When it come to working on tube amps, what should the general practice be to check before powering on? I just replaced the big electrolytics in the PS section of a Heathkit, and the fuse blew when I powered up- There's a lot of things that I could have done, but what should I check?
The fuse blew instantly, as though there were a short circuit. I put a new fuse in, plucked out the rectifier and power was sustained- all of the tubes lit up, and all of my power transformer voltages are correct.
So; Rectifier IN=short, Rectifier OUT=sustained.
Problem is obviously in the PS circuit somewhere, but I made every effort to follow the schematic, so now I have to troubleshoot; and I'm not sure what to look for.
The original rectifier was what blew out in the first place; I plugged a new one in, but everything started to get very hot; VERY hot- untouchable. One of the electrolytics was sizzling, another started smoking- so I went and replaced the lot of em'
I thought that the caps were the source of the problem, but maybe I'm wrong?
Please help if you can.
PS- I am astute at working with high voltages- coming from the electric motor trade. But tubes are still new to me; I am but an amateur in the learning process.
The fuse blew instantly, as though there were a short circuit. I put a new fuse in, plucked out the rectifier and power was sustained- all of the tubes lit up, and all of my power transformer voltages are correct.
So; Rectifier IN=short, Rectifier OUT=sustained.
Problem is obviously in the PS circuit somewhere, but I made every effort to follow the schematic, so now I have to troubleshoot; and I'm not sure what to look for.
The original rectifier was what blew out in the first place; I plugged a new one in, but everything started to get very hot; VERY hot- untouchable. One of the electrolytics was sizzling, another started smoking- so I went and replaced the lot of em'
I thought that the caps were the source of the problem, but maybe I'm wrong?
Please help if you can.
PS- I am astute at working with high voltages- coming from the electric motor trade. But tubes are still new to me; I am but an amateur in the learning process.
Have you got a solid state or valve rectifier?
I am very new to valves, so if that is what you have, I cannot help.
If solid state, sounds like you may have a polarity issue. Have you got a one piece bridge rectifier or one you made from individual components? Check the polarity of the outputs, you might have wired it backwards.
My last chip amp was built with the help of my father, an electronics engineer. Normally I build the amp from power supply to amp unit, checking each stage along the way very carefully. As I assumed my dad knew what he was doing, we just wired it up and turned it on. Polarity was wrong. Blew all my power supply caps and the power supply regulators and the amp chips! Not fun, but a valuable lesson!
I am very new to valves, so if that is what you have, I cannot help.
If solid state, sounds like you may have a polarity issue. Have you got a one piece bridge rectifier or one you made from individual components? Check the polarity of the outputs, you might have wired it backwards.
My last chip amp was built with the help of my father, an electronics engineer. Normally I build the amp from power supply to amp unit, checking each stage along the way very carefully. As I assumed my dad knew what he was doing, we just wired it up and turned it on. Polarity was wrong. Blew all my power supply caps and the power supply regulators and the amp chips! Not fun, but a valuable lesson!
When I breadboard tube amplifiers I usually:
1. Make a chassis
2. Bolt everything I need
3. Connect power socket, switch, fuse, primaries of the power transformer,
4. Switch on and check AC voltages on secondaries.
5. Unplug. Wire PSU including rectifiers, voltage regulators, etc...
6. Turn on and test voltages
7. Turn off. Unplug. Discharge capacitors using resistors if no bleeding resistors exist.
8. Wire the rest.
9. Connect dummy loads with speakers through 100 Ohm resistors in parallel. Fight against parasitics. Tweak parameters...
10. Connect speakers, signal source, enjoy the music...
I understand that you are probably on the stage 6 and got some errors such as wrong polarity of caps or a bridge. The tranny should be connected to "~" leads of the bridge. "+" lead should be connected to the "+" lead of the capacitor, "-" from the bridge should be connected to the "-" of the capacitor.
1. Make a chassis
2. Bolt everything I need
3. Connect power socket, switch, fuse, primaries of the power transformer,
4. Switch on and check AC voltages on secondaries.
5. Unplug. Wire PSU including rectifiers, voltage regulators, etc...
6. Turn on and test voltages
7. Turn off. Unplug. Discharge capacitors using resistors if no bleeding resistors exist.
8. Wire the rest.
9. Connect dummy loads with speakers through 100 Ohm resistors in parallel. Fight against parasitics. Tweak parameters...
