Jose Plexi mods

[Spencerleehorton]

I’ve made the transition over to SLO50 now and everything sounds pretty good apart from the bias, I changed the dual bias to this version


Can’t seem to dial in the left hand EL34?
I have several different tubes so will move the tubes and see if the problem swaps sockets or stays where it is.

 

[Spencerleehorton]

Have got two JJ EL34 and have got 35mv on one and 28mv on other with the above bias, will try changing bias again to dual bias.
Going to change the 1uf cathode caps to 2.2uf to bring a bit more bottom end back as there doesn’t seem to be much bottom end, when I turn the bass pot nothing much happens.
Will try and improve the OD channels noise floor.

 

[BIASKING]

Sounds like you may have a leaky coupling cap ofsetting your bias voltage. Did you check (with no output tubes) the DC voltage at pin 5?

 

[Spencerleehorton]

I will thank you
I’ve just spotted my mistake before the tone stack, pretty sure that’s why the tone stack isn’t really working but the presence is!!

 

[Spencerleehorton]

Having a nightmare with this bias again!!
Have gone back to 150k then diode then 15k, 56k then 50k pot, 2 x 10uf/100v either side of 15k to gnd.
Getting -50v at bias output so this seems right but when measuring at the 1R resistors I don’t get any readings!!
I’m going to go over it all tomorrow again as I’m sure I’ve made some mistakes in the wiring.
I have only put in one cathode follower as I have a metro fx loop pcb which I put last in line before going into the phase inverter, just hoping it’s ok to do that with this SLO50 design?

 

[BIASKING]

Since you're having balance issues, have you tried wiring it with the bias supply directly feeding both grids, bypassing the balance circuit for troubleshooting purposes?
Use matched current draw tubes and the balance pot wouldn't be necessary.

 

[Spencerleehorton]

ive actually gone back to the standard arrangement with 220k into diode into 15k into 56k then 50k pot, with 10uf/ 100v either side of 15k to gnd, i take my ac feed from the HV winding.
getting 34mv one side and 20mv the other now, the tubes im using are used and abused so cant really be trusted so i'll get a new matched pair and then measure again.
now ive changed a few wrong wired bits its sounding pretty good, although its a bit too gainy!!! could do with mellowing it out a bit, will take off some of the extra cathode bypass caps ive put on there and change a few grid resistors from 470k to 1meg as im using 1meg pots.

 

[BIASKING]

Is the voltage at pin 5 the same on all output sockets without the tubes installed?

 

[Spencerleehorton]

Is the voltage at pin 5 the same on all output sockets without the tubes installed?

I will check this tomorrow, back on night shift tonight!!
Will change 2.2uf cathode bypass back to 1uf as the OD channel does sound a bit too bass heavy now.

 

[Spencerleehorton]

Have changed the bias to dual bias and works fine now.
Also changed the 2.2uf cathode bypass caps to 1uf with 1k and bottoms end on OD is much improved, I might try going a little further and reducing the caps to 0.68uf on the OD channel as I think I can cut a little more, still plenty of bass to dial in on the bass control.
But sounds fantastic now, it’s got the bite that I like and the OD channel is a beast!!

 

[Spencerleehorton]

For anyone who’s interested this is pretty much what I’ve done.
You should be able to understand where I have put some ranges of resistors and caps as to be able to play around and get it to where I would like.
The main big difference is I have left out a cathode follower stage.

 

[Spencerleehorton]

Wanting to up the B+ so it’s as close to the SLO voltages but wanted to ask specifically what would the differences be, would it effect the headroom, as in, node 1 is 445v, node 2 is 309v after first 10k, then node 3 after 2nd 10k is 250v, then node 4 after 10k is 214v.
So I’m thinking I can change node 2 to 400v, node 3 to 375v then node 4 to 350v.
By my maths of ohms law dropping 45v for node 2 at 13.6ma I would need a 3.3k
Node 3 dropping 25v at 5.9mA 4.2k and node 4 dropping 25v is 6.9k.
Would it improve noise? Would this effect gain? More or less?

