Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Archives of favorite topics => Topic started by: gregolson65 on December 31, 2007, 07:45:51 am
-
I like to ask the very smart folks here about powering up an built amplifier for the first time. This is more to solve some lessons learned and get some straight answers from the techs. I guess I have been lucky in my first attempt to power one on but reading from this tech here: http://paulrubyamplifiers.com/info.html#FirstPowerUp
it seems I did things the wrong way. Is the article correct in saying taking readings with the transformers disconnected? I would assume you take readings when you have everything connected but I guess I am not an expert at this.
The real question is.
What is the easiest way to power up an amplifier for the first time without the chance of burning your circuits or transformers up. It seems the article explains way too many things to do to check your amplifier build. Just my opinion only. Nothing against all you high tech guys at all. that is why I am asking.
-
Actually I use a Light bulb current limiter (http://www.el34world.com/Forum/yabb2/YaBB.pl?num=1168723173). First with the tubes out and check all the voltages knowing they will all be lower than normal. Then I put the tubes in and use the same procedure.
** Quite a bit of information on light buld current limiters here: http://ampgarage.com/forum/viewtopic.php?f=6&t=20341
-
Paul Ruby's startup guide seems like a good place to start ;) You might also check out the "Initial Testing" and " Final Test" on the Ampmaker site: http://www.ampmaker.com/pp-18-final-testing-1045-0.html. This is for a specific 18-Watt clone, but the basic ideas are solid IMHO.
Chip
-
I'm just a casual hobbyist, having built only 6 amps in the past 30 years, 5 of which were built in the last 2 years, as a result of renewed interests. I also recommend the current limiter for most people. I have never used one and don't have plans to use one, but as said, I'm just a casual hobbyist. There are plenty of good methods for powering up and Paul Ruby's is one of them. I haven't read it, but I've heard it referenced many times.
Here's what I do... There is never only one 'big power on' event. Instead, I [highlight]power up in stages [/highlight]as the build progresses. LOOK, LISTEN, FEEL, AND SMELL during all phases of the build when power is applied!
PHASE 1. The PT primary circuit is the first thing I wire up. As soon as the power cord is in place (but before the secondaries are connected) I plug in for the first time. I make sure the power switch, fuse, indicator light, etc. work and that the unloaded secondary voltages are as expected.
PHASE 2. Next I wire the 6.3v filament circuit. (If you're fond of the Fender style overhead filament string, this step would come later.) Then power on again and measure voltage across the filament pins of each socket. All tubes are then plugged in and visually checked for glowing filaments. REMOVE ALL TUBES.
PHASE 3. Rectifier tube is wired (nothing else). Check for proper filament and HT voltages on the recto socket. Then plug in the recto tube and check for B+ at the cathode. There are no filter caps at this time, so readings will be low. At this point the entire PT circuit has been checked, except for a possible bias winding that's still flying loose.
PHASE 4. Wire in the board(s), pots, input/output jacks, etc. IOW, complete all remaining wiring. Use the schematic and an ohm meter to verify that grid and cathode resistances measure expected values to ground and that plates are correctly connected to the proper B+ nodes. Visually trace the schematic (not the layout) and compare to your actual build. Verify polarity of ALL electrolytic capacitors and bias diode.
PHASE 5. No tubes in yet. If you have a fixed bias amp, power up and verify proper negative bias voltage is present on the control grid pin of every output tube socket. If adjustable, be sure the range is adequate and then set it for max negative voltage. DON'T PROCEED UNTIL THIS IS RIGHT!
PHASE 6. Plug in the rectifier tube ONLY. Power up and check for plate voltages on every empty tube socket. They will be high.
PHASE 7. Plug in all tubes and connect a speaker. Power on and see how it sounds. If you've successfully gotten this far, there shouldn't be too many surprises. Watch for red plates, smoke, etc, and listen for howl, hum, buzzes, bacon frying, etc.
If this is a NFB amp, there's a 50/50 chance you may need to reverse the OT plate leads. There may be an obvious howl, or you may just have a bad or weird sound, possible squeal/howl on certain notes, the kind of things you can't explain. Just reverse 'em. Hey, you may find out that it was right to begin with, but at least now you'll know for sure. I recommend not trimming the OT plate leads to length until you know for sure.