10. Connect speakers, signal source, enjoy the music...
I understand that you are probably on the stage 6 and got some errors such as wrong polarity of caps or a bridge. The tranny should be connected to "~" leads of the bridge. "+" lead should be connected to the "+" lead of the capacitor, "-" from the bridge should be connected to the "-" of the capacitor.
Hi Sir Trefor,
Sounds like you might have a bad rectifier tube since the fuse blew immediately at power up and didn't with it removed. It is also possible that you have wired a cap backwards, particularly if some small amount of time passes before the fuse blows.
Usually I do a careful visual inspection, probe key points with the dvm to make sure there are no shorts, and use a variac and/or ballast lamp to bring up the device I am testing.
What I most strongly recommend is the use of a ballast lamp in series with the primary of your power transformer during the trouble shooting phase. This will prevent fried power transformers at the minimum and in a lot of cases prevent a lot of other damage as well. At its simplest this can just be a hacked extension cord with a lamp socket wired in series with the hot conductor. You can use a 100W bulb to limit current to a safe value in the event of a short. Higher and lower wattage bulbs can be used depending on the normal operating current of the device you are testing. Usually the voltage will be quite low, but you should be able to measure some % of normal operating voltage. The bulb will illuminate to full brightness in the event of a short and limit current in the primary to the current required to light the bulb. (Slightly less than 0.9A@120V)
Take advantage of friends with electronic backgrounds when available and you are stumped, very often they will notice things you may have missed repeatedly.
Make sure to discharge supply capacitors prior to working on equipment, and always check to make sure they are discharged whether or not there is a bleeder resistor installed. (They can fail)
Slightly off topic safety recommendations:
Protect yourself and other equipment with a ground fault interrupter (GFI/GFCI) that powers everything on the bench. Note that the GFI will not protect you from contact incidents with anything other than the mains voltage.
Make sure the chair you sit in when working on any HV circuit is not conductive! (Exact opposite of good esd practices) I use a wooden chair.
No bare feet..
Anti-static mats and ground planes are nice when working on low voltage electronics, but not a good idea around high voltages, and make sure that you cannot come in contact easily or accidentally with grounded bogie objects (excluding the d.u.t which most likely will be and should be grounded) while probing high voltages.
It's a really good idea to have someone else around when working with tube circuits in the event of an accident.
If possible set up probe points in HV or very high current circuits prior to applying power - this does require a couple of meters to be practical.
Insulate all but the very end of probe tips.
I am sure lots of people can add to this list, but hopefully you find this useful and not too paranoid..
Have fun!
Sounds like you might have a bad rectifier tube since the fuse blew immediately at power up and didn't with it removed. It is also possible that you have wired a cap backwards, particularly if some small amount of time passes before the fuse blows.
Usually I do a careful visual inspection, probe key points with the dvm to make sure there are no shorts, and use a variac and/or ballast lamp to bring up the device I am testing.
What I most strongly recommend is the use of a ballast lamp in series with the primary of your power transformer during the trouble shooting phase. This will prevent fried power transformers at the minimum and in a lot of cases prevent a lot of other damage as well. At its simplest this can just be a hacked extension cord with a lamp socket wired in series with the hot conductor. You can use a 100W bulb to limit current to a safe value in the event of a short. Higher and lower wattage bulbs can be used depending on the normal operating current of the device you are testing. Usually the voltage will be quite low, but you should be able to measure some % of normal operating voltage. The bulb will illuminate to full brightness in the event of a short and limit current in the primary to the current required to light the bulb. (Slightly less than 0.9A@120V)
Take advantage of friends with electronic backgrounds when available and you are stumped, very often they will notice things you may have missed repeatedly.
Make sure to discharge supply capacitors prior to working on equipment, and always check to make sure they are discharged whether or not there is a bleeder resistor installed. (They can fail)
Slightly off topic safety recommendations:
Protect yourself and other equipment with a ground fault interrupter (GFI/GFCI) that powers everything on the bench. Note that the GFI will not protect you from contact incidents with anything other than the mains voltage.
Make sure the chair you sit in when working on any HV circuit is not conductive! (Exact opposite of good esd practices) I use a wooden chair.
No bare feet..
Anti-static mats and ground planes are nice when working on low voltage electronics, but not a good idea around high voltages, and make sure that you cannot come in contact easily or accidentally with grounded bogie objects (excluding the d.u.t which most likely will be and should be grounded) while probing high voltages.
