Hammond AO-43-1 to "something" 2x (12AX7 + EL84)

Discussion in 'Shock Brother's DIY Amps' started by Digital Larry, May 11, 2019.

  1. Digital Larry

    Digital Larry Tele-Holic

    Posts:
    997
    Joined:
    May 30, 2017
    Location:
    Silicon Valley, CA
    I have the aforementioned Hammond amp which is the "reverberation" amp for an L-100 series organ.

    AO-43-schematic.JPG
    As such it has a 12BH7 tube to drive the reverb spring and a choke to act as a high-boost for the reverb drive as well (it's the plate circuit load, so its impedance goes up with frequency and so therefore does the voltage gain).

    I was thinking to sell this but the more I think about it (and nobody's buying it) maybe I would like to try my hand at adapting this to my own nefarious purposes. My own background is suitable. I have a double-E degree from a noted university, recently involved in the admissions scandal, but I had nothing to do with that your honor! My electronics textbook's first 8 chapters are all about tubes but we skipped over that part because it was no longer relevant (cough). I know all about ohm's law and have a nice scope and the workbench is here somewhere under all this crap. But I don't have a good "feel" for tubes in spite of reading all sorts of stuff about them over the years.

    I'm thinking about how to get started with this project. Let me sketch out my initial ideas and you tell me where I'm mistaken.

    Step #1: Get the power supply/rectifier going. That section is not shown in the schematic above, but you can see that B+ over on the output transformer is +340V. As shipped this thing specifies a 5U4 as the rectifier. I don't actually have a 5U4 in hand - I have a couple other octal base rectifier tubes, but they are all old. Tube Depot sells a JJ version for about $15, doesn't sound like an excessive risk.

    I've read a few things hinting that the higher line voltages today compared to when this thing was designed may result in a too-high B+. Looks like there is a way to insert a high power zener diode into the B+ ground return in such a way that you can "lift" the voltages by virtue of a bias voltage, but I'm not super clear on how that works and how you might adapt a circuit such as the one above to using that, if it becomes necessary.

    Other thoughts, summarized as this is already TL;DR --
    a) Use the power amp section intact. It has NFB "Factory Volume Adjustment" that might allow insertion of a presence control. Being as I'm lazy and the power amp section is already built (modulo electrolytic cap replacement) I am really leaning heavily in this direction.

    b) Use 1 left over 12AX7 for the input and maybe subsequent cathode follower to drive a standard passive TMB tone control section.

    c) I've looked at some things like the Phaez Daisycutter, which uses this tube complement, and as far as sound samples it's all well into the sizzling lead and hot crunchy rhythm zone. That's great but I also use clean tones a lot and I'm wondering what it takes to get something like this to work at least "OK" in the clean zone.
     
  2. sds1

    sds1 Tele-Holic

    Age:
    42
    Posts:
    961
    Joined:
    May 4, 2017
    Location:
    Orlando, FL, USA
    Until you hit 400vdc your b+ isn't a concern. As a matter of fact I'd be looking to raise that b+ personally. Use GZ34 instead of 5U4GB for higher b+.
     
  3. Digital Larry

    Digital Larry Tele-Holic

    Posts:
    997
    Joined:
    May 30, 2017
    Location:
    Silicon Valley, CA
    Thanks for your response.

    a) Ok, do you have any idea why people might be concerned about B+ too high (> 365 V) for EL-84s then? Seems to be a fairly common perspective.

    b) What is your rationale for suggesting higher B+? I'm sure I won't need more volume.
     
    sds1 likes this.
  4. sds1

    sds1 Tele-Holic

    Age:
    42
    Posts:
    961
    Joined:
    May 4, 2017
    Location:
    Orlando, FL, USA
    Yeah sorry man, I was kinda locked onto the 12BH7/12AX7 tubes when I made that comment.

    The EL84 datasheet indicates max plate voltage to be 300VDC, so that's surely the origin of any concerns you've come across.

    On the other hand, there are amps that run them closer to 400VDC (the Peavey Classic 50 comes to mind).

    You're in grey area here with respect to maybe running the tubes on the hot side, therefore I must rescind my previous comment as "isn't a concern" is not true.

    A GZ34 rectifier will get you B+ of around 375VDC.

    See what other folks say, tube life aside the SE cathode bias may not even like the higher B+ voltage.
    I'm thinking more about the 12AX7 than the EL84, the higher plate voltage enables an increase in the voltage swing between the saturation/cutoff regions. To the player this adds up to increased dynamic response and what many refer to as "touch sensitivity".
     