Play guitar for a short time. Then measure plate, grid, and cathode voltages for all tubes. Measure on the socket.
That's basically what I do. You may have to adapt details for different amps, but the idea is to [highlight]power up in stages[/highlight]. Divide and conquer. You'll have confidence built up for that final complete power up.
EDIT... Highlighted an important phrase. Corrected spelling errors (I hope).
-
> it seems I did things the wrong way.
There's no "right" way. No matter how you do it, there is always some way it can be screwed-up and you won't know until disaster strikes.
> Is the article correct in saying taking readings with the transformers disconnected?
There's parts that are not clear on first reading. Maybe not on second reading.
He's taking you through a logical sequence. Do you really have 120V inside the chassis? Can You Turn It OFF?? OK, energize the PT with secondaries disconnected. Do you get 6V-7VAC on the 6.3V windings? Etc, etc, etc. Skip a step, that's "probably" (Murphy's Law) the place you put the right wire in the wrong place.
> What is the easiest way to power up an amplifier
"Try it" is easiest way to power-up.
If "try it" turns to smoke and flame and an unknown number of ruined parts, it is NOT "easiest". Debugging is HARD WORK. Easily the hardest thing about electronics. If you have built 999 identical amps correctly, odds are that your 1,000th will also fire-up fine. But if you just puzzled your way through a new-to-you design, odds of it being perfect the first time are nearly zero. Odds of it being imperfect enough to smoke and burn.... I dunno. Only a few mistakes are DISASTER, but they stand out.
> It seems the article explains way too many things to do to check your amplifier build.
There are MANY-MANY-MANY-MANY ways to build an amp wrong. I've had years of experience, but hanging around here I see some mistakes that just dumbfound me. They are not stupid, or not stupider than my many mistakes. Places where you would not even think you could get it wrong, until you do it wrong with ease.
Do you want to check 99% of the more likely goof-ups beforehand?
When I build my 1,200V SE 813 amp, with more custom iron than I can lift, more than I care to buy twice, I sure as shooting will do a LOT of pre-flight checks.
> I usually get a long stick and flip the on switch from far away
I've also done that. What's the worst can happen? Set off the smoke detector, bring the fire trucks down on me? The hassle of that makes $50 of smoked-iron seem trivial.
It's not rocket science. It is unlikely to kill spectators.
A Lamp Limiter is good for "Let's just GO!" fire-ups. Then "the worst can happen" is probably a half-bright lamp and a warm amplifier which may not smoke for an hour. Beats "clik-BOOM!!".
My own plan is a reversal of the one you cite. I will often start with the power supply, even feed 6VAC into the 120V input to see if I get 1/20th of expected voltages before a full 120V test. Maybe wire one heater and see if a tube lights right. Then stages one by one. Since I have test gear, I may wire up the Power Bottle alone and test with signal generator and dummy-load to verify that I'm getting the power output I expect.
Like building a house. At each floor, you walk on what you just did, before you move on to the second floor. In fact every beam and stud, you step/lean on it in the process. If any piece feels goofy, or falls-off, you fix that before you move on. Electronics is different, because you can't feel it. But after a few pieces are "up", you can stop and test things as you go along.
-
> If this is a NFB amp, there's a 50/50 chance you may need to reverse the OT plate leads.
On "most" guitar amps with NFB, start with NFB disconnected. Amp should come up, play OK, be a bit too high-gain and a bit too brash. It shouldn't suck, or smoke, or burn.
If it "works", then connect NFB. If you understand the plan, use "too much" NFB. If there's 47K from OT to driver, tack 22K. Wear earmuffs. Power-up but keep finger on switch. 50:50 it howls, kill it, swap leads, try other way. When you find a way which is dead-stable, but a bit "tame" due to "too much" NFB, go to your target NFB value. This won't be your final value, it may want to be different depending on your speakers and style.
This works because most guitar amps have "very little" NFB. It's a tuning trick, not essential to the way they work.
This is harder to do with Hi-Fi (and early Sunn) amps. They are designed with a LOT of NFB. They may be nearly unusable without NFB. Like 20 times more gain than you know what to do with. It can be done, with extra shielding around input and careful measurements. Sometimes it helps to start with light NFB, 330K instead of 47K, then see if 220K makes gain go up or down.
