18W Amp Build - Few Minor Issues

[Littlewyan]

Hello everyone

You may or may not remember I mentioned that I was looking at building myself a valve amp a few months back, well I did and it works! I've attached some pictures and a few sound clips. I have taken photos throughout the entire build process but obviously cannot post them all here so I've just posted a few of the parts pre build and then the amp post build.

Now this is an Ampmaker PP-18 amp kit which sounds alright as it is but has a slight issue with the overdrive. When the amp goes into overdrive it seems to have a fizzy/buzzyness as the note decays. You can hear it in the sound clips I've attached to this post, especially when I switch to the neck pickup during the Tone 0 clip. Unfortunately I haven't got immediate access to a scope to see where this is happening and have read MANY different causes of buzzing and fizzing in these 18watt builds. So I thought I'd let you guys have a listen to see what you think. Some people believe that this sound is a characteristic of these amps, HOWEVER, I have heard 18w Amps not have this problem so I'm sure it can be fixed.

I should also mention that I changed the V1 Coupling Cap to a .01uF which made the issue worse (I am going to change this to a smaller value but thats a story for another day).

Here is the schematic.

http://www.ampmaker.com/images/ak03kit/ak03sc2.jpg

Any input would be greatly appreciated.

Ryan

P.S. I've also attached the schematic I drew up of the Pre Amp in LTSpice along with the Triode ASC File that I used. I've also used a potentiometer symbol in LTSpice but I imagine a lot of people already have this.

[phatt]

The Signal likely gets way big and as there is no series resistance after V1 then it's just asking for trouble into PI section.
I've worked on and built a few glass amps and it seems to be a common problem,
you get more gain but you pay with crap tone.

Also if the AC voltage going to bridge is 275 Volts then you have in excess of 380 VDC and that won't help tone either.

IMHO, once past ~330 VDC EL84's just become room heaters an my bet is valve life will be reduced no matter how cool you bias it all.

The pro Junior schemo might give some clues. It's fixed bias but the preamp is likely a better more controlled setup for any 2 knob small Amp.

My own build which is a PJ circuit except output is Cathode biased and has no cathode cap. (C9 on your schemo)
I found it mucked up the tone at full tilt,, my guess is small Amps can't reproduce the extended bass anyway so deleted it. much sweeter drive. 

Others here may know more.
Phil.

[Littlewyan]

Phil

Now this makes complete sense as I have been looking at schematics for the Marshall 1974 (Pretty much the same amp with a few exceptions) and the 2061. Now the 1974 doesn't have grid stoppers but it only has 330VDC on the Power Amp and the Power Amp is biased a bit different (125 resistor instead of 150). The 2061 has a higher voltage on the Power Amp BUT it has grid stoppers and the PI is a bit different.

I am going to reduce the 47uF Cathode Bypass Cap to 1uF as at the moment there is so much bass even with the tone on 10 which should be really trebly! If I still have too much bass then I'm going to reduce the Coupling Capacitor back to .005uF or worst case .0022uF. I shall then look at putting in a grid stopper of 470K with a 470pF Bypass Capacitor. I plan to do it in this order as the cathode bypass cap and coupling cap are easier to change than adding in a grid stopper and should reduce the gain a bit. What do you think?

Ryan

[Littlewyan]

Although Phil it was my understanding that Grid Stoppers dont tend to reduce gain due to the miller capacitance. They might reduce treble slightly but not normally gain.

Changed the Cathode Bypass cap btw to 1uF and now the tone is great! Its just the fizz thats the issue now -_-

[Cpt. FixIt]

Although Phil it was my understanding that Grid Stoppers dont tend to reduce gain due to the miller capacitance. They might reduce treble slightly but not normally gain.

Not really, rather Ultrasonic/RF: Let's assume your first 12AX7 stage's gainfactor to be about 60(12AX7 with bypassed cathode resistor), its Miller capacitance equates to Cm=Cgk + (Cga * gainfactor). The datasheet says Cgk=Cga=1.6pF, so we get Cm= 1.6 + (1.6*60) = 97.6pF(~100pF).
With a 56k Gridstopper fc of the resulting lowpass calculates to fc= 1/(2 * PI * R * C)= ~28kHz.