    Digital Larry likes this.
  5. Digital Larry

    Digital Larry Tele-Holic

    Posts:
    997
    Joined:
    May 30, 2017
    Location:
    Silicon Valley, CA
    Since I also have an old octal 6V6 amp to mess with, I think I'm going to proceed with this one, getting the power supply and power amp sections running as designed. Then I can strategize what kind of preamp to slap on there. I don't actually have the 12BH7 tube and I'm pretty sure it's only in there for the reverb. I think it's a common reverb driver tube.

    Trying not to fall into "analysis paralysis" again which is how I got to this point - thing's been sitting there for ten years! If I get it working that's at least a place to start.
     
  6. clintj

    clintj Friend of Leo's

    Posts:
    4,916
    Joined:
    Apr 4, 2015
    Location:
    Idaho
    That PT with a 5U4GB will supply about 365V under load. Not grossly out of line for EL84 plates, but stressful to the screen grids. I'd add enough dropping resistor to get those down to under 330V, even better would be 300V, and upgrade the screen resistors to 1K/5W for added protection against overdissipation at high drive levels. That choke also does well as a power supply filter, BTW.

    That chassis is an excellent candidate for a gut and conversion to a Matchless Lightning clone or similar, you get three preamp tubes to play with for gain staging.

    Sent from my Pixel 2 using Tapatalk
     
    Snfoilhat and Digital Larry like this.
  7. Digital Larry

    Digital Larry Tele-Holic

    Posts:
    997
    Joined:
    May 30, 2017
    Location:
    Silicon Valley, CA
    Hmmm something else to consider now! I better make a project binder. I bought one of those cheap variacs from eBay for starters.
     
  8. Digital Larry

    Digital Larry Tele-Holic

    Posts:
    997
    Joined:
    May 30, 2017
    Location:
    Silicon Valley, CA
    Started looking at the actual amp last night. The Hammond schematic calls for fuses in every single one of the secondaries, i.e. for CT windings that's 2. So a total of 5 fuses on the secondaries. In the L-100 organ those were all off in a separate location so there are no fuse holders at all in this chassis. I think I am gonna cheap out and just put one on the primary side.

    I'm leaning towards a Daisycutter as described here. I'm not 100% convinced that I am gonna like an EL-84 based amp and then there's the variable of which speaker to use with it. At the moment I can't tell if I will want to keep this or sell it after I get it working. My motivation is more of a technical learning exercise than needing to have a rawkin' toob amp, if that makes any sense.

    I will probably change my mind 15 times in the next day.
     
  9. Digital Larry

    Digital Larry Tele-Holic

    Posts:
    997
    Joined:
    May 30, 2017
    Location:
    Silicon Valley, CA
    Let's talk about phase invertors shall we? I see the Hammond and Daisycutter both use a cathodyne or concertina type, where a single triode stage is used with roughly equal loads in cathode and plate circuits, so that the current through the tube makes one go up and the other go down, hopefully at about the same amplitude. The Matchless Lightning uses 2 triode sections.

    Here's an article I was hoping would shed light on the subject and it sort of does, but degenerates into a word salad that doesn't really explain what is going on. I will admit that the concept of using a higher current handling tube as the PI makes sense so it doesn't run out of steam (current that is) so to speak.

    http://www.guitaramplifierblueprinting.com/files/Phaseinverter.pdf

    My goal is to have an amp that works well over a range of gains. It's hard to tell with the Daisycutter video demos because most of those are cranked and I have no idea what it sounds like clean.

    The Matchless Lightning

    https://el34world.com/charts/Schematics/files/Matchless/Matchless_lightning.pdf

    looks like it uses a "long tailed pair" if I am interpreting that correctly.

    Call me demanding but I am not sure I'm going to be able to sustain interest long enough to try all possible variations.

    One thing that looks like it sounds unpleasant is crossover distortion. I know that in the solid state world, crossover distortion is dealt with by having tons of open loop gain along with lots of negative feedback. Some tube amps uses NFB (like the Hammond) and some don't (like the Matchless). I'd probably like to be able to start with inherent low crossover distortion so I can try both approaches on the NFB angle.

