PL400 schematic. I'm keen to open up my two units...
I've recently picked up a chinese one off the bay, AUD16? for assessment. Was going to build my own then saw this. I believe in full recapping if you want the
most out of your vintage gear, so support a recap of the PL400. The White Oak modes look interesting but a little costly, will need to sleep on that. Certainly
order more than one transistor. I buy quite a lot of stuff from mouse, shipped by UPS. Order monday pm, delivered thursday afternoon. Makes a mess of
Aust Post.
I think the 2N3439 is the way to go but would be nice to get confirmation.
First time I've looked at the PL400 schematic. I'm keen to open up my two units...
Yes, even if you stick with original board, no WO upgrades a recap will make a big difference in how she sounds, at least it did for me on an original series 1 700 that I recapped. I now have a 400 and a 700 WO'd, and one of each on the shelf that still need it. And with the output relays in place don't have to mess with inline fuses anymore.
What I liked about it most was that I could stop working on the damn amp and simply listen to it!
What I liked about it most was that I could stop working on the damn amp and simply listen to it!
I replaced the main filter caps and the bridge rectifier in my Pro 700 (its a 700 series 2 with a black faceplate) and honestly couldn't believe the difference it made. I had previously rebuilt the driver board. No real help but it didn't hurt. I also made sure to replace any off value resistors and set the bias while I was in there, but it was the main filters that truly made the difference. I upped the capacity of the rectifier because the new caps are larger, and the part was like 4 bucks.
The fuses are actually power supply rail supply fuses to each channel, not speaker fuses. In theory if the protect relay does its job, the rail fuses probably aren't going to blow, but don't bypass them.
The fuses are actually power supply rail supply fuses to each channel, not speaker fuses. In theory if the protect relay does its job, the rail fuses probably aren't going to blow, but don't bypass them.
Didn't mean those fuses. I fuse in line between speaker and amp as per old PL bulletins.
$100 and change for a WO new control board is pretty cheap, just figure out how much of your time and parts changing is costing you. You will have fun setting the bias with the old finger wheels. The old boards have major issue with traces lifting, a very good adjustable heat solder station is required for any type of work on the old PL boards and silver solder is just about mandatory
The WO control board is a tempting prospect but once you factor in the International price, exchange rate and postage, it will be over AU$200. Postage from US to Aust is quite variable but often surprisingly expensive relative to EU. And if I were doing that, I'd go for speaker protection and improved output configuration at the same time, so no longer a low cost project. I wanted to see if I could get the PL400 going without spending to much and see if I bond with it before pimping it. But it's something I'll be looking at down the track.
While we're discussing it, on other posts it looked like you need more than basic equipment to do the WO mod (signal generator etc). Is this mandatory or are there simpler options for set-up?
Gadget - can you recall the approx voltage/capacitance you used for the filters and the rating of the replacement rectifier. Even if I ultimately replace the control board, this will be worthwhile. But I'll check resistors and replace electrolytics on the control board.
Thanks all.
While we're discussing it, on other posts it looked like you need more than basic equipment to do the WO mod (signal generator etc). Is this mandatory or are there simpler options for set-up?
Gadget - can you recall the approx voltage/capacitance you used for the filters and the rating of the replacement rectifier. Even if I ultimately replace the control board, this will be worthwhile. But I'll check resistors and replace electrolytics on the control board.
Thanks all.
The 400 stuff is smaller/different. My 700 uses a 16000uf 160v cap to replace a 9800uf 100v cap. Rectifier I don't exactly remember, it was something from Mouser. I think the part number came from a thread over on the Pheonix board. The 400 has a more readily available cap setup available here:
http://www.whiteoakaudio.com/PL400_Cap_Assy.aspx
No idea if there are other options besides that. No affiliation, etc, etc.
My 700's caps were a group buy special run sort of thing that Northwinds organized a couple years ago. I bought them, moved them around for a long while, then got sick of moving the box around and just stuck the parts in the amp. I was not expecting much, but I have to say I was pleasantly surprised at the results.
http://www.whiteoakaudio.com/PL400_Cap_Assy.aspx
No idea if there are other options besides that. No affiliation, etc, etc.
My 700's caps were a group buy special run sort of thing that Northwinds organized a couple years ago. I bought them, moved them around for a long while, then got sick of moving the box around and just stuck the parts in the amp. I was not expecting much, but I have to say I was pleasantly surprised at the results.
Parts are on their way.
mbz - I've also bought a Chinese speaker protection circuit and I'm reasonably impressed with it. It's off topic here, so I've started a new thread:
http://audiokarma.org/forums/index.php?threads/chinese-speaker-protection-circuit-kits.781588/
mbz - I've also bought a Chinese speaker protection circuit and I'm reasonably impressed with it. It's off topic here, so I've started a new thread:
http://audiokarma.org/forums/index.php?threads/chinese-speaker-protection-circuit-kits.781588/
The parts arrived and I installed the Q10 sub - 2N3439. Q10 was not the problem and it loos like the problem is upstream.
