[BUILD] GIX-51X tube preamp

[gemini86]

Sorry, Bruce, I've been busy tonight.

Looks like I'll need to update the BOM,

C5 and C2, Those should be 100V parts at least, best to go 250V. The original G9 BOM says 250, but it also specifies 250V for C10, which doesn't have any dc on it. In the 51X world, real estate is expensive, but I left footprints for 5mm,7.5mm, and 10mm  (I accidentally grouped C2 and C10 together on the BOM.)

The input stage is a cathodyne type, where the tube has a 47k resistor on plate and cathode, then a biasing resistor to bring the grid down below the cathode. The grid is actually up at around 80V relative to ground, or, your guitar. In a normal gain stage, with a grounded grid, there would be no need for a cap at all, as you see in your LCMP circuit.

Also need to sort out some HV converter details in the BOM- I did the math for the feedback resistors to give an output of 240V. When I did stuff those resistors, I ended up with something more like 265V, which is too high. I had to dial back the resistors some to compensate. (I'll have to look to see what's I ended up with there, I'm unable to get to my work bench right now.) The original had a trim pot which was nixed due to possible instability problems.

 

[bruce0]

Hey.. it is DIY... no apology necessary ever (the Y in DIY is me)!

Thank you for the clear explanation, I looked at the LCMP schematic and this one, and I understand a bit better.

I will wait to hear on the HV board change, I assume that is affecting the R502,R503,R504 divider?  At 265 Volts it puts several caps beyond their ratings (speaking of which... do we have to allow for tolerance on the voltage rating or does that just apply to the capacitance?  Is a 20% 250V Electrolytic potentially only rated for 200V?)
.
I already ordered from Mouser today for the 4 kits, and I did order high voltage parts for C5 and C2.


b

 

[gemini86]

The tolerance is for the capacitance. I would like to use 400V caps in these places, since they're cheap and will last longer (less stress on the dielectric), but they're huge. If you can find a 300V cap that will fit, go ahead and use it. The 250V caps will work great for quite a long time.

 

[bruce0]

Yet another warning:

C3 I ordered works but I really had to bend the leads to make it not touch the bottoms of the tube sockets which can get hot.  So another choice would be better.  I noted this in the attached partial BOM but left my part number in because it did fit.

Another Warning:
This is an update to the list, stuffing boards and ...

I ordered BFC230348474 for C5.  Unfortunately I misread the lead spacing, and it is 15mm instead of 10.  So without bending the leads this won't fit.  Sorry.  I have removed it from the list attached.  In my case I plan on trying to form the leads to make it fit. If you have not already, don't order BFC230348474 .


Warning:
This is an update to the list, I just got my order and ...

You want switches ending in 2QE-EVX (PCB pin) not 1QE-EVX (solder lug)
SW4 in this list wrong in my original list.

Sorry (SW4 still reaches, you just need to clip the lug in half so it fits in the hole.) Sheepish


In case it helps...
Here is a mouser list of what I ordered -  use at your own risk.
There may be typo's
I ordered flat bat handle switches  (don't know if they fit, bushing is shorter)
I chose PP film over PET if I thought it would fit.
The list excludes all the common resistors, I had most of them, and anything I had (certain caps).  Not sure the parts I chose will fit.

It is an excel file in tab delimited text format.  I couldn't figure how to upload excel.

 

[bruce0]

Warning:

You want switches ending in 2QE-EVX (PCB pin) not 1QE-EVX (solder lug)
SW4 in this list wrong in my original list, I will try to update.

Sorry (it still reaches, you just need to clip the lug in half so it fits in the hole.)


The text file looks like a mess, here is a pdf.

 

[bkbirge]

Beautiful start to this thread. I'm going to go put together a mouser BOM now...

 

[tonycamp]

hi Rodney, will we be notified of shipment, or should we keep an eye on the mailbox? No pressure, just wondering

Thanx
T

 

[gemini86]

I'll try to notify you as I send them out. I've been busy doing some intensive job hunting, since my current boss has decided to start treating my wife and I (we work together) less like people, and more like shop equipment. I'm still hoping to get everything shipped by the end of the week, but no promises there. (sorry)

 

[mitsos]

Hey! nothing to add to the build info, but I wanted to say that my kits arrived safely, and the PCBs and metal look great. Good job overall, can't wait to get them done.  Like you, I'm super busy for the next month or so but I should have everything here, so I'll try to sneak in some DIY at nighttime!! Thanks again for getting this done, if it sounds like it looks, it'll be awesome!

cheers!