It's a really good idea to have someone else around when working with tube circuits in the event of an accident.
If possible set up probe points in HV or very high current circuits prior to applying power - this does require a couple of meters to be practical.
Insulate all but the very end of probe tips.
I am sure lots of people can add to this list, but hopefully you find this useful and not too paranoid..
Have fun!
In "short" (no joke inside !)
- use a variac/auto-Xformer, always. Rise slowly, monitor
- differential protection, always (mains supply rail w breaker available)
- left hand in pocket when you take readings on line
- check for caps b/w live or neutral and chassis, remove if any
(applies for instance to old "all supplies" vintage radios)
- remove all tubes, check voltages
- then insert rectifier, recheck voltages
- whenever all is correct, replace the tubes
- be patient, keep cool, breathe !
Hervé
Grenoble
- use a variac/auto-Xformer, always. Rise slowly, monitor
- differential protection, always (mains supply rail w breaker available)
- left hand in pocket when you take readings on line
- check for caps b/w live or neutral and chassis, remove if any
(applies for instance to old "all supplies" vintage radios)
- remove all tubes, check voltages
- then insert rectifier, recheck voltages
- whenever all is correct, replace the tubes
- be patient, keep cool, breathe !
Hervé
Grenoble
Sir Trefor said:
I dunnow. Never heard of any "general practice", just what I do.
If using a recycled power xfmr, I always inspect carefully, and make sure the thing works right before using it. Regardless of whether it's solid state or hollow state, complete the power supply first. After completion, but before power up, check all connections, then check 'em a couple more times. Occasionally, I've caught mistakes before they went up in smoke and flames. It's also a real good idea to use the resistance function of a DMM (power off, of course!) to see what resistance the PS is looking into. If it says "0R" then it's probably not a real swell idea to power up just yet.
Next, I wire a lightbulb in series with the primary. This gives you a good idea if something isn't right, since the bulb shouldn't light up brightly (except for a second or two as the filter capacitors charge up). That way, you just might prevent some inconvenient poofage.
It's also a good idea to test each stage as it's completed. That way, if something's not right, you know where to look. Again, make certain that everything's wired the way it should be: all electrolytics properly polarized, all connections going to the right socket pins.
That's not a good thing. If you got the polarities of the electrolytics wrong, that's what can happen whether it's a low voltage, high current SS PS or a high voltage, low current hollow state PS.
Hi Sir Trefor,
Did you use brand new capacitors, or those "NOS" bombs?
If your fuse blew immediately, I would suspect that your rectifier got damaged. Try another. I generally use a variac to power things up. In your case I might pull the rectifier and power up the heaters. Then I'd use a variable high voltage supply to get the rest working. This is fairly safe and smokeless.
-Chris
Did you use brand new capacitors, or those "NOS" bombs?
If your fuse blew immediately, I would suspect that your rectifier got damaged. Try another. I generally use a variac to power things up. In your case I might pull the rectifier and power up the heaters. Then I'd use a variable high voltage supply to get the rest working. This is fairly safe and smokeless.
-Chris
Yes !
I completely agree with the expression "NOS bombs"
Energy stored varies as square of voltage, which is HIGH
Electrolytics are ageing even in the drawer and they need to be reformed gently before use, you may find how at the following link or ask Google with "reforming capacitors" :
Reforming capacitors
I completely agree with the expression "NOS bombs"
Energy stored varies as square of voltage, which is HIGH
Electrolytics are ageing even in the drawer and they need to be reformed gently before use, you may find how at the following link or ask Google with "reforming capacitors" :
Reforming capacitors
Wow, I am surprised and grateful at the response to my question!
One thing that I did NOT mention was that I lost the schematic I drew as I took everything apart- I was putting it back together the way I "remembered" it. But, I found a true schematic at the public library(Sam's photofact), and I am currently wiring it properly; I had made a few mistakes- and the supply caps were not properly connected to ground.
Thanks to everybody for what should be common sense. I'll be sure to let you know how it goes when I'm finished.
PS-Might I use a rheostat in place of the variac?
No, the old multi-section supply "cans" were replaced with new Sprague Atoms- all voltages correct.
Polarity of the caps has been double checked- it is correct. The Rectifier tube may be the problem; as it is an old, used 5AR4. It "was working" before being taken out of service, but, then again...