    Here's another resource which is a bit heavier on specifics, which I like.

    http://www.valvewizard.co.uk/cathodyne.html
     
    Last edited: May 13, 2019
  10. Digital Larry

    Digital Larry Tele-Holic

    Posts:
    997
    Joined:
    May 30, 2017
    Location:
    Silicon Valley, CA
    OK I took the first step on this which was to check out the power section. I measured the DC resistance on all transformer windings with the tubes removed. They all seemed kind of reasonable, although I was curious about the 110V primary, at only 3.5 ohms that would pull about 30 amps if it was resistive!

    Checked the resistance on the existing electolytics in the power supply section. No shorts. Although my plan is definitely to replace them.

    Fortunately I got a 5 amp variac and dialed up the voltage gradually. Checked the AC secondary voltage on the high voltage winding. When I got the input to the 117V mark that was about 750 volts AC. Then I powered off and popped in the 5U4GB I just bought, then brought the voltage up slowly again. Got up to about 280 VDC on the first cap at around 50 VAC in. Turned off the AC and waited until the cap voltage was below 10 VDC, then I added the 2 EL84s I just bought.

    Ramped up the AC again while watching the +340 tap (output transformer center tap). It got to 340 VDC with the input voltage about 90 or 100 VAC. THAT's interesting, as the schematic does show 340 VDC with a 5U4. I wonder if the 5U4GB rectifier tube is not dropping as much as expected with a 5U4?

    So next I will see about using the 12AX7 currently in there as a cathodyne phase invertor and a preamp stage immediately prior. This is all using the existing design which is designed as a reverb return amplifier and seems to have some shelving EQ built into some of the plate and cathode loads.

    hammond ao-43 power supply.png
     
    Last edited: May 18, 2019
  11. Nickfl

    Nickfl Tele-Afflicted

    Posts:
    1,411
    Joined:
    May 24, 2016
    Location:
    Florida
    Those organs often had several different amplifier modules all being supplied by the same power transformer. Depending on what that unit was designed for, with only 2 EL84 tubes you may be drawing a good bit less current than it is in the schematic that specifies 340v, in which case you would see a higher rectified voltage.

    Edit: Its on that power transformer schematic in your post: 160ma is what the PT is designed for, but the 2 EL84 tubes in your power section schematic are only drawing about 80ma.
     
    Last edited: May 19, 2019
  12. Digital Larry

    Digital Larry Tele-Holic

    Posts:
    997
    Joined:
    May 30, 2017
    Location:
    Silicon Valley, CA
    Fair enough. This did obviously send the supply voltages elsewhere on a block connector which I just ripped out!

    Also the test voltages on the schematic assume a signal going into a load. In any case I will use the variac to keep things under control during development. Maybe when everything's up and running I'll need to add a higher value resistor inline or a power zener in the ground return. Also thinking about moving the choke which is part of the reverb driver plate load into the power supply section, which would drop a little bit more.
     
  13. Nickfl

    Nickfl Tele-Afflicted

    Posts:
    1,411
    Joined:
    May 24, 2016
    Location:
    Florida
    Do you have a lightbulb limiter? The variac is well and good but a current limiter to check for and protect from shorts is more important IMO.

    The variac is only really useful if you're trying to go easy on old electrolytic caps like the ones that you've got in there while you're testing it right now. Or if you want to temporarily change the voltage to test it at a different voltage than what your transformer is giving you, but that's not really something to worry about until later in the process and as you said you can set that with a zenner or dropping resistors or whatever you like.
     
  14. Digital Larry

    Digital Larry Tele-Holic

    Posts:
    997
    Joined:
    May 30, 2017
    Location:
    Silicon Valley, CA
    No, I don't have something like that set up yet. It's a good idea. The variac is fused and we've managed to get this thing powered up also now with the phase invertor and previous preamp stage but I'm not getting anything (not even hum) on my speaker but I'm not sure it works. It's an old "Ampro" speaker that goes with this 16mm film projector I got from my grandfather. 1X12 convertible open/closed back cabinet! Also holds the take up reel.

    By the way the choke used in the reverb circuit has about 500 DC ohms, so if I used it in the power supply, at 160 mA nominal current (as marked on the schematic) that would be an 80 volt drop, also let's see P = V^2/R, 80 x 80 = 6400/500 = 12.80 watts, ouch, not sure that's practical. I might give that a pass for the time being.
     