I checked Q10 out of circuit before replacing it. It was not conductive c-e and my component tester correctly assigned terminals, so it looks like it is probably OK.
With the new component in place the voltage c-e on Q10 was 2.3V, instead of the previous 4.0V - so no real change. With the new transistor I was able to measure the voltage directly from the b-terminal and it measured nearly +62V (relative to earth/chassis). I then re-checked my measurement of the voltage at the Q8/D9 junction (relative to earth/chassis) and also got a voltage just under +62V. This is very different to my previous measurement of 1.28V. I'm not sure if this has changed or I measured incorrectly last time. It is a difficult/tight place to measure from. But if Q10 had not failed, the high b-voltage may be consistent with the low c-e voltage at Q10.
mbz - if you're still following this thread, any thought on where the high +ve voltage around D9 might be coming from (as previously reported, C11 has been replaced).
Cheers, Jon
I checked Q10 out of circuit before replacing it. It was not conductive c-e and my component tester correctly assigned terminals, so it looks like it is probably OK.
With the new component in place the voltage c-e on Q10 was 2.3V, instead of the previous 4.0V - so no real change. With the new transistor I was able to measure the voltage directly from the b-terminal and it measured nearly +62V (relative to earth/chassis). I then re-checked my measurement of the voltage at the Q8/D9 junction (relative to earth/chassis) and also got a voltage just under +62V. This is very different to my previous measurement of 1.28V. I'm not sure if this has changed or I measured incorrectly last time. It is a difficult/tight place to measure from. But if Q10 had not failed, the high b-voltage may be consistent with the low c-e voltage at Q10.
mbz - if you're still following this thread, any thought on where the high +ve voltage around D9 might be coming from (as previously reported, C11 has been replaced).
Cheers, Jon
Only got a few minutes. A high Q8/D9 junction is not good. Suggests no current flow through branch R17,8, D6,7,8.
No current flow means no voltage drop so Q10 base is at near rails voltage. Need to check D6,7,8, you can try in circuit
however you probably need to measure the voltages at R18/D6, then D6/D7, then D7/D8 then D8/R19. Suspect one of these
diodes has failed open circuit.
No current flow means no voltage drop so Q10 base is at near rails voltage. Need to check D6,7,8, you can try in circuit
however you probably need to measure the voltages at R18/D6, then D6/D7, then D7/D8 then D8/R19. Suspect one of these
diodes has failed open circuit.
A failed Q6 also fits the scenario.
Almost no current is flowing down R17 & R18 so we don't have the expected volt drop at Q10b
Possibility that D6,7 or 8 have failed open, so Q6 is not turning on/conducting.
Other possibility that Q6 has failed so unable to conduct.
Try incircuit diode test of D6,7,8.
C8 and C9 will cause incorrect readings for in circuit test of Q6 (however try). Measure Q6 c/b/e voltages with respect to chassis.
R20 bias trimmer OK?
Almost no current is flowing down R17 & R18 so we don't have the expected volt drop at Q10b
Possibility that D6,7 or 8 have failed open, so Q6 is not turning on/conducting.
Other possibility that Q6 has failed so unable to conduct.
Try incircuit diode test of D6,7,8.
C8 and C9 will cause incorrect readings for in circuit test of Q6 (however try). Measure Q6 c/b/e voltages with respect to chassis.
R20 bias trimmer OK?
The measurements of voltages at each side of the components (relative to ground) on the branch beginning with R17 follow. They don't appear to point to a cause. The current constriction might be on the negative side of D8.
R17: 64.0->62.1V
R18: 62.1->61.9V
D6: 61.7->61.4V
D7: 61.7->61.2V
D8: 61.1->60.8
The work bench is outside, so I'll try to check the voltages across R19 and R20 tomorrow. The copy of the SM I'm working from (from HiFi Engine) appears to be a scan of a hard copy. The components on the branch beginning with R17 in Diagram III-5 appear to have been crossed through, as though they were being checked/eliminated by someone - perhaps working on a similar problem. Whoever it was checked R19 and 20, which makes sense. I'll do the same.
PS - just saw your edit - will also do the other tests you suggest.
R17: 64.0->62.1V
R18: 62.1->61.9V
D6: 61.7->61.4V
D7: 61.7->61.2V
D8: 61.1->60.8
The work bench is outside, so I'll try to check the voltages across R19 and R20 tomorrow. The copy of the SM I'm working from (from HiFi Engine) appears to be a scan of a hard copy. The components on the branch beginning with R17 in Diagram III-5 appear to have been crossed through, as though they were being checked/eliminated by someone - perhaps working on a similar problem. Whoever it was checked R19 and 20, which makes sense. I'll do the same.
PS - just saw your edit - will also do the other tests you suggest.