BUILD: Micro-Style With MOSFET Tremolo & Boost

Discussion in 'Shock Brother's DIY Amps' started by sds1, Jan 18, 2019.

  1. sds1

    sds1 Tele-Holic

    Age:
    42
    Posts:
    991
    Joined:
    May 4, 2017
    Location:
    Orlando, FL, USA
    A friend asked me to build him a small amp. I've been looking forward to a Micro build, after considering my friend's requirements and some ideas of my own, we arrived at a design with the following:

    - Micro Bassman PP power amp and phase inverter using 12BH7 for output
    - solid state rectifier
    - Tone control
    - switchable parallel boost stage
    - bias-wiggle tremolo w/ MOSFET oscillator and blinking LED
    - Master volume
    - relay switching with 2-button footswtich

    Here's the schematic:

    [​IMG]

    Faceplate design:

    [​IMG]

    Faceplate is front-engraved acrylic. Chassis is a Hammond 15"x7"x3". All the transformers are designated on the schematic, but I chose the larger Hammond 125C output transformer (8W) and a power transformer with 315-0-315 HT secondary for a target B+ of 420VDC. I'll be building this on terminal strips only. I will probably do a quick DIYLC layout before assembly.

    I expect to do some gain stage balancing, tweak coupling caps, and blow up some MOSFET's before calling this build final. It's the most "original" build I've done so far. Most of it copy-and-paste from other builds of course, but most unique overall and therefore some uncertainties. The tremolo circuit is the biggest wildcard, I believe I can get it to work but ready to roll with the punches!

    Parts are on their way. I got some new tools I get to use for the first time -- a little nibbler, vintage radio punches, a drill press vise, some new cobalt drill bits, and a riveter. Stoked!
     
    Teleterr, Nickadermis, robrob and 2 others like this.
  2. jonrpick

    jonrpick Friend of Leo's

    Posts:
    2,238
    Joined:
    Mar 17, 2014
    Location:
    Marietta, GA
    Subscribed!
     
    sds1 likes this.
  3. jonrpick

    jonrpick Friend of Leo's

    Posts:
    2,238
    Joined:
    Mar 17, 2014
    Location:
    Marietta, GA
    I'm really curious about the parallel boost stage. I have an unused V1b in my amp and it needs a boost when I cut a lot of bass.
     
    sds1 likes this.
  4. robrob

    robrob Poster Extraordinaire Ad Free Member

    Posts:
    7,499
    Joined:
    Dec 29, 2012
    Location:
    United States
    Nice!
     
    sds1 likes this.
  5. Mexitele Blues

    Mexitele Blues Tele-Meister

    Age:
    41
    Posts:
    329
    Joined:
    Mar 7, 2018
    Location:
    Westminster, CO
    The 12bh7 is cathode biased but the tremolo wiggles the grid voltage. I'm sure that's on old trick but I dont think I've seen it before.

    It's impressive how simple a tremolo circuit can be. Curious, why does the intensity pot wiper connect to C9?
     
    John How and sds1 like this.
  6. sds1

    sds1 Tele-Holic

    Age:
    42
    Posts:
    991
    Joined:
    May 4, 2017
    Location:
    Orlando, FL, USA
    Indeed, this tremolo circuit is adapted directly from the Gibson GA-19RVT:
    http://www.prowessamplifiers.com/schematics/Gibson/ga19rvt.pdf

    It shouldn't, I've deleted that wire junction at least once already. I shall do so once more. Nice catch, thanks. :)

    Yeah I figured this would be a great opportunity to get to know a tremolo circuit. Uncle Doug has a nice video on the topic. Most of the components are with respect to phase shifting in the feedback loop. Once you remove the 3 CR filters, there's not much left to talk about that's above and beyond what you might find in any other gain stage.
     
    Last edited: Jan 20, 2019
    Mexitele Blues likes this.
  7. Lynxtrap

    Lynxtrap Tele-Holic

    Posts:
    733
    Joined:
    Feb 29, 2016
    Location:
    EU
    Interesting!
    I'm planning a micro myself, so I'm a little curious about the output stage.

    How much more wattage do you get out of the BH7 compared to an AU7? Do they need >400V on the plates to work optimally? And how does the output impedance compare between these tube types?

