OK. About a month ago I reported on an issue with a Blues Jr that buzzed when coming from Standby to On. I did the Frommel mods to this amp for a buddy. One of these was to add a standby switch.
Now when you turn it on and let it sit a few seconds and then switch from standby to on after about 5 seconds it lets out a fairly loud burst of what sounds like 60 Hz for a few seconds.
Both he and I contacted Frommel and they eard of this issue. So I finally got him to bring the amp back for me to look at. Other than this issue the Frommel mods really improved the amp.
So I hooked my scope up across the speaker. We ha already troubleshot to discover that removing the V3 12ax7 tube which is the phase splitter makes theproblem go away. Anyway, when I captured it on the scope I was pretty amazed. The voltage at the speaker shows a burst of high frequency oscillation. Actually it is a series og high frequency bursts with a rep rate right around 50 Hz. Note 50 Hz, and this is a 60Hz ac system. So it is not AC ripple or anything.
The amplitude is huge. It is about 50-100 volts peak across the speaker! The speaker doesn't blow because what you hear is the 50 Hz rep rate of the bursts of oscillation. The high frequency is not reproduced by the speaker, or by my ears. I can get some screen shots of this, but I thought I would give people a chance to consider it.
My theory is that the phase inverter which is where the feedback is coupled back is somehow going into oscillation because of the sudden application of the B+ everywhere, after the tubes are warmed up.
So what would force the amp to oscillate for 4 seconds and then stop and behave normally? This only happens if you use the standby switch. If you just go from off to on it does not do this.
He also discovered that using his attenuator changes the way it sounds. We trid it with the attenuator at minimum attenuation, and th turn on oscillation burst was lower in volume and only lasted about 1 second instead of the normal 4 seconds.
Ok guys, I need some help here! This is a tough one.
Now when you turn it on and let it sit a few seconds and then switch from standby to on after about 5 seconds it lets out a fairly loud burst of what sounds like 60 Hz for a few seconds.
Both he and I contacted Frommel and they eard of this issue. So I finally got him to bring the amp back for me to look at. Other than this issue the Frommel mods really improved the amp.
So I hooked my scope up across the speaker. We ha already troubleshot to discover that removing the V3 12ax7 tube which is the phase splitter makes theproblem go away. Anyway, when I captured it on the scope I was pretty amazed. The voltage at the speaker shows a burst of high frequency oscillation. Actually it is a series og high frequency bursts with a rep rate right around 50 Hz. Note 50 Hz, and this is a 60Hz ac system. So it is not AC ripple or anything.
The amplitude is huge. It is about 50-100 volts peak across the speaker! The speaker doesn't blow because what you hear is the 50 Hz rep rate of the bursts of oscillation. The high frequency is not reproduced by the speaker, or by my ears. I can get some screen shots of this, but I thought I would give people a chance to consider it.
My theory is that the phase inverter which is where the feedback is coupled back is somehow going into oscillation because of the sudden application of the B+ everywhere, after the tubes are warmed up.
So what would force the amp to oscillate for 4 seconds and then stop and behave normally? This only happens if you use the standby switch. If you just go from off to on it does not do this.
He also discovered that using his attenuator changes the way it sounds. We trid it with the attenuator at minimum attenuation, and th turn on oscillation burst was lower in volume and only lasted about 1 second instead of the normal 4 seconds.
Ok guys, I need some help here! This is a tough one.
Where is the standby switch, in series with the HT or switching the screen grid off as it should.
It actually is between the hv transformer winding and the rectifier diodes, so it is switching ac not dc. So in standby the heaters are on and the main power tube bias is there. But nothing has any hv.
If you wire it in series with R32, you will get the difference. There will be no more charging main smoothers to give problems. Unusual to be unstable though. You mention 50 to 100volts HF noise across the speaker, you do realise that if that was the case then (8R secondary and 7kR primary is a turns ratio of 875:1) there must be between 4 and 875 thousand volts on the primary! Where can that come from as the HT is a maximum of 330Volts I wonder. Could it be the probe settings?
Yep
The amplitude doesn't make sense to me either.
Putting it in series with R32 still puts hv on the power tubes, which kind of defeats the purpose of standby altogether. Of course most people think standby is pretty useless anyway.
I am going to check my work and make sure I didn't crack a pcb trace when I replaced the caps. I will also look at the hv waveforms to see if there is anything interesting going on there.
At this point I will just be looking for something broken or done incorrectly, because a lot of people have made these mods with turning their BJr into an RF transmitter.
The amplitude doesn't make sense to me either.
Putting it in series with R32 still puts hv on the power tubes, which kind of defeats the purpose of standby altogether. Of course most people think standby is pretty useless anyway.
I am going to check my work and make sure I didn't crack a pcb trace when I replaced the caps. I will also look at the hv waveforms to see if there is anything interesting going on there.
At this point I will just be looking for something broken or done incorrectly, because a lot of people have made these mods with turning their BJr into an RF transmitter.
Here are some pics of the voltage across the speaker coming from standby to run.
