Hoffman 5F6A build issue

[Apexelectric]

Been trying to sort out an issue with a new build. Got me a bit perplexed at the moment.

After firing int up for the first time without tubes everything checked out. After installing rectifier and the pre amp tubes everything checked out. After installing the output tubes the fuse blew and a bad solder joint on the output tube filament string surfaced. This was in standby mode. There was also some arcing occurring in the rectifier tube.

I fire everything up through my current limiter first with a solid state rectifier installed until I’m pretty sure that there are no issues then I install the tube rectifier and plug it in directly. Not sure why I would be blowing fuses in standby mode with the output tubes installed. No issues with the preamp section and all voltages test as would be expected. I’ve gone through the whole board, jumpers, tube sockets and controls multiple times with no luck in finding any bad joints, wrong values or  miswires.

Looking for a direction on which way to go now.

[tubenit]

arcing occurring in the rectifier tube

I could not tell if you have replaced the tube rectifier?  I've only had one go bad ever, but I think it would be worth replacing it or use your solid state plug in with tubes in (if you have not tried that already). I'd try different power tubes also if changing the rectifier doesn't help.

Since you're not finding anything else wrong, maybe try different tubes?

With respect, Tubenit

[bnwitt]

I think some photos of the tube sockets and the controls would be helpful

[Apexelectric]

I did test it he tubes after the problem arose. All new ones too. They all tested ok. Will post more pics later this afternoon.

[bnwitt]

How did you test the rectifier tube??

[mresistor]

Out of curiosity - does your 6.3 volt filament winding have a center tap? I'm assuming it does and you're using it as the 100 ohm resistors look unused.

[sluckey]

With rectifier only plugged in, measure voltages on ***ALL*** pins of both output tubes. Do this in standby mode and also operate mode. Post readings here.

[Apexelectric]

With rectifier only plugged in, measure voltages on ***ALL*** pins of both output tubes. Do this in standby mode and also operate mode. Post readings here.

Will do Sluckey.

Mresistor, correct on the center tap

[Apexelectric]

Sorry for the delayed response. I haven’t checked the tube rectifier but am using the solid state Weber replacement. Output tubes check ok. All preamp tubes installed and output tubes removed.

Voltage to transformer ground on output tube pins in standby as follows (same readings on both tubes)

1- 0
2- 3.3 VAC
3- varies as capacitor charges and drains
4- varies as capacitor charges and drains
5- -48.6
6- -48:6
7- 3.3 VAC
8- O

Voltages in operating mode are the same except for pins 3 and 4 which read 487 and 485 respectively.

[sluckey]

3- varies as capacitor charges and drains
4- varies as capacitor charges and drains

What oes that mean? There should be no voltage on pins 3 and 4 when in standby.

[Apexelectric]

Reading the stored voltage in the filter capacitors.

[sluckey]

Plug in your SS rectifier and the two power tubes. No little tubes. Connect a speaker. Turn the amp on. Does the fuse hold?

[Apexelectric]

Fuse holds that way. As well as with the tube rectifier.

[sluckey]

OK. Now plug in all the tubes with a speaker connected. What happens?

[Apexelectric]

Fuse blows but only after filaments warm up. Seems to hold ok in standby now even though I thought it blew a fuse before.

[sluckey]

Disconnect the negative feedback wire from the speaker jack. Fuse still blow?

Do you have a current limiter lamp bulb? If so, use it until the fuse blowing issue is resolver. If you don't have one, build it.

[Apexelectric]

The disconnect of the NFB lead cured the fuse blowing. Why?

Yes I have a current limiter that I use. Been using it during this process. Except for the last couple tests you requested me do.

[sluckey]

The disconnect of the NFB lead cured the fuse blowing. Why?

Because the phase of the NFB is wrong and causing your amp to oscillate. The fix is to swap the OT primary plate wires at the tube sockets pins 3.

[Apexelectric]

Guess I’ll store that one in the memory bank for the future. Thanks so much

[dude]

So, switching the OT plates, fixed the problem?

[Apexelectric]

Had to take off for a while. Will confirm the resolution later tonight or tomorrow?

Thanks again

[Apexelectric]

Swapped the OT phase conductors at the tube sockets and reconnected the NFB and got the out of phase squeal this time. Must be an issue in the NFB loop and not the phasing? There is a switchable NFB resistor option in the system.

[sluckey]

There is a switchable NFB resistor option in the system.

That's not standard on Hoffman's 5F6a. Tell us more. Schematic would be nice to understand what this mod does and how it is actually wired.

