PROGRESS!!!!!
Good call, Dave Ciqna:
I put the 22u cap on the V3 Cathode resistor and now I can turn the reverb all the way and no feedback!
Interestingly, putting the reverb driver tube back in brought back my fizzy distortion (including with the reverb all the way down). The gain has to be almost off to avoid it... What would V2 have to to with the fizz?
The high frequency is still there when MV and Gain pots are up (with or without the V2 installed)...
Note: John, I actually have the NFB disconnected right now and it is almost no different from having the presence turned up.
Good call, Dave Ciqna:
I put the 22u cap on the V3 Cathode resistor and now I can turn the reverb all the way and no feedback!
Interestingly, putting the reverb driver tube back in brought back my fizzy distortion (including with the reverb all the way down). The gain has to be almost off to avoid it... What would V2 have to to with the fizz?
The high frequency is still there when MV and Gain pots are up (with or without the V2 installed)...
Note: John, I actually have the NFB disconnected right now and it is almost no different from having the presence turned up.
Well, it's been a very successful couple of days. A while ago, I had changed the tube-driver cathode resistance to about 500ohms. Having seen something similar in the TUT books by Kevin O'Connor, I gave it a try on the reverb problem I had. It actually did sound a little better, so I left it. The problem with that is it made the stage much less resistive, so after having dealt with some of the other problems, when I put the tube back in, the V+ in the entire preamp was dropped to about 175V. Bumping that back up to about 2.2k fixed the V+.
Also, it occurred to me there was too much bass being amplified in the preamp. Unlike the Fender tone-stack, the Baxandahl stack I'm using doesn't attenuate/tighten the bass until the knob is rolled almost all the way off. By modifying coupling cap and cathode resistor values I managed to get rid of the fizzy sound entirely.
The final problem was the oscillation at about 15khz. The 1645 OT has 4 primaries. The 2 that I'm not using, I taped and stowed. Poking around with a screw-driver, I discovered I could make the oscillation happen by moving the shaft near those leads and one of the more sensitive connections. So I tried to move them to a less sensitive area...I thought. It made it worse. Finally, I just clipped them off. That helped. I also
shielded the connection that appeared to be most sensitive to the noise. That also helped. Unfortunately, because this layout is on fiber-board, it's not too easy to modify the layout. So some of this problem seems to be inherent in the layout (My next amp will be point-to-point). Turns out though, now that the rest of the amp is a bit more stable, by putting the snubber cap back on the load of V3b, I managed to kill the oscillation entirely.
Thankyou all for helping me get through this. It really helps the learning process.
Now I get to add the salt and pepper...
So as it sits, it's stable. As Kevin mentioned above however, If I clip the stages before the reverb, there are problems. I can't put the cathode cap on V1b or I get all my other problems back. So considering most of my clipping appears to be after the MV, the MV doesn't really serve a purpose. Is there a simple way to get more overdrive before the MV?
Also, it occurred to me there was too much bass being amplified in the preamp. Unlike the Fender tone-stack, the Baxandahl stack I'm using doesn't attenuate/tighten the bass until the knob is rolled almost all the way off. By modifying coupling cap and cathode resistor values I managed to get rid of the fizzy sound entirely.
The final problem was the oscillation at about 15khz. The 1645 OT has 4 primaries. The 2 that I'm not using, I taped and stowed. Poking around with a screw-driver, I discovered I could make the oscillation happen by moving the shaft near those leads and one of the more sensitive connections. So I tried to move them to a less sensitive area...I thought. It made it worse. Finally, I just clipped them off. That helped. I also
shielded the connection that appeared to be most sensitive to the noise. That also helped. Unfortunately, because this layout is on fiber-board, it's not too easy to modify the layout. So some of this problem seems to be inherent in the layout (My next amp will be point-to-point). Turns out though, now that the rest of the amp is a bit more stable, by putting the snubber cap back on the load of V3b, I managed to kill the oscillation entirely.
