DIY Gates 5215

Help Support GroupDIY:

RSRecords

Well-known member
Joined
Jun 8, 2009
Messages
284
Thinking about trying a 5215 build. I've got a few 5879s and some transformers that I think will work Stancor WF-20 (A-10) and a UTC a25. Any suggestions on where to start for a power supply? Planning to use DC heaters. Unsure of the current requirements.
 

Attachments

  • gates 5215.jpg
    gates 5215.jpg
    112.7 KB · Views: 165
Last edited:

emrr

Well-known member
Joined
Apr 12, 2006
Messages
8,054
Location
NC, USA
Save the A-25 for a direct feed output. Plenty of other options for this thing. Current is a couple mA B+.
 

RSRecords

Well-known member
Joined
Jun 8, 2009
Messages
284
I've got a few other 600ohm output transformers. Since there no dc what should I be looking at for spec's?
 

RSRecords

Well-known member
Joined
Jun 8, 2009
Messages
284
Thanks! I guess that gives me some room for experimentation.

I did a quick board layout. First attempt at a tube PCB so welcome to any feedback. Probably do a seperate board for the psu
 

Attachments

  • 2022_01_23m5215 .png
    2022_01_23m5215 .png
    28.3 KB · Views: 126

rock soderstrom

⚽️⚽️⚽️⚽️⚽️
Joined
Oct 14, 2009
Messages
1,471
Location
Berlin
Without a deeper analysis, the following points stand out to me.

The solder pads and traces could be larger/wider.

The footprints of the HV (de)coupling capacitors are too small. Due to the high voltage, these are already quite large even at small uF values.
 

RSRecords

Well-known member
Joined
Jun 8, 2009
Messages
284
Good points I’m planning to make the B+ and filament 50-60 mil. That enough?
Should I go I’ll figure out a BOM to make sure the capacitors will fit and beef up the pads. What about the GND plane on the bottom? I generally do a gnd plane on all SS designs so stuck with what I know. Plan is to keep the gain fixed and use a -20dB pad on the input and a variable pad on the output. I might do a connector but I’ll stick with wire pads for now
 

rock soderstrom

⚽️⚽️⚽️⚽️⚽️
Joined
Oct 14, 2009
Messages
1,471
Location
Berlin
Disclaimer: I am not an experienced PCB designer, maybe one step ahead of you! ;)
I don't use ground planes for tube circuits, I prefer ultra fat traces up to 3mm and solderpads for the big components with a diameter up to 4,5mm.

Here as an example a project of mine. A headphone amp called Schwarzwald, as an idea how I do it. There are certainly better PCB designers, I find the SilentArts PCBs for example very well designed.

Screenshot_Schwarzwald HPA2.png

Screenshot_Schwarzwald HPA.png

PS: the filaments were freely wired here for flexibility reasons. Regarding the part sizes: I always print out my layout on paper first and check if my parts fit before I go ahead. My designs are always one sided, because I am "self etcher"
 
Last edited:

RSRecords

Well-known member
Joined
Jun 8, 2009
Messages
284
What voltage should I expect at Pin7 V1? roughly

Traces and Caps sized up.
 

Attachments

  • 2022_01_23 m5215b.png
    2022_01_23 m5215b.png
    36.8 KB · Views: 70

rock soderstrom

⚽️⚽️⚽️⚽️⚽️
Joined
Oct 14, 2009
Messages
1,471
Location
Berlin
What is your planned B+ voltage? Let's assume about 250V, then I would also take that as a minimum for the capacitors, although at pin7 the voltage will be lower, depending on the flowing screen grid current. Think of it as a safety reserve.

It is interesting that the V1 screen grid goes to the cathode of V2 via the capacitor instead directly to ground. Why is this done? Some kind of feedback?
 
Last edited:

RSRecords

Well-known member
Joined
Jun 8, 2009
Messages
284
Yes it’s some kind of feedback. I thought that was strange too but there’s a blurb about it the gates m2515 brochure. I’ll see if I can find it later.
edit: here's the link to the doc describing the feedback.m2515a doc
"to compensate for differences in tube characteristics, reduces distortion and noise"
This is for the version with ef86s but I think the schematic is relatively unchanged..