Can you easily bypass your mod and connect the NFB exactly as done on Hoffman's layout?

[Apexelectric]

It’s just a mod to have the option for two different resistor values for the NFB or none at all via a spdt switch.  I’ll put it back to the stock to see if that’s the issue.

[sluckey]

We may be on a wild goose chase. 

[Apexelectric]

Swapped the OT leads back and disconnected the NFB again. No more squeal. I get 7.4 VDC on the NFB lead going to the speaker output. Does that seem to make sense before I reconnect it again?

[sluckey]

Swapped the OT leads back and disconnected the NFB again. No more squeal. I get 7.4 VDC on the NFB lead going to the speaker output. Does that seem to make sense before I reconnect it again?

yes.

If I had known about your NFB switch mod I would have just suggested turn the NFB off.

Have you even tried to play guitar through the amp? What does the amp sound like?

[Apexelectric]

Haven’t plugged into it yet. I’ll get the NFB reconnected and try it. Let you know soon.

[Apexelectric]

Can’t quite figure out why the addition of the 10k NFB resistor option caused an issue but it did. Must have created a path I didn’t realize. Pulled the resistor and just left it with the 27k and all seems fine now. Amp sounds good through the shop test speaker. I’ll run it through some more paces when I get the speaker cab finished. Got started on covering the head cabinet though.

[Apexelectric]

Well I got the speaker cabinet done and had a chance to plug into it with the head and it sounded great at first and played well for a while but then after some hard power chords the volume dropped in half and the sound got tinny. After letting it rest the sound quality comes back until you start pushing it again.

I went through the board, tube sockets and controls and reflowed anything that looked remotely suspicious. No change. All the idle voltages seem to be in line and the output bias is set to only 30W dissipation. Can’t get it any hotter than that. Need to change the bias resistor to get the range proper.

Earlier in this process I did find that my rectifier tube was defective and was arcing then blowing fuses when the amp was taken off standby. The amp doesn’t blow fuses with the Weber rectifier. I tested the current draw of the amp and it only draws about 0.5 amps at idle.

Problem is the same for either bright or normal channels and all the tubes test ok.

What should I test next to try to isolate this issue.

[shooter]

I tested the current draw of the amp and it only draws about 0.5 amps at idle.

I might be out to lunch here, but, say each PA tube draws 100mA, 200mA PA, 50mA for all the small ones, that leaves 250mA somebody is sucking

[sluckey]

I'm gonna guess the .5A is on the PT primary.

[shooter]

I'm gonna guess

AH, makes sense, can't remember last time I measured AC current for a DC circuit

[sluckey]

I 'think' apex is an electrician. Probably used to measuring AC current.

[Apexelectric]

Yes, I’m used to measuring A/C. Trying to illustrate that the amp is not pulling anything close to what the fuse can handle and that it’s unrelated to the previous issue. 

I read 34mv across the 1 ohm resistors on the output tubes @440V.

Does this sound like a tube behaving oddly under load?

[sluckey]

Tubes should always be high on the suspect list because they are so easy to change. They need to be eliminated as suspects.

Take a full set of voltage readings when the amp is behaving properly. Then take another set of readings when it is acting up. Compare readings to see if there are any clues.

[Apexelectric]

Not sure if I’m going to be able to get the voltage readings on the amp while it’s not working properly due to the short time it takes to recover from the power loss issue. Was going to try replacing V2 and V3 one at a time as it seems to me to be somewhere in the preamp and not V1 due to the problem being the same on both channels. Just trying to check my logic with those of you who have more experience. Not to mention speed up the process of correcting the issue.

Would a bad cap produce these symptoms?

[MFowler]

Why does your filter caps ground wire go under the board rather to the left to ground near the PT?


Need artificial center tap two 100R resistors to ground off the filament supply wires.

[sluckey]

Bad cap could cause this. So could a lot of things, including bad solder connection. A scope would help you quickly determine where the signal goes bad. If you have one connect it to the preamp output (wiper of the treble pot). If the signal displayed on the scope drops in amplitude when the problem occurs, then the problem is in the preamp, otherwise the problem is in the power amp.

Go ahead and get a full set of voltage reading while the amp is working. This becomes your known good baseline. Then monitor a voltage at one of those points to see if it changes when the amp fails. Check it off the list if no change and monitor another point. Etc., etc...

[Apexelectric]

Ok. Just got a scope not that long ago and haven’t really had a chance to familiarize myself with it. Picked up a good used Tektronics model with a couple different probes that should be more than sufficient for this type of work.

