Deluxe Reverb Tremolo Delayed Start

[jim dresser]

Just replaced the opto in 1966 Deluxe Reverb.  Tremolo start with ft sw is delayed 20 seconds when starting first time, after which off/on works every time.  The opto is completely sealed in heat shrink.  Put original opto back in, instant start but it drops out every 10-12 pulses but starts up on its own.  Replaced all drifted upward resistors, suspect bypass caps, tried different 12AX7 tubes, tried a shorting RCA plug, and voltages all measure close to AB763 schematic.  Also tried another opto that is open on each end, not totally sealed, it has instant start but loads down the amp volume of the guitar signal when on with intensity at 5 or greater, but volume goes to normal at intensity of 1.  So I'm out of ideas.  Any suggestions are welcome.

[ginger]

Hard to find a broken 22ga wire, hidden under the cloth . Check continuity for every wire going to the Trem ... ground wire to the pedal jack too .

[shooter]

suspect bypass caps

Good choice
not sure your schematic, the one I found here showed ~2.1vdc on left cathode, ~ 5vdc on right cathode
since you tried other roach's that changed symptoms wouldn't rule out your new one.

[PRR]

An opto with one end open, in normal room light, WILL reduce signal. Turn off the lights to prove it. Then seal the end with something opaque.

The resistor end of the opto has no effect on the oscillator start-up. The string of 0.02-0.01-0.01 caps is a likely suspect. After that: note the connection to the power tube bias supply. A simple LFO can take seconds for oscillation to build-up. This rig takes -50V to cut-off, which throws a 50V kickstart into the works when LFO is enabled, which kicks the first cycle into near full strength. Bad parts or mis-repairs can happen here.

[jim dresser]

Continuing to hunt down the cause of tremolo delayed start.  When the amp has been off for 24hrs, caps drained, the delayed start is present.  Sequence:  power on, warm up 2-3 min, flip stand by sw to play, turn up vol to 3, play guitar, click Vib ft sw (which is new BTW), tremolo starts in several seconds.  It sounds good at all speeds & intensity.  Once started it will work fine, off/on. 

Back tracked by installing original .01, .01, .02 disk caps, waited all day with amp off, still a delayed trem start again.  I can try another new sealed since I bought two.  The kick start -51v is always present at the 220K and 1M resistors and goes to 0v when the ft sw is clicked to start the tremolo.  Maybe the sealed roach requires an increase in 12AX7 bias ??? 

So, replacing all drifted resistors and new caps in the tremolo ckt didn't have any change on this operation with the sealed roach. Suggestions?

[tubeswell]

Read what PRR said above about the trem kick start circuit voltage. Check that the large resistor (connected to the power tube bias) is connected etc.

[shooter]

12AX7 bias

that's sorta basically the cathode voltages which from schematic 2.1v 5V
and ya, roll in more 12ax, try 12AT, new roach, what PRR said

[jim dresser]

Problem started with the install of the new sealed roach in an effort to correct the original roach drop out. 

The kick start -51v is at the 2.2M, 1M, 1M junction and goes to 0v when the ft sw is clicked, just fine.  No issues at bias board, -51v there , too.  Bias pot works ok for power tubes.  The cathode bypass caps have been replaced along with all resistors in the circuit.  No single or complete R or C replacement has changed this delayed start issue.

Once started, the effect works fine and sounds great, but why is there a delayed start upon amp turn on after having been off for some time?  Cap charge time for LFO to start its cycle??  Do roaches of different types have different loading on the circuit? 

1. I'll try more 12AX7 tubes on cold amp
2. I can put in another sealed roach and try it on a cold amp

Thanks for all suggestions.

[shooter]

why is there a delayed start upon amp turn on

connect a meter VDC to the plate of your AX.  make sure amps been off long enough to discharge everyone, hit power switch and stopwatch together, stop watch when AX is at normal B+ and stable. 

[sluckey]

The kick start -51v is at the 2.2M, 1M, 1M junction and goes to 0v when the ft sw is clicked, just fine.

Measure at pin 2 of V5 (meter probe directly on the socket pin. What have you with trem on and trem off?

[jim dresser]

Sluckey - thanks for the reply. 

V5 pin 2 with tremolo off is -38.3v

V5 pin 2 with tremolo on is oscillating reading of 1.3v to ~3v, hard to see on DVM

I'll measure voltage again tomorrow after all caps drain to see what is at pin 2 when ft sw is pressed and if there is a delayed tremolo start.

[jim dresser]

After a few days off, I'm back on the delayed tremolo start issue.  All of the R, C components and roach in the tremolo ckt except speed and intensity pots have been replaced.  All solder connections are solid.  V5 pins and solder connections are sound.  Continuity checks found no poor grounds.  Ft sw is new and connections are sound.  Tried a new JJ 12AX7 for V5.  The kick start voltage -51v drops to -39v at V5 pin 2 prior to clicking the ft sw to start the tremolo.  When the ft sw is clicked, voltage at V5 pin 2 decreases to ~ 0v then begins to oscillate between -1.3 and 3.0v.  The decrease in kick start voltage takes a few seconds to drop prior to the oscillation.  Again the 5-10 sec tremolo start delay only happens upon 1st start attempt when the amp has been off for a long time.  After is does start, switching on/off/on is instant every time.  I'll try another sealed roach, even though no delayed start when the original roach was re-installed.  Stumped.

