Ike power supply "difficulties"

[erichayes]

Hi All,

Barrie, I'm going to let Shannon take this one, as it's his baby, but I just want you not to get too discouraged with the DC resistance measurements you got on the power transformer. Alternating current is interesting stuff; what would appear to be a dead short when measuring with an ohmmeter could actually have several (or even dozens) ohms impedance when measured at 50 or 60 cycles, or Hertz.

One quick and inexpensive test you can make requires a 15 or 25 watt 120 volt light bulb and socket, a line cord and an AC receptacle (you brethren living in countries with 240V mains should obviously use a 240V lamp). Wire the lamp socket in series with the receptacle and the line cord, so that the black (or brown) lead of the cord goes to one terminal of the socket, the other socket terminal is connected to one terminal (say brass) of the receptacle, and the other (silver) terminal of the recept. is connected to the white (blue) lead of the line cord. Forget about the green or green/yellow lead on the line cord if you're using a three wire cord; if you're using a 2 wire cord, follow the above instructions but ignore the colors.

THE FOLLOWING PROCEDURE INVOLVES DANGEROUS, PERHAPS LETHAL, VOLTAGES. IF YOU HAVE ANY DOUBTS OR QUESTIONS REGARDING THE TEST PROCEDURE, DO NOT PERFORM IT YOURSELF. SEEK HELP FROM A PERSON WHO IS EXPERIENCED WITH WORKING AROUND HIGH VOLTAGES. THIS DISCLAIMER IS TO PROTECT SHANNON'S AND MY RESPECTIVE REAR ENDS.

Screw the lamp into the socket, disconnect ALL the secondary wires from the circuit board and make sure none of them can contact one another, plug the amp into the recept. and plug the line cord into the wall. If the lamp lights brightly, you have a shorted turn(s) on one of the windings and are proud owner of an expensive paperweight. If the lamp lights up faintly, or not at all, it's probably OK. You can verify this by CAREFULLY--use clip leads and not your fingers--shorting one winding at a time with power removed, and reconnecting power each time.

The lamp should light at near to full brilliance in each case, indicating that excessive current is flowing in the secondary winding, and, therefore, also in the primary. The lamp acts as a ballast, soaking up the excess current and preventing damage to the transformer (assuming it passed muster in the first place).

[WA4SWJ]

Barrie,

I once turned one of my eiclones on and off too fast and blew a rectifier diode. Check your power supply diodes. They can cause the problem you describe if one or more of them are shorted. It might be a much simpler fix than you think. I agree with the comments Eric made on transformer impedances.

Although it's a remote possibility, check your 6SL7 tube. It might be shorted and drawing too much current. Again, that's a remote possiblity. More likely there is a solder splash somewhere on the PCB somewhere in that vicinity causing a short circuit. Or perhaps something is unsoldered causing the tube to draw too much current.

Your center tap fuse is a good idea. I did that on mine and I'm glad. I rather doubt your power transformer is bad.

Best of luck in trouble shooting. The eiclone is a great sounding amp. I have 3 in a home theater setup and they are working great!

All the best,

[WA4SWJ]

Barrie,

One more comment - when you placed a 1/2 amp fuse in the primary of the transformer and powered it up, there could be enough inrush current when the transformer is first energized to blow a fast blow fuse of that rating. That would be normal. Again, the transformer is probably OK. Try it again with a slow-blo fuse with a slightly higher rating - maybe 1 amp or so. It should be fine.

Also, although this is probably not the problem, check the polarity of the electrolytic capacitors on the board. Make sure they are proper but it sounds like they are OK.

I really do suspect the power supply diodes - but something caused them to fail, so if one or more are shorted check for other root causes on the board like the solder splashes I mentioned in the previous post.

Hope all goes well,

[Shannon Parks]

Hi Barrie,

As Ed said, do check your diodes and replace them. One of them is blown. I'll bet a dollar to a loonie that your transformer is A-OK. The smoking inrush current limiter occurance seems to happen when a diode is blown and max current flows through the limiter.

The only blown diodes I have heard of in the Ike always seemed to be the result of over voltage on the diodes, directly related to the use of a standby switch. See the standby switch thread, but bascially, when the DC reference at the center tap is removed you can see very high voltages (over 1kV) on the secondary, which breaks down the diode creating a short. My thought here is that your center tap fuse blew first (very possibly from startup surges - I haven't measured the current here). At that point the secondary voltages surged and broke down a diode (or two). Go ahead and directly connect your center tap to the PCB ground and fix the diodes, then we'll hammer out a solution on the voltages (not related I think, but I'll look this over today).

Not related to your problem, but Ed Long mentioned he blew a fuse while cycling power once. While this may have lowered the effectiveness of the inrush current limiter, I'm not happy that this happened. I'll see if I can spec out a 1.5A replacement diode. Thanks for this info, Ed, as I've not seen this.

Shannon

[EWBrown]

I've popped fuses when cycling power off/on on both the ST35 and Ikes.
Reason is the CL90 is still "hot" and therefore has low resistance. The inrush current limiting function isn't there. So far I've been lucky and not popped any uF4007s. I tend to fuse "conservatively" and don't use the standby switch - as that can lead to other problems mentioned elsewhere.

It would take a major effort to blow up one of these power trannies, the fuses should blow out well before that happens. A long, slow overcurrent situation would be far more harmful than would be a quick fuse pop.

ed B in NH

[erichayes]

Hi All,

Hoo boy, did I drop the ball on my advice to Barrie.