Changed the Cathode Bypass cap btw to 1uF and now the tone is great! Its just the fizz thats the issue now -_-

There's two 47u cathode bypass caps(first stage and powertubes) - which one are you talking about?
Don't worry about using small values for C2 - even .001uF might be necessary here for a tight bass response.
C5 is obsolete btw. - looks like somebody did a sloppy job simplifying the 1974 schematic.
To get rid of the fizz I'd try a cap(100pF - 500pF) between the PI anodes first.

Regards,
Toby

[Littlewyan]

Sorry I had part of that conversation in my head, I was talking about the 470K Grid Stoppers you sometimes have before the PI. Am I right about grid stoppers not reducing gain?

I changed the first stage Cathode Bypass Cap to 1uF and now the amp has just the right amount of bass, although I've only tested it at bedroom volume levels so far with an attenuator. I've left C2 as .01uF for the moment. I see what you mean about C5, not entirely sure about that as at .022uF its not really doing anything in the way of tone shaping as its letting the entire guitar freq range through it.

The guy who designed the amp is sending me the parts for a conjunctive filter on the OT as he believes this should tame the fizz. I have asked if he could also send me a 100pF Cap in case the conjunctive filter doesn't do the trick and I've asked for a 120ohm resistor in case I want to experiment with the bias of the power amp.

Ryan

[Cpt. FixIt]

Hello Ryan,
now that I've listened to your clips over decent speakers I think that fizz sounds like intermodulation.
I have attached a zoom into the waveform(the decaying chord in your first clip where the fizz is pretty audible) that makes the suspicion even stronger.
Something in your amp might be( on the edge of) oscillating(or picking up EMI).
Do you have a scope or can you borrow one?
I don't think the problem lies in the OT(it had to be a real POS if that were the case...) and therefore don't believe in putting a filter there. That would be like buying earplugs because your muffler's shot...

Toby

[Littlewyan]

Toby

My grandad has a scope, i'll see if i can pop over there tomorrow night to use it. How did you put a scope on my audio clip?

Ryan

[Cpt. FixIt]

Now that sounds like a good option - grandpa with a scope.
Does he have a signalgenerator, too? If not, just download a sinewave soundfile and use your iPod or burn a CD.
I did not scope your clip, I just opened it in Cool Edit and zoomed in on the waveform view, I guess I'm not fast enough to photograph the scope at the right moment...

Toby

[Littlewyan]

Yes he has a signal generator as well. I'm actually looking at getting a Picoscope 2204 for xmas as these have a signal generator built in. Do you have any experience with these?

[Cpt. FixIt]

Not only do the picos have an inbuilt signalgenerator, they also have this really cool sweep feature, looks like it saves a lot of setup time...
I don't have any personal experience with these devices - I use Hitachi scopes in the shop and a Tektronix TDS 210 to carry around - but my buddy owns a 2206A and so far seems to love it.

Toby

[g1]

  Most of the picoscopes can only handle small voltages (20V or less?) so if you are wanting to work on tube amps you will probably need 100:1 probes.
  Recommend doing further research on their limitations before you purchase one.

[Littlewyan]

Ah yes that was what I needed to look at, what probes I can get for it. I would go for a proper scope to sit on my desk but I just haven't the room and I don't think the cheaper ones include as many features as the USB Picoscope.

[Littlewyan]

Right we scoped the amp and there is no oscillation. Altho we did get it to oscillate a few times with no input and the volume turned up but that disappeared as soon as we plugged something into the input. We found The signal didnt distort until the PI output and then again a bit more on the Power Amp output. Going to try the conjunctive filter when ampmaker send the parts.

[Cpt. FixIt]

Right we scoped the amp and there is no oscillation. Altho we did get it to oscillate a few times with no input and the volume turned up but that disappeared as soon as we plugged something into the input.

That leaves us with two possibilities. Either something in there is on the edge of instability - or you forgot to wire your input jacks as shown in the schematic, so that the inputs are grounded if nothing is plugged in.

We found The signal didnt distort until the PI output and then again a bit more on the Power Amp output.

That doesn't come as a surprise - the PI is just the second stage in this design, you'd need a really beefy booster to make your input stage distort...

I personally think you just have a minor leaddress issue - but check your input jacks first.

Toby