[sluckey]

The decrease in kick start voltage takes a few seconds to drop prior to the oscillation.

That should happen instantly. Changing roaches should have no effect on how fast the oscillator starts.


Not saying this is a ground issue but you can't check the quality of grounds with a meter set to continuity mode. Most meters will beep with anything < approx 70Ω.

[ginger]

Use a alligator clip jumper , to every connection

[jim dresser]

Ok I think it's DVM reaction time as the voltage a V5 pin 2 changes from -39v to the oscillating -3v to +1.3v.  Anyway, there are three scenarios, all else being constant:
Start up condition - amp powered off for 24 hrs or longer
1. Original opto - instant trem start but random drop out, trem is weak
2. New sealed opto - delayed trem start but instant on/off/on thereafter, and best trem sound
3. New open ended opto - instant start, instant on/off/on thereafter, but amp volume is lower when tremolo is on

I can live with #2 scenario, but will continue to find the cause.  Thanks for all comments.

[jim dresser]

8/5/19 update -
Start up condition - been off >24hrs, switch on, warm up 3 min, take off standby, monitor V5 pin 2 = -39v.  It is -51v coming from the bias board and drops to -41v across 2.2M, drops to -39v across .01 cap and lead wire to V5 pin2.

Press vib ft sw, appears that LFO starts, -38v at V5 pin 2 goes to 0v but oscillates between -3v to 1v on DMM. 
Tremolo working.

If V5 is in cutoff, should the voltage at V5 pin 2 be oscillating or solid 0v? 
Is this oscillating voltage at V5 pin 2 an indication of fiber board conduction?

I've read old posts on other forums describing this issue and none of the posts ended in a root cause or solution.  Thoughts?

[sluckey]

If V5 is in cutoff, should the voltage at V5 pin 2 be oscillating or solid 0v? 

Neither. The tube is in cutoff when pin 2 is -38vdc.

Is this oscillating voltage at V5 pin 2 an indication of fiber board conduction?

No. It's an indication that the oscillator is working properly. If you look at pin 2 with a scope you will see a small low frequency sine wave that has a center baseline of zero volts. The sine wave will be swinging positive and negative around this zero volt baseline.


Everything you said today indicates the oscillator is working just fine like it should. You didn't mention any startup delay? BTW, startup delay would be hard to see with a dmm because of the slow meter response. Need a scope or just use your ears.


As for your roach that has a slow startup, I'd cut a small hole through the shrink tubing so you can observe the neon bulb flashing. Do you see any delay in the bulb flashing? You can use some nail polish to seal up the hole after testing. Someone you know may even have some black nail polish. 

[jim dresser]

Sluckey - thanks for the reply regarding the tube cutoff question. 

Yes the tremolo started without detectable delay and I could hear it at the speaker. 

The oscillating voltage monitored at V5 pin 2 was instant or as fast as the DMM can display.  The amp was on the bench under a bright desk lamp during this test.  The instant start may have been just luck this time.

I'll try the next test with the chassis inserted into the cab.

[jim dresser]

8/7/19 update -
Tried it with chassis inserted into the cab, delayed start as usual.  On the bench took some measurements:

 - kick start voltage now only -48v at bias bd, dropping to -38v at V5 pin 2
 - 470 ohm resistor on bias bd measures 496 ohms
 - only 48v is available when bias pot is turned fully CCW, 10K bias pot range is 1 ohm to 9.86K

I suspect the 470 ohm resistor has drifted too high and is limiting the kick start volts at V5.  Should the bias bd voltage range go higher, let's say -60v???    What about the 1N4007 diode causing low voltage on tap??

So after 20sec the tremolo starts, slow start vs kick start.  Then, after it starts it will click on instantly every time.  Power amp off for 10 - 24 hrs, it's delayed start 1st time again.  Once started the trem ckt seems to work fine.

What should the bias ckt voltage be to produce enough kick start voltage at V5, pin 2  ???  Anyone have verified measuremet vs specs for Deluxe Reverb AB763?

Thanks

[shooter]

What should the bias ckt voltage be

It should be the correct voltage to put your 6V6's in a happy place, kickin the trem is like a 2fr 

[sluckey]

-38V is  exactly what you should have at V5-2. It's plenty. The tube will be completely cut off and zero current will flow. Going to -50 ain't gonna do anything.


The kickstart occurs when that -38v is instantly switched to zero. That instant transition from -38 to zero is like slapping that tube into conduction/oscillation. This circuit will slowly build up to oscillation even without any outside help, but it takes several seconds. The kickstart simply ensures the oscillator starts immediately. No one likes to stomp a pedal then wait several seconds for the effect. Musicians are especially impatient.   


