Vox AC100/2 build

[pompeiisneaks]

Hello all,

I had another thread going in the schematics area about what I could do to reproduce a vox AC100/2 '65 circuit.  That thread is here:

http://el34world.com/Forum/index.php?topic=19426.0

I've actually completed the build, and am making a video of it, for all to see, but before I do, I have something going on, and I need to have some input on what may be up.

I've got the tubes out right now, and am slowly bringing the power up on a variac to make sure nothing goes wrong.  As I've been doing this, I noted one of my 2W resistors is starting to smoke, so I shut it down.  This is a resistor that goes to ground off of the B+ rail, (the thread above has the circuit diagram, but if you look at my schematic I'm attaching to this post, it shows the current build very close to how I made it. 

Is this resistor getting a lot of current because the tubes aren't in it, so all the current is going down that resistor?  The resistor on the circuit I'm attaching is R23, on the B+ rail between B and C. 

Any thoughts about why there would be such current going through that? 

Here's the measurements I made:

225V and an 18k resistor gives 0.0125A or 12.5mA and watts is about 2.8, that's more than the 2watter wants obviously so I see why its getting hot.  (This is still way below the max wattage of the amp too, but again, I'm thinking it is just because the tubes arent' there to conduct the current right?)

Is there any point at which I should be stopping voltage tests as things are 'ok' and then connect in the tubes?  Or do most people just do one quick test with a little current to make sure its not grounding out and then put in the tubes?

~Phil

[sluckey]

You need to replace that 18K/2W with a 18K/25W resistor. You're already smoking it with the B+ only up to 225V. Guess what's gonna happen when you take it off the variac and the B+ comes up to 450V!

[pompeiisneaks]

Ok, is there any reason that the schematic would put it as a 2 watter (the original Vox one that I copied for my schematic?)  why was that schematic so far off?  You'd think they tested the design out? Or is this original '65 schematic also not that great?  I thought this was the one the Beattles used no?

BTW link to original schematic I used here: http://el34world.com/charts/Schematics/files/Vox/Vox_ac100.pdf

~Phil

[sluckey]

I don't see any mention of that resistor being only 2 watts. Where do you see that? Do the math. 450²/18000 = 11.25 watts.

I looked at several pics of '60s AC100s. That resistor is about the size of a 10W sand block resistor (maybe a bit larger), only it's round.

[pompeiisneaks]

D'oh! You're right, okay, I'll get that replaced :) 

[pompeiisneaks]

Okay so finding 25watt 18k resistors isn't proving easy.  I guess I could just do 4 18k 10W in series and then parallel to get 18k at 20 watt right?
 (i.e. two parallel sets, then put into series with one another)

[sluckey]

The resistance value is not critical. I would just use this...

http://www.mouser.com/ProductDetail/Vishay-Dale/RH02520K00FE02/?qs=sGAEpiMZZMtlubZbdhIBIMow1WQ9W2ngj8pozixWgko%3d

[pompeiisneaks]

Ok cool, 'not critical' as in 'at least 18k' or how much variance can you get by with on something like that?  is say 10k or 15k too low etc?  More for my knowledge, not for this specific case.

~Phil

[sluckey]

I would not go lower because the bleed current will increase and you would need a higher wattage resistor. 20K is only 11% more than 18K. Slightly less bleed current. 20K is an easier value to find.

The reason I initially suggested 18K @ 25W was because of an old rule of thumb that I was taught a long time ago. Calculate power and double that for the resistor wattage to provide a safety margin. Then choose the next higher wattage that is commonly available. Specifically, that 18K will dissipate 11.25 watts when 450V is applied. So, double 11.25W is 22.5W. Next commonly available size is 25W.

[pompeiisneaks]

Excellent, makes perfect sense, thanks!  I'll get one and proceed after I've replaced it.

[pompeiisneaks]

I had one of those AHA! moments after this reply.  I got all the science and math of it just fine, but for whatever reason this simple fact was eluding me... there was nothing wrong with the circuit, the simple fact that an 18k resistor in a 450V circuit will always generate 25 mA and 11.25 Watts is unchangeable no matter how you modify the circuit.  I don't know why I was trying to over complicate it.  I had done all the math already but somehow was thinking that the other components of the circuit would have some play in this.  They don't. 

Thanks Sluckey!

[sluckey]

I tend to over think stuff all the time too!

That amp will work without that resistor if you want to continue testing while waiting for a big resistor.

