Need pro opinion on some fine tuning my build

[hesamadman]

Hey all,


Ive been tinkering with this 30 watt for some time now. Basically its an ac/30 power amp with a 2204ish pre amp. Here are my issues.


The amp sounds ok on the higher frequencies for the most part. But the Low e up to the first A note sounds incredibly muddy. Almost muffled. On top of that...the type of distortion I have is rather fuzzy and not what I am wanting to get. I wanted to what see we could figure out to fine tune it.

I should note....the resistors are flame proof metal oxide. They are being used in plate resistor locations also which means they see the B+. Maybe these are not rated for that voltage?


Heres one thing....and I know that this is a whole other topic BUT the resistors that I used....are all metal oxide resistors from parts express. I have a ton of these for cross over builds for speaker cabinets. Could these use of these be a reason for the undesired tone? Or is this really such a matter?  Other than that....everything is either from hoffman. I think I got a dual 16uf cap for the power supply offline somewhere.

[Willabe]

Can you post a schemo of the power supply (PSU) with B+dcv's nodes and all tubes voltages too?


                  Brad   

[hesamadman]

Ill measure the DCV tomorrow 

[Willabe]

How many H's is the choke?


               Brad   

[hesamadman]

How many H's is the choke?


               Brad   

Alrighty...so.


I have:
350vdc after rectifier
350vdc after choke
at [4] 16uf filter cap 320vdc
at [5] 8uf filter cap 320 vdc
at [3] 8uf filter cap 310 vdc (10 less....which is odd to me....same value resistor same power source as the other 8uf filter cap)


The choke is a 5 henry....ac30 reissue

[hesamadman]

still getting tube voltages

[VMS]

First thing I would try is decreasing the value of the V2-a cathode cap. You could start from 100nF and increase the value until it sounds good. Some sort of decade box would be handy here.

If the the first mod doesn't work then I probably would decrease the coupling caps to power tubes. What value is that master volume?

[VMS]

Third thing to consider is that AC30 type power amp needs only 1/4 signal for full power compared to 2204 power amp. So adding a 1M resistor between treble pot wiper and volume pot might help.

Grid stopper on V2-a might help on the fizzyness, maybe start with 220k and decrease to taste.

Do one mod at a time.

[DummyLoad]

some NFB would help tighten up the LF.


--pete

[hesamadman]

I made a mistake when posting my schematics. I reposted them. The pre showed MV at tone stack and output showed it as a PPIMV. It IS a PPIMV dual 500k. No longer at the tone stack.

Can you post a schemo of the power supply (PSU) with B+dcv's nodes and all tubes voltages too?


                  Brad   

Here are the voltages. Originally I was thinking it could have been a problem in the power section as I know that can add muddiness. Let me know what you think.


V1:
1= 248v
3=2.8v
6= 185v
8= 1.4v


V2:
1= 237v
3= 2.2v
6= 315v
7= 237v
8=237v


V3:
1= 240
2=43
3= 65
6= 255
7= 46
8=65


V4:
3= 10.7v
7= 345v
9= 345


V5:
3= 10.7v
7= 345v
9= 345v


V6:
3= 10.7v
7= 345v
9= 345v


V7:
3= 10.7v
7= 345v
9= 345v

First thing I would try is decreasing the value of the V2-a cathode cap. You could start from 100nF and increase the value until it sounds good. Some sort of decade box would be handy here.

If the the first mod doesn't work then I probably would decrease the coupling caps to power tubes. What value is that master volume?

I had done both of these. Reduction of CC to el84's helped some.

Third thing to consider is that AC30 type power amp needs only 1/4 signal for full power compared to 2204 power amp. So adding a 1M resistor between treble pot wiper and volume pot might help.

Grid stopper on V2-a might help on the fizzyness, maybe start with 220k and decrease to taste.

I did the grid stopper with a 220k. Seemed to shape the overdrive a little better. Could you explain the 1m resistor again. I messed up the schematic when I posted. I no longer have a MV following the treble.

some NFB would help tighten up the LF.

I havent done the NFB yet. Not really sure where to start there.


After doing these changes my problem is less but still there. One way to describe my problem is, if I play say an E (one octave higher than the open E string. Or even an A (an octave higher than the open A string) I get the note its suppose to produce......but I almost feel as if I am getting a very faint low end tone of that note. Maybe even an octave lower than the note being played. Thats the only way I can describe it. It doesnt sound like an octave pedal by any means but it just have some slight low end to notes that shouldnt have it. Ill see what you guys say then I will post some audio clips later.

[shooter]

try gator clipping another 220uf in || with the cathode cap on the PA section, also try raising the 100r G2's to 470r.  fwiw

[hesamadman]

try gator clipping another 220uf in || with the cathode cap on the PA section, also try raising the 100r G2's to 470r.  fwiw

I dont have any 3w 470's..... would I get by for testing purposes if I used 1w 470 for a short time?

