Twisted Snowflake

[dorrisant]

I got the idea for this a while back.

Henry's Snowflake seemed like a great circuit... High plate Skyliner with a built-in Dumbleator. Some say the 6V6s are too weak to reproduce the full bandwidth. But what about the JJ6V6S? What if the cathode driver stage was sandwiched in there? Sound interesting? You bet!

SnowFlake Layout.jpg

So... I had a Fender SKB 100 (100 watt solid state bass amp) from '85 or so. Not working, not even gonna try to get it going.... but that chassis and headshell though... that part is a win. It was was given to me for free... "Here, you throw this away if you can't fix it." Score!!

I had a set of Silverface Deluxe Reverb PT/OT on the shelf. Removed the SKB PT... bolt hole pattern is the same as the Deluxe PT. Yes! I nibbled out about 1/16" across two sides of the laydown hole and it fit perfectly! Another win. Its gonna happen for sure now. Ordered a Hammond 194A choke to complete the TX set.

SKB Chassis.jpg

The red circles show where the existing holes are in front and back... this way I know how to lay out what I have and avoid a bunch of off-center drilling for the front and back holes (avoiding this PITA).

For my version of this main board I added an HT fuse and dropped all of the bias supply components in lieu of the bipolar supply (thank you Martin) which is on a separate board. The tube used for the cathode driver has no on-board components, so nothing to add to the main board for that. Henry had included the relays on the main board, no need to deviate there.

Deputy Preamp.jpg

You can see the shape of the board on the chassis layout... it is stripped to just an outline and mount holes so that I know what to transfer to the chassis after placement is done. Yes, I put the chassis under the mill for all holes. It makes the whole thing go together like a kit when done this way.

I made provision for the relay supply rectifier like Henry had laid out, but this soon changed for the better. I used the 5v rectifier taps on the PT for a much better supply... I wanted to use a footswitch with LEDs anyway.

[talbany]

I actually built a few of these amps (6l6 versions) after Henry posted it here. Henry was the king at cram mage (everything on one board and in a small chassis) i also built a few of his mini D-lators as well. The result?
In my case none of these amps sounded very good. the clean channel was do-able but the OD side always sounded rough and a bit noisy and difficult to control? so I sold them immediately. I cannot tell you why these amps didn't sound good. My only conclusion that made sense was? somewhere in the layout. That, and simply (in my case) too many components crammed into to small a chassis.


Hopefully you will have better luck

Tony

[bepone]

in high gain amps as wide as possible is the ticket with layout.. squeezing things cannot lead to good.. or one need to think a lot about lead dress why, where, shielded not shielded..

[dorrisant]

I didn't mean to mislead... I already built this thing. It is a monster. It sings very easily. No extra hum, no oscillations, etc.. I would build this one again if needed. I superimposed an ODS preamp board over my layout of the Snowflake and it is very much the same spacing of the components. It does have the Dumbleator components sandwiched between the OD and PI... this lengthens the preamp board a bit to accommodate but I don't have any issues with interactivity or the like. The cathode driver stage added makes for a bandwidth that rivals a 50 Watt 6L6 power section.
I had to change the dropping string to get the plates reasonable. I will probably change a little more there to get them down a bit more. Not much else to change to get it to function as expected.
Yes 483v on the power tubes but they are JJs. I would suggest to all to try this. It really punches with authority. I had used a MOSFET B+ Reducer to get the voltages down to approximately 400v, but it was not working with the bipolar supply for the cathode driver. It essentially unbalanced the voltage doubler and left one side of it with negative voltage across one of the filter caps... I decided to just remove it since I am using JJs anyway. I reserve the right to change my mind on this in the future.
As mentioned before, I changed the relay supply board for another circuit. Too much relay chatter. This is what I ended up using:

5v Relay Rec & Reg v2.jpg

I had some leftover tolex from a Bogner re-wrap. I think it looks pretty sharp. Maybe suede next time. Here is a clip to analyze... https://youtu.be/9aVHLWoE8zs

Here is an unedited track (sorry about the first part) that shows off some of the OD. Casey had never realized how the ODS circuits can sing. He held a note at about 1:50 and couldn't believe what he was getting. Turn it up very loud and you can hear him say something about it at about 2:15. SM-57 into a Behringer UMC202HD using Cakewalk.
https://soundcloud.com/anthony-dorris-2 ... al_sharing

Also of note, I put in a switch on the back panel to remove the LNFB on V1. It works nicely and I would suggest trying it. It makes a noticeable difference. I added a dropping resistor between the screen node and the individual screen resistors. Screen voltage was above plate voltage otherwise.