ATTENTION: If you are going to test your power transformer using the "shorted secondaries" method I described, you MUST replace the 15 watt bulb with one of at least 150 watts. The reason is simple: you want the brunt of any possible current flow to go through the path of least resistance (worst case: shorted xfr winding/best case: high wattage light bulb that doesn't light).

You guys who are mentors on this forum need to keep close watch on Heathkitesque step-by-step instructions posted by anyone, and send up a flare if there's a problem with the method involved.

[EWBrown]

Fusing conservatively - that is just using the lowest current rated fuse that will still poerate reliably. Using the CL90 eliminates the power-on surge, so a smaller rated fuse can be used. Use a fast blow instead of slow blow, if possible, and I usually try to get away using a fuse of approximately 50% higher than the actual operating current. Like of the amp draws 2 amps, I'll use a 3A FB fuse. YMMV, etc...

Just too many times, I've seen used amps, or those that needed repair, and the original user had simply stuffed in a 15 amp "no-blow" fuse, and if they were really lucky, nothing too serious would happen. Worst case usually meant that a new power tranny was needed.

/ed B in NH

[Shannon Parks]

I lifted the cathodes on all three diodes on the board and, with the DMM on the diode setting, they all showed 0.45V or so forward biased, and infinity reverse. Should I just replace them anyway?

Hi Barrie,

Yes, I would go ahead and replace them. Maybe the low DMM voltage check across it isn't causing it to breakdown completely (guess here). If you don't have any spares then you can use some from Radio Shack if that is more convenient. PN# 276-1114 package of 3 diodes, rated at 2.5A & 1kV.

I would lift R27 & R28 (or disconnect the choke) before my first power up. Also, have the output tranny disconnected from J4-8. We'll take it one step at a time.

Shannon, I had read the thread on the standby switch, so I didn't include it. But I'm wondering about your thoughts on the current surge at turn-on. Maybe just replacing both fuses with slow-blo ones would be my answer.

The blown fuse emulated a switch in the off position - not really intuitive until it happens. I had some inline fuses that I purchased a couple years ago specifically for this purpose (fusing CT), but see issues with it now. I realize standby switches and fuses in center taps are accepted practises, but I am now of the opinion that they can do more harm than good. I think moving the fuse to the B+ line of the output or to the output tubes' cathodes might be a better solution. But I haven't done any tests with those methods.

A pic would be great if you can mail me one, Barrie. I'll post it here. Some good up close ones! Yellow_Light_Colorz_PDT_18

Shannon

[Shannon Parks]

Okay, I've followed what you've said Shannon. I've replaced all three diodes and did the resistance checks. They looked close to what I originally got. I turned it on without J4-8 or the choke connected. The voltages were close to the originals as well.

Hi Barrie,

Barrie's Amp Pix
http://www.diytube.com/forumpix/barrie1.jpg
http://www.diytube.com/forumpix/barrie2.jpg
http://www.diytube.com/forumpix/barrie3.jpg

As I mentioned in the email, be sure your transformers are securely grounded to chassis ground using star washers. I'm not sure if yours are isolated, but I remember one of the faithful here having strange behavior before he 'strapped them down'. Also, maybe you have already done this, but have the safety ground connected to a power transformer bolt and run a jumper from J4-4 or J4-5 to the same bolt. Technically speaking, UL expects the safety ground to have it's own dedicated chassis bolt very close to the AC input. We can tweak for low noise later.

Go ahead and connect the choke. Short the RCA input. Power up with all your tubes in and check the voltages. Make sure the 6SL7 looks good. Ditto with the 6SN7. Voltages should be rock solid with no jitter. Measure the volatge drop across the choke and divide by its DCR to see how much current the system is using. Maybe 10mA? Check your biasing now to make sure you will be over biased so you can avoid a runaway tube on start up.

Shut it down and look over the output connects again. Hammonds connections can get a little goofy. Leave your negative feedback at J1-3 disconnected at this step - you can immediately reconnect it after this step. Double check your hookups to J2 and J3. Hook it up to J4-8 after she's bled down. Fire it up.

Shannon

[Shannon Parks]

I do have a couple of questions that I'm hoping someone can comment on.
I have this buzz from the power transformer and it seems to running hot (no signal - the input is shorted).

Hi Barrie,

Good work!!

Is this mechanical buzz emanating from the PT with your ears up close or does it manifest as humm/buzz coming from the speakers?

The PT will get pretty warm (eg can touch for 2.5 seconds) just sitting there as it is. This is normal.

Your voltages look fine. It looks like you could remove one or both of the filament dropping resistors and replace them with jumpers.

Shannon

[EWBrown]

Some of those Hammond power trannies can get pretty "buzzy", it's just a matter of how they are assembled. Try tightening the screws that hold the end bells on , usually 4 of them. Another cause is that the windings or internal laminations can vibrate, and there isn't much that can be easily done about that.

FWIW, the Hammond 27X series power trannies have a 115VAC rated primary, so they tend to run a little hot and hard on today's more common 120-127 VAC power mains. The ICL-90 helps dissipate some of this AC input excess during normal operation. The more expensive 37Xseries have "universal" primaries which allow for 120/240 VAC operation, as well as a dedicated bias tap on the HV secondary.

Also, a steel chassis can pick up the magnetic field from the power trannie and cause its own vibrations. A simple fix for that is to put a piece of thick rubber between the tranny's core and the top of the chassis. that should damp out the induced chassis vibes.

/ed B in NH