One thing that I did NOT mention was that I lost the schematic I drew as I took everything apart- I was putting it back together the way I "remembered" it. But, I found a true schematic at the public library(Sam's photofact), and I am currently wiring it properly; I had made a few mistakes- and the supply caps were not properly connected to ground.
THIS is a great suggestion; I think I even have a socket out in the garage...
Excellent advice, thank you.
Thanks to everybody for what should be common sense. I'll be sure to let you know how it goes when I'm finished.
PS-Might I use a rheostat in place of the variac?
Hi Sir Trefor,
If you do any kind of work like this you should pick up a variac off of ebay. A 2A model will be fine for most of your work. If you find a 5A model, great. 10 A models are getting large and heavy.
An AC voltmeter and a suitable current meter on the output will be valuable. You'll be very glad you did this, try it.
-Chris
Yes, but it will be difficult to use.
If you do any kind of work like this you should pick up a variac off of ebay. A 2A model will be fine for most of your work. If you find a 5A model, great. 10 A models are getting large and heavy.
An AC voltmeter and a suitable current meter on the output will be valuable. You'll be very glad you did this, try it.
-Chris
I did this a few minutes ago with my variac. I have another problem seperate from the power supply; the bias voltage.hd38 said:
This particular amplifier employs a grid-leak biasing system, with a 20uF capacitor. Bias voltage supply is supposed to be -15V. When I was checking everything out while the tubes were unplugged I had my temperature probe out to make sure nothing was going to explode- and I was right. After less than 30 seconds this capcitor reached 90 degrees C
I immediately shut everything down. I gonna take it out tomorrow and check everything again, as I am tired and have pushed my brain past it's useable "uptime". I think I may have installed a faulty(as in "the wrong one") silicon diode- maybe the capacitor is receiving too much voltage?
Hi Sir Trefor,
If you replaced a selenium rectifier with a silicon diode, your voltage may run too high. I wouldn't expect it to be that high. Confirm you're getting the correct polarity. Then use a series resistor with the diode to bring your bias voltages in line. Your B+ will also drop as your output tubes will now be drawing current.
-Chris
Cool! A little overkill. I have one too, but I normally use a 2A unit mounted on the bench. The 10A unit is too large.
If you replaced a selenium rectifier with a silicon diode, your voltage may run too high. I wouldn't expect it to be that high. Confirm you're getting the correct polarity. Then use a series resistor with the diode to bring your bias voltages in line. Your B+ will also drop as your output tubes will now be drawing current.
-Chris
Thanks, Anatech. The capacitor I described is actually connected to the diode as a smoothing capacitor; sorry. The supply voltage is 55VAC, there is already an 8.2 kOhm in series with the diode to drop the voltage to the correct -15V. Both of these voltages read 55 and 15 respectively on my multimeter. I quadruple checked the polarity of the electrolytic; it is correct- (-)connected to cathode band, (+)connected to earth.
Would the leaking in of AC to the cap cause it to overheat like this? Would the simplest solution to this problem be a malfunctioning diode? When I check the voltage with my meter I cannot get a DC reading; I assumed it is because with this being a half-wave rectifier that the ripple is to great for the meter to see it as DC...
(leaves computer to check unit, comes back, continues typing)
...Yep, the diode is no good- current flows both directions.
I grabbed the wrong part at the electronics store; I checked the bag and the maximum PRV is only 40 V. But it's also supposed to be rated for 1Amp; is going 15V over at .001Amp really what made it fail?
I don't want anymore mishaps; I want to get this thing done, and I want to get it done RIGHT. I will get a new diode, but I want to make sure that it was ignorance on my part that caused failure before going any further.
PS- I did change the capacitance value of the electrolytic from 20 to 22 uF; I didn't think it would be a problem. Should I not have done that?
Would the leaking in of AC to the cap cause it to overheat like this? Would the simplest solution to this problem be a malfunctioning diode? When I check the voltage with my meter I cannot get a DC reading; I assumed it is because with this being a half-wave rectifier that the ripple is to great for the meter to see it as DC...
(leaves computer to check unit, comes back, continues typing)
...Yep, the diode is no good- current flows both directions.
I grabbed the wrong part at the electronics store; I checked the bag and the maximum PRV is only 40 V. But it's also supposed to be rated for 1Amp; is going 15V over at .001Amp really what made it fail?
I don't want anymore mishaps; I want to get this thing done, and I want to get it done RIGHT. I will get a new diode, but I want to make sure that it was ignorance on my part that caused failure before going any further.