Thankyou all for helping me get through this. It really helps the learning process.
Now I get to add the salt and pepper...
So as it sits, it's stable. As Kevin mentioned above however, If I clip the stages before the reverb, there are problems. I can't put the cathode cap on V1b or I get all my other problems back. So considering most of my clipping appears to be after the MV, the MV doesn't really serve a purpose. Is there a simple way to get more overdrive before the MV?
Mostly, I don't have a clear picture yet how to size the transformers for these projects and the 1645 was the one in my TUT book that K. O'connor uses for amps similar to this one. If I remember right, it was also significantly cheaper than the Fender replacement. I'd love to hear them side by side, but the Hammond does sound good, so I don't have any complaints.
Oops. I jumped the gun. There's still fizz. It's different though. It's not a farting out crackly fizz. It's just a fizzy distortion that sounds like a really cheap distortion pedal played through a tiny amp...then attenuated...lol
Anyhow, the reason why I missed it before is because it doesn't change volume with the MV (even all the way down), so the real signal completely overpowers it when it's up, but at lower volumes it's definitely noticable. Interestingly, much like the oscillation I eventually defeated, this fizz starts at about 1 on the gain dial.
I'm not sure where to start on this one.
Anyhow, the reason why I missed it before is because it doesn't change volume with the MV (even all the way down), so the real signal completely overpowers it when it's up, but at lower volumes it's definitely noticable. Interestingly, much like the oscillation I eventually defeated, this fizz starts at about 1 on the gain dial.
I'm not sure where to start on this one.
Most of them are carbon film.
I've been doing a bit more experimentation:
The distortion apears to be occuring in the reverb driver circuit. The only thing that seems non-standard about it to me is the transformer. In TUT, there's an example of one of these circuits and the primary impedence of the reverb transformer is 10k. Mine is rated as 25k. Figuring I was getting too much gain/clipping out of it, I tried raising the cathode resistor to 10k, removing the bypass cap, lowering the signal from the previous stage... All I really accomplished is worse sounding (and less) reverb. Removing the tube gets rid of the problem...and the reverb of course.
I've been doing a bit more experimentation:
The distortion apears to be occuring in the reverb driver circuit. The only thing that seems non-standard about it to me is the transformer. In TUT, there's an example of one of these circuits and the primary impedence of the reverb transformer is 10k. Mine is rated as 25k. Figuring I was getting too much gain/clipping out of it, I tried raising the cathode resistor to 10k, removing the bypass cap, lowering the signal from the previous stage... All I really accomplished is worse sounding (and less) reverb. Removing the tube gets rid of the problem...and the reverb of course.
My guess is that the main problem is not the reverb driver transformer or anything major to do with the basic circuit itself, but simply layout, possibly (probably?) with more than one problem. The way to move forward from here is to examine all of your leads and find the one(s) that is causing the trouble. The easiest way is to put a scope on the output (speaker jack) and feed a cleanish sine wave into the circuit somewhere. You might start by feeding the signal into the phase splitter, correcting any problems from there to the output, and then work backwards to the input.
At this point I think you just need to play around with the wires and see what causes any change. Then you'll know where the problem(s) is. You need to pay close attention to all of your leads and what's near what, especially what runs right along side what, and especially leads that lead to grids or screens.
-- Dave
At this point I think you just need to play around with the wires and see what causes any change. Then you'll know where the problem(s) is. You need to pay close attention to all of your leads and what's near what, especially what runs right along side what, and especially leads that lead to grids or screens.
-- Dave
You've got too much gain
Hi,
While I do think that your problem is lead dress. If you posted a picture of the inside of your amp, maybe we could help.
But, in addition to lead dress, there are a number of elements in your schematic that surprised me...especially if you're going for a basic Fender kind of sound.
First, your Phase Inverter tube is a 12AX7. On my Fender Deleuxe Reverb Reissue, it's a 12AT7. That 12AX7 may have too much gain in that position. It may cause instability or bad tone in this circuit...especially if it is already receiving a signal that is too strong.