    Any particular reason for not bypassing the cathode resistor?
     
    sds1 likes this.
  8. Nickfl

    Nickfl Tele-Afflicted

    Posts:
    1,469
    Joined:
    May 24, 2016
    Location:
    Florida
    I see a potential issue with the way you've implemented the switchable boost. If you have the boost pot turned up all the way up and you switch the boost off, you ground the signal for the whole amp, so it basically becomes a mute switch. Even if its not turned up all the way the boost pot will act as a sort of volume control when its switched "off" because it still becomes the tail of a voltage divider and lowers your output considerably if its turned up much at all, which seems to negate the point of having it on a switch.

    It would probably be better to just disconnect the grid from the pot wiper or if you don't want to leave that grid hanging, use a SPDT switch and have it switch the grid between the pot wiper and ground. You can use a 1M resistor across the switch to prevent popping during switching.

    BTW, where did you get vintage radio punches? I've been making due with a cheap electrical box punch set and it would be cool to have some that were actually sized right for tube socket holes!
     
  9. sds1

    sds1 Tele-Holic

    Age:
    42
    Posts:
    991
    Joined:
    May 4, 2017
    Location:
    Orlando, FL, USA
    According to @robrob, whose design the power amp is copied from, the 12BH7 will provide ~30% additional output power over the 12AU7.

    Regarding the other questions, I would have to refer you to Rob. :)

    Bassman Micro project page:
    https://robrobinette.com/Bassman_Micro.htm
     
  10. sds1

    sds1 Tele-Holic

    Age:
    42
    Posts:
    991
    Joined:
    May 4, 2017
    Location:
    Orlando, FL, USA
    Thanks Nick you're right, I copied this stage from a known working design. Unfortunately I copied it wrong! This is the correct drawing:

    [​IMG]

    Regarding the radio punches, I got them on eBay! Search "vintage radio punch" and they are all about $25. I quickly see both a 3/4" punch and 1-3/16" (separate listings) available for less than $50 shipped. Run!
     
    Nickfl likes this.
  11. Teleterr

    Teleterr Friend of Leo's

    Posts:
    4,325
    Joined:
    Feb 7, 2011
    Location:
    Lewes De.
    Very cool. Are the positive and negative wave heights going into the BH7 the same size ?
     
    sds1 likes this.
  12. sds1

    sds1 Tele-Holic

    Age:
    42
    Posts:
    991
    Joined:
    May 4, 2017
    Location:
    Orlando, FL, USA
    According to my reading, the phase inverter is very well balanced - - until the power tube goes into overdrive.

    I look forward to observing this behavior for myself once Lil' Timmy is up and running
     
    Teleterr likes this.
  13. sds1

    sds1 Tele-Holic

    Age:
    42
    Posts:
    991
    Joined:
    May 4, 2017
    Location:
    Orlando, FL, USA
    I finished drilling and assembling the chassis. I had to drill a lot of holes! The layout was planned in advance on the computer and then printed to paper templates. I riveted in all of the terminal strips using my new cheapo hand riveter. I used a nibbler and machining file (both also new tools) for the IEC cutout. There are some open side holes there that will get #10-24 rivnuts as soon as they arrive, these are used for fastening to the cabinet.

    [​IMG]
    [​IMG]
    [​IMG]
     
    Paul-T, robrob and jonrpick like this.
  14. sds1

    sds1 Tele-Holic

    Age:
    42
    Posts:
    991
    Joined:
    May 4, 2017
    Location:
    Orlando, FL, USA
    It's basically working. I need to go thru the details tomorrow, balance the gain stages, etc.

    Also the tremolo isn't working. I think it's wired correctly, and I think I saw the bias wiggle on the scope. But I'm not hearing anything and I never did see a good sine wave coming from the oscillator. I suspect it's a gain thing with the MOSFET, I'll trying reconfiguring it tomorrow for better results.

    [​IMG]
     
    robrob, Mexitele Blues and Paul-T like this.
  15. sds1

    sds1 Tele-Holic

    Age:
    42
    Posts:
    991
    Joined:
    May 4, 2017
    Location:
    Orlando, FL, USA
    Just wanted to drop a final schematic by as I decided to draw out all of the low voltage stuff which includes the relays and the tremolo LED.

    The LED was a fun challenge and learning experience, in the end feed the tremolo oscillation waveform thru a voltage divider set up so it switches Q3 (a BC547 transistor) at peak signal (30VDC down to around 0.7VDC). This blinks the LED on and off in sync with the tremolo. Good times!