PS- I did change the capacitance value of the electrolytic from 20 to 22 uF; I didn't think it would be a problem. Should I not have done that?
Sir Trefor said:
Your description makes me suspect that the diode is backwards, the cathode end of the diode is positive in operation and you need negative. It also would explain why that cap gets so hot. Replace both before continuing if the above is true. You also need a 200PIV diode minimum. 22uF is not a problem.
So if its an electron valve we call the NEGATIVE electrode the cathode; if it's a semiconductor we call the POSITIVE electrode the cathode?? 
I wired it as I read the schematic; unless I am forgetting my electronics course the arrow in the [diode] symbol represents the anode and the straight line it points at is the cathode(?) IF that were the case, then as the schematic reads the negative terminal of the capacitor IS connected to the cathode. However, you have corrected me, and apparently I don't even know what little I thought I knew.
I wired it as I read the schematic; unless I am forgetting my electronics course the arrow in the [diode] symbol represents the anode and the straight line it points at is the cathode(?) IF that were the case, then as the schematic reads the negative terminal of the capacitor IS connected to the cathode. However, you have corrected me, and apparently I don't even know what little I thought I knew.
Hi Sir Trefor,
Your memory is correct. The diode is breaking down as the PIV is only 40 VDC. Your meter should have read some DC unless the diode finally shorted and you now have a wire that looks like a diode.
I normally only stock the highest voltage diodes, in your case this would be a 1N4007. I'd select a 1N4002 at the minimum, but why bother? 1N4007's are often the same price - or very, very close.
22 uF is fine for a 20 uF. The tolerances used to be +100%, -50%. These days they are often within 10 % either way. Replace the cap. It's been horribly abused.
-Chris
Your memory is correct. The diode is breaking down as the PIV is only 40 VDC. Your meter should have read some DC unless the diode finally shorted and you now have a wire that looks like a diode.
I normally only stock the highest voltage diodes, in your case this would be a 1N4007. I'd select a 1N4002 at the minimum, but why bother? 1N4007's are often the same price - or very, very close.
22 uF is fine for a 20 uF. The tolerances used to be +100%, -50%. These days they are often within 10 % either way. Replace the cap. It's been horribly abused.
-Chris
For a negative supply the cathode band faces the transformer secondary. For a positive supply it faces towards the load.. (In simple terms)
I use either 1N4007 or UF4007 diodes.
The 40V piv rating was far too low and it would have failed almost immediately regardless of orientation.
Your capacitor was wired correctly, if you measured a positive voltage at the cap's negative terminal relative to the chassis the rectifier is wired backwards.
Think of a tube rectifier, the operation of an ss diode is absolutely analogous. (Exclude filaments in tubes and minority carriers in ss diodes) The cathode is always the more negative electrode in either a ss or tube rectifier. The rectifying action is by virtue of the fact that the diode allows current to flow only in one direction. Reversing the diode results in a current of the opposite polarity. I'm trying not to confuse you or me, it isn't all that trivial to explain apparently..
All I am trying to say is in your negative bias supply the banded end (cathode) of the diode should face away from the negative terminal of your cap.
To avoid further confusion can you post the schematic or details of the actual supply circuits? (Errors sometime happened in old schematics, the diode might even be drawn facing the wrong way.)
I have posted a simple schematic of what I was talking about, hopefully it will be helpful.
I use either 1N4007 or UF4007 diodes.
The 40V piv rating was far too low and it would have failed almost immediately regardless of orientation.
Your capacitor was wired correctly, if you measured a positive voltage at the cap's negative terminal relative to the chassis the rectifier is wired backwards.
Think of a tube rectifier, the operation of an ss diode is absolutely analogous. (Exclude filaments in tubes and minority carriers in ss diodes) The cathode is always the more negative electrode in either a ss or tube rectifier. The rectifying action is by virtue of the fact that the diode allows current to flow only in one direction. Reversing the diode results in a current of the opposite polarity. I'm trying not to confuse you or me, it isn't all that trivial to explain apparently..
All I am trying to say is in your negative bias supply the banded end (cathode) of the diode should face away from the negative terminal of your cap.
To avoid further confusion can you post the schematic or details of the actual supply circuits? (Errors sometime happened in old schematics, the diode might even be drawn facing the wrong way.)
I have posted a simple schematic of what I was talking about, hopefully it will be helpful.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.