High speed switch in psu..I have better test equipment now

[heathtech]

Hey folks,
I am repairing a power supply and am reasonably sure I see the problem, but before I replace any parts, I'd like some opinions. I am going to post a simplified schem. of an NTE 2310 npn high V high speed switch in between two transformers. As I read it, T1 induces an AC signal at Xhz through the secondary to the base of the transistor, which switches at Xhz, producing elevated alternating DC through the primary of T2, which induces the same into T2 secondary. I am reading approx 320VDC across the E-C pins of the transisor with a VM, shouldn't I be reading AC? I am also reading voltage on both legs of T2 primary with respect to ground, but no voltage drop across the primary, and no voltage is induced into the T2 secondary. I am suspect of the transistor. Any opinions?

 

[Nigel Goodwin]

Re: Question about High speed switch in power supply

Hey folks,
I am repairing a power supply and am reasonably sure I see the problem, but before I replace any parts, I'd like some opinions. I am going to post a simplified schem. of an NTE 2310 npn high V high speed switch in between two transformers. As I read it, T1 induces an AC signal at Xhz through the secondary to the base of the transistor, which switches at Xhz, producing elevated alternating DC through the primary of T2, which induces the same into T2 secondary. I am reading approx 320VDC across the E-C pins of the transisor with a VM, shouldn't I be reading AC? I am also reading voltage on both legs of T2 primary with respect to ground, but no voltage drop across the primary, and no voltage is induced into the T2 secondary. I am suspect of the transistor. Any opinions?

Presumably this is a crude simplified version of the actual circuit?.

I'm puzzled by your 320V DC reading, that's the kind of voltage you get from rectified 230V mains, 110V should be more like 160V - is there a voltage doubler on the incoming supply?. Is this a PC PSU?, they often do that.

It's fairly unusual for the base to be fed from a transformer?, it 'may' be to isolate feedback from the secondary?.

My first thought though, when presented with a dead switchmode supply, where the chopper isn't short circuit - is the startup resistor, almost all SM supplies have one, and they commonly go high or O/C. This prevents the supply starting up, and should be the first thing you look for - be careful though, the large electrolytic will probably be left charged up!.

Look for a large value resistor (usually between 50K - 2.2M), it probably connects to the 320V supply, and feeds the base of the chopper transistor.

 

[heathtech]

Yes, it is simplified without Caps, resistors, diodes, etc.
I have more information....T1 is signaled via an IC (manufacturer's spec i imagine) labeled HYB9912. This is an older Boschert PSU later produced by Artesyn...model XL125/4602. It is an either/or 110/220 VAC to quad-DC power supply. It came out of a lab instrument manufacured by LECO corp. The issue is, the IC chops the transistor, but the voltage required to power the IC comes off a tap from T2, which is not inducing. It is kind of a loop that isn't working. And, oh yes, the cap is very hot!! hehe. I got that tingly feeling once and started discharging it each time I depower.

 

[heathtech]

BTW, I am reading .6 VDC at the base of the transistor with no measurable AC freq when measured with a freq. counter. There is no voltage whatsoever at the IC. There is no resistor located in series with the base of the transistor. It connects directly to the transformer secondary. I wish I could post the schematic, but all I have is a very fuzzy facsimile version that looks like it is a copy of a copy of a copy. I was very lucky to get it from the tech support of the Leco instruments company. I find it unusual to get schematics from vendors.

 

[Nigel Goodwin]

Try and find the startup components, as you have a circuit (even though poor) it shouldn't be too hard.

As I said before, usually it's a high value resistor (and high value resistors are well reknowned for going higher or O/C).

The next most common component would be an electrolytic capacitor, these almost always go low ESR (probably the most common failure mode in the 21st century?). It could be used to give a 'kick' to start the PSU, or it could be a decoupler on the supply rail for the IC - a third possibility is a coupling capacitor to the chopper, but not if it uses a transformer!.

 

[heathtech]

I am going to go home and scan the schematic. It is fuzzy enough as it is, but maybe it could shed some light for Nigel or another more experienced guru than myself.

 

[heathtech]

I hope I can upload this schematic.

 

[Nigel Goodwin]

I would be inclined to check R2 and C9, if you don't have an ESR meter check C9 by replacement!.

 

[eblc1388]

The next most common component would be an electrolytic capacitor, these almost always go low ESR (probably the most common failure mode in the 21st century?).

Many repair technicians would be out of work if capacitors goes LOW ESR. 8)

 

[eblc1388]

And, oh yes, the cap is very hot!! hehe.

Which capacitor(s) get very hot?

 

[Nigel Goodwin]

The next most common component would be an electrolytic capacitor, these almost always go low ESR (probably the most common failure mode in the 21st century?).

Many repair technicians would be out of work if capacitors goes LOW ESR. 8)

Oops! - slight typo! - I meant, of course, HIGH ESR.

 

[heathtech]

eblc1388 Posted: Fri Jul 01, 2005 4:25 am Post subject:

--------------------------------------------------------------------------------

heathtech wrote:
And, oh yes, the cap is very hot!! hehe.


