Vox AC10 build issues

[Jalmeida]

Hey guys! I don’t post here, but read here often. I recently built a Vox JMI period AC10 with scratch built chassis to fit into a trainwreck Cambridge reverb 30 amp cabinet that also uses it’s speakers.

I also incorporated a direct coupled single 12DW7 reverb circuit. I am admittedly have alot to learn about circuit design and have been getting some advise from a local tech who gave me some tips for the mixing resistor network. I built the circuit using smaller daughterboard using tagboard. I set it up to be able to be removed from the circuit using jumpers on a small TB. The reverb works, but when I integrate the circuit the normal channel volume drops by almost half.

Here is a poorly drawn diagram I put together showing where the new circuit interfaces with C11 of the normal channel.

Perhaps Swap the 40k and 180k resistors- putting 180k between the coupling cap and the tube seems like it would drop alot of current to the Phase inverter.
Perhaps go from the 40k/180k to 120k/100k. Or in 100k/120k order
Playing with these resistor values should be easy to experiment with as long as I make sure I have the total 220k in series to the V2A tube plate. I must admit I am also curious what playing with the 220k mixing resistor values may do also. Any thoughts?

[Jalmeida]

More pics

[Jalmeida]

And..

[Jalmeida]

Bump

[sluckey]

Why did you split the 220K plate load for V2A? You lost a tremendous amount of normal dry signal by doing so. I would use a 220K just like the original circuit and reevaluate.

[Jalmeida]

Why did you split the 220K plate load for V2A? You lost a tremendous amount of normal dry signal by doing so. I would use a 220K just like the original circuit and reevaluate.

It was advised by the guy helping me with integrating the circuit in the wet signal to have resistance between the where the circuit drive signal picks off and where it gets reintroduced. I have no experience designing amp stages. I have alot to learn. So forgive me.

[Jalmeida]

Why did you split the 220K plate load for V2A? You lost a tremendous amount of normal dry signal by doing so. I would use a 220K just like the original circuit and reevaluate.

So I put C11 coupling cap back at pin 1 of the tube and left the 40k and 180k in series seeing they are the same plate resistance value. Volume is back to where it was. But integrating my reverb back into the circuit does not drive the reverb circuit. Curious if it’s due to the impedence between V1 and the 12DW7? I also tried tying in the between the 40k and the 180k with a .0022 cap and no reverb. The only way I have been able to get the reverb to drive is wire per the diagram. But the normal channel is then too quiet.

[kagliostro]

Give a look to the VOX Cambridge schematic and see you can find inspiration

https://el34world.com/charts/Schematics/files/Vox/Vox_cambridge_reverb.pdf

Franco

[Williamblake]

In your schematic the reverb goes to the wrong side of the mixer. Maybe wrong is not the right word but i think it may be worth a try if you split the driver because reverb is to weak.

[sluckey]

Try this...

[Jalmeida]

Thanks for all the help, guys! So to get this straight….

Go back to standard dry signal resistor as factory plate resistor C11. Then integrate as shown above?

[Williamblake]

Integrate as shown above and revaluate. You haz reverb? Enough reverb? Reverb souds good?

[sluckey]

I predict you will need one more triode.

[Jalmeida]

I predict you will need one more triode.

Drive or recovery?

[sluckey]

Recovery.

Meant to ask this earlier... Which tank are you using?

[Jalmeida]

Recovery.

Meant to ask this earlier... Which tank are you using?

4FB3A1B

[Jalmeida]

Recovery.

Meant to ask this earlier... Which tank are you using?

So your suggestion worked. I can get alot of reverb, but I also get a feedback squeeze the gradually grows to unbearable if the reverb level is above 5/10. Click off the reverb and it’s gone. So originally there was a 10k resistor between the reverb level pot and C6(.0022uf) coupling cap that was advised I remove. Curious if putting that 10k resistor may help the squeal? I am merely spitballing.

[sluckey]

May be V1B has too much gain. Replace R8 with 100K. Any joy?

What is connected to the cathode of V1B?

[Jalmeida]

May be V1B has too much gain. Replace R8 with 100K. Any joy?

What is connected to the cathode of V1B?

22uf cap and 470R resistor.

Possible phase issue from the reverb circuit?

