Crazy talk yields next build project...Revibe!

[PRR]

......I see this as a playground for gremlins. ..............

[BobSmith]

As I understand, the revibe works great when the oscillator is using the R/C on the cathode of the oscillator. Right? If so, the solution seems obvious to me... Dump the LED and switch then connect the R/C directly to the tube socket like Hoffman did.

There should be a solution for using the LED also. I know my unit works fine with the LED, but I did several mods that deal with tweaking the LFO. If I were to replace the LED with the R/C, I'd probably have to re-tweak the LFO. It may be that trying to switch between R/C and LED is futile.

At first I though that your extra wide chassis may be a factor. It's still a concern for me because it requires your wires to the tube sockets to be very long. I see this as a playground for gremlins.

Adding the LED is mainly for eye candy with the revibe, but something else happens when you use the LED. Mainly, the LFO output will be considerably larger (and as you've noticed, the frequency range shifts slightly). I think this increase in LFO signal is at the root of your "weirdness". I'd like for you to look at that LFO signal. Connect your scope to the hot side of the INT pot. Set the scope to view a 5Hz sine wave. I also like to set the speed exactly for 5Hz as measured on the scope just for a reference. You should have a clean sine wave about 18V_pp when using the LED. Now switch to R/C. What is the amplitude and freq of the LFO now? The freq shift is probably slight, maybe only a fraction of 1Hz.

You also liked a 3M pot better than the 10M pot. There's a 4.7M resistor in series with the INT pot. This forms a voltage divider. The 10M pot will give you a much bigger LFO signal to send to the modulator. The 3M pot will give a considerably smaller LFO signal.

I think all this means you can probably tweak the LFO for good results with either LED or the R/C but probably not both. As John Sebastian once said, "Say yes to one and let the other one ride."

Still very curious about the LFO signal amplitude/freq changes when you flip the switch.

Indeed R/C sounds good, but the LED is really cool.  I think if I can get the LED in a happy place, the R/C will probably work too.  (I just have that feeling...)

So I did the test you suggested.  I dialed in a speed of 5hz on the R/C (about noon on the pot) monitoring the wiper and hot side of intensity.  I had the intensity down all the way (because as noticed that as I bring the intensity up on the dial, the intensity voltage dropped...not sure if that’s normal).  I didn’t get 18v peak to peak, but rather only 7.2v (?problem?). See picture 1. 

Then I switched to the LED. Settings on knobs same (noon on the speed, off intensity).  The speed dropped down a bit to 4.3hz.  Intensity went up to 7.3v (up about 0.1v).  Still a nice sine wave.  See picture 2.

I made a video of this test and some other tests.  I demonstrate the intensity voltage drop when turning up (?problem?) and also the noise I hear (and see on scope) with high speed and intensity on LED switch.  I’m not sure what the noise is coming from.  Is it maybe a tube?  It’s like something is discharging voltage to ground.  Even monitoring the reverb side seems to pickup some noise when this happens?

I’m a little puzzled about where to go.  I’m happy to try other test cases you may propose.  I’m hoping to keep the LED and R/C circuits together.  The Flashing light is very cool.

I’ll post a link to YouTube when it uploads...

[BobSmith]

9+ minutes of me narrating the test above. 

[sluckey]

Enjoyed the video. Wish I could do that. Nice test equipment too. I have the grand daddy of your function generator, 3310B, knobs rather than push buttons. BTW, what is your day job and electronics background. I suspect we may have similar backgrounds.

I made some measurements on my revibe this morning for you to compare.

My reference was made at the INT pot high side with Speed set to 5Hz and INT pot max CCW (zero). My amplitude was 16.8Vpp. With INT set to max CW the amplitude dropped to 8.8Vpp. This is normal operation because now there is 2.47M (470K + 1M +1M) resistance in parallel with the 10M INT pot. This changes the voltage divider ratio set up by the 4.7M resistor and 10M pot such as to reduce the signal available at the high side of the pot. Nothing unusual about that.

Next, I checked the frequency/amplitude range with INT set to max CCW. Low freq = 3.09Hz, amplitude = 14.6Vpp. High freq = 10.17Hz, amplitude = 11.2Vpp.

Finally, I checked the frequency/amplitude range with INT set to max CW. Low freq = 3.09Hz, amplitude = 8.16Vpp. High freq = 10.17Hz, amplitude = 5.76Vpp.

As you can see, the LFO amplitude drops off at higher frequencies. This is normal but I changed C27 and C18 to allow higher amplitude at higher frequency. This will increase your max speed. (NOTE... the oscillator will stop oscillating if the amplitude drops too much.) You may want to change these caps also. Refer to the notes on my schematic.

I never had any of those hash spikes on my waveform, regardless of the SPEED and/or INT settings. Clean all the way. I believe your hash spikes and low amplitude are related and hopefully will be an easy fix. Start by replacing the tube. Roll several 12AX7s and keep the one that gives the highest amplitude. If still low amplitude and/or hash spikes, replace the LED. I am using a 5mm red LED. If still no joy, check DC voltages on the tube pins 1,2,3 and 6,7,8. Disable the oscillator with your footswitch so the voltages will be steady. Also measure the B+ node voltage that feeds the oscillator tube.