[kagliostro]

If I can say, the AC100 circuit is known as to have problems about the bias section of the circuit


if you can, give particoular attention to this section, also, sometime you can find mods on the web


Franco

[pompeiisneaks]

Franco,

Thanks, I actually modified that circuit, if you look at my schematics I have a different bias in there.  Please let me know if that looks okay. 

So next steps (I was able to buy a 22.5k 25w resistor locally), I'm seeing something quite bad/odd in the power.  I've got my variac up to about 60v AC and the power on the B+ is already up to 500 VDC.  That's not good.  Does this mean somethings wrong with the PT?  Why would it be putting out such high voltages at half power?  Even the Vox AC100CPH schematic that has the voltages for this tranny show it at 497V right before the caps/choke at A, and then at B its down to 488. 

When PT's go bad do they sometimes increase their power output? on the other side?

The heaters even are running at like 70V instead of 6!  All this at 60ish VAC.

~Phil

[kagliostro]

Which PT are you using ?

Are you sure about primary connections ?

Franco

[pompeiisneaks]

I'm using the PT from the Vox AC100CPH, linked in the other thread at the top, This is obviously the right rails, or it would be one of the lower windings right?  (I've removed the boards from the 100CPH and am making my own circuit in there, just reusing the PT/OT/Chokes from the existing one)

[sluckey]

I'm using the PT from the Vox AC100CPH

WHOA!!!

That PT is meant to be used with a full wave conventional rectifier that uses a center tap to ground circuit. But you have connected a full wave bridge rectifier to that PT, so it will produce exactly twice the B+ you desire.

If you want to use that PT then copy the rectifier circuit from the AC100CPH. The filters can be the same as the orig '60s AC100.

[kagliostro]

Ditto

Franco

[pompeiisneaks]

Oh my, I'm glad

a. I used a variac to bring it up
b. I learned this before I did something really bad :)  Okay, I'll sort that out now. 

So why does a full wave only output half the power vs bridge rectifier?

~Phil

[pompeiisneaks]

Okay so this is what you mean, I'm attaching a shot of the circuit with the rectifier.  The problem I have now is that I don't have a board to attach the diodes to.  Do I also need the 1n 1kv caps on this?  I have the 22R 7W ones still from the old one, I just don't know if I'll have those caps.  (I can buy some of course).  What is the purpose of the caps in a situation like this?  Does it protect the diodes?

Edit: Helps if I actually attach it... side note I DO have all the parts from this schematic so I'm golden :)
~Phil

[shooter]

What is the purpose of the caps in a situation like this?  Does it protect the diodes?

diode rectifiers tend to generate hi-freq spikes, the caps tend to suppress those spikes

[HotBluePlates]

So why does a full wave only output half the power vs bridge rectifier?

Both rectifier circuits output the same power. But the bridge will output twice the voltage (at half the current, for same-power).

The non-bridge rectifier has a grounded center-tap, and only half the secondary winding is passing current at any one time. The bridge passes current from the entire winding, all the time. So if you take a center-tapped secondary and use a bridge rectifier, you will wind up with twice the normal output voltage.

[pompeiisneaks]

Cool thanks to both of you, makes sense. 

~Phil

[pompeiisneaks]

Oh I also figured out why one of my heaters was at 70ish VAC.  There is a power rail running 150V that I'm not going to need, but I confused it with the heaters and connected that, I'll be able to desolder those and hook it up correctly now as well.  what fun :)

[birt]

it's very good that you go easy on the first start-up and find these things before anything puffs out it's last smoke 

[sluckey]

You will also need to use a different bias circuit with that PT. I'd just copy the AC100CPH circuit.

[Willabe]

it's very good that you go easy on the first start-up and find these things before anything puffs out it's last smoke 

Also use your light bulb limiter with the variac.       

Variac > light bulb limiter > amp.

[pompeiisneaks]

You will also need to use a different bias circuit with that PT. I'd just copy the AC100CPH circuit.

Why is that?  So the circuit I have made for the other doesn't work with the PT?  I thought it was just about having a variable resistor to adjust the bias resistance, and the other parts provide circuit protection.  Is that current setup not going to provide the protection I need?  Even with the class X cap there?  I guess I'm needing to learn a lot more about how PT interacts with the circuit.  I was thinking that it just provides the input voltage and the other parts of the circuit were what modified how it worked in each stage?  (Not sure if that makes sense?)

[pompeiisneaks]

it's very good that you go easy on the first start-up and find these things before anything puffs out it's last smoke 

Also use your light bulb limiter with the variac.       

Variac > light bulb limiter > amp.

