Can’t find the pulse - Selmer tremolo is trem-o-no

[Jennings]

Recently picked up a Selmer New Standard TV8/T from 1960 as an untested but potential runner. Looked really untouched and undamaged, and had the original Mullard valves too. The Mercury 5 is the same circuit, but with added valve for trem pulse rate indication.


Replaced all the electrolytics today, and tested the valves. All were strong, bar a tired ECC83. Replaced that, cleaned the sockets and jacks, and fired up on a lightbulb limiter. Worked great…very bright amp really, but a cool tone to it.


However…no tremolo! Plate voltage for V2b looks OK against the annotated schematic, grounds all tested OK, checked my connections and they traced OK, measured resistances were mostly a little high (as I expected due to age…see below) but the only one that was way off was R19. I measured 6,66M against a spec of 3.3M. I added a 5M6 in parallel and measured 3.38M now.


So I guess I’m either getting no pulsing, or all the trem is being shunted to ground somewhere? I can hear a slight tonal difference when I rotate the depth and speed controls, but no vib/trem. I’m really scratching my head on this, especially as I’ve replaced only a couple of components in the trem circuit, and all solder joints looked to be virgin red-stained from factory. The trem circuit caps look fine, but I haven’t removed them to test. The only ones I replaced were the dreaded black, crumbly Hunts. Grateful for any suggestions re troubleshooting. Pots? Reflow the virgin joints perhaps?


R9   = 1.139M (1M)
R19 = 3.38M   (3.3M)
R20 = 3.94K   (3.3K)
R21 = 207K    (220K)
R22 = 2.84M   (2.2M)
R23 = 679K     (510K)
R24 = 53.1K    (47K)

[sluckey]

VC3 should be 2.5M, not 220K

[Jennings]

VC3 should be 2.5M, not 220K

Thanks…definitely a typo 👍


I’m going to work through the caps in the next few days.

[pdf64]

VC3 should be 2.5M, not 220K

That’s possible, but I think there’s a fair chance the pot really should be around 220k.
Bear in mind that the modulation oscillator output signal will be maybe 50 - 100V, and it needs bringing down to a level suitable for a 12AX7 CC stage grid.

[sluckey]

I was just going by the info in the parts list.

[Jennings]

Well after replacing the three feedback caps in the trem section I have a pulse! However it’s still not quite right yet. At max depth I can get an increasing pulse on the speed knob up to half way, where instead of being rapid vibrato it sort of gradually stops pulsing. No discernible oscillation…and to restart the throb again you have reduce depth to zero, reset the speed to either minimum or maybe a quarter turn, abs then turn up the depth again. Progress, but not fixed yet.

[sluckey]

Replacing the oscillator cathode resistor and bypass cap with a 5mm red LED just might be an easy fix.

[PRR]

What he said: a SOLID cathode network overcomes lame-brain design and parts aging.

Engaging brain: the 47k speed-stopper is a heavy load on a 12AX7 with 220k plate resistor. Normally we want load larger than plate resistor. But there my brain stops. Instead I would plagiarize values from a simple Fender trem oscillator.

[Jennings]

Thanks for the pointers guys…much appreciated.


Current status is that I’ve replaced all the ERO caps in the trem circuit, and now have a nice, strong throb at low frequency. It still stops oscillating at halfway through n the speed knob…which is way before vibrato type speeds, but definitely a reasonably rapid trem.


I tried subbing the cathode resistor and cap for a red LED, no other changes, and got no trem at all. I’m tempted to do as suggested next by changing the cap, resistor and pot values using the Valve Wizard site as a guide:
http://www.valvewizard.co.uk/trem1.html


Seems like I just either lose too much gain or dump too much to ground as the speed knob is increased.

[sluckey]

I tried subbing the cathode resistor and cap for a red LED, no other changes, and got no trem at all.

Maybe the LED was connected backwards. The cathode must connect to ground. And the LED must be naked, IOW, no series resistor.

[Jennings]

That’s was my first thought too…checked it, and visibly it was the correct way round, but I even tried switching it just to be sure. It does go directly to ground too, from cathode pin to directly chassis attached tagboard tag. Switching back to the 50uF cap with parallel 3.3k resistor gets it tremming again. A bit of head scratcher!

[sluckey]

You should probably just throw that LED in the trash since you have connected the anode to ground. The LED will quickly die when connected backwards.

[Jennings]

It was only an odd spare trial one anyway that I had laying about, so it was always destined for the trash. It just stayed permanently lit when it was round the correct way, and as expected not at all the wrong way round.

[kagliostro]

Just curious

if a LED can be feed with AC (and It lights) why he die if connected backward ????

Franco

[sluckey]

if a LED can be feed with AC (and It lights) why he die if connected backward ????

Hopefully you will use a current limiting resistor in series with the LED. For one half cycle of the AC, the LED will be forward biased and it lights. The next half cycle the LED will be reverse biased and will not light. But most small LEDs only have about a 5V rating for reverse voltage. Exceed this reverse rating and the LED dies.

[PRR]

if a LED can be feed with AC (and It lights) why he die if connected backward ????

The old red LEDs would take a lot of abuse. As long as I had a series resistor so the average current was less than half the rating, they always worked for me.

The newer ones are much brighter but apparently not as tough. (So maybe just scale-up your series resistor?)

Also the White is a significantly different device which may be more delicate.

[Jennings]

It’s an old red I had laying around (can’t even remember where that came from) and tried out. Odd that just swapping the cap/resistor for that resulted in always-on, no oscillation. Must have upset the bias and killed the feedback loop. I’ll get hold of some decent LEDs and try another, as the one I found in my parts try could easily have been some kind of oddball.


If the LED swap along doesn’t get oscillation more stable across the range of the speed sweep I’ll probably go the route of rebuilding using the Fender values and the Valve Wizard details as a guide.