Good luck and report back. I've attached a scope pic but it just visually states some of what I said above. I think your sine wave looks better than mine. Mine seems a little bent on the negative half cycle. Oh yeah, all I did was scope my revibe this morning. I did not play guitar through it.

[shooter]

Nice test equipment too

+1
I got the ugly second cousin to his scope 1012B 

[BobSmith]

Once upon a time I was an electrical engineer in the semiconductor industry, but I retired from that and now doing something in a totally different industry.  Let’s just say I fix things for a living...  ;)

So, I started getting desperate and cast a wide net on causes.  I rolled the tubes, (all three 12ax7 performed the same), tried another reverb tank (I thought I was hearing the springs rattling, and I thought maybe I was too rough with it when removing all the glue from the small foam block above the spring).  All the same.  I had previously chopsticked everything pretty hard, but I tried again, and think I found the problem. 

The wire going to the grid of 5B crosses over the turret lacing that has the 5B plate voltage, when lifted out of the way, the noise ceases.  That wire is very sensitive.  After playing with it enough, it has to be pretty far from the line to remove all traces of the noise from all combinations of sweeps of speed and intensity.  See picture below.

Although Hoffman wire could probably hold itself like that, I can’t leave it that way.  I could run it the other direction and then under the board and hope to avoid the trouble spot, or (and I’m thinking this is the right way to go) use shielded cable with the ground going to my bus bar.

I made another brief video showing how the wire creates the problem on the scope.  I don’t know know why I’m getting the lower voltage than the Sluckey Revibe, but the performance is good otherwise and not related to the noise solution it seems.  I’m thinking that our differences in C18 and C27 may be attenuating more of the signal.

I’m also posting my DC voltages on the tubes.  (I’m using all Mercury transformers, same as for the Reverb unit).

V1
1-182
3-1.48
6-181
8-1.46

V2
1-141
3-1.90
6-147
8-1.99

V3
1-23
3-318
4-327
8-23

V4
1-288
3-3.28
6-293
8-3.28

V5
1-128 (R/C), 147 (LED)
3-1.31 (R/C), 1.73 (LED)
6-170
8-1.36

[sluckey]

Yay! I would use shielded wire grounded on one end. The buss should be fine. (long wires, gremlins' playground  )

I'm pretty sure that if you change those two caps our voltages would be about the same. Not really a reason though since you are happy with it as is. I changed them because they were attenuating the higher freq range (I wanted to go up to 10HZ) so much that the oscillator would occasionally drop out. Of course they also attenuate the low freq range )not as much though) but I was not having problems with drop out at 3Hz. Anyhow, once I boosted the signal amplitude I could run from 3Hz to 10Hz with no drop out. I don't recall what the amplitude was before I changed the caps. The 5.1M across the speed pot was all part of this too.

This has been a good refresher for me. I see a couple of my notes no longer make sense to me so I may explore that a bit more and correct the notes to be more meaningful. As a bonus, it gave me a real world exercise for my new chinese scope. Can only look at a function gen before getting bored!

BTW, I've been working radar systems since about '68. My last 40 years were with FAA. Retired since 2012. Enjoy the revibe. IMO it's one of those projects that any hobbyist should build, much more useful than just one more amp!   

PS... I lied about my LED. It's actually orange. Still gives 1.7v drop.

[shooter]

Let’s just say I fix things for a living...  ;)

 
sounds like the guys in the 'lil house above my Radar when I was In.
I did most of the board repair for the Ship, they would hand me a scrap of paper for a schematic n tell me "that's ALL you get" 
moved on to Ct and MR scanners in civie cloths 

[BobSmith]

I did a combination of solutions for the noisy grid wire.  I exchanged it with a shielded cable and ran it the other way so it wouldn’t cross over the high voltage line.  Very good results!  No noise!

So the insides have been buttoned up.  I only need to attach a ground wire from chassis to cage top and we’re done!

A very satisfying build, and fabulous sound to listen to!  I like the switch to the LED, because it does give a slightly broader range of speed (lower frequency with LED, higher with R/C).  there is some small tonal difference between the 2 when accounting for speed, but it’s really subtle and I couldn’t say really what’s “better”.  Both swamply good!

Here are some build shots...

[sluckey]

Very nice job. Lot's of layout planning and it shows.   

[BobSmith]

I was thinking of building a second one of these. 

I’m very happy with how this one turned out.  But I think for my next one I will simplify a few areas.  I will lose the doghouse and put the caps inside.  The filter cap board should fit on the opposite side of the chassis.  That will leave more room on the top, shorter wires and less holes to drill (mounts for doghouse).  I will remove the lift switch, it’s always in the lift position, so why bother.  I’m still not sure about the LED switch.  LED always seems to be on.  Perhaps the “Sluckey” slow switch is a more useful feature.  Again, simpler.