[/url]IMG_1277 by Dennis Kelley, on Flickr[/IMG]
[/url]IMG_1278 by Dennis Kelley, on Flickr[/IMG]
[/url]IMG_1279 by Dennis Kelley, on Flickr[/IMG]
The oscillation frequency during the hash is about 57kHz. But it has a 50Hz rep rate at the beginning, eventually becoming continuous. All you hear is the 50Hz part of it.
The oscillation frequency during the hash is about 57kHz. But it has a 50Hz rep rate at the beginning, eventually becoming continuous. All you hear is the 50Hz part of it.
This was a long time after power up.
[/url]IMG_1280 by Dennis Kelley, on Flickr[/IMG]
It doesn't matter if you go through standby or not, as far as this oscillation. The only difference is when going from off to on you can't hear the first 3-4 seconds of the oscillation because the 50Hz part is already done.
But it is still oscillating at 57 kHz.
It doesn't matter if you go through standby or not, as far as this oscillation. The only difference is when going from off to on you can't hear the first 3-4 seconds of the oscillation because the 50Hz part is already done.
But it is still oscillating at 57 kHz.
across?
Across, as in plate to ? GND, B+, or does it not really matter?
One other thing, he also upgraded the output transformer before it got to me. It seems like that could cause some issues. I tried swapping the output transformer's output leads. Now that was some oscillation!!!!
I have read of putting a cap on the gate to cathode of the power tubes. That is pins 2-3 of the power tubes, so that would be pretty easy. It needs to have V3 to make it oscillate, but basically if you alter the phase or amplitude anywhere along the forward or feedback path it should fix it.
I have some hundreds of pF caps in 1000V so I may try them in the g-c first, since that is pretty common.
I tried putting one across the resistor divider in the feedback path (R16 in this schematic) , but it had no effect.
http://www.blueguitar.org/new/schem/fender/blues_jr.gif
I may try the caps, or maybe just open the feedback to make sure that is the source of the oscillation.
Across, as in plate to ? GND, B+, or does it not really matter?
One other thing, he also upgraded the output transformer before it got to me. It seems like that could cause some issues. I tried swapping the output transformer's output leads. Now that was some oscillation!!!!
I have read of putting a cap on the gate to cathode of the power tubes. That is pins 2-3 of the power tubes, so that would be pretty easy. It needs to have V3 to make it oscillate, but basically if you alter the phase or amplitude anywhere along the forward or feedback path it should fix it.
I have some hundreds of pF caps in 1000V so I may try them in the g-c first, since that is pretty common.
I tried putting one across the resistor divider in the feedback path (R16 in this schematic) , but it had no effect.
http://www.blueguitar.org/new/schem/fender/blues_jr.gif
I may try the caps, or maybe just open the feedback to make sure that is the source of the oscillation.
There is a 47pF cap across the plates of the PI tube already. Is that where you mean to put a cap?
Seems like a good place to put a cap to filter HF differential noise or oscillation doesn't it?
Perhaps changing the transformer increased the hf gain and made it oscillate?
It looks like the R of the RC filter is around 100K there, so a 500pF cap will give me a break freq of 20KHz. For a guitar amp it could probably be 10KHz without a problem, but I will start there.
Thanks for the reply. Will report back later.
Seems like a good place to put a cap to filter HF differential noise or oscillation doesn't it?
Perhaps changing the transformer increased the hf gain and made it oscillate?
It looks like the R of the RC filter is around 100K there, so a 500pF cap will give me a break freq of 20KHz. For a guitar amp it could probably be 10KHz without a problem, but I will start there.
Thanks for the reply. Will report back later.
Thanks a bunch Mozz, a treasure trove of info there.
What a loss that was to lose BillM. But then I guess he made some amazing contributions too, so it wasn't lost.
What a loss that was to lose BillM. But then I guess he made some amazing contributions too, so it wasn't lost.
I went to the BillM site and read up on bjr oscillations. I moved the cables as suggested and added a 150pF silver mica cap I just happened to have and guess what, no 50 Hz, no 54kHz, just awesome tube amp sounds!
Thanks again Mozz.
I remembered reading all about that, but at the time I wasn't aware there was an issue with oscillation. If we hadn't added the standby switch which allowed the amp to amplify the 50 Hz modulation we might never have realized there was an issue. I wonder how many of these amps are burning themselves up at ultrasonic frequencies, with nobody knowing the cause.
Thanks again Mozz.
I remembered reading all about that, but at the time I wasn't aware there was an issue with oscillation. If we hadn't added the standby switch which allowed the amp to amplify the 50 Hz modulation we might never have realized there was an issue. I wonder how many of these amps are burning themselves up at ultrasonic frequencies, with nobody knowing the cause.
I had loads of issues with those ribbon cables connecting to the tube socket board. And the soldering of the sockets to that PCB. Crackling sounds, and occasionally a similar 60Hz (or possibly 50 Hz) very loud output like you describe, on power up. No standby switch on mine. Got all the cable connectioms resoldered as well as the tube sockets to the PCB. Zero problems with it since then, about 9 months of regular use. If you haven't already redone those solder joints it maybe a good plan to do so.
- Status
- Not open for further replies.