I still haven't built one heh, maybe I need to do that first thing tomorrow before I go to next steps.

[sluckey]

Your bias circuit with the class x cap works well with a Bridge rectifier. But since you will not be using a bridge with that PT you need to use a more common bias supply. It's easy to modify your existing circuit. Just remove R35 and replace C16 with a 220K 1W resistor. You may need to fiddle with the value of this resistor to get a bias range of about -25 to -50v. Everything else in your bias circuit remains the same. This is a very standard circuit used by many Marshall amps.

[pompeiisneaks]

Oh ok cool, thanks.  I'll do that.

~Phil

[pompeiisneaks]

Ok got that sorted out today, all seems fine so far, with one new twist.  I'm attaching the heater schematic information, maybe i'm missing something, but I don't think so.  In bringing it up again, I ran into a smoking resistor (This time just a 330Ohm on my power LED line, that part was pretty non important), but then I measured heater voltage, and the ac voltage at about 60VAC is about 11VAC when its supposed to be 6.3VAC at full 120VAC.  I reviewed the circuit and I don't see anything 'wrong' here, I've used the right wires that were connected only to the preamp before, but now have them to the power amp heaters as well.  In the section of the original schematic I'm attaching, it shows that both preamp and power amp heaters c ome from the same line, but that it splits off the preamp to a bridged rectifier so its dc power.  There are a few other caps etc.  I don't see that I'd really need that, as all the other amps I've seen on this forum so far just use AC for heaters.

Is there something else I'm missing?  Why are my heaters running so high voltage?  (I added 2 100 Ohm 1W resistors to ground on the circuit like I see many current ones do, but that doesn't increase voltage right?). 

Ideas?

~Phil

[pompeiisneaks]

D'oh! Nevermind, I  figured it out, I disconnected the 'wrong' one, and then disconnected the 'right' one to mix the two together, but accidentally closed off the 'right' one and reconnected them all to the 'wrong' one  Oh well, I'll rewire those now and then see where I'm at ;)

[pompeiisneaks]

Ok sanity restored, but now I think I may need the full bridge rectifier that I was using before.  They sent it as DC because they wanted to convert it to 6VDC because it seems to run about 3VAC.  Is there any other reason I'd only have 3VAC on the heaters?  If I look at the schematic I posted back two comments ago, it doesn't have a rectifier for the power amp heaters, only for the Preamp.  Anyone have ideas?  Or is it that I'm using the resistors early in the circuit and that is causing the voltage to drop?  Are the resistors to ground supposed to only be at the end of the ac rails or is putting them in the start okay so long as they are there?

[shooter]

Is there any other reason I'd only have 3VAC on the heaters?

How are you measuring them? 1htr wire to ground?  That = about 3vac.  If you measure one htr wire to the other htr wire, THAT should = about 6vac

[sluckey]

If you are measuring filament voltage with one probe on chassis ground, then 3VAC is what you should measure. You should measure filament voltage with one probe on a filament pin of a socket and the other probe on the other filament pin of a socket. IOW measure filament voltage BETWEEN pins 2 and 7 of one of the EL34s, or BETWEEN pins 4/5 and 9 of a 12AX7.

I would not bother with the DC heater supply in that amp.

Those two resistors can be connected anywhere along the filament string. Makes no difference whether at the start or finish.

[pompeiisneaks]

Oh excellent, then I'm in great shape.  I'm going to give the tubes a shot now, and hook er up to the oscilloscope and give it a test

[pompeiisneaks]

Yup that was the case, when I tested against the two it came up 6.3, we're golden.  I've run it with the oscilloscope at full power, and it seems to have some really odd distortion going on, I'm even seeing it in what appears to be in the volume stage where I'm getting a lot of really weird distortion that seems to come and go.  If I have the volume at a very low setting it seems mostly gone, but as soon as I get it up even a little, maybe 3/10 it starts really doing odd stuff. 

Also, the input voltage to the grid is at -50ish and when I biased it down a bit I got it to about -35 I think was the lowest I could go, it seems like from the EL34 datasheet it really only wants about -36 max, is that right?  Does that mean that maybe its just biased too hot and getting really odd distortion?  I think maybe you mentioned I may have to modify the main resistor to get it to vary between --25 and -50, right now its about -55 or -60 down to about -35. 

I've yet to hook it up to speakers and a guitar, just using dummy load, and a phone app sending in a sine wave.  It def shows a lot of distrortion on the scope as well.

Any ideas on where to poke first?