  15. Nickfl

    Nickfl Tele-Afflicted

    Posts:
    1,411
    Joined:
    May 24, 2016
    Location:
    Florida
    If you are only going to use the 2 EL84 and 2 12ax7 you are probably only going to be at 80-90ma total current draw FWIW. Still probably pushing it to use that choke on the whole supply though. In the schematic it is seeing about 25ma of current supplying that 12BH7, so I'd guess its not rated for a whole lot more than that, maybe 50ma at the most? You could still use it in the standard Fender position, putting it in the power supply between the EL84 plates and screens, then it would only have to handle the current for the power tube screens and the preamp tubes, so maybe 10ma.

    But, you won't see a lot of voltage drop across it then and as ClintJ pointed out you should probably get the screen voltage down closer to 300v (and definitely put some screen resistors on there too). If you have 365v at the OT center tap and you want 300 on the screens, you are going to need something around 6.8K for the dropping resistor to the screen grid node, so maybe better to just save that choke for another project.
     
  16. Digital Larry

    Digital Larry Tele-Holic

    Posts:
    997
    Joined:
    May 30, 2017
    Location:
    Silicon Valley, CA
    Normally I'd be too proud to ask and just go look it up (which I will do anyway), but I'm gonna get beyond that and ask:

    a) What does the screen circuit do? My level 1 understanding is that if the screen is held at a constant voltage then the grid is isolated from the voltage movement of the plate, reducing the Miller effect.

    b) How do you decide which voltage the screen should be at? And then how do you calculated the resistor which does that?

    c) Should the screen be bypassed and if so how does one calculate the required cap?

    I read some 15 year old forum post (different forum) from a guy who was experimenting with different screen resistor values. He concluded that little or no screen resistor was best, and that terrible awful distortion came with increasing it, but it doesn't sound like he bypassed it at the screen.
     
  17. Nickfl

    Nickfl Tele-Afflicted

    Posts:
    1,411
    Joined:
    May 24, 2016
    Location:
    Florida
    Some of your questions answered here and here, though in reference to small signal pentodes, not power tubes. I highly recommend you get a copy of Merlin's book, it goes into all of this in more detail and you'd probably get a lot out of it.
     
    Digital Larry likes this.
  18. peteb

    peteb Friend of Leo's

    Posts:
    3,836
    Joined:
    Apr 25, 2003
    Location:
    Cascadia

    right, the high DC voltage on the plate or the screen pulls the electrons from the cathode. the signal voltage on the plate detracts from this. when the signal voltage is at its highest point on the grid, the signal on the plate is at its lowest, the plate and grid are out of phase with each other. the screen, screens or masks what is going on at the plate. The DC on the screen and plate are affected by power supply sag, but when the AC on the plate drops, the screen does not drop in voltage. the screen has no AC voltage on it. The screen is AC grounded to the B+ for this reason.
     
    Digital Larry likes this.
  19. clintj

    clintj Friend of Leo's

    Posts:
    4,916
    Joined:
    Apr 4, 2015
    Location:
    Idaho
    The screen resistors serve to limit max screen dissipation under overdrive conditions. When plate voltage dives towards zero, the screens will, because of their positive charge, start drawing the lion's share of the tube current. Having a resistor at each socket causes the screen voltage to sag when this happens, limiting the dissipation to a safe value and preventing them from getting hot enough to sag and short the tube out. One of those articles may describe that under the idea of "sliding screen" operation. Limiting the screen node voltage helps do the same thing.

    1k is pretty transparent on an EL84. Larger values can cause added compression and slightly reduced power/efficiency due to more screen sag. Vox and others used 100r, and really only got away with it because of how rugged NOS Mullard tubes were and, like old Fenders, weren't designed with an eye to modern rock distortion.

    As far as current numbers to play with for design, figure the screen idle current to be roughly 7% of the plate when calculating your dropping resistor values.

    Sent from my Pixel 2 using Tapatalk
     
    Digital Larry likes this.
  20. peteb

    peteb Friend of Leo's

    Posts:
    3,836
    Joined:
    Apr 25, 2003
    Location:
    Cascadia
    From the Blencowe site: from Nicks post #17

    “The screen voltage is usually held more-or-less constant so that electrons are accelerated towards it, and after they pass between the grid wires it is too late for them to slow down and do anything else, so they crash into the anode, whatever the anode voltage happens to be”

    And

    “Since the screen grid is fairly fragile, it can only take so much before it melts. It is therefore essential not to exceed the rated maximum permitted screen voltage or average power dissipation!”
     
    Last edited: May 20, 2019
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.