Most transistor amps are hopeless without their NFB. Fortunately most don't use transformers so NFB polarity can be determined by inspection. That's not the end of the "fun", but not our concern here.
NOTE: this is from a later post of PRR's about start up of an amp and checking the fixed bias negative voltage.
different ways of achieving a similar outcome?
That's what I see. Different ways to skin the cat.
Hoffman's connection will work. Fender's is more common. You also see the "pot" wired as a 2-terminal rheostat.
Pick a plan and follow it. Don't put tubes in. It would be good to use a Lamp Limiter.
Stand back, plug in, hope for no smoke.
Turn OFF.
Put rectifier tube in.
CAREFULLY use clip-leads to connect a voltmeter to ground and the first filter cap (or standby switch).
Stand back, plug in, hope for no smoke. If no excitement, read the meter, and shut-down.
Because there is no load, the normal DC voltage (probably 440V) will read high (maybe 500V). Just confirm it happens.
Un-plug!!!
CAREFULLY use clip-leads to connect a voltmeter, Black to ground and Red to the point where the 220K resistors come together.
Plug-in, turn on. This bias voltage should go to NEGATIVE 40V or 50V, and it should be adjustable with the trimmer. Don't linger, the unloaded voltage may be hard on the filter caps. But you MUST be sure you have LOTS of NEGATIVE bias voltage, before you put tubes in.
If you can't get good bias, come back.
Note which way gave the most bias voltage, and leave it set that way.
Now you can put one 6L6 in, monitor the 1 ohm cathode resistor under it. About 10 seconds after turn-on, it should rise from zero to maybe 20 milliVolts (mV). The general goal is to get both tubes up to 40mV or 50mV, but the StartUp Goal is to see both 6L6 tubes working "a little". I like to start one at a time, verify I can turn each one below 20mV and up towards 40mV. And no horrible noise from the speaker.
At that point you can load the whole amp up, see if it plays, then trim the bias for around 30mV-40mV in each 6L6. Let it run some hours, burn the funk off the tubes, watching for any severe bias change, before you get fussy about bias.
Back to top
-
Along theses lines, I totally fried an expensive KT-88 power tube and may have damaged its companion which severely red-plated due to lack of negative fixed bias in an older amp. I was tricked! The old bias voltage caps must have been bad, but charged after a while. So, if I left the amp on for a while, and then got around to testing bias voltage it was there. But from a cold power-on, there was no bias voltage for a while - long enough for the power tubes to fry.
So, now I check to see that there is bias voltage first, immediately upon turn-on, with no delay. I even use 2 meters, one for ea tube socket, to avoid more unpleasant surprises. Then I check the voltage on the other pins.
-
Re: solid state amps.
I usually start with the driver board and feedback the output into the differential amplifier. I set the output bias to as low as it will go.
Once the driver board is proved i then connect the output transistors and set the bias to get rid of cross over distortion.
-
From the Amp Garage when someone asked about starting up a Dumble amp for the first time.
1) Right before start-up I go through and check all my grounds 1 more time on each cap in the Power supply.
2) Then meter out the + side of each cap make sure there are no dead shorts to gound.
3)Then do the same for the bias supply and the rectifier diodes for the B+ make sure I don't have a shorted diode.
4) From there if you like you can move to the preamp and check those grounds as well.
5) You are then ready to bring it up slow with a variac and a bulb limiter. I generally attach my meter right off the plate supply cap and monitor that on up. After I reach about 100 Vdc or so there I will go ahead and meter out each node again.If everything looks even. I'll go ahead and bring everything up again monitoring the plate cap supply. Once everything is full up and you have a dim bulb and everything looks OK.
6)Adjust the bias supply to max - voltage (I generally like to see around -65/-70 V) from there you can load in the preamp tubes make sure they light up and check all the voltages again.
7) Now you're ready for the output tubes and the smoke test. Make sure you have a load (speaker or load resistor) on the amp. 8) Pop in the tubes bring down the negative voltage until the tubes ae properly biased.
9) If the amp makes a great big squeal noise you need to reverse the output transformers primary wires.
10) Make one more sweep of all the voltages if everything is in spec or close.
11) Plug in a guitar bring the vol up slow pray the thing works. :lol: