Dual Amp With Mosfet CF

[zendragon63]

Hey gang, not been building much lately but was inspired by seeing sluckey's dual light (humbuckers with a JTM45 sound right but need something else for the Strat). Anyway I cobbled together a draft schematic. I know that this is primarily a TUBE forum but just wanted to ask y'all for a little advice on the merits of using 1) Mosfet CFs and 2) using a LND150 to warm up the front end of the 5879 (dial the gain back to maybe to 15-20x).

Anyway, any thoughts on these two items are appreciated. Thanks in advance. Regards

dennis

[tubenit]

I've always been a fan of Dennis's innovative work!  Check this out, he has a VVR on this amp also.  Since not everyone has ExpressSCH,  I put this into a GIF file for viewing.  The SCH file is on one page.

I like the design!   

With respect, Tubenit

[tubenit]

Dennis,

Can you comment on the LND150 to warm up the front end, please?  Not familiar with that?  IF you have used that before, how does it sound to you?

Don't know if you are familiar with the FET design on the Dumblish amps?  I think this was originally designed for an acoustic guitar but I've heard some YouTubes of D-style amps with FET and using an electric guitar and I thought they
sounded really great to me.

With respect, Tubenit

[2deaf]

The bias VVR has a current-limiting circuit (10R, 12V zener, 100K), but no inrush to justify its existence.  Meanwhile, the HT VVR has a significant inrush straight into the 40uf filter cap., but no external current-limiting circuit. 

During the initial HT inrush, the MOSFET source is essentially grounded while the gate is held at a higher voltage by the 1M pot.  The STP6NK90Z has internal gate to source zeners to prevent MOSFET death from excessive gate-source voltage, but we don't know what their parameters are.  Assuming that they can handle the current at maximum inrush, the 10K resistor on the wiper is not large enough to limit the current through the presumably 1/2W 1M pot., possibly resulting in damage and/or failure.

[2deaf]

In the bias VVR, I would like to see a capacitor on the MOSFET source for an AC path to ground for the signal coming off of the PI.  The cap. would have to be large enough for low frequencies, but not so large that the charge/discharge causes a lag when the HT VVR is adjusted.

I designed a similar circuit a couple of weeks ago with a PNP BJT, but then decided that an NPN BJT had some advantages.  The main advantage is that it reverts to full bias upon BJT failing open.  1M wiper failure also results in full bias whereas the PNP version resulted in no bias in both cases.

[zendragon63]

Thanks tubenit for your response.  I have seen the D-style FET of which you speak and think that essentially it buffers, boosts and shapes—I suppose I am trying to do the same thing but certainly not as gracefully. The D-style FET front is indeed a nice sound especially for hi-z.
 
Point taken 2deaf—a cap to ground on the output of the bias VVR is a good idea.  I have yet to have a failure on the STP6NK90Z HT VVR but it is more likely luck than skill—limiting the inrush of current on solid state stuff is a good idea period. Thanks for sharing your insights and your VVR NPN bias circuit; hope you don’t mind that there exists the probability of plagiarizing some of it….

The LND150 is kind of unique but isn’t necessarily new; it is a higher voltage JFET and can literally directly replace a triode so there is not the need for lower voltage supply lines. I built a LND150 ‘preamp’ preamp a couple of years ago on a board that wasn’t much bigger than a postage stamp. It was essentially a built in boost pedal. It had a lot of gain—something like 100x even without the source cap—and sounded fine if you need/like a lot of boost. I didn’t spend much time fine tuning it and just moved along to another experiment. The thought here is to bias it to get a very slight amount of color, significantly cut the gain to 15-20x, feed it into the 5879 to get just enough for a little more color and then selectively roll off some lows.

The Mosfet source followers as a CF seem like a good idea. No filaments to heat. Other thoughts: 280-0-280V PT may be too soft for EL34 though I very rarely crank the VVR up to full voltage on any of my amps; 5K OT may or may not pan out the sound I would like of either or both front ends.  And I will second guess this thing until it is built. The cabinet is Canary wood so at least it will look nice.

Thanks again and I’ll let you know how it turns out. Regards

dennis

[tubenit]

The LND150 is kind of unique but isn’t necessarily new; it is a higher voltage JFET and can literally directly replace a triode so there is not the need for lower voltage supply lines. I built a LND150 ‘preamp’ preamp a couple of years ago on a board that wasn’t much bigger than a postage stamp. It was essentially a built in boost pedal. It had a lot of gain—something like 100x even without the source cap—and sounded fine if you need/like a lot of boost. I didn’t spend much time fine tuning it and just moved along to another experiment. The thought here is to bias it to get a very slight amount of color, significantly cut the gain to 15-20x, feed it into the 5879 to get just enough for a little more color and then selectively roll off some lows.

Dennis,  PLEASE share how this works out for you.  I am very interested in your designs and this idea is very intriguing to me.

Thanks and respect,  Tubenit

[2deaf]

I was wondering where you were going to connect the 12V:117V transformer to.  I ran some quick calc.'s to see if the unused 5V winding would be feasible and it looks like it is.  So I altered your schematic to show my idea.  I tried to put the changes in red to show that they were not original, but my jSchem would only change colors on some components and not others.

With 5V going into that transformer, I would expect maybe 48V coming out the other end.  With a full-wave bridge and a filter cap, I would expect a maximum of 68V DC.  With a load of a couple of ma's and various losses, I guess 60V DC minimum.  10uf filter cap.'s are plenty with the full-wave bridge and low current and they will charge up faster than the 47uf's.