[sluckey]

Too much gain.

Replace the 220K plate resistor with 100K.

Replace the 470Ω cathode resistor with 1.5K.

[Jalmeida]

Too much gain.

Replace the 220K plate resistor with 100K.

Replace the 470Ω cathode resistor with 1.5K.

Feedback squeal was fixed by that, but now what sounds like a 60hz hum is present in the reverb channel. Don’t have an O-scope at home. I rechecked and reflowed the solder to ensure no resistive solder joints. Odd. I’m trading problems. I am guessing that this was present before and changing other things brought it forward. Perhaps changing the 22uf cathode bypass caps for another value may be a good thing to try?

[Jennings]

Worth double checking layout and grounding at this point?

[Jalmeida]

Worth double checking layout and grounding at this point?

True. Good point. I will say that it is odd that the hum started adter changing the gain on the 12DW7 revovery stage. Before I had no hum. Just a squeal when the reverb was turned above 50%. The hum goes away when the reverb level is all the way down or switched off. So I retraced all of the reverb circuits and reflowed the solder connections. No wire strands shorting. No change there.

Has me curious if I need to change my cathode caps to another value? 22uf is what they are currently. Perhaps they got damaged changing out the cathode resistor? I did probe both sides of the caps checking for voltages and all seemed normal. Perhaps a 1.5k cathode resistor paired with a 22uf cap isn’t filtering out frequencies as they were with the 470R cathode resistor? 

[sluckey]

OK, put the 220K and 470Ω back in and we'll try to lower the reverb signal another way.

Put that resistor back between C6 and the reverb pot, but rather than a 10K, use a 100K or even larger. Experiment with different values until you get a usable reverb level without squealing.

[Jalmeida]

OK, put the 220K and 470Ω back in and we'll try to lower the reverb signal another way.

Put that resistor back between C6 and the reverb pot, but rather than a 10K, use a 100K or even larger. Experiment with different values until you get a usable reverb level without squealing.

Sounds good. I did try putting the 10k in past night and the hum remained. But that was with the 100k/1.5k resistors. Curious what current numbers I should expect on that resistor near C6? I have a decade box I can pipe in and play with real time. But it is not made for significant current. But eould make for easy testing. Definitely wouldn’t want it for the plate resistors.

[sluckey]

Good idea. Decade box will do fine. You're putting a reverb circuit into an amp that never had that circuit before so there will have to be a bit of experimenting. You're getting close to the joy.   

I could be wrong, but I don't think your hum is due to using that 100K/1.5K to lower the gain of the recovery stage.

[Jalmeida]

I could be wrong, but I don't think your hum is due to using that 100K/1.5K to lower the gain of the recovery stage.

I agree. It was likely already there and just brought forward.

[Jalmeida]

Good idea. Decade box will do fine. You're putting a reverb circuit into an amp that never had that circuit before so there will have to be a bit of experimenting. You're getting close to the joy.   

I could be wrong, but I don't think your hum is due to using that 100K/1.5K to lower the gain of the recovery stage.

Okay, original resistor values back in with 10k resistor, and have less squeal than without the 10k. So pulled the resistor and jumpered in a decade box and dialed values until squeal was gone. 220k to be precise. The hum was still present, so I tapped around with chopsticks and found my TB jumpers that allows me to remove thereverb circuit all together had a jumper that was resistive and would crackle and hun would come and ho. The terminal block clamp screw was not precisely on the wire so folding the jumper ends corrected that mostly. Still have some reverb hum. About on par with my AC15C1 hum so not gonna chase too much. All else seems good. Thanks for the helps guys. I’ll play it a bit and see how the hum translates with use not hyper focused in troubleshooting brain. Then re-address latter if needed.

That said, curious if the remaining  noise is heater noise? The heaters are AC 6.5v. I could install a pot hum eliminator setup.

[sluckey]

Your schematic does not show a center tap on the heater winding so you probably are geting some filament induced hum. A humdinger pot or artificial center tap with two 100Ω resistors in series across the heater winding with junction of resistors connected to ground or the cathode (pin 3) of the EL84s will reduce filament hum.

