Line 6 HD-147

[ricach]

Please do some research before you dive into the SMPS.
The Primary is HOT to line voltage.
The 'ground' reference is the -170 volts point when attempting any voltage measurements.(on the primary side)
The secondary is totally disconnected from the primary (sort of) so voltage readings there can be referenced to chassis ground.

What role does Q5 play? The line voltage hits that before the Primary side of the xformer. Feel free to explain switching technology to me.

BTW Juan: T3 primary is shown in the Primary Supply Diagram, at the very right of the diagram. Q5 &Q6 drive it.

[J M Fahey]

Dear Ricach:
no, that is secondary T3c and it drives Q5/6 , the same way as secondary T3b drives Q3/4.
I'd love to find T3a which I guess is the primary.
I find real stupid spreading the PSU over several pages, real STUPID.
It should be all together on a single page.
If they need to make that page tabloid newspaper size, so be it.
Worst is the PSU designer must be thinking "hey"" , it's EASY !!!! "
Not only that, the amp itself is also a nightmare to me.
Funny thing is, *my* 300W amp also uses 4 x IRFP250 , has 1/4 the parts count and works like a charm ... and can be repaired
As of the switcher: thge 120VAC gets rectified into 170V DC, so C36/37 get 85V each.
Q3/4 and Q5/6 form a "half bridge inverter" providing a 170Vpp squarewave to T5 primary.
It has a lot of secondaries which get rectified by high speed dual diodes, seen on the secondary side.
I guess (too tired/bored/angry to check) that there is another smaller PSU which feeds the "logic/control" part of this PSU and also powers the preamp.
I also guess that *maybe* the main PSU is fine, but the power amp is dead and some sensing circuit is shitting its power off.
As you see, way too many unknown variables.
I'd unhook the power amp and check whether those "Low" and "High" rails reappear.
You *might* get lucky.

[ricach]

Juan, if you see one of my early threads I did check the voltages with it disconnected from the power amp. It was about half. half of about nothing is pretty much nothing. lol.

[ricach]

I think my logical next move is to take voltage readings from the primary side first.
After the first full-bridge rectifier, should I see AC anywhere on the Primary side?
What do I use for a ground reference? -170V rail?
Should I see 340V from hi to lo rail?
Where should my scope pick up a squarewave?
What other expected results should I see?

If this all checks out, then next logical place is the secondary. But I already know coming off the half-wave rectifiers on the secondary side I have little to no voltage. These half-wave rectifiers are seperate from each other - they all can't be bad. But the secondary sides of the xformer is (obviously) one component, as well as the 4 inductors being one component. If primary side checks out ok, then it HAS to be a faulty xformer or inductor. Correct?

[J M Fahey]

Consider *nothing* on the primary side is Ground.
*Everything* there is hot, so you can't connect a scope there or any other instrument.
You can only use a multimeter , very carefully, as if you were checking a live appliance, which is exactly the case.
You can only check that you have 170V from the top of C37 to the bottom of C36.
You will not be able to check for oscillation on the transformer primary (20KHz to 100 KHz possible) but you already know that indirectly because you heve not DC on the secondary.
Anyway, supposing you have the primary 170V, my real suspect is that those MosFets are not being driven, simple as that.
Why?
Either the driver stage has failed or more probable, has auto shut down because of some other problem it detected.
With amp unplugged, check all diodes on secondary side.
Even one of them shorted should turn the PSU off.

[ricach]

Anyway, supposing you have the primary 170V, my real suspect is that those MosFets are not being driven, simple as that.
Why?
Either the driver stage has failed or more probable, has auto shut down because of some other problem it detected.

This is what needs further explanation for me. What driver stage? What has auto shut down? Is all the switching done on the primary side? Do I assume the secondary follows all the normal rules? Are is there logic elsewhere still making the decisions for the secondary side as well?

[J M Fahey]

Yes, of course.
A conventional PSU has a very simple primary side because the transformer works with the power line AC voltage and a simple secondary stage with diodes and capacitors.
A SMPS is very complex.
It first rectifies wall voltage, then switches it at high frequency then a special transformer and rectifiers, *but* the switches are driven by a circuit which not only does that, but also controls and regulates ouitput voltage, and adds safetry featuresm such as soft turn on and short protection.
I have not found this 3rd stage, although it obviously is there.
What I *hate* is that the SMPS is split and spread over many pages.

This is a very simple SMPS:

Look at the grey rectangle labelled Control Drive SMPS IC

[ricach]

Now I'm beginning to understand why you are preaching 'doom and gloom' in regards to this SMPS.