[terminalgs]

 regarding resistor voltage rating: of course you shouldn't exceed the max voltage rating of your resistors, but that rating is from one end of the resistor to the other (not from B+ side to ground).

your MV setup doesn't need the 220K grid-leak resistors after it.  What you get at Max MV position is 500K||220K, or 152K to ground.  That resistance  influences the amp in several ways: (#1) high-pass filter with preceeding CC, (#2) grid leak for the el84s, and (#3) signal load for the preceding triode stages (the PI).   152K is too low for any of these, although you mitigated #1 with a .1uf CC.  I'd remove the 220K grid leaks,, or the MV. #3 or #2 might be the fizzy and muddy bottom end.

[hesamadman]

   I'd remove the 220K grid leaks,, or the MV. #3 or #2 might be the fizzy and muddy bottom end.

That makes sense. I'll remove those and see what happens.

So basically I have some serious grid leak going on. Since the MV should have taken place of the original grid leaks?

Ultimately a lot of signal loss too correct? Leaking some highs. Could be why my tone is almost dead sounding.

[VMS]

I did the grid stopper with a 220k. Seemed to shape the overdrive a little better. Could you explain the 1m resistor again. I messed up the schematic when I posted. I no longer have a MV following the treble.



After doing these changes my problem is less but still there. One way to describe my problem is, if I play say an E (one octave higher than the open E string. Or even an A (an octave higher than the open A string) I get the note its suppose to produce......but I almost feel as if I am getting a very faint low end tone of that note. Maybe even an octave lower than the note being played. Thats the only way I can describe it. It doesnt sound like an octave pedal by any means but it just have some slight low end to notes that shouldnt have it. Ill see what you guys say then I will post some audio clips later.

1M resistor suggestion was just to make a voltage divider with the volume pot to reduce the signal that hits the power tubes, since EL84 tubes needs much less signal compared to EL34 tubes.

Have you chopsticked the wires to see if that makes a difference to the distortion? Specially the first volume pot wires shouldn't be too close to the tone stack wires.

[2deaf]

Too much bass causes the symptoms you describe.  Attached is a suggested circuit that limits bass.  Also, turn the bass control in the tone stack way down or off when overdriving.

[2deaf]

Attached is the standard NFB circuit for these kind of amps.  The PPIMV renders NFB less effective as you turn it down.

[hesamadman]

Attached is the standard NFB circuit for these kind of amps.  The PPIMV renders NFB less effective as you turn it down.

Thanks a ton! What value resistor do you suggest in the NFB?


Also....removal of the 220k grid leak resistors was a HUGE improvement!

[VMS]

Soldano atomic 16 is pretty similar to your design so maybe you can take some inspiration from that amp. You can keep the power amp cathode biased.

[2deaf]

What value resistor do you suggest in the NFB?

The resistor hooked to the speaker is typically ten times as big as the resistor hooked to ground for this size amp.  Soldano uses a 4.7K hooked to ground because of the presence control he uses, so he put a 47K to the speaker output.  You probably wouldn't put a presence control in because the PPIMV ruins it, so you have a wide array of values to choose from.

Since you have the PPIMV that you probably don't put on 10 very often, you might want to go with a lower ratio to compensate for decreased feedback at lower volumes.  Maybe 4.7K / 1K  or 27K / 4.7K  or whatever you have on hand that gives around a 5 : 1 ratio.

[terminalgs]

Thanks a ton! What value resistor do you suggest in the NFB?

Depends 100% on the output transformer secondary that you get the signal from. since the Vrms is different for each sec. tap (16/8/4/2ohm)  you need to calculate the two resistors (one is usually a fixed value in the PI cathode circuit) that form the NFB voltage divider based on the Vrms, the open loop gain of the output stage, and the amount of NFB you want to supply.  It's not too hard to figure:


http://www.aikenamps.com/index.php/designing-for-global-negative-feedback


Also, part of the Vox magic is the absence of an NFB loop, so make it switchable and then decide. 

[hesamadman]

Also, part of the Vox magic is the absence of an NFB loop, so make it switchable and then decide.

After doing some research and reading about Brian mays setup along with some videos, I have decided NFB is not what I need for this build. :)

[hesamadman]

Have you chopsticked the wires to see if that makes a difference to the distortion? Specially the first volume pot wires shouldn't be too close to the tone stack wires.

Do you mean just move some wires around in the chassis? I haven't really. I have like wires routed together though. Cathodes together. Ect. I don't really have any noise thought. With the 1m resistor, since I do not have a volume at the tone stack, should I just put one off the treble in series with the 220k headed to PI?

[VMS]

Do you mean just move some wires around in the chassis? I haven't really. I have like wires routed together though. Cathodes together. Ect. I don't really have any noise thought.

Yes. When you have two gain stages between signal wires those signals are in-phase and can feedback. I had this problem in my higain amp and when I turned the gain up (first volume in your amp) the feedback was canceling the signal and it didn't get any louder just buzzier.

With the 1m resistor, since I do not have a volume at the tone stack, should I just put one off the treble in series with the 220k headed to PI?

No, it doesn't really help if you don't have something that goes to ground.

In atomic there is 470k and 150k voltage divider after the tone stack but I'm not sure if this will do any good in your amp.


What settings are you using when you play? Do you get the overdrive sound from your preamp or do you crank the master volume and get the overdrive sound from your power amp?