However, if this reverb circuit is just temporarily connected to the amp using long jumper wires and no shielded cable, this is a great way for hum to enter the circuit. The output of the reverb tank is the lowest signal level of the entire amp and requires shielding and neat layout/wiring, especially to the grid of the recovery tube.

Post a hi-rez pic of the amp with the circuit jumpered in.

[Jalmeida]

Your schematic does not show a center tap on the heater winding so you probably are geting some filament induced hum. A humdinger pot or artificial center tap with two 100Ω resistors in series across the heater winding with junction of resistors connected to ground or the cathode (pin 3) of the EL84s will reduce filament hum.

However, if this reverb circuit is just temporarily connected to the amp using long jumper wires and no shielded cable, this is a great way for hum to enter the circuit. The output of the reverb tank is the lowest signal level of the entire amp and requires shielding and neat layout/wiring, especially to the grid of the recovery tube.

Post a hi-rez pic of the amp with the circuit jumpered in.

I will get pictures later.

Thanks for the thoughts on the heater wire. I will clip in some resistors to my heater TB temporarily to see if that helps.

For the integration, i use shielded wire for the reverb tank signals to and from.

Where I tie in for the recovery at the normal volume pot 220k resistor is not shielded. Nor is drive portion. That is what I have connected to the TBs. So if I wanted it out of circuit I would pull the jumpers and tube.

[Jalmeida]

So tonight I had band practice and tried out the amp at appropriate volumes. The permanent hum was fixed.

However the feedback I had prior that was a higher pitch is now a lower pitch and takes longer to start. Like after being on for say 30mins you cycle on reverb and in about 30 seconds it comes on gradually as a lower frequency feedback building. My kneejerk is a capacitor it damaged and leaking.

[sluckey]

Okay, original resistor values back in with 10k resistor, and have less squeal than without the 10k. So pulled the resistor and jumpered in a decade box and dialed values until squeal was gone. 220k to be precise.

However the feedback I had prior that was a higher pitch is now a lower pitch and takes longer to start.

Make that 220K larger, even 1M if necessary. Use the smallest value that kills the oscillation and still yields enough reverb. You may consider using a 250KA reverb pot. If so you'll need to experiment with this resistor again.

[Jalmeida]

Okay, original resistor values back in with 10k resistor, and have less squeal than without the 10k. So pulled the resistor and jumpered in a decade box and dialed values until squeal was gone. 220k to be precise.

However the feedback I had prior that was a higher pitch is now a lower pitch and takes longer to start.

Make that 220K larger, even 1M if necessary. Use the smallest value that kills the oscillation and still yields enough reverb. You may consider using a 250KA reverb pot. If so you'll need to experiment with this resistor again.

I did some digging since my last post and came across some very sulimilar symptoms. So before I go down that path again I am going to play with the reverb tank shock mounting a bit too. It is mounted on dampeners, but they are rather thin. Maybe just power up and then pull the reverb out of the cabinet to see if it does the same. I want to rule out feedback loops created by speaker magnet interference and/or speaker cabinet vibrations.

It may be wishful thinking at best. But I am hoping to not crank down the recovery stage levels with a hogher resistor as I am already loosing a ton of reverb wet signal versus dry now with the 220k in there. I may also get another 12DW7 to rule that out too. So I will try a few things and then circle back as you suggested, Sluckey. I can’t thank you enough for your help.

[sluckey]

Put the reverb tank in a bag and mount in the bottom of the cab if possible. Keep the sensitive output jack far away from the power and output transformers.

[Jalmeida]

Okay, i feel like I am going backwards.
 

I have along the line got this odd low frequency noise now. Not sure how to describe. I have checked reverb circuit forwards and backwards.

Here is what I have done so far:

1. Rewicked all solder connections
2. Checked for DC voltages at both sides of the capacitors against the drawing.
3. Checked all connections for stray wire strand connections or solder shorts
4. Installed humdinger pot and adjusted listening for noise.
5. Installed 100R resistors between filaments and EL84 cathode.(initially done for other noise)
6. Re terminated and verified reverb circuit wire and integration terminal block(screw type)

The noise DOES goes away when:

1. I remove the phase inverter tube
2. Turn the reverb level down
3. Pull the input and output sign of the reverb circuit where it integrate into the factory amp curcuit. As soon as I plug it back in the noise returns.