Which brings me to my next point, by bypassing the tone stack the way you did, you're going to have a wickedly strong signal that will probably overdrive the later gain stages in your amp. Are you intending to get a strongly overdriven/distorted tone?
In most amps (Fender, Marshall, Boogie, etc), the tone stack yields a a massive loss of signal voltage. Basically, whatever gain stage follows the tone stack is only there to make up the loss. Therefore, if you've bypassed your tone stack but left in the following gain stage, you've provided massive boost to your signal. In my opinion, You will overdrive everything downstream of your first gain stage. In your current setup, your 1M "gain" pot probably cannot be turned down low enough to prevent this from happening.
So, in addition to switching your PI to a 12AT7 (just pull the 12AX7 and stick in the 12AT7), you have to do something about having too much gain. You could (1) cut out the first gain stage (V1a) and the tone stack and insert your guitar at the top of the "gain" pot, or you could (2) cut out the second gain stage by disconnecting the 0.022uF cap from V1B and jumpering from behind the 0.0022uF cap on V1a all the way to after the now-lifted 0.022uF cap on V1b, or you could (3) insert a simple voltage divider in place of your tone stack...try 1M in series followed by 100K to ground...that may or may not be enough attenuation to simulate the loss of the tone stack.
If you eventually fully debug your amp and go to put your tone stack back in, note that it might not have the same overall loss of gain as the traditional Fender tone stack. Therefore, you might always be overdriving your amp...which could be cool...or not...depends upon what you're going for.
OK, that's two issues...the third thing that I saw is that the voltage rail supplying your Reverb transformer (Point E) is in common with the Point E supplying your V3b gain stage. I have little experience here, but I don't like that connection at all. Nope, not at all. That common point of connection means that you could have some feedback from the V3b stage back earlier to the the reverb stage. Kinda sounds like what your amp is doing?
In my Fender DRRI, the reverb transformer is attached instead to the rail supplying the screens on the 6V6...about 394V vs the 260V supplying all the pre-amp tubes. Putting it on the supply for the screens may reduce your oscillation problems by removing this feedback path.
Finally (whew), I noticed that the first coupling cap off V1a is 0.0047uF. A traditional Fender uses a much higher value here. I think mine is 0.047uF not 0.0047uF. Your amp will work just fine with your value, but it'll have less bass response (more like a Marshall). This absolutely isn't a problem...in fact if you're going for an amp that's going to have high gain, you want to cut the bass in this way. But, if you're going for a Fender sound, this isn't it.
Good Luck!
Chip
Hi,
While I do think that your problem is lead dress. If you posted a picture of the inside of your amp, maybe we could help.
But, in addition to lead dress, there are a number of elements in your schematic that surprised me...especially if you're going for a basic Fender kind of sound.
First, your Phase Inverter tube is a 12AX7. On my Fender Deleuxe Reverb Reissue, it's a 12AT7. That 12AX7 may have too much gain in that position. It may cause instability or bad tone in this circuit...especially if it is already receiving a signal that is too strong.
Which brings me to my next point, by bypassing the tone stack the way you did, you're going to have a wickedly strong signal that will probably overdrive the later gain stages in your amp. Are you intending to get a strongly overdriven/distorted tone?
In most amps (Fender, Marshall, Boogie, etc), the tone stack yields a a massive loss of signal voltage. Basically, whatever gain stage follows the tone stack is only there to make up the loss. Therefore, if you've bypassed your tone stack but left in the following gain stage, you've provided massive boost to your signal. In my opinion, You will overdrive everything downstream of your first gain stage. In your current setup, your 1M "gain" pot probably cannot be turned down low enough to prevent this from happening.