I was thinking of adding pads for cathode bypass caps too for some experimentation.
 
Last edited:

lassoharp

Well-known member
Joined
Jan 3, 2009
Messages
2,085
Location
USA
May as well try the Edcor 15K:600 2W for this circuit . . . . you almost can't afford not to.

If you make a pair and put a 600ohm attenuator (the hairball T pad or a vintage ladder) on the output, and run them in series you will have a single mic pre path that will give you enough gain to cover using a ribbon on light picking acoustic guitar and other soft sources while being able to maintain pretty good S/N ratios.
 

RSRecords

Well-known member
Joined
Jun 8, 2009
Messages
284
May as well try the Edcor 15K:600 2W for this circuit . . . . you almost can't afford not to.

If you make a pair and put a 600ohm attenuator (the hairball T pad or a vintage ladder) on the output, and run them in series you will have a single mic pre path that will give you enough gain to cover using a ribbon on light picking acoustic guitar and other soft sources while being able to maintain pretty good S/N ratios.
I was actually thinking about that. I may morph this into an M5382. Aside from pinout I wonder what would need to change from the ef86s in this schematic. The output iron would need to handle dc now
M5382
 

emrr

Well-known member
Joined
Apr 12, 2006
Messages
8,054
Location
NC, USA
You can’t get that output transformer, for one thing. Not sure anyone makes SE w/ feedback.

there are 5879 versions of that amp. Dig through all the 9 pin schematics including the consoles, they’re all more or less the same.
 
Last edited:

rmburrow

Well-known member
Joined
Oct 8, 2013
Messages
197
What is your planned B+ voltage? Let's assume about 250V, then I would also take that as a minimum for the capacitors, although at pin7 the voltage will be lower, depending on the flowing screen grid current. Think of it as a safety reserve.

It is interesting that the V1 screen grid goes to the cathode of V2 via the capacitor instead directly to ground. Why is this done? Some kind of feedback?
Rick. the Gates supply for the Yard and Gatesway consoles using the M5215 preamp puts out 310 volts dc regulated. The heaters are run on AC in the console. Using DC for the heaters may get improvement on hum or noise. Quiet 5879 tubes can be found and are less expensive than the EF86.

You might like the choice of UTC A-10 or A11 for input, and UTC A-24 for output transformers for this preamp.
 

rock soderstrom

⚽️⚽️⚽️⚽️⚽️
Joined
Oct 14, 2009
Messages
1,471
Location
Berlin
Yes it’s some kind of feedback. I thought that was strange too but there’s a blurb about it the gates m2515 brochure. I’ll see if I can find it later.
edit: here's the link to the doc describing the feedback.m2515a doc

This is for the version with ef86s but I think the schematic is relatively unchanged..

I was thinking of adding pads for cathode bypass caps too for some experimentation.
Thanks for the link but this text describes only the typical V2 anode to V1 cathode negative feedback.
The V1 screen grid to the V2 cathode reference is still mysterious to me. Any further ideas?

In my Gates SA70 build (similiar two stage amp but with both pentodes in triode mode) I made the cathode capacitors switchable.(see attachment) This allows you to make the amp cleaner, which makes sense for certain applications. A switchable input pad is also quite important.

On the subject of tubes the EF86/5879 are fine pentodes, the predecessors 6J7/6SJ7, their HQ variants and 6C5/6J5 triodes are very good alternatives. I love them and as long as you stay with the steel tubes they are still very cheap. (y)
 

Attachments

  • ST70 Layout.png
    ST70 Layout.png
    158.6 KB · Views: 56
  • 20220125_211952.jpg
    20220125_211952.jpg
    181.5 KB · Views: 95
Last edited:

emrr

Well-known member
Joined
Apr 12, 2006
Messages
8,054
Location
NC, USA
There's some similar strange links in some variations of some Gates amps, and not in others. Lost wisdoms of the ancients, tweezes we no longer care much about for whatever reason. Frequently I find no real difference if I switch to a more standard connection with these things. Curious.
 
Top