The noise DOES NOT go away when:

1. I unplug the reverb tank
2. Change phase inverter tubes
3. Change rectifier tubes
4. Utilize elevated(EL84 cathode) filaments
5. Remove 12DW7 reverb drive/recovery tube
6. Unplug instrument.
7. Turn either channel volume down

The noise is not effected or changed by manipulating the shielded reverb tank drive or return cables or solder connections. I triple and quadruple checked all solder connections for resistive connections and resoldered every integration point for good measure. I have put the reverb tank in a bag.

Based on the fact the noise goes away when I turn the reverb all the way down can deduce that the issue is between the normal preamp tube and the driver side of the 12DW7. The one thing on my list I haven’t ruled out is to replace the 470pf coupling cap C1. I will get another of those and rule that out. Beyond that I am truly at a loss at this point. I have walked away from it at this point as I am not getting anywhere. Any ideas would be appreciated.

Here’s some videos.

[Jalmeida]

Okay, so my noise is called motorboating. Didn’t know that. Add to the long list of things to learn about tube amps.


So, reading a bit more and researching my schematics I see a disparity between the three different and common AC10 schematics I have. The presence of a 200R 4w B+ decoupling resistor. My amp does not have the decoupling resistor installed. So if my motorboating is due to the pack of decoupling reaistor, that would be an easy fix. However I am admittedly skeptical as the noise is only present on the reverb part of the circuit. But given the inclusion of the resistor is not difficult, that is easy enough to rule out. I will also replace the 470pf reverb coupling capacitor while I am at it.

However I am curious about noise being phase related

[PRR]

> 100R 4w

Where do you see that?

Do we have a complete As-Built schematic? I see factory plans and your snippets but not a complete plan? (I may be blind.)

[Jalmeida]

> 100R 4w

Where do you see that?

Do we have a complete As-Built schematic? I see factory plans and your snippets but not a complete plan? (I may be blind.)

That was a typo. I corrected the post. Sorry. The three schematics shown above are two different years of Vox factory schematics and one from Stephen Grosvener’s Service engineer’s guide for the AC30(covers multiple AC models).

As far as how it was actually but, I doubt they were all the same. Here are two pictures of original AC10 with different resistors between the 16uf can cap and pin 3 of the EZ81 rectifier. One can be read as 300ohm. I am betting that the earlier drawing that shows the lack of a resistor was something added latter to remedy a design issue. Or it’s anyone’s guess on what different values it may have had.

That said, I installed a 200R 5w resistor and it did not remedy the motorboating. I changed the mixing resistors to 470K from 220K and no difference. Changed the 470pf reverb coupling capacitor out for new 500pf.

As of right now, i know that when I probe a DMM to check voltage between the .01uf C15 capacitor and the 12AX7 Phase Inverter, the motorboating goes away. And seeing I have already swapped in another PI I believe my issue is design related. 

[Esquirefreak]

I've had motorboating issues in one amp due to 2 stages sharing K bypass caps. Might be worth a try to split the cathodes and give them each a dedicated resistor and cap. Double the resistance to maintain the same bias.

/Max

[Jalmeida]

I've had motorboating issues in one amp due to 2 stages sharing K bypass caps. Might be worth a try to split the cathodes and give them each a dedicated resistor and cap. Double the resistance to maintain the same bias.

/Max

Thanks for the recommendation. The reverb circuit has their own separate cathode bypass caps. The vibrato and oscillator and the 2x EL84 are the tubes that use a shared cathode capacitor. Not sure I can do anything about either of those though. Do you think these would cause this?

[Esquirefreak]

Couldn't tell for certain, but I guess it's possible. Motorboating is (correct me if I'm wrong) a form of oscillation. So signal is probably leaking into or being pickuped up, somewhere it isn't supposed to.

Someone who knows this stuff might be able to help you more than I can.

The EL84 cK can be split into two 250Ω resistors, for what it's worth.

/Max

[sluckey]

You're working with an unproven circuit so you should expect some adaptation challenges.

I don't think you will ever solve these issues as long as that sensitive reverb recovery circuit is just dangling outside the chassis.

[Jalmeida]

You're working with an unproven circuit so you should expect some adaptation challenges.