So, in addition to switching your PI to a 12AT7 (just pull the 12AX7 and stick in the 12AT7), you have to do something about having too much gain. You could (1) cut out the first gain stage (V1a) and the tone stack and insert your guitar at the top of the "gain" pot, or you could (2) cut out the second gain stage by disconnecting the 0.022uF cap from V1B and jumpering from behind the 0.0022uF cap on V1a all the way to after the now-lifted 0.022uF cap on V1b, or you could (3) insert a simple voltage divider in place of your tone stack...try 1M in series followed by 100K to ground...that may or may not be enough attenuation to simulate the loss of the tone stack.
If you eventually fully debug your amp and go to put your tone stack back in, note that it might not have the same overall loss of gain as the traditional Fender tone stack. Therefore, you might always be overdriving your amp...which could be cool...or not...depends upon what you're going for.
OK, that's two issues...the third thing that I saw is that the voltage rail supplying your Reverb transformer (Point E) is in common with the Point E supplying your V3b gain stage. I have little experience here, but I don't like that connection at all. Nope, not at all. That common point of connection means that you could have some feedback from the V3b stage back earlier to the the reverb stage. Kinda sounds like what your amp is doing?
In my Fender DRRI, the reverb transformer is attached instead to the rail supplying the screens on the 6V6...about 394V vs the 260V supplying all the pre-amp tubes. Putting it on the supply for the screens may reduce your oscillation problems by removing this feedback path.
Finally (whew), I noticed that the first coupling cap off V1a is 0.0047uF. A traditional Fender uses a much higher value here. I think mine is 0.047uF not 0.0047uF. Your amp will work just fine with your value, but it'll have less bass response (more like a Marshall). This absolutely isn't a problem...in fact if you're going for an amp that's going to have high gain, you want to cut the bass in this way. But, if you're going for a Fender sound, this isn't it.
Good Luck!
Chip
The gain in this amp is a little odd. I would expect to get some good clipping before all the instability. But the moment I clip the V1 even a little, the fizz comes back. I think the output of V1b is too hot, but I was under the impression that I could get pre-amp distortion out of this configuration. For some reason though, it appears there’s instability if I overdrive anything before the PI.
You’re right about the tone-stack having different characteristics. The bass is pretty massive. That’s why the small coupling cap. I’ve reattached it because I’m pretty convinced at this point that it’s not the problem. I’ll probably add a voltage divider anyway because there’s very little loss - much less than the Fender stack.
The point where the reverb transformer is power is actually not the same as V3b… Sorry: that’s a drawing error. There’s actually one more filter step I devoted to entirely to the reverb transformer.
I finally decided to get myself a signal generator and received it today. I tracked down the point where I first find signal that shouldn’t be there. It’s at the output of the PI (nothing at the grid). The rail is clean (point D). I’m wondering if the 0.1u caps might be picking it up. Is there a type of cap that is less parasitic I could try?
By swapping the PI 12AX7 for a 12AT7 parasitic signal is cut about in half.
When I install the reverb tube, the parasitic signal is distorted. The wave has a smaller inverted signal near the peak rather than being flattened.
I’ll try to post a pic of the layout soon.
You’re right about the tone-stack having different characteristics. The bass is pretty massive. That’s why the small coupling cap. I’ve reattached it because I’m pretty convinced at this point that it’s not the problem. I’ll probably add a voltage divider anyway because there’s very little loss - much less than the Fender stack.
The point where the reverb transformer is power is actually not the same as V3b… Sorry: that’s a drawing error. There’s actually one more filter step I devoted to entirely to the reverb transformer.
I finally decided to get myself a signal generator and received it today. I tracked down the point where I first find signal that shouldn’t be there. It’s at the output of the PI (nothing at the grid). The rail is clean (point D). I’m wondering if the 0.1u caps might be picking it up. Is there a type of cap that is less parasitic I could try?
By swapping the PI 12AX7 for a 12AT7 parasitic signal is cut about in half.
When I install the reverb tube, the parasitic signal is distorted. The wave has a smaller inverted signal near the peak rather than being flattened.
I’ll try to post a pic of the layout soon.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.