    The boost relay circuit includes an MPSA13 darlington transistor which is used disable the Boost panel switch when the footswitch is plugged in.

    The MIDI cable I bought for the footswitch only had 3 of the 5 wires connected, which adheres to specification but was a surprise to me! MIDI cables I've purchased in the past had all 5 pins connected end to end.

    [​IMG]
     
    Mr Ridesglide likes this.
  16. printer2

    printer2 Poster Extraordinaire

    Posts:
    6,263
    Joined:
    May 24, 2010
    Location:
    Canada
    Do you think you will you be getting much boost from your arrangement? Also the second stage has a 100k on the grid along with the 250k pot in parallel, will that not limit the drive to the stage?
     
    sds1 likes this.
  17. sds1

    sds1 Tele-Holic

    Age:
    42
    Posts:
    991
    Joined:
    May 4, 2017
    Location:
    Orlando, FL, USA
    Well, the parallel gain stage is a fairly subtle addition to the circuit. I've used it before. It thickens up the signal more than it boosts it, although there is some additional gain.

    However, I see that parallel path to ground now. I'm lacking the knowledge to confidently sort it out. Maybe you can help me.

    There's too much signal coming into V2A (no tone stack loss), you'd have to set the Volume pot pretty low to keep that stage from clipping. To keep it cleaner on the dial, I installed voltage divider R6 and R10. Just so you know the origins of those 2 resistors. They come with no other design intent.

    The 100K grid leak R10 causes effective resistance of the Boost pot to drop to around 71K. I see this now, thanks! But does this have an effect on the voltage into the V3B (Boost) grid? I'm thinking no, right?

    You specifically used the word "drive", were you more referring to an effect on impedance?
     
  18. printer2

    printer2 Poster Extraordinaire

    Posts:
    6,263
    Joined:
    May 24, 2010
    Location:
    Canada
    Non-master amps have a gain structure where the output distorts before the previous stage, the previous one before that stage and so on. So you won't be getting clipping of the second stage before the output is already there. Effectively the same structure as the Princeton. Whether you feel the knob is too low is up to you.



    There is a reason for the 1M pots that we use, not just because they look pretty. You try to have the next stage only load the preceding stage less than 10%. The preceding stage has an output impedance of about 38k, with the volume control full up the voltage out of the first stage will be split between the 38k of the stage and the 71k of your combination. So you get less than 70% of your signal to the next input. Drive in the context of voltage, sorry.
     
    sds1 likes this.
  19. jonrpick

    jonrpick Friend of Leo's

    Posts:
    2,238
    Joined:
    Mar 17, 2014
    Location:
    Marietta, GA
    Not to hijack, but I have been wondering about volume pot values. I used 250k for the gain pot and master, because the grid leaks would otherwise be 220k. Am I inhibiting gain by using such a low value?

    Hijack off...
     
    sds1 likes this.
  20. sds1

    sds1 Tele-Holic

    Age:
    42
    Posts:
    991
    Joined:
    May 4, 2017
    Location:
    Orlando, FL, USA
    No hijack, we gonna talk about grid leak resistance now. :)

    So prompted by @printer2 's post above, I did some reading and came up with 2 major considerations with respect to inter-stage grid leak resistance:

    1) Higher resistance to ground means less load on the previous gain stage.

    2) For maximum efficiency a rule of thumb is input impedance to be minimum 5X higher than the output impedance of the previous stage. As printer2 notes above, the output impedance of a 12A_7 gain stage is ~40k, so a minimum 200k input impedance.

    So in my case above the effective grid leave resistance in V2A (Volume Pot 1M, Boost pot 250k, R10 100k) is 67k -- too low to satisfy either of the considerations above.

    If I change R10 to 1M, the best I can get to is 167k. Changing the Boost pot is an option as well. Seems like ideally, all 3 resistors in play should be 1M.

    Also, I wonder if adding a capacitor before the Boost pot would decouple and essentially eliminate that pot from V2A's grid lead resistance altogether?

    Also, it's worth mentioning if you take a look at AB763 for example, you don't see much (any?) parallel grid leak resistance, so the values stay really high.

    The master volume wouldn't be part of this specific discussion (same with like a Presence control, for example), and the 250k gain pot may not be out of sorts if it's the only resistance to ground there. But it seems clear now that using a 1M would always be the most appropriate choice.
     
IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.


  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.