I don't think you will ever solve these issues as long as that sensitive reverb recovery circuit is just dangling outside the chassis.

Agreed. Alot to sort out, experiment with and learn. I do appreciate allthe help. So forgive my misunderstanding and learning curve.

Not sure what you mean dangling outside the chassis. Which part dangling were you referring to. The chassis is out of the cabinet, but all of the reverb recovery is attached to the same chassis as the rest of the amp circuit. The only part not affixed and dangling is the footswitch jack plate and tank send and return which are all shielded and get their grounds in the chassis. They are plate mounted and screw to wood for mounting, so no ground there missing. The reverb tank cable is all shielded cable. Where the dry signal for the reverb circuit picks up in the chassis. And where the wet gets back in to the mixing resistor network is also in the chassis.

Do you think the chassis being out of the cabinet would cause the motorboating?

Keep in mind that I have reassembled the amp at numerous stages to rule out outside noises and have had the same noises. I truly think after everything I have gone over that I have a design issue. But I am far from an expert and hopeful to learn.

[Jalmeida]

Here is a cleaned up and updated schematic showing integration.

[PRR]

> showing integration.

Motorboating is very often through the power supplies. Despite requests, you are not showing a fully integrated signal and power path. Maybe if/when you do, it will be clear.

[Jalmeida]

> showing integration.

Motorboating is very often through the power supplies. Despite requests, you are not showing a fully integrated signal and power path. Maybe if/when you do, it will be clear.

Pictures of the wires? Circuit? What specifically would help? Are we trying to determine the wire routing? Wire type? I have been updating schematics for exactly how my amp is currently to try and get a complete picture. I am not trying to be reluctant to provide info. I just am not sure exactly what it is you want to see. Or what will help?

[Jalmeida]

Here is the entire print updated…

[PRR]

Now we see relationships. V2b plate is throwing huge DC into V1A grid. Just a drawing error or in real life? The plate output of V1 is ignored, signal is taken from the cathode bypass cap. The reverb pot wiper is unemployed.

Are there two "V1"s? Whatever...

Now I see how the power runs. Taking so many stages on the one "255v" filter stage is, not surely bad, but risky. I would at least make one more filter node called "250V" (it may be higher, just a name) for the V1B stage. Also in this day of cheap electrolytics I would like to see much more uFd at "225V", 50 or 100uFd.

[Jalmeida]

Now we see relationships. V2b plate is throwing huge DC into V1A grid. Just a drawing error or in real life? The plate output of V1 is ignored, signal is taken from the cathode bypass cap. The reverb pot wiper is unemployed.

Are there two "V1"s? Whatever...

Now I see how the power runs. Taking so many stages on the one "255v" filter stage is, not surely bad, but risky. I would at least make one more filter node called "250V" (it may be higher, just a name) for the V1B stage. Also in this day of cheap electrolytics I would like to see much more uFd at "225V", 50 or 100uFd.

Drawing error.. sorry. I will edit to replace. Here is the updated print correcting:

1. Missing 500pf coupling cap
2. Incorrect Reverb pot wiper
3. Added Reverb circuit TB tie points
4. Reverb circuit tagboard layout.

[Jalmeida]

Are there two "V1"s? Whatever...

Now I see how the power runs. Taking so many stages on the one "255v" filter stage is, not surely bad, but risky. I would at least make one more filter node called "250V" (it may be higher, just a name) for the V1B stage. Also in this day of cheap electrolytics I would like to see much more uFd at "225V", 50 or 100uFd.

The V1 that is the 12DW7 is all the new reverb add-on circuit. I kept the original AC10 valve assignments in tact because the reverb add-on was an afterthought after the amp was mostly built. Hence the tagboard integration with TB in case I couldn’t get it working as desired.

As for the 255v usage, that is pretty much 100% orignal AC10 design. I left the 255v alone and picked off the 305V. Which has me suspecting that getting the dry signal feed to the 12DW7 from the ECF82 that gets 255v and then driving the tank and recovery from the 305V is where the oscillation is rooted as a phase disparity. But I am so far from being in my wheelhouse with valve amps. So, I may just try the revovery stage on the 255v to see if the motorboating changes in any way. Luckily I have a 100uf Big ass cap if I need to add another node.