[Jazz P Bass]

Here is a previous post to help with the SMPS theory.
http://music-electronics-forum.com/t23974/
As to the Line 6 schematic, page 10 is the SMPS 'hardware' & page 12 shows the 'drive logic'.
You can see many points on the drive logic schematic that can 'shut down' the main switcher.
So not only does the main switcher have to be sound but also the drive logic.
I have removed the main switcher mosfets already when trying to track down a drive problem.
Less chance of going poof.
I have noted an 'Enzo trick' where you mearly pass the scope probe over the primary transformer to see if it is doing anything at all.
By the way, we are talking somewhere around a 50K switching frequency when all is working.

[J M Fahey]

Thanks JPBass (and thanks to Enzo for his invaluable capacitive test Enzo trick of the day).
Different to most (99.9%) modern SMPS which rely on a dedicated Controller IC (which, worst case , can be swapped) ; Line 6 chose to make a discrete Controller which only they understand.
This one has over 40 (forty !!!) Test Points (TP1/2/3 .../42 and counting) which obviously are needed to check some aspect or another.
Clearly indicated are a Current Sense and 2 DC fault sense , one per channel, which will turn this amp off if they don't like something, but I'm sure there are other internal watchdogs doing their job.
Of course, it *is* repairable, and there are guys who make a living out of this, but they have learned along time a very specific set of skills.
Is it worth learning them yourself?
Just as an example: imagine a $1000 thingie is broken, and you have a service manual ... in Korean ... and it can't be machine translated because it's a .jpg or something ... will you learn Korean for that one repair?.
Worst of all is that I'm sure Line 6 provides (or used to) a full board replacement, and not even the official service technician repairs it at component level.

[Enzo]

On that secondary side, I see four dual diodes. Are ANY of them shorted anode to cathode on either side?

Your SMPS is not running. One of the most common reasons for that is a shorted rectifier.

And I just have to chime in that the SMPS is EXTREMELY dangerous to work on. On the primary side, there is no "ground." A schematic might use a ground symbol, but it refers to the -170v rail. These things rectify and filter the mains DIRECTLY. That means either 170 or 340vDC at all the current the wall outlet has to give. You cannot connecty your scope ground to it. You really need to use an isolation transformer to work on it.

[ricach]

First of all I'd like to thank all of you for taking the time to help, and/or dissuade, me. My hope/intent was more to learn about SMPS in general, and not this apparent 'mandarin chinese' version that only LINE 6 can understand. lol.

Enzo, I have not had time to check those diodes.

Is it normal to still see voltages (however small) at the outputs even if the logic has 'shut down' the power? And remember that those voltages are different (by 50%) depending on whether its connected to the power amp or not.

LINE 6 will not sell me a SMPS. Only to an authorized service place. I plan to contact one and ask (nicely) if they would mind purchasing one for me. Does anybody have a good working relationship with one of these?

[Jazz P Bass]

Well it has not been determined if the SMPS is at fault. (or the only fault)
There may be a problem with the power amp itself.
I realize that the power amp header was disconnected but this in itself does not 'prove' that it is the smps that is at fault.
There may be a signal that is required 'from' the power amp to the smps.
Just a thought that I had when I saw you that were looking at purchasing a new smps.

[J M Fahey]

Just *suppose* Line 6 sells you a brand new SMPS for U$ 200 .
You hook it up and it shuts off ... because the (unrepairable) power amp is dead (the original failure).
Another $200.
The amp can be bought on EBay (working) for about $400.
Tough decision, huh?
This is something I posted in another Forum, where I'm trying to help a guy with a dead Line Spider IV, 120W 2x10" combo:

Personal opinion: I'm seeing it more and more all over the place: wonderful programmable amps which because of any, often minor problem, have to be junked, because, of course, info is unavailable and even if you get it, it's too complex to be used.
I'm getting into the mindset that the sensible option is to get a multi emulator *pedal* or POD and plug it into a regular amp.
Failure of one does not kill the other, separate parts tend to be more robust and worst case, one dead means you can still play with the other and save the night.
POD dead?: plug straight into the amp and play somewhat simpler, less effects.
I bet the *music* won't suffer that much.
Amp dead?: plug the POD straight into the PA.

While in the full digital combo amp, any problem means very frustrating not playing at all.

[Enzo]

Forget this 50% stuff. If a 40v supply is sitting at 5mv or 10mv is irrelevant, it is not at 40. SOme tiny residual voltage doesn't matter if it changes by "50%."

If a SMPS tries to start but can't, there is often a small spike, which is enough to put a few millivolts or whatever into the filter caps.

Imagine you all come out to your car, and the tire is flat. Really flat. Now the rest of your family goes ahead and gets into the car. Your flat tire is now even 50% flatter. The important fact is that there is no air in the tire, not the dimension of the squished rubber.