 

[Alva Goldbook]

Wanting to up the B+ so it’s as close to the SLO voltages but wanted to ask specifically what would the differences be, would it effect the headroom, as in, node 1 is 445v, node 2 is 309v after first 10k, then node 3 after 2nd 10k is 250v, then node 4 after 10k is 214v.
So I’m thinking I can change node 2 to 400v, node 3 to 375v then node 4 to 350v.
By my maths of ohms law dropping 45v for node 2 at 13.6ma I would need a 3.3k
Node 3 dropping 25v at 5.9mA 4.2k and node 4 dropping 25v is 6.9k.
Would it improve noise? Would this effect gain? More or less?

Why change the B+ voltage? It’s not going to change noise. The amp’s over all lead dress is why the noise is there.

Tubes can run over a very wide range of B+ voltage. The biggest impact of higher voltage is more biasing headroom. A 12ax7 triode with a 100k plate and 1k5 cathode bias resistor will produce a cleaner gain stage if the plate voltage is higher. You might notice a difference in distortion from a plate voltage of 150 to 250, but adjustments to the cathode bias resistor can still produce the same amount of clipping.

The bigger issue is the B+ dropping resistor value. B+ is not 100% silent. Some DC noise is present. The trick is to make that noise happen below the human hearing range of 20Hz. A simple RC calculation between your filter cap value and the dropping resistor value will tell you if these two components are optimal choices for a low noise circuit design.

 

[john12ax7]

Higher B+ for more headroom and a tighter sound. Lower B+ for more of that "brown" sound.

 

[Spencerleehorton]

Yes that is perfect, I like the brown sound so will stick where I’m at and continue to tidy up and do little tweaks

 

[Spencerleehorton]

ok, now im sort of happy with the sound minus a few minor tweaks i want to look at the noise floor and routing the cables in the correct way and cable tieing them where necessary.
i will need to move the 5v psu over to the other side away from input and PPIMV needs screened leads.
I will try and put together a diagram showing ideal routes for cables to verify im going to do it correctly.

 

[Spencerleehorton]

I wondering now that i should take out the HI / LOW switch as it only add hum and ive got full control over the amp now with the PPIMV.
i will post up what i think are the best cabling routes, suggestions very welcome.

 

[Spencerleehorton]

here is the gnd scheme as it stands

 

[Spencerleehorton]

hopefully this can be understood!! the only cables i havent put on there are the 5v ones which would run either under the centre of the turret board or over the top.
at the moment i have the mains txf CT to the 500ma fuse and this is showing the other method of putting the fuse after the standby switch, is there a preferred way?
i was thinking to try and keep all the cables on the right so i can cable tie them.
all suggestions welcome.

 

[Alva Goldbook]

View attachment 139571
here is the gnd scheme as it stands

In principle, there should only be two earth connections in the amp. One by the AC 240 volt inlet, on a dedicated ground wire. This ground wire should be longer than the blue and brown 240v wires, so that if ever the IEC jack gets pulled out of the amp, the ground wire is the last wire to break connection.

The other should be at the input jack, either through a ground connection on the jack itself (Fender alpha jack designs), or by a dedicated terminal with a screw and bolt which is connected to the input jack ground (Marshall CLIFF Jack designs).

Each filter cap should have the grounds from their respective nodes connected. For example if the ground from a cathode follower is powered by Node D then the grounds on that cathode follower circuit should go to the ground of the filter cap on Node D. Then each filter cap should be connected together starting from the B+ rectifier to the last filter cap node.

Exceptions to this rule is if your build style just connects grounds everywhere on the chassis, such as when using terminal strips, as well as on Marshall designs, connecting a ground locally on the Presence circuit. If using a low voltage relay circuit, they can have their own ground, but most low voltage circuits use a bipolar supply like +15v and -15v.