[hesamadman]

Everything you guys suggested has been a huge help. I did have one question. As far as wire routing goes, I have all cathodes if v1 and v2 routed and zip tied together. Is there any reason I shouldn't do this? Keeping my wires neat on this build is a slight pain due to the large chassis. Every know and then a low frequency motor boating will occur and I'm trying to make sure it's not in my routing. I have a switched 25k mid pot that has a slight mid boost mod on it. I THINK that pot is faulty, as it seams low quality but it's all I could find. But I'm looking into further possibilities.

[ac427v]

I'm scratching my head and trying to understand the MV. It seems that as you turn the shaft clockwise, the output from V3a would get quieter but the output from V3b would get louder. That would cause interesting but unpredictable phasing and output variations. Do you have a photo of the MV pot wiring you can post?
Craig

[sluckey]

I'm scratching my head and trying to understand the MV.

Does pic this help?

[ac427v]

Steve, your picture works for my brain. Without your dual pot arrows going the opposite directions, I would mistakenly wire the right terminal from one pot to ground...and the left terminal from the other pot to ground. Creating the loss of signal and dead tone described in the first post. I get easily confused transferring a schematic of a dual pot into real components and wires.
Craig

[sluckey]

I get easily confused transferring a schematic of a dual pot into real components and wires.

Me too! Not just dual pots either. Every time I see an unfamiliar tone pot I have to do a big brain exercise to figure out which way to physically connect the pot lugs. And then I have to think in reverse notation if I'm looking at the back of the pot! Sometimes I get it right. Lately I've just resorted to tack soldering the outer lugs and just accepting that I will do the final solder joint after I know if it's right or wrong.   

[jojokeo]

madman, you've got a LOT of big coupling caps going on that could very likely be lowered starting with replacing the .1uf's after the pi to .022uF's. This will help a lot with low end woofiness. I'd lower the caps down to .01uf or .022 prior to the pi also. Lastly, on the 2nd triode labeled v1a, loose the 10k resistor (very cold bias not what you want). Put a 1k in there w/out bypassing and see how it changes your gain & tone. Then try a 1uF or 2.2uF bypass cap and see the gain increase this creates. You may like it? since you're not bypassing the cathode follower. Everything else should be fine at the moment not messed with until you try the above tweaks. More following tweaks can then be made (but are minor in comparison) if you want a bit more high end or want to shave some off.

(BTW - terminalags' tip on the 220k resistors is what helped you the most with your initial issues I'm sure.)

[hesamadman]

madman, you've got a LOT of big coupling caps going on that could very likely be lowered

Thanks. I will try the resistor swapping. For the most part I am happy with the sound after removing the unnecessary grid leak in the output.


I did find out that the resistors I am using are rated for 300v. I have about 250vdc on my plates. I am curious as to what affect these resistors have on the amp if they are being nearly maxed out. Since they are not really costly I would assume that thy have a rather large tolerance. So I guess the question is, when a resistor is being maxed out voltage wise, would it increase the chance of unwanted noise. Where I am now is trying to quiet the amp up a tad bit so Im looking into options there.


~minus the power supply and power amp. All resistors are sufficient.

[sluckey]

I did find out that the resistors I am using are rated for 300v. I have about 250vdc on my plates. I am curious as to what affect these resistors have on the amp if they are being nearly maxed out.

The voltage rating of a resistor refers to the voltage ACROSS the resistor. For example, in a typical gain stage you may have a plate voltage of 250v and a plate supply voltage of 350v. The voltage that is felt ACROSS the plate resistor would actually be 350-250=100v. So, your resistors are probably not being used near max rating.

 

[hesamadman]

I did find out that the resistors I am using are rated for 300v. I have about 250vdc on my plates. I am curious as to what affect these resistors have on the amp if they are being nearly maxed out.

The voltage rating of a resistor refers to the voltage ACROSS the resistor. For example, in a typical gain stage you may have a plate voltage of 250v and a plate supply voltage of 350v. The voltage that is felt ACROSS the plate resistor would actually be 350-250=100v. So, your resistors are probably not being used near max rating.

Aaaah. Got ya! 

[jojokeo]

madman, you've got a LOT of big coupling caps going on that could very likely be lowered

Thanks. I will try the resistor swapping. For the most part I am happy with the sound after removing the unnecessary grid leak in the output.


I did find out that the resistors I am using are rated for 300v. I have about 250vdc on my plates. I am curious as to what affect these resistors have on the amp if they are being nearly maxed out. Since they are not really costly I would assume that thy have a rather large tolerance. So I guess the question is, when a resistor is being maxed out voltage wise, would it increase the chance of unwanted noise. Where I am now is trying to quiet the amp up a tad bit so Im looking into options there.


~minus the power supply and power amp. All resistors are sufficient.

I'm not talking about resistor types, values, or ratings - or resistors at all (except V1a). I am talking about capacitor values which is what shapes your sound from one stage to the next and finally to the output valves.

[hesamadman]

I know I'm sorry. That post was untrelated.

[jojokeo]

suggestion: now that you're close, this is not the time to stop experimenting. This is where the good stuff happens and the learning truly begins. Congrats on getting it figured out too.