But you say your oscillator starts instantly or as fast as your meter can tell. That says to me that the delay is happening somewhere else, probably in the LDR inside your roach. That's why I want you to cut a window in the shrink tube covering the neon lamp and LDR. Then you can see if the neon flashes immediately. Nail polish will quickly/easily close the window when you're through testing, or you can put another piece of shrink tubing on it.

[jim dresser]

Deluxe Reverb Delayed Tremolo Start Update  -

On 8/8/19 I snipped the opto heat shrink to create a small hole on the lamp end only.  I powered on the amp, clicked on tremolo, and observed the instant start of the neon flicker thru the small opening in the heat shrink.  V5-2 voltage measurements as before, -38 to 0.  I installed the chassis into the cab and waited 6-7 hrs.  Warmed up the amp, put it to play, clicked the Vib ft sw, instant neon flicker as seen thru the ext spk jack and instant tremolo!  Intensity up/down does not alter guitar loudness.  Speed works across the range of the pot. 

On 8/9/19 tried it this am and again this pm, still working OK!  No additional changes or parts swapped.  All seems fine.  A small opening in the heat shrink allowed the instant tremolo start when completely sealed it was delayed start .  Why???

[jim dresser]

Deluxe Reverb Delayed Tremolo Start -

Thanks to Sluckey and everyone who offered suggestions and encouragement to assist me with this issue. 

Regards

[sluckey]

I think your roach is cursed. Time to get out the Bengal's.   


I don't think exposing the neon bulb is a cure. Seal it up again, unless you are tired of chasing gremlins. Play some Duane Eddie.

[jim dresser]

I don't understand why it is working now.  Previous posts on the issue I listed the following regarding different roaches:
- original open ended - instant start, random drop out, but starts up again on it own
- 1st replacement sealed both ends - delayed start 1st time, after which works/sounds great.  Tried two fully sealed, both same symptom
- 2nd replacement open ended - instant start, reduced program volume, less intensity
- modified 1st replacement with small hole on lamp end, LDR end sealed, instant start, sounds great

So why does the circuit prefer an open vs sealed roach?  None of the replaced components or swapped tubes changed the delayed start with the fully sealed roach.  I'll continue to test the amp after 24hrs.  I'll seal the hole and try again.  Not giving up, yet.

[sluckey]

2nd replacement open ended - instant start, reduced program volume, less intensity

Just a guess... The ambient room light was affecting the LDR causing it's resistance to decrease. Steady ambient light equals steady lower resistance. Lower resistance means less intensity and reduced program volume. I bet if you put this roach back in and make the room totally dark (or tape the roach up so no room light can get in) it will work fine with plenty of intensity and plenty of program volume.

So why does the circuit prefer an open vs sealed roach?

I don't think it does. How could it know anyhow?

I'm coming to believe that the roach is the problem, not the neon lamp part, but the LDR part. Different LDRs have different characteristics. They have different resistance ranges. Some may have an ON (illuminated) resistance of just a few hundreds of ohms and an OFF (dark) resistance of many megohms. Others may go from several K ohms to several 100K ohms. They also have different delay characteristics, ie dark to illuminated times vary widely between LDRs. And illuminated to dark times are usually different from dark to illuminated times.

If these LDRs are operated with ANY ambient light, all this stuff changes unpredictably. The LDR needs to be in complete darkness so that the only influence on it will be the intended light source.

The experiment with cutting a hole near the neon light was not meant to see how much ambient light affects the roach. I had only one objective... to see if the neon begins to flash immediately. And it does! That means the delay is not caused by anything in the circuit. It must be caused by the characteristics of the LDR used in these new roaches. Your only solution may be to buy a dozen "Fender" brand roaches and hand pick the one you like best. BTW, where are you getting your roaches? And who actually made them?

I know this is an old amp and you probably want to keep it stock. But if you are more interested in a strong, noise free, hassle free trem, let me know. I'll introduce you to the Trem-O-Nator! It'll blow your roach out the door... 

[PRR]

Conversely, I will theorize that the neon is old, "weak", does not want to ionize. When new, a pinch of radioactive is added to neons. The half-life is on the order of decades, so old neons get hard to start. Light will also help to start them-- I have old pilot lights which go out in the dark and start to flicker when the room light hits them.

Between neons and PbS photoresistors, it is a wonder roaches ever work. I was just reading Philips' 1958 article on photoconductors. The chemistry is amazingly fussy and not well understood. Hours of critical cooking. Impurities Copper and Gallium play a part but must be in some range of proportion to the CdS and the other impurity. The least touch of Iron poisons the effect. The breakdown of Neon at not-huge voltages is a statistical event- there is some chance it will never start without some kick of ionization.

So I dunno. New neon from anonymous asian factory? LED with totally different V/I? A THAT VCA chip with a TapFlo LFO PIC?