With the 50V zener on the far side of the 10K resistor, 10V will drop across the resistor which will be 1ma.  Now I failed to mention that you would need a resistor from the MOSFET Source to ground when I suggested placing a capacitor in the same position.  It will probably take a 10uf cap. to get a flat response to 82Hz, and with the very low current in the bias circuit it would take a long, long time for that cap. to discharge when you adjust the VVR so that the bias isn't tracking with the HT.  It will charge up real fast, however, so there is no danger of red-plating.

We want the MOSFET Source resistor to be as small as possible for fast response, but we don't want it to draw more current than the 1ma going through the 10K resistor.  If we exceed 1ma, the voltage at the MOSFET Drain will drop below 50V and the bias will no longer have a linear relationship with the HT when the VVR is adjusted.  The worst case scenario for the voltage across the Source resistor would be 50V, so we need a resistor a little larger than 50K.

I put the classic butt-to-butt zeners from the Source to the Gate as cheap protection for the MOSFET.   

   

[zendragon63]

I really appreciate the input 2deaf--I have been unexpectedly tied up with corporate folks this week but will dig into your response in the next couple of days. Again thanks! Regards

dennis

[zendragon63]

OK, I think I updated about most all the prelims. Bonehead mistakes not withstanding. I have a 325-0325 w/ 50v tap Hammond so I think I'll go ahead and use it. I wanted as few controls on the panel as I could get away with but having second guessed the master for both preamps, it made more sense to have one on each. Added the LEDs for both the pedal and the panel.

Again, thanks 2deaf and all; I often let my initial inspirations to get out in from of the technical detail skis. Board layouts completed. Now is the challenge of getting the chassis layout with some sembence of logic and order.... you may not see much from me in the next week or two but I'll post an update once I get the iron, sockets and panel assembled. Regards

dennis

[zendragon63]

Ok gang, 'tis done. Turned out pretty good overall if I do say so myself. I changed just a few values and it ended up delivering the sounds that I had intended--wonders never cease.  The Marshall side is JTM in spades and the 5879 pentode side gives a fatter sound for the Strat. The KT66 sound is pretty hard to beat in my never-to-be-humble opinion. Thanks to 2Deaf and tubenit for their initial feedback

Had one bonehead startup goof--the relay/LED power supply requires the 6.3v filament AC but those windings had a center tap which I grounded that wasn't going to work no matter how much I wished it so. Caught it with the light bulb limiter though. I ended up using the 5v windings and though it only gave me ~4.3v out, it works fine.

I had the idea to shock mounted the tubes with those silicon washer/dampners that you see on reverb tanks. I don't know how well it will work in the long run but the intention is to ease tube abuse inside the torture chamber known as 'the combo' and reduce the onset of the rattle and microphonics.

The Canarywood worked out in that I got lucky and stumbled into a 10" wide piece locally that was just long enough to do the cab.

A few thoughts on this are 1) the MOSFET source follower really works well 2) the LND150 preamp to the 5879 is really, really nice; don't be afraid to try it in this capacity. It is more of a boost rather than OD; the gain is indeed dialed back to about 30% or so and warms up the 5879 just enough within the span of the 500K volume 3) the speaker compliment is a 12" scumnico and a Weber 10A125 which work well together and 4)Doug Hoffman is a phenomenal purveyor of components.

Lastly--while I don't post much, I appreciate those of you who do and draw out the technical knowledge of the PRRs, the Sluckeys, the Hotblueplates, etc. These kinds of projects wouldn't turn out as successful without a great community like this. Regards

dennis

 

[zendragon63]

A couple more pictures:

[HotBluePlates]

... Lastly--while I don't post much, I appreciate those of you who do and draw out the technical knowledge of the PRRs, the Sluckeys, the Hotblueplates, etc. These kinds of projects wouldn't turn out as successful without a great community like this. ...

Thanks, but 2deaf supplied the technical horsepower on this one!

I hope your new amp sounds as great as it looks!

[tubenit]

Dennis,

That's an incredibly innovative and nice amp!  Fantastic!  I appreciate your sharing the schematic and photos.

I am a fan of your builds and your soundclips also.  IF you can post sound soundclips, I'd love to hear the amp (& your playing).

I have one question and that is about the "trim" pot for the LND150 preamp.  It looks like there is high voltage going right into that pot?
Is that correct?

Would there be another effective way of doing that?  I've never used something like that. I was thinking maybe there should be a capacitor between the high voltage and the pot?

The Canary wood is beautiful! Nice cabinetry work. How is the Canary wood to work with?

With respect, Tubenit

[zendragon63]

Thanks tubenit, HBP. Yes, there is 245v going into the drain resistor. I figured the voltage drop at 245v-118v=127v and then the current across it at 127v/100K=.00127 and then .00127*127, I got .16 watts and I have a .25 w pot there. I could have that completely wrong but that is my logic. It is certainly handy to vary the amount fed to the initial 500k volume pot. I suspect that there are several ways to skin that cat.

I have done a few MP3 clips some time ago but I think these forum rules are to directed toward the soundcloud or equivalent and I know little about it. I will just need to get my archaic mind into the 21st century. And I am no Jimmy Page for sure either .

Canarywood is easy enough to work with. It is a little denser and not as soft as mahogany but as long as your bits are sharp, the cuts will be clean.

And right you are HBP--my thanks to 2Deaf on the solid state recommendations. Regards

dennis

[Willabe]

Agree, the Canary wood is beautiful! 

How pricey is it, where do you get it and is it hard to get?

[zendragon63]

Willabe, only place locally (Boise) that had any Canary Wood was WoodCraft--and it was slim pickins'. $14.25/lf which seemed reasonable enough considering other outlets in the region (SLC, Portland) were significantly more expensive. That it was only piece 10" wide and 15/16" and 89" long pretty much sealed it's fate.... Regards

dennis

[Willabe]

Hmmm, maybe I'll have to go take a look at a local wood craft. 

Thanks.