Voltages:
Clean
V1-1 194.9
V1-3 1.768
V1-3 205.4
V1-8 1.771

OD
V2-1 207.5
V2-3 1.650
V2-6 214.9
V2-8 1.684

FX Loop
V3-1 299.0
V3-3 17.98
V3-6 135.7
V3-8 1.215

PI
V4-1 310.1
V4-3 61.69
V4-6 301.8
V4-8 61.69

Cathode Driver
V5-1 137.2
V5-3 -45.64
V5-6 137.2
V5-8 -45.69

JJ6V6S @ 24mA
V6-3 483.7
V6-4 482.0
V6-5 -45.60

JJ6V6S @ 24mA
V7-3 483.6
V7-4 482.10
V7-5 -47.67

[martin manning]

That sounds really great! I think maybe the high voltage is a good thing!

[Colossal]

Very nice work, Tony!

[Guy77]

Nice build! I see you also like to run your cathode follow plates low ( below 200v) I find that when they get too high the amp gets a little too tight for my taste and losses some of that nice compression and organic feel.

Cheers
G

[dorrisant]

Thanks guys!

Yes the voltages on the cathode driver work better now that I got the B+ up higher which gave about 200v across each triode. I have been using Martin's design quite a bit. Nothing but good things to report. Also, a bit more headroom works out well with the higher B+ as well, which is a nice change.

If anyone needs boards for one of these, just let me know.

[martin manning]

Yes 483v on the power tubes but they are JJs. I would suggest to all to try this. It really punches with authority. I had used a MOSFET B+ Reducer to get the voltages down to approximately 400v, but it was not working with the bipolar supply for the cathode driver. It essentially unbalanced the voltage doubler and left one side of it with negative voltage across one of the filter caps... I decided to just remove it since I am using JJs anyway. I reserve the right to change my mind on this in the future.

I do not think the bipolar supply and a MOSFET B+ reducer on the PT CT will ever work together since the AC voltage at the bias tap is shifted down by whatever the B+ reduction is, and it is no longer centered around zero.

BTW, did you make a schematic for this amp?

[brewdude]

Do you have a schematic?

[dorrisant]

I will have to draw one up... cobbled schematic fragments as per usual. Let me see what I can do. I think Jeff posted the Snowflake schematic in editable format. I'll find it, edit and post.

Edit: On second thought, I will not repost a corrected .sch schematic. I hate the Express program. I will do it with Kicad and post from that instead.

[dorrisant]

]Here's what I have for a schematic. I've stared at it too much... Let me know if you spot any errors, I will correct them.

Deputy Schematic.jpg

Also, Henry's schematic for reference.

[attachment=1]SnowFlake Schem.GIF[/attachment

EDIT: Changed the mid cap position to correct location, added 270k to ground at the power tube grids. Last edit was to correct the relay schematic symbols.

[WiderGates]

Hello,
cathodes of 6L6 in Henrys schem are not grounded.

[martin manning]

Thanks, Tony, very nice! I would add a 220k to ground from each of the CF driver cathodes to keep the Vh-k down at start-up.

[dorrisant]

Hello,
cathodes of 6L6 in Henrys schem are not grounded.

Should I do the same thing with my version? I didn't draw that one up. You can take it up with Henry... That is an obvious omission. I drew the one above. I have no control over the one that Henry drew.

Thanks, Tony, very nice! I would add a 220k to ground from each of the CF driver cathodes to keep the Vh-k down at start-up.

Is this in any of the layouts or drawings? I haven't seen that anywhere that I recall. Not that I would ignore your advice, but I have at least one amp built with this, the JM Sig that you helped me add the CF driver too. It has countless hours out in the gigging world as I type. No problems reported.

If you don't mind, can you explain how this helps?