~Phil

[sluckey]

Plug in a speaker and guitar.

If you want the bias voltage to be smaller, increase the value of that resistor I mentioned.

[pompeiisneaks]

Sounds good, I'll give that all a go tomorrow.

~Phil

[sluckey]

Another easy way to lower the bias voltage is to LOWER the value of R33 from 47K. You have it labeled as 'bias range' on your schematic.

[shooter]

really odd distortion going on

Make sure you *verify* your *source* signal after plugging it in, I got burnt because I started at the spkr, chased the problem back to a flakey cable bringing in the sine wave

[pompeiisneaks]

Shooter, that's exactly what I started thinking too, but I tested on the input jack with the 'scope and it was perfect there and only there.  I always do taht to eliminate input as the source of the issue.  I do think it could be possible though, that my dummy load was being silly, (not sure if having poor connection there can cause bad behavior) so I think as sluckey said, I'll plug in the speakers and a guitar and see what it 'sounds' like for real :)

Edit:  In fact that's part of the reason I bought a 4 input 'scope so I can keep one on the input to ensure its not 'off' :)

[sluckey]

What is the voltage level of your input signal?

[pompeiisneaks]

I'll have to take a look when I get home, I'm trying to remember, I think it was 200mV /div and only covered maybe 4 divisions, so 800mV to maybe 1V?  Is that too hot?

Edit: Googled it and it seems most guitars that are really hot will not exceed 1V, but I'll have to double check, it may be that I was a bit over it, and this input wasn't happy with that volume level. 

I'll let you know what I see on the 'scope at home.

[sluckey]

I usually use 200mV peak to peak applied to the input jack.

[pompeiisneaks]

Ok I was definitely pretty hot on my phone.,  (forget that headphone output is pretty hot).  I lowered it down to 200mV peak to peak and it was going up to almost 5 or more before the weirdness started happening.  Also this sine wave or input cable may be adding some noise that I'm not able to hear without speakers so after work I'm going to take the speakers to the garage, and a guitar, and give er a whirl.

[pompeiisneaks]

First test drive, and it works,

BUUUTTTT... at low volumes its sounding fine except the massive ground hum.  I'll monkey with that a bit. 

The other issue is still that as soon as I get it up to maybe 5 or so there is a horrible pop type noise that seems cyclical, is that a kind of parasitic oscillation?  Or what else would make a cycling pop/chirp sound?  I also tapped around the circuitry with a chopstick and near one area where the exit of the phase inverter and it makes a nice crackle noise.  Is that just that the solder joints in that area may need touch ups? It is odd that i can't get it to happen when I tap one of the specific areas, but if I hit two things at once it seems to make the pop noise. 

Any pointers on what would cause that?  It does sound pretty soft still up til I hit that point and then it just becomes overwhelming with the repeated clicky/pop type (or maybe even chirp like I said). 

~Phil

[sluckey]

Recheck your drawings. Make sure your schematic agrees with the original (except for the mods you've had to make to use that PT). Make sure your layout drawing agrees with your schematic. Make sure your amp is wired IAW your schematic and layout.

Recheck all solder joints.

Swap tubes.

Measure (and post) voltage readings for all filter cap nodes and every tube pin (even the pins that should be zero volts). What voltage is on pin 5 of each EL34? Pull the B+ fuse and measure bias voltage at pin 5 with the bias pot cranked to each end. What voltages do you get?

Divide and conquer. If you have added reverb, disconnect it. Any better? Pull V1 and V2. Temporarily connect an input jack to the treble control wiper and plug a guitar in. This connects the guitar directly to the input of the power amp. It won't be nearly as loud but should be hum free and very clean sounding. Is it? If so, the problem is in the preamp. If not, the problem is in the power amp.

Post some hi rez pics of your amp. Post your layout drawing.

PS... Please put all pertinent layout and schematic drawing in this thread so we don't have to go chasing them down in other threads.

[pompeiisneaks]

Oh sorry I'll get the schematic, (i'll update it first to the latest changes though) and post here.  I'll give all of that a go as well, and post here,  I do recall the power at A,B,C,D,E all being very close to what I recall it should have been from memory, but do need to go down that logical path of writing it all down etc. 

I'll get the pin readings as well, and such. 

Thanks, I should have thought of a lot of this stuff from all the reading I've done but didn't recall seeing a specific sticky thread that talks about some of this kind of tribal knowledge, or did I just miss it? 

If there isn't one, maybe I'll start one up to start distilling you guys wisdom into something more central for quick review by newbs like me :)