Which capacitor(s) get very hot?

BY "Very Hot" I meant charged up, not thermally hot. It is a slang expression I use. The capacitors that hold the charge are C12,C13. I have been discharging them each time I depower the circuit. By the way, Nigel. I did in fact already replace R2 because I found it open. This did not fix the problem, but it was in fact open. It is supposed to be 91kohm (fuzzy in scheme.) It was reading infinity on the Megohm scale of my meter. So, by logical deduction, you think C9 is the kick start cap? I will try replacing it and report! thanks.

 

[heathtech]

I replaced C9 and now I have unfiltered AC 56v on the low voltage side of the PSU!! C5 is bulging a little, I'm going to look at replacing it...any other ideas? BTW, the AC is at line frequency of 60hz.

 

[Nigel Goodwin]

If you have ANYTHING on the secondary side, then the PSU has to be running (even if it's not running correctly, it still has to running!).

However, I'm confused by your measurement of 56V AC?, where and how did you measure it?.

 

[heathtech]

I am measuring it from the -neg pin on the line rectifier to the 24,12,and 5v outputs. I also get it from neutral to those points. I don't however get it from the return terminal to those points. kind of confusing. I tried an experiment by swapping the 470uF 200 V C5 capacitor with a 500uF 200V(I have no 470 at hand) It changed the outputs to 52 Vac. interesting but may be what one should expect...am I painting a good picture?

 

[heathtech]

This might be important, I am getting a 60hz signal at the base of the transistor Q5. I suspect that the A1 IC is generating 60 hz maybe because the low voltage side isn't filtered properly?

 

[heathtech]

Another helpful fellow named Bjorn on Electronic lab Forum suggested the crowbar CR16 might be troublesome...I am going to check on that. Thanks for the help, BTW...the more heads I can get helping my peabrain the better!...In my work I only get to troubleshoot PCB's maybe once a coupla months, so I ain't the most experienced guy around.

I am reading Cr16 as an C122f-1 SCR reverse Blocking Thyristor...Does that sound right to you?

 

[heathtech]

Another observation is that there is a Power fail detect terminal on the output terminal strip (pin one) that is outputing 2.5 volts DC

Oh, yes, this signal is controlled it seems by comparitor U1

OK, without a shadow of a doubt, Switch Q5 is doing its thing right now. The primary side of the main transformer is working (though incorrectly I think.) The voltage drops as would be expected across T1, and drops even further across the switch. There is definately conduction occuring. I seem to be seeing this when I measure the T1 with my meter. On several taps I read 300+ vdc, with 11 volts of AC...I believe we are narrowing it down to a cap on the primary side. Maybe C4?

Just as I feel sure of something, my mind becomes doubtful. I am going to try to find the switching transistor at a local shop.

Well, I have a physicians apointment in 45 minutes, so I'll be continuing this investigation later. Thanks for the help!

 

[Nigel Goodwin]

I am measuring it from the -neg pin on the line rectifier to the 24,12,and 5v outputs. I also get it from neutral to those points. I don't however get it from the return terminal to those points. kind of confusing. I tried an experiment by swapping the 470uF 200 V C5 capacitor with a 500uF 200V(I have no 470 at hand) It changed the outputs to 52 Vac. interesting but may be what one should expect...am I painting a good picture?

If I understand you correctly?, you have the negative side of the meter on the primary, and the positive side on the secondary?. You can't do that! - the primary and secondary are isolated, that's the point of the transformer. So any readings you get are totally meaningless - and I mean TOTALLY!.

 

[heathtech]

Yeah, I realize that now. I guess I just got excited when I read SOMETHING on the low voltage side. As I have said before in other posts. I know about 60% of what I'd like to know about electronics, and even the best only ever know about 99%. In other words, I make mistakes and learn something every time I see a new board. I only get the chance to make PCB repairs maybe once a month at my job. Most problems I see are mere fuses or pitted contacts on relays and switches. This PSU I find particularly challenging. But, me thinks that is the point isn't it!! I love troubleshooting because I learn something new and get better each time! Anyway, enough of the excuses. I am posting from home and will be back at work after our Yankee (LOL) independence day holiday. I am going to study the schematic in my free time and next week i will blow the dust off my oscilloscope at work and get serious about fixing this supply. The Equipment i will save by repairing it is worth about $50,000 US dollars, and I am not allowed due to legal liabilities to replace it with anything but an exact replacement, and the manufacturer doesn't support this instrument anymore. Your suggestions are much appreciated!
....I think at this time I'll summarize the facts. I have replaced C9,C5,C4, and R2 and, though I was fooled before, i am not seeing activity on the low V side. I do now, however, that there is 320 V on the primary side and I have seen about 11 V of AC ripple at certain points, so I think I am still seeing a filter cap problem. At this point, because of the value of this board, I am going to also replace Q5. Can't hurt, huh? It has been my experience in the short decade I have been around electronics that heat is the enemy as we all know! ( so a heatsinked solid state component such as this is propably just about ready to go anyway) til next week, thanks again.