I’ll give it a shot and check in again this weekend on what I find.

[shooter]

Earlier in this process I did find that my rectifier tube was defective and was arcing then blowing fuses

Bad cap could cause this. So could a lot of things, including bad solder connection

+1

due to the short time it takes to recover from the power loss issue.

set up the equipment 1st, then it's "monitoring" before it acts up. I like to "watch" with my scope and "record" with my meter for short duration issues

[PRR]

A missing grid-to-ground resistor may play OK at first soft sounds, then go way out of bounds on a Power Chord, then leak back to happy less or more quickly.

[Apexelectric]

A missing grid-to-ground resistor may play OK at first soft sounds, then go way out of bounds on a Power Chord, then leak back to happy less or more quickly.

Not 100% sure what you’re referring to. Are you talking about the grid resistors on the output tube terminals?

[sluckey]

Not 100% sure what you’re referring to. Are you talking about the grid resistors on the output tube terminals?

Probably not, since those grid resistors connect to the bias supply. Just measure resistance of all grids to chassis (meter connected directly to the socket pin) and compare to the resistance path on the schematic. For example, V1 pin 2 should read 33K if nothing is plugged into J1 or it would read 1M if a dummy plug is inserted into J1. No need to check V2 pin 7 since it connects directly to the plate of V2A. Even the output tube grids should measure more than 220K to ground.
 

[Apexelectric]

Probably not, since those grid resistors connect to the bias supply. Just measure resistance of all grids to chassis (meter connected directly to the socket pin) and compare to the resistance path on the schematic. For example, V1 pin 2 should read 33K if nothing is plugged into J1 or it would read 1M if a dummy plug is inserted into J1. No need to check V2 pin 7 since it connects directly to the plate of V2A. Even the output tube grids should measure more than 220K to ground.
 
[/quote]

Gotcha!

[Apexelectric]

Swapped V2 for V3 and that seemed to resolve the issue. Only had a chance to play it for 10 minutes but that was more than long enough to make it surface last time. Resistances on the grids seems to match up with what you discribed.

Will test it a bit more thoroughly over the next few days and report back.

Sill need to resolve the bias range. Last time I adjusted the bias to get it in the proper range on my last build, I swapped the smaller resistor after the pot, which worked fine.  Should I swap the 3W resistor instead? It would be easier to swap that one on the board. I’m maxing out at about 30W dissipation right now so I would need to bump the value up a bit.

[sluckey]

Changing either resistor will change the bias voltage. Decreasing the 1K/3W will increase the negative bias voltage which will cause the tubes to run cooler. But decreasing the 47K after the pot will decrease the negative bias voltage which will cause the tubes to run hotter. Sometimes you may have to tweak both resistors to get what you want.

I like to have the bias pot set near the middle of rotation be able to adjust above or below my desired bias point (usually 70% max idle dissipation). Sometimes the schematic values work out just right. But since most of the stuff I build uses donor transformers, more than likely I will need to tweak the bias resistors. Here's the procedure I use to tweak so the bias pot will be at the middle of rotation when I have my desired bias point...

Set the bias pot to provide the desired bias point. Sometimes the range may be off enough that I have to tweak a resistor just to get my desired bias point. Once I have the bias pot set to my desired bias point, I connect a voltmeter to the grid of one of the output tubes and write down the measured negative voltage. Now put the amp in standby mode, or pull the rectifier tube, or pull both output tubes. Turn the bias pot to the center of rotation and leave it there. Now start changing the bias resistor(s) until the negative bias voltage on the meter is the same as what I wrote down. Now turning the bias pot should adjust the bias voltage above and below my bias point. Finally set the bias pot so the meter reads the same as the voltage I initially wrote down. Recheck the bias current through the 1Ω cathode resistors and tweak slightly if required.

This little procedure will result in having the bias pot centered to provide your desired bias point and you will be able to adjust hotter or cooler.

[Apexelectric]

Thanks again for the help.

In what instance would you change both resistors instead of just one?

Also wondering if my bad solder joint on the filament string at one of th output tubes lead to my issues with the preamp tube and the rectifier?

If not then I need to figure out how to source more reliable tubes. Been buying TAD.

[Apexelectric]

Swapped the 3W bias resistor for a 3.6K and got the range to max out at about 60ma per tube. I’ll try to put some time in playing it later tonight.

[Apexelectric]

Got her all done and she sounds great!

Thanks again for the assistance!