 

[bruce0]

Rodney

I am holding off populating the Hi Volt board resistors till you let us know what tweaks you made to get it below 265 Volts...

thanks

 

[gemini86]

Crap I forgot about that...

Looking again, at it, it seems that I actually did correct the schematic. My final resistors are:

R502 - 1M
R503 - 5.6K
R504 - 560R

This has me at about 245 volts. (I am using 5% resistors, though, so you may want to measure before you stuff them, so you don't end up with a net gain of 5% low, which would end up causing a higher Vout)

 

[bruce0]

ok, I will check. 

I don't really understand how the voltage is set, but if it is too high I will consult the Max_1771 datasheet again, or ask here.

bb

 

[gemini86]

The feedback resistor network (r502,503,504) form a voltage divider that send 1.5V to the FB pin, pin3. This FB pin must always see 1.5V and will adjust the output until it's happy.

On a related note, I went to do some quick recording with my GIX when I got home from work, only to find a loud 7hz noise on the HV output. I don't know what's changed, or if my chip has gone bad, or what... I'm looking into it, but not sure if I'll have time to tonight.

I really wish we had an SMPS guy on the forum...

 

[irfrench]

Hi Rodney,

I'm looking forward to getting stuck into this once my stuff arrives.

Have you had any luck with the 7Hz issue?  I'm happy to offer help but it did take me a few seconds to work out what SMPS was...  :-\


Thanks again for sharing the project.

Ian

 

[mitsos]

7Hz? Does this make it into the audio?  Can we put a HPF filter on the DC-DC output? 

 

[gemini86]

I messed with it a bit last night. I thought maybe I had over-volted a cap or something, so I started adding parallel caps to the output. This changed the frequency of the ripple, and also the amplitude. I'm still investigating. There's not a ton of room on the converter pcb, but there's a filter cap after L503 that you may be able to fit something bigger like 47uF there.

I'm going to build up another converter today and make some comparisons.

 

[mitsos]

cool, thanks for the info.  I'll try to do some tests when I start mine, but I probably won't be able to touch it this week.

Did the extra capacitance lower the F?  Shooting from the hip here, maybe we can add an LC or RC filter? 

But again, is this thing audible (not clear from your first post about it)?  Does it affect the audio or are you just seeing it by scoping the converter output?

 

[gemini86]

Yes it's audible.

Increasing C increases the ripple frequency. A diode and a large cap would kill most of it.

I'd really like to find the source of the problem, rather than trying to put a band-aid on it.

 

[irfrench]

Hi,

I've done a bit of blind searching (I don't really know what I'm looking for) but a recurring theme seems to be the inductor/zener combination.

eg: http://www.electro-tech-online.com/repairing-electronics/113123-repairing-smps-i-lost-component.html

My apologies if this is hindering rather than helping.

Ian

 

[gemini86]

This is just a simple boost converter, no zener diodes.

The basic principle is this:

24 is applied to one side of the inductor (the big one), which flows through the diode and on to the output. The feedback resistors are connected to the output and feed a scaled down voltage to the feedback pin3 on the controller. The controller sees the voltage and if that voltage is low, it pulses the FET on. This redirects the 24V supply through the inductor to ground. The inductor builds a magnetic field in the core and starts to saturate. The controller IC has a current sense pin that reads the voltage drop across the 0.25 ohm resistor. When this voltage builds high enough to overcome the controllers internal comparitor, the fet is turned off and all that stored energy is the inductor is released into the diode in one huge blast. The voltage is dumped in the output capacitor and the voltage is again read by the feedback pin. This process repeats until the output is constant.

Right now I'm thinking the problem may be in the current sense resistor. I'll have to do some experimenting because there's not a good answer for what this value should be.

Another cause could be the feedback is getting radiated interference from the inductor. This design has been condensed into a tight